Stanford Geothermal WorkshopFebruary 8-10, 2027

Predicting the thermal performance of enhanced geothermal systems requires understanding the distribution of inter-well flow paths, the mechanism sustaining fracture conductivity, and the rock volume accessed for heat sweep. At Utah FORGE, integrated diagnostic datasets provide an opportunity to connect observed hydraulic behavior with its mechanical controls and thermal performance. This study integrates field diagnostics, geomechanical analysis, and semi-analytical modeling to characterize the inter-well connectivity network between wells 16A and 16B and constrain thermal breakthrough predictions. Fiber optic crossflow strain measurements, microseismicity, production logging surveys, thermal slug analysis, tracer responses, and circulation measurements are combined to construct an inter-well connectivity map and assign flow contributions to candidate pathways. Geomechanical analysis evaluates fracture slip tendency under a representative normal-faulting stress regime and pressure conditions relevant to injection and circulation. The inferred connectivity map and flow allocations inform an infinite-plane semi-analytical model of two-dimensional fluid flow and heat transfer, reducing dependence on prescribed finite fracture dimensions and heat-exchange surface area. The integrated diagnostics framework complements multiple observations, and interpretations to identify preferential flow pathways. Fracture orientations inferred from frac-hit observations align with orientations favorable for shear reactivation in the Utah FORGE stress field. This agreement supports reactivations of preexisting fractures as a mechanism governing inter-well connectivity. In addition, sustaining the critically stressed state for these fractures during circulation provides a physical interpretation for hydraulic communication. Consequently, the infinite plane model links the inferred flow connectivity map to thermal breakthrough predictions while mitigating the uncertainty in the heat exchange surface area. Together, these results establish an informed connection between reservoir diagnostics, fracture mechanics, and thermal performance. The approach provides a basis for evaluating how alternative connectivity interpretations affect long-term heat extraction and for guiding reservoir characterization and circulation strategies in enhanced geothermal systems.

Topic: FORGE

Geothermal well architecture is becoming substantially more complex. Horizontal and multilateral enhanced geothermal systems and closed-loop configurations require downhole tools, including packers, sliding sleeves, flow control valves and casing wear, to be slid over long, deviated, and thermally demanding intervals before reaching setting depth. As a result, the downhole tools arrive at depth in a condition different from the one in which they were qualified, and the wear they accumulate during the run-in remains largely undocumented. This paper introduces a new testing apparatus developed to evaluate the wear resistance of downhole tools. The complete tool assembly is attached to a reciprocating system that moves it into and out of a casing section, traverses a given length of the section, and returns, so wear accumulates over repeated passes. A dead weight of several hundred kilograms can be applied above the downhole tool, allowing the contact severity to be varied and the influence of applied load on element degradation to be examined. During reciprocation, a single face of the element remains in contact with the casing. The apparatus also permits the downhole tool to be indexed rotationally between runs, so that additional faces can be exposed in separate tests. The apparatus forms part of a broader laboratory capability for downhole tool qualification. The authors also operate a high-pressure, high-temperature (HPHT) downhole tool testing facility, described in previously published work, in which a worn element can subsequently be qualified.

Topic: Drilling

Deploying downhole tools in geothermal wells requires testing under representative high-pressure, high-temperature (HPHT) conditions that conventional oil and gas standards do not follow. This paper presents a suite of experimental setups developed and designed at the Well Construction Technology Center (WCTC) at the University of Oklahoma (OU) to evaluate downhole tool performance and survivability before field deployment. This work was partially funded by two FORGE projects over the past 5 years. This paper describes four geothermal simulation tests of downhole tools conducted at the WCTC testing facility at OU. The first experimental setup is an HPHT testing setup capable of simulating downhole geothermal conditions at 10,000 psi and 500 °C. An induction unit delivers heat, while pneumatic pumps control the pressure distribution within the testing system. An array of sensors interfaced with a DAQ card and DASYLab acquisition software monitored pressure and temperature in real time, enabling continuous data logging. The second setup is a pressure chamber used to qualify the downhole tool under pressure. This pressure chamber is rated to 30,000 psi and provides a faster screening option for seal and housing integrity. The third setup addresses mechanical wear of downhole tools under representative loads, which can take place when the tool slides against the casing to reach the desired depth in the well. The fourth setup employs casing containing localized and non-localized defects to evaluate the detection capability of wireline inspection tools under a standardized inspection procedure. The setup can also be reconfigured as a concentric cemented casing to evaluate detection through the cement sheath. These testing setups provide a stage-qualification pathway, from component-level pressure screening to wear testing, and then to downhole tool evaluation under HPHT conditions to assess integrity. Moreover, this testing suite verifies the inspection tool's operation under simulated geothermal conditions.

Topic: FORGE

Nitrogen-doped carbon quantum dots (NCQDs) were synthesised from spent coffee grounds via solventless mechanochemical method at three different N/C ratios (0.5:2, 1:2 and 1:1) to study the effect of N-doping amount on the thermal properties of the dots. NCQDs were examined using FTIR and UV-Vis spectroscopy, SEM, TEM and DLS and their thermal conductivity, thermal resistivity, volumetric heat capacity and thermal diffusivity were measured in aqueous dispersions (0.3, 0.5 and 1.0 wt%) and natural soil (0.5 wt%) via transient line-heat-source analyser (KD2 Pro). The response is highly dependent and strictly monotonic to the amount of doping. The dots with the lowest amount of doping (0.5N:2C) show an increase in the thermal conductivity of water depending on concentration up to 0.69-1.22 W m−1 K−1 at 1.0% (+15-100% relative to water), 1N:2C dots show a smaller increase (0.60-0.87 W m−1 K−1 at 1.0%) and 1N:1C dots do not change conductivity of water or even decrease it down to 0.51-0.59 W m−1 K−1 and enhance the resistivity with concentration. In soil, the dots with 1N:2C ratio have higher thermal conductivity (1.26-1.81 W m−1 K−1) and diffusivity compared to 1N:1C dots (1.29 W m−1 K−1, 0.57 mm2 s−1). The negative correlation between the amount of nitrogen and enhancement of conductivity can be explained by phonon scattering due to substitutional nitrogen and associated lattice defects in the sp2 core, which reduces the intrinsic conductivity and raises interfacial thermal resistance of dots in liquid environment, hence providing another way to tune the thermal properties of coffee-waste-derived NCQDs either as thermal conductivity enhancers or thermally inert additives.

Topic: Low Temperature

This paper presents a hybrid physics-informed deep learning framework for modeling and forecasting subsurface temperature distributions in a real Aquifer Thermal Energy Storage (ATES) system in Stockholm, Sweden. Operating since 2017, the system serves two commercial buildings (~30,000 m²) through three warm and two cold wells. Continuous temperature and flow measurements, complemented by distributed temperature sensing (DTS) at 0.25 m vertical resolution, provide a detailed spatiotemporal thermal dataset. Deep neural network (DNN) and physics-informed neural network (PINN) architectures are developed to model and forecast the three-dimensional aquifer temperature field. Hyperparameter optimization and model comparisons evaluate predictive accuracy, computational efficiency, and generalization. The models predict weekly temperature evolution with up to 95% accuracy under varying heating and cooling demands. The results demonstrate that physics-informed deep learning can complement numerical groundwater models, enhance ATES operational control and thermal recovery efficiency, and provide a scalable framework for data-driven geothermal energy management.

Topic: Low Temperature

This paper presents a component-based OpenModelica model and techno-economic assessment of an aquifer thermal energy storage (ATES) system integrated with heat pumps and district heating for commercial buildings. Four scenarios compare constrained and unconstrained aquifer extraction temperatures and heat pump/district-heating shares of 60/40% and 90/10%. ATES supplies 29–50% of annual heating and 41–57% of cooling demand. The 90/10% scenarios achieve improved thermal balance, with monthly performance factors reaching 8 for heating and 55 for cooling, and up to 25% greater free cooling. Mass-flow modulation improves heat-exchanger effectiveness to 0.95. Techno-economic analysis evaluates operating costs, energy expenditures, investment feasibility, and sensitivity to electricity and district heating prices. By linking energy performance with economic indicators, the study examines trade-offs between operating strategies, energy costs, and long-term aquifer sustainability. The framework supports cost-effective ATES operation, reduced district heating dependence, and improved geothermal energy utilization.

Topic: Low Temperature

Accurate wellbore positioning in geothermal drilling is frequently complicated by the presence of highly magnetic igneous and metamorphic basement rocks. This paper demonstrates the viability of utilizing high-resolution In-Field Referencing (IFR) to correct Measurement While Drilling (MWD) magnetic surveys in these magnetically challenging environments. Specifically, this study addresses the downward continuation problem, the complex geophysical challenge of projecting surface magnetic anomaly data to depth when drilling near or directly into magnetic basement rock, and whether currently used downward continuation transformations are sufficient for these geothermal wellbores. In a collaboration between a directional drilling service provider and a geomagnetic modeling company, a localized, high-resolution IFR magnetic model was developed and deployed for a directional geothermal well drilled in Beaver County, Utah. Traditional global magnetic models often fail to account for the steep magnetic gradients associated with basement structures, leading to significant positional uncertainty. By integrating the IFR model, the drilling team was able to accurately characterize the local crustal anomalies and mitigate the majority of magnetic interference expected from the geologic basement. The MWD survey data was continuously corrected using the depth-specific IFR values, and the results were compared against standard survey management protocols to evaluate the applied toolcode and reduction in the Ellipse of Uncertainty (EOU). The application of the customized IFR model sufficiently resolved the localized magnetic field while drilling horizontally within 1000 feet of the granitic basement rock in Beaver County. Incorporating the IFR data resulted in a quantifiable reduction in azimuthal uncertainty, allowing for precise wellbore placement within the targeted geothermal reservoir. The data suggests that the magnetically challenging drilling environment can be managed through accurate crustal modeling, limiting the positional variations typically induced by highly magnetic formations. Consequently, the reliance on specialized, high-temperature-rated gyroscopic surveying tools was minimized, optimizing both operational time and drilling costs. While IFR is a proven technology in conventional oil and gas operations, its application in deep, high-temperature geothermal wells penetrating basement rock remains limited due to concerns over crustal magnetic unpredictability. This paper provides empirical evidence that, with precise modeling, IFR magnetic services offer a reliable and cost-effective alternative for wellbore positioning in the geothermal sector.

Topic: Drilling

Public understanding plays an important role in geothermal development in Indonesia. Existing geothermal education initiatives are often done through project socialization, school visits, or education centers within geothermal power plant areas. While useful, these approaches may have limited public reach due to access restrictions to the power plants and scheduled visits. This paper proposes integrating geothermal education centers into existing hot spring bathing and resort facilities as an alternative public engagement platform. These facilities provide accessible locations where visitors can directly experience geothermal heat, creating opportunities to introduce geothermal systems, electricity generation, direct-use applications, and environmental and safety considerations. The study reviews selected geothermal education center concepts, site observations at education facilities including Kamojang, and hot spring tourism facilities in West Java, particularly Subang and Garut. Preliminary mapping of geothermal power plants, development prospects, and nearby hot spring destinations is also conducted to identify potential locations for implementation. The proposed model involves collaboration between geothermal operators or developers, hot spring facility owners, and local governments. It considers stakeholder roles, potential funding mechanisms, educational content, exhibit options, implementation costs, and site selection criteria. The approach may support geothermal literacy and community engagement while providing additional educational value to existing tourism facilities. The paper discusses implementation considerations and limitations, including content accuracy, facility maintenance, and coordination among stakeholders. The proposed education centers are not intended to replace existing community engagement programs or directly resolve social acceptance issues, but to complement them by making geothermal energy more accessible and understandable to the public.

Topic: Direct Use

Lost circulation is a common challenge in geothermal drilling, particularly in non-productive intervals above the reservoir. When lost circulation material (LCM) is ineffective, cement plug placement may be required to restore circulation or achieve wellbore isolation. In contrast, cement placement in reservoir intervals is generally avoided to prevent damage to potentially productive fractures. Under total lost circulation conditions, cement placement is complicated by uncertain loss-zone geometry, slurry losses into fractures, and the risk of cementing the steel stinger in place, particularly when faster-setting cement systems are used. This paper assesses the potential application of fiber-reinforced plastic (FRP) pipe as a sacrificial stinger or tailpipe for cement plug placement in geothermal wells. Previously applied in oil and gas drilling, this method allows the lower FRP section to be disconnected from the steel drillstring and left in the cement plug for subsequent drill-out. Published oil and gas case studies are reviewed to assess its applicability to geothermal drilling, particularly in upper hole sections (26 in. and 17-1/2 in.) above the reservoir. The assessment considers geothermal-specific constraints, including temperature exposure during cement placement and wait-on-cement (WOC), pressure and mechanical loading, connection integrity, cement compatibility, and drill-out performance in hard volcanic formations. The findings suggest that sacrificial FRP stingers may be an alternative for selected cement plug operations where conventional lost-circulation treatments are ineffective and the risk of cementing the steel tailpipe in place is high due to fast-set cement. Further field-specific testing, equipment design, and pilot trials are required to establish operational limitations and applicability under representative geothermal conditions.

Topic: Drilling

Acidic fluids in high-temperature geothermal systems create reservoir and well integrity management challenges beyond corrosion control because acidity can affect well integrity, long-term production sustainability, and optimum reinjection strategy. Although field experience exists, lessons are often fragmented across geochemistry, reservoir engineering, and casing or cement materials studies, limiting their transferability between fields. This study reviews reservoir and well integrity management lessons from geothermal systems where acidic fluids have affected field development and operation. A comparative review examines three case studies: Rotokawa (New Zealand), Krafla (Iceland), and Los Humeros (Mexico). These fields represent different acidic-fluid conditions, including magmatic inputs and acidic steam condensation. Conceptual models, geochemistry and mineralogy studies, reinjection histories, monitoring data, and well integrity reports were evaluated to compare acid-fluid origins, operational impacts, mitigation responses, and trade-offs. The review indicates that acidic-fluid risks are site-specific. At Rotokawa, a CO2-rich aquifer above the deep reservoir contributed to severe external casing corrosion in several production wells. Early shallow reinjection helped resaturate and partly neutralize this corrosive aquifer by diluting or displacing CO2-rich fluids and suppressing boiling that exsolves CO2. However, this strategy also raised concerns over ground uplift. Reinjection was therefore shifted deeper and farther from production areas, improving reservoir pressure support, reducing uplift, and limiting cooling breakthrough, while allowing corrosive conditions in the CO2-rich aquifer to re-emerge. At Krafla, the IDDP-1 and IDDP-2 wells demonstrated the high energy potential of superhot geothermal systems but also revealed severe operational constraints from corrosion, material degradation, and casing thermal shock associated with shallow magma intrusion, acidic HCl- and HF-bearing steam condensate, and extremely high temperatures. At Los Humeros, acidic conditions are linked to acidic steam condensation and possible magmatic fluid inputs. Collectively, these cases show that acidic-fluid effects should be managed as a coupled reservoir and well integrity problem rather than as separate or solely surface-facility issues. Effective management requires site-specific characterization of fluid sources and chemistry, adaptive reinjection strategies that account for reservoir response and well integrity, appropriate well and cement material selection, and continuous reservoir and well surveillance from early field development through long-term operation as reservoir conditions evolve.

Topic: Field Studies

Faults and fractures can provide the permeability required for geothermal circulation, but the same structures can accelerate thermal breakthrough, compartmentalize flow, or reactivate under injection- and cooling-induced stress changes. Previous studies have predominantly represented wells as source/sink terms, reservoir-grid connections, or lower-dimensional inflow elements. Relatively few studies resolve the well trajectory together with explicit fault geometry, and those that do generally focus on reservoir-to-well inflow rather than coupled segment-by-segment wellbore hydraulics. Consequently, the interaction between explicit fault-controlled reservoir flow and a multisegment wellbore network, including internal pressure losses, hydrostatic effects, completion intervals, and complex well trajectories, remains comparatively underexplored. The current wellbore literature shows that friction, gravity, heat exchange, and transient well pressure-temperature behavior can materially affect geothermal system performance, while the fault literature shows that well placement and fault connectivity can control thermal sweep, seismic risk, and reservoir lifetime. To our knowledge, no prior study has explicitly coupled detailed fault representation with multisegment wellbore modeling in a geothermal reservoir framework. Here, we demonstrate the impact of fault geometry and damage-zone properties on geothermal production using this coupled approach.

Topic: General

Current geothermal techno-economic models, such as the Geothermal Electricity Technology Evaluation Model (GETEM), GEOPHIRES, and the System Advisor Model (SAM), typically represent stimulation costs as a single lump‑sum per well cost. These models do not resolve the discrete fixed and variable – time, energy, and materials – costs for stimulating EGS wells nor do they capture the operational granularity required for accurate project‑level cost estimation. To address these gaps, we have developed a bottom-up stimulation cost model that provides a mechanistic, component‑level representation of EGS stimulation economics. The model development followed a two‑phase workflow: initial formulation of a pseudocode architecture in Excel to define model inputs, outputs, and cost‑driving parameters, followed by implementation of the fully integrated workflow in Python. The modeling methodology decomposes stimulation operations into discrete cost units grouped into higher‑level cost categories that aggregate into the total stimulation cost. Cost categories are designated as required (e.g., water acquisition) or conditional (e.g., chemical additives). They are also classified either on a per‑well basis (e.g., total proppant cost) or on a per‑pad basis (e.g., mobilization and demobilization) to ensure accurate scaling for multi‑well pad development scenarios. The model runs on default engineering and economic parameters, all of which can be replaced with project‑specific data reflecting reservoir properties, equipment specifications, material types, logistics constraints, and local cost structures. Preliminary outcomes using the model defaults indicate that pumping obligations (including rental, operation, and energy), proppant requirements for propped stimulation designs, and zonal isolation plus plug drill‑out constitute the dominant cost drivers, collectively representing about two‑thirds of the total stimulation cost per well. Multivariate sensitivity analyses show that stimulation cost variability is primarily controlled by the number of wells per pad, pumping energy, proppant selection (e.g., local sand, Northern White sand, ceramic, resin‑coated), and type of fluid additives (e.g., slickwater, linear gels, cross‑linked gels, viscoelastic surfactants).

Topic: General

Indonesia holds vast geothermal potential essential for achieving its baseload renewable energy targets under the National Electricity General Plan (RUPTL); however, geothermal expansion faces severe capital expenditure (CAPEX) bottlenecks during exploration and development, where well construction accounts for 40% to 50% of total project investment. Because geothermal developers rarely own drilling equipment, they must compete directly for land rigs and specialized service crews with the financially dominant upstream oil and gas industry. When crude oil prices rise, petroleum operators absorb market capacity, creating severe equipment scarcity, dayrate inflation, and forcing geothermal projects to contract over-capacitated, high-cost rigs. To resolve these structural procurement failures, this study examines empirical well cost records and adapts proven supply chain mechanisms from Indonesia's mature upstream petroleum sector. The study analyzes a normalized historical dataset of 51 geothermal wells drilled in Indonesia between 1997 and 2023, adjusted to 2023 USD using the United States Producer Price Index for drilling services (reflecting a 1.3 cumulative inflation factor). Statistical evaluation establishes a typical well baseline cost of $6.64 million, revealing that rig services constitute the single largest financial burden, representing 26% to 45% of total well expenditures (median ~29%). Comparative field validation in Java confirms this dominance, with rig services consuming 44.52% of well costs in Field X (Central Java) and 41.95% in Field Y (East Java), followed by cementing, casing tangibles, and directional drilling. To lower risk-adjusted total evaluated costs and eliminate single-bidder tender failures, this paper synthesizes four strategic cross-industry procurement pillars. First, adapting the upstream petroleum Centralized Integrated Vendor Database (CIVD) operated by SKK Migas into a geothermal-specific registry (CIVD-G) standardizes administrative pre-qualification, eliminates redundant tender paperwork, and increases market transparency. Second, executing multi-operator campaign bundling consolidates well demand across operators to amortize mobilization and demobilization expenses—reducing unit well costs by up to 15%—while enforcing technical rig-sizing, such as selecting 1,200 HP compact mast units to reduce civil pad footprint. Third, transitioning to alliance contracting replaces rigid dayrate or traditional EPC contracts with collaborative 'pain-share/gain-share' models that equitably distribute subsurface geological uncertainties between operators and contractors. Fourth, structuring combined service packages under updated local content regulations—specifically Ministry of Industry Regulation No. 35 of 2025 and MEMR Regulation No. 11 of 2024—maintains legal TKDN compliance while preserving access to specialized high-temperature logging and directional tools. Ultimately, transferring these operational and commercial practices from the petroleum sector shifts geothermal procurement from reactive sourcing to strategic market-shaping, significantly reducing drilling CAPEX and accelerating national renewable energy deployment.

Topic: Drilling

The application of geothermal direct use within Indonesia's thermal bathing industry offers substantial potential for advancing sustainable tourism and empowering local economies. To guarantee environmental sustainability, public safety, and the standardization of facilities, the Indonesian government has instituted the Operational Worthiness Certificate (Sertifikat Laik Operasi, or SLO) requirement for industry operators. As this sector moves toward comprehensive formalization, a critical opportunity arises to streamline the SLO application process and expedite regulatory compliance. Utilizing field data and stakeholder insights gathered from local governments and thermal spring operators across Garut and greater West Java, this study pinpoints crucial areas for procedural improvement and capacity development. Key recommendations involve aligning Central and Regional notification systems within the Online Single Submission (OSS) platform, streamlining Geothermal Situation Map prerequisites for Micro and Small Enterprises (MSEs), and establishing uniform Technical Guidelines (Juknis) to assist local authorities in executing accurate field inspections.Rather than delivering a fully developed software solution, this research proposes a multidisciplinary conceptual framework integrating Geothermal Management, Public Policy, and Information Systems. Initially, we examine the potential of an integrated API architecture—drawing inspiration from local platforms like Subang's SINANAS—to theoretically facilitate seamless data exchange between regional and central permitting databases. This concept demonstrates how enhanced system interoperability can reduce administrative bottlenecks and misrouted permits. Furthermore, informed by an empirical capacity-gap assessment, we outline a standardized field verification methodology calibrated to the existing capabilities of regional governments.To catalyze future progress in the industry, this paper outlines a practical toolkit aimed at assisting both operators and regulators. The core strategic proposals include: (1) a comprehensive Technical Guideline (Juknis) framework that defines minimum field-testing criteria (such as temperature, pH, and H2S levels) and evaluates the integration of Regional Health Laboratories (Labkesda) for equipment validation; (2) the deployment of affordable, standardized Situation Map templates designed to alleviate financial pressures on MSEs; and (3) an intuitive procedural roadmap to guide operators through the existing OSS framework. Supported by conceptual system architecture diagrams, user-journey flowcharts, and capacity-building matrices, this research delivers a strategic roadmap to guide policymakers, support local enterprises, and drive the sustainable expansion of geothermal direct-use applications in Indonesia.

Topic: Direct Use

The Eagle Ford Shale gas condensate reservoirs represent a heat-extraction candidate because of their high temperature range (125 °C to 175 °C). These reservoirs were already fractured for unconventional hydrocarbon recovery, enabling Enhanced Geothermal Systems (EGS) development at lower cost than in other impermeable media such as crystalline rock. Interest in shale-to-geothermal demonstrations in formations such as the Utica Shale in the United States has grown recently. To the best of our knowledge, no previous studies have quantified the rock and engineering parameters that control heat recovery from repurposed condensate-window wells. This study aims to use publicly available data from the Eagle Ford condensate window to develop a simulation framework that ranks the rock and engineering parameters controlling heat recovery and tests the influence of residual condensate on recovery performance from the geothermal reservoir. Constant volume depletion (CVD) and constant composition expansion (CCE) data were collected from Karnes County to develop the thermodynamic model using WinProp. The thermodynamic model is integrated with a thermal solver (CMG-STARS) to conduct a sensitivity analysis of rock and engineering parameters that affect thermal recovery from a depleted gas condensate reservoir. The model used is a box model with one water injection well and one production well. The fractures are incorporated using a discrete fracture network (DFN) tool in CMG. This study investigates the sensitivity of rock parameters such as matrix porosity, permeability, natural fracture heterogeneity, initial reservoir temperature, rock heat capacity, and thermal conductivity to thermal recovery from the Eagle Ford condensate window. The study assesses engineering parameters such as fracture half-length, fracture conductivity, fracture spacing, and well placement. The study also tests the potential blockage of thermal recovery due to residual condensate drop-out. Sobol indices from a proxy-based sensitivity analysis are used to rank parameters by their influence on produced temperature and cumulative heat. The simulations are designed to determine whether thermal rock properties outweigh engineering parameters in controlling thermal recovery from the condensate window. Condensate blockage is expected to affect thermal recovery, as residual condensate may reduce injectivity and heat sweep. The study marks an initial stage in repurposing shale gas reservoirs for heat extraction, and the Eagle Ford condensate window is a relevant analog to the Jafurah sub-basin in the Middle East; hence, the study can be used as a reference case for prospective shale-to-geothermal demonstrations in the Middle East and North America.

Topic: Modeling

Emplacing small modular reactors (SMRs) at depth in boreholes replaces engineered surface containment with lithostatic and hydrostatic confinement from the host rock and fluid column. The concept rests on two questions. Can the host formation maintain containment under sustained thermal loading? And how reliably can the perturbed subsurface temperature field be predicted? Both are heat flow and thermal conductivity problems, the same ones that underpin geothermal resource assessment. The thermal characterization needed to qualify a site for containment therefore also quantifies the heat available for recovery. A single thermal model can then evaluate a deep borehole reactor both as a containment system and as an engineered geothermal resource. We develop this framework for the Williston Basin. Decades of hydrocarbon exploration and prior heat flow studies there, including temperature measurements in temporarily abandoned wells, provide a dense and well-constrained thermal dataset. Starting from an analytical heat flow solution in which temperature increases linearly with the natural logarithm of depth, we build a temperature- and pressure-dependent thermal conductivity model. We then use it to compute the thermal field around an emplaced reactor. From this field, we define containment criteria: conductive heat transport away from the source, thermal stability of the host formation, and long-term temperature evolution in the near-borehole region. The Williston is a conductive basin with gradients that are generally subeconomic for conventional geothermal power. An emplaced reactor supplies the high-grade heat the basin lacks, and the natural gradient adds to the thermal budget at depth. We outline a cascaded recovery scheme with a supercritical CO2 topping cycle, an organic Rankine bottoming cycle, and direct-use loads for agricultural and process heat. Borehole depth is treated as a design variable. Greater depth increases containment margin and recoverable enthalpy but also raises drilling cost, so depth selection becomes an optimization problem rather than a fixed target. The result is a consistent framework in which basin-scale thermal characterization serves as the primary screening criterion for deep borehole nuclear deployment.

Topic: Emerging Technology

In situ combustion generates thermal energy directly within petroleum reservoirs through oxidation of a fraction of the in place hydrocarbons. Although this energy is conventionally used to mobilize oil and enhance hydrocarbon recovery, substantial thermal energy may remain stored in the reservoir rock and fluids during and after combustion. This study investigates whether this residual heat can constitute a technically meaningful secondary geothermal resource. The objectives are to quantify the thermal energy generated and retained during in situ combustion, experimentally determine the fraction of stored heat that can subsequently be recovered by fluid circulation, and translate laboratory measurements into reservoir scale metrics of geothermal energy per acre foot. Ramp Temperature Oxidation, thermogravimetric analysis and differential scanning calorimetry, and laboratory scale combustion tube experiments are conducted using a synthetic porous medium containing Ottawa sand, 2 wt percent illite, crude oil, and reservoir brine. Twenty thermocouples distributed along the combustion tube monitor combustion front propagation and spatial and temporal temperature evolution. Produced gas composition and oxygen consumption are used to characterize oxidation behavior and constrain fuel consumption and combustion generated energy. A transient energy balance quantifies sensible heat retained by the rock and reservoir fluids and distinguishes stored energy from energy transported by produced fluids and gases and heat lost to the surroundings. Following combustion, temperature decline is monitored to characterize natural thermal depletion. Controlled water circulation through the heated porous medium is then used to directly measure transient thermal power, cumulative recovered energy, and thermal recovery efficiency. A self sustained combustion front was established at approximately 400 degrees C and propagated at an average velocity of 17.7 ft per day. Ramp Temperature Oxidation measurements identified low temperature oxidation at approximately 200 to 350 degrees C followed by high temperature oxidation above approximately 400 degrees C, while reactor temperatures exceeded 500 degrees C. The experimentally constrained thermal energy density is scaled to an acre foot of heated reservoir and evaluated under ideal and heat loss constrained recovery scenarios. This framework follows the complete energy pathway from combustion generated heat to reservoir thermal storage, experimentally recoverable heat, and potential surface thermal energy. The approach provides a quantitative basis for determining whether heat remaining after in situ combustion can support subsequent geothermal or direct use applications.

Topic: Reservoir Engineering

Enhanced Geothermal Systems (EGS) offer significant potential for expanding geothermal energy beyond traditional hydrothermal systems by engineering reservoirs that access heat from hot, dry rock. In EGS, subsurface operations encompass well drilling, reservoir creation and stimulation, and fluid circulation between wells to transport thermal energy to the surface. Surface facilities then convert this thermal energy into electricity. Sustained and reliable electrical generation from EGS requires a comprehensive understanding of subsurface behavior, wellbore integrity, fluid chemistry, and surface energy conversion. Mazama Energy's Newberry EGS pilot in Oregon represents the hottest EGS pilot demonstrated to date, featuring circulation between a pair of deviated injector-producer wells at a bottomhole temperature of approximately 331°C. The system demonstrates hydraulic connectivity between the wells and high thermal-density geothermal fluid recovery. Building on this foundation, Mazama is expanding the pilot to integrate surface power-generation facilities and evaluate overall system performance. The project investigates reservoir response, thermal recovery, circulation stability, and electrical generation efficiency. Key measurements include injection and production rates, wellhead pressures and temperatures, fluid recovery, and power output. Geochemical monitoring assesses changes in fluid composition, fluid-rock interactions, and potential scaling or corrosion mechanisms. Wellbore integrity assessments examine the effects of sustained high-temperature operation and thermal cycling on well performance and reliability. The integration of surface power-generation facilities provides additional insight into energy conversion efficiency, cooling requirements, parasitic power consumption, and operational stability. This project represents an important milestone in establishing the viability of reliable power generation from high-temperature EGS, while providing field data and operational insights that support commercial-scale geothermal development.

Topic: Enhanced Geothermal Systems

The Hawaiian Archipelago overlies an active mantle hotspot and contains substantial geothermal resource potential. Although commercial geothermal power production has been established on the Island of Hawaiʻi, current knowledge of geothermal resources across the islands remains geographically uneven. The University of Hawaiʻi's (UH) Hawaiʻi Groundwater and Geothermal Resources Center (HGGRC) leads a multi-year effort, supported by partners including Idaho National Laboratory (INL) and the National Laboratory of the Rockies (NLR), to address these gaps by building on UH's prior Hawaiʻi Play Fairway (HPF) work. Previous investigations, including the HPF studies, have produced valuable geologic, geophysical, thermal, hydrologic, and geochemical information, but these datasets vary in spatial coverage, analytical scope, data quality, and accessibility, leaving some areas outside known geothermal systems sparsely characterized and limiting regional assessments of subsurface heat, fluid circulation, and resource potential. The UH-led project encompasses geophysical surveys, slim-hole drilling, and integration of new data into the HPF statewide probability model. INL supports this effort through dataset review and standardization, hydrogeochemical sampling and analysis, petrographic and fluid inclusion analysis of drill core, power generation potential estimates, and drilling advisory. Work to date has focused on compiling existing aqueous geochemistry, field parameters, well and spring records, stable isotope measurements, and noble gas data. These data are used to identify geographic and analytical gaps and prioritize locations for future field sampling. Here, we present the data framework along with preliminary spatial and hydrogeochemical findings. This work establishes a foundation to reduce exploration uncertainty and guides continued sampling, geophysical investigation, drilling, and geothermal resource characterization across Hawaiʻi.

Topic: Geochemistry

Characterizing subsurface heterogeneity and transferring that information into numerical reservoir simulations remain significant challenges in a variety of subsurface feasibility studies from geothermal to energy storage applications. This work presents an integrated workflow for progressing from three-dimensional static geologic models developed in Seequent’s Leapfrog software to coupled thermal–hydrological–mechanical–chemical (THMC) and reactive-transport simulations using TOUGH numerical codes. Geologic, geophysical, borehole, and petrophysical data are first integrated in Leapfrog to construct stratigraphic surfaces, faults, and reservoir geometries. The static model is then simplified and discretized at a resolution appropriate for numerical simulation while preserving the geologic features expected to control fluid and heat transport. Porosity, permeability, thermal properties, mineralogy, initial temperature, and pressure conditions are assigned using measured data where available and literature-derived ranges where site-specific information is limited. These data are transferred to a TOUGH-compatible computational model for evaluation of fluid circulation, pressure response, heat recovery, and thermal breakthrough. Reactive-transport simulations are incorporated to evaluate temperature-driven mineral dissolution and precipitation, changes in fluid chemistry, scaling potential, and associated evolution of porosity and permeability. Sensitivity and uncertainty analyses identify which geologic assumptions and material properties exert the greatest influence on predicted reservoir performance. Application of the workflow to several GeoTES (Geologic Thermal Energy Storage) and potential Lithium extraction sites demonstrates how iterative exchange between static geologic modeling and process simulation can improve conceptual-model consistency. The resulting framework provides a reproducible approach for evaluating geothermal reservoirs, designing injection and production strategies, and determining which additional data would most effectively reduce uncertainty in long-term thermal and hydraulic performance.

Topic: Modeling

Enhanced geothermal systems represent the next frontier of geothermal energy, yet reservoir characterization in EGS settings remains challenging due to sparse labels, complex fracture networks, and heterogeneous lithologies. The Utah FORGE site provides a unique, fully open dataset combining drilling records, microseismic catalogs, well logs, and temperature measurements from a purpose-built EGS research site, creating an exceptional opportunity for rigorous machine learning evaluation in a controlled field setting. This study applies and s a suite of machine learning methods to two coupled EGS problems: fracture network characterization from microseismic event clustering and well log signatures and subsurface temperature distribution prediction integrating drilling and geophysical observations. We evaluate classical ensemble methods, physics-informed neural networks that embed heat conduction equations directly into the learning objective, and graph neural networks that represent the fracture network as a structured graph to capture spatial connectivity. Results demonstrate that physics-informed approaches reduce data requirements by over ~ ~40% compared to purely data-driven baselines while maintaining competitive predictive accuracy, a finding with direct operational significance for EGS sites where labeled measurements are expensive. Graph-based fracture representation captures connectivity patterns that conventional feature-engineering approaches miss, improving microseismic cluster characterization. Temperature prediction accuracy is strongly governed by the spatial density of conditioning measurements rather than model complexity, pointing to an optimal data acquisition strategy for future EGS sites. These findings translate into concrete recommendations for EGS site characterization workflows and establish a reproducible on the FORGE dataset that the broader community can build upon as EGS deployment accelerates globally.

Topic: FORGE

The Salton Sea geothermal field represents one of the world's major geothermal resources and is characterized by high-temperature, high-salinity brine containing significant concentrations of critical elements, such as lithium. This creates an opportunity for the co-production of renewable energy and critical minerals. However, the complex chemistry of these brines presents considerable challenges for surface processing and subsequent reinjection into the geothermal reservoir. Several reservoir studies have simulated lithium transport and recovery; others have examined long-term lithium evolution and water-rock interactions via reactive-transport studies. These studies have treated lithium recovery and reservoir geochemical response as separate processes, leaving a clear gap in the existing work. The coupled effects of direct lithium extraction (DLE), modified reinjection-brine chemistry, thermal recovery, and reservoir injectivity have not been evaluated within a unified reservoir-scale framework. This study presents an integrated reservoir-geochemical framework to identify the operational tradeoffs between geothermal energy production, lithium recovery, and long-term reservoir sustainability. A geocellular model is designed to represent the reservoir intervals of the Salton Sea geothermal system, with production and reinjection focused on the deeper Palm Spring–Imperial reservoir sequence. Using the Intersect reservoir simulator, the ionic composition of the produced brine is tracked and combined with demonstration-scale DLE processing data to define the composition of the reinjected brine. Mineral precipitation and dissolution reactions are represented through a simplified equilibrium reaction scheme. Reaction stoichiometry and equilibrium states are constrained using published thermodynamic data and PHREEQC calculations, while fluid properties are obtained from the International Association for the Properties of Water and Steam (IAPWS) tables. To isolate the impact of DLE on reservoir performance, multiple scenarios are evaluated. A baseline geothermal case without lithium extraction is compared with a lithium-removal-only case to establish reference thermal and geochemical behavior. Additional cases incorporate measured post-treatment brine chemistry to assess how DLE-induced changes in fluid composition influence reservoir response. The simulations quantify thermal recovery, lithium production, and thermal and chemical breakthrough behavior. Sensitivity analyses examine the influence of extraction efficiency, reinjection temperature, and operating strategy. The results will help to identify the conditions under which DLE-modified brine reinjection affects scaling, injectivity, and the long-term recovery of heat and lithium from the reservoir.

Topic: Reservoir Engineering

Closed-loop geothermal borefields are designed on the assumption of vertical boreholes at nominal coordinates, and long-term thermal capacity is computed from that idealized geometry. In practice bores deviate, and because deviation surveys are rarely economic at ground-source depths, designers have no empirical basis for the geometric uncertainty they are carrying. Conservative margin substitutes for measurement, and the resulting capacity estimate is a point value with no confidence interval attached. This paper combines field measurement with probabilistic thermal analysis to address that gap. We report measured trajectories from multiple closed-loop geothermal sites in eastern and central Massachusetts, drilled by independent contractors in crystalline bedrock and surveyed trajectory and temperature at depth. From these surveys we fit a stochastic trajectory model describing build rate and walk azimuth, and sample synthetic realizations of as-drilled borefield geometry consistent with the observed statistics. Each realization is evaluated with a g-function model that accounts for inclined and curved boreholes, yielding a distribution of long-term thermal capacity relative to the nominal vertical design. Results are reported as a distribution rather than a mean, with emphasis on the lower tail that governs whether a field meets its design load. We examine how this distribution depends on nominal bore spacing, since the thermal consequence of deviation is expected to differ between dense fields where boreholes already interact strongly and conventional spacings where deviation introduces interference that was not designed for. Bore-to-bore minimum separation is reported from the same geometry as a secondary output. The contribution is a method for replacing a point estimate of borefield capacity with a measured-geometry confidence interval, and an initial empirical prior on shallow geothermal borehole deviation in New England bedrock, where published trajectory data are currently almost absent.

Topic: Low Temperature

Objectives/Scope: Every produced sample carries reservoir intelligence — equilibrium temperatures, boiling and mixing signals, injection returns, scaling precursors — that geochemical surveillance reads for the cost of laboratory analysis. This review consolidates geothermal fluid geochemistry as an operational tool: solute and gas geothermometry reliability, tracking reservoir processes through time-series chemistry, injection-breakthrough detection, integration with physical surveillance, and the operating record of decisions geochemistry demonstrably informed. Methods, Procedures, Process: Geothermometer families are reviewed with their equilibrium assumptions graded against the disequilibrium field fluids actually present — re-equilibration during ascent, mixing, boiling corrections. Time-series interpretation methods are compiled from operating fields: chloride-enthalpy tracking, gas-ratio evolution, isotopic shifts under injection load. Breakthrough-detection performance is assembled against tracer-confirmed cases, multivariate and machine-learning treatments of chemistry streams are assessed, and decision case studies are compiled. Results, Observations, Conclusions: Geothermometry earns its keep in trends, not absolutes: compiled time-series show solute and gas thermometers tracking reservoir cooling and boiling-zone evolution reliably even where absolute temperatures scatter — the derivative is trustworthy, the intercept negotiable. Chloride-enthalpy diagrams remain the workhorse for process discrimination. Chemistry detected injection returns before thermal breakthrough in most compiled confirmed cases, at years of warning in several — surveillance value the sampling budget rarely reflects. Novel/Additive Information: An operations-focused rehabilitation of geochemical surveillance — trend-mode interpretation standards replacing absolute-temperature overreach, breakthrough-warning statistics quantifying the lead time chemistry buys over thermal signals. The integration framework assigns chemistry its role alongside pressure and temperature surveillance, and the sampling-design guidance converts the compiled decision record into monitoring-plan arithmetic for the fields now committing to sustained-yield management

Topic: Emerging Technology

Tubular strings in superhot geothermal wells undergo extreme temperatures and repeated thermal cycles during production, shut-in, and stimulation. Under these conditions, casings yield. Their structural mechanics depend on steel hardening, creep/stress relaxation, damage accumulation, and contact interaction with the cement sheath. Elastic stress analysis and design checks do not capture these load- and time-dependent mechanisms. This study discusses data needs for post-yield material modeling in superhot geothermal casing design. Casing specifications mainly set room-temperature yield and tensile strength, and temperature effects are usually covered by thermal derating of the yield strength. Advanced post-yield modeling needs much more than this, and for most casing grades the additional data are not published and are unavailable. The data needs differ sharply between models. Hardening needs complete stress-strain curves beyond yield, at several temperatures; reverse-loading tests to capture the Bauschinger effect, cyclic hardening/softening, and ratcheting. Creep needs stress relaxation tests, while damage needs tests carried to large strain to measure loss of stiffness. Connections’ low-cycle fatigue requires strain-life curves and proprietary notch geometry stress localization, which is generally kept confidential. Last but not least, casing-cement contact debonding requires strength and fracture energy of the interface in tension and shear, ideally under thermal cycling.

Topic: Drilling

Geothermal energy has been recognized by Spain and the European Union as a promising renewable energy source to improve energy security, grid stability, and sustainable economic growth. The City of Melilla is a Spanish enclave located in the north coast of Africa, and thus particularly dependent on imported fossil fuels for its energy needs. In result, Earth Energy Explorers and Teverra are evaluating the potential of geothermal energy to provide a source of renewable heat and electricity to the city. Melilla is located on the southern margin of the Alboran Sea, within the Cenozoic Melilla-Nador Basin, which has been studied for water supply and hydrocarbon exploration. The basin is underlain by a fractured Paleozoic to Triassic metasedimentary basement. Acidic volcanism in the area began north of Melilla and was later concentrated immediately to the south, forming the Gourougou stratovolcano. This volcano exhibits a complex eruptive history from 9.5 to 2.6 Ma, including basaltic lava flows extending into Melilla. The tectonic history of the region is complex and tied to the evolution of the Rif Mountains during the Alpine Orogeny, which produced broad deformation, fracturing, and partial exhumation of Neogene strata. Geothermal information near Melilla is limited to a hot spring on the flank of the Gourougou volcano, and a thermal anomaly in a well to the southeast, both within a 25 km radius. Temperature data from nearby offshore wells reveals a geothermal gradient of more than 40ºC/km, but no deep thermal data is available beneath the city. Hence, the geological conditions beneath the City of Melilla are compatible with the presence of blind geothermal systems, but exploration activities face unique challenges associated with its dense urban environment. A multidisciplinary geothermal exploration program has been conducted consisting of the application of geochemical, geological, and geophysical methods (magnetotelluric, gravity, and reflection seismic), along with temperature measurements of water wells in the area. However, the logistical complexities of conducting surveys in a dense urban and militarized environment present unique challenges to the application of geophysical methods to geothermal exploration. These limitations can be addressed by integrating all available geoscientific datasets within a 3D model, and understanding their impact on the uncertainty of subsurface knowledge. Multiple conceptual models for geothermal resources in the subsurface of Melilla have been developed, from conduction dominated aquifers, to convection dominated hydrothermal systems along major fault zones, and will be tested through an exploratory drilling campaign. Once geothermal resources of sufficient quality for heating, cooling, or electricity generation are found in the City of Melilla, their development could provide a local, stable, reliable, and renewable source of energy to the city’s residents and facilities.

Topic: Field Studies

There has been extensive effort to identify geological hydrogen prospects across the United States. Metrics such as in-situ hydrogen production rate and purity are important inputs for techno-economic assessment of these prospects. These inputs are estimated through geochemical simulation of different conceptual rock models and the chemical composition of the recharge fluid. We present a novel workflow integrating geochemical modeling with geostatistical interpolation to constrain redox conditions in mafic and ultramafic igneous basement domains across the United States and generate probabilistic forecasts of hydrogen occurrence, purity and solubility. Our workflow builds an ensemble of facies distributions across the U.S. basement, followed by hydrogen occurrence prediction. Facies modeling was performed using a hierarchical Bayesian framework that identifies basement domain groups, then rock types within each domain (mafic and ultramafic), and finally facies for each rock type, including basalt, harzburgite, and related lithologies. Bayesian model inputs include geochemical vectors from the PetDB, GEOROC, and NAVDAT databases, together with a raster of 3D resistivity and basement domain map across the United States. For geochemical simulations, basement temperatures were distributed using recent temperature maps, and pressures were estimated using a hydrostatic gradient. To distribute source-water composition across the basement, we developed a water chemistry model that assigns salinity and dissolved gas compositions. The model uses a hierarchical framework that assumes meteoric water in all domains and then overlays mixing regions associated with magmatic, ultramafic, and basinal fluids. Geochemical simulations at each point follow three pathways: recharge, during which source water is heated; reaction, where fluids reach peak temperature and hydrogen production; and discharge, during which buoyancy and diffusion transport reacted fluid upward while continuing to alter the surrounding rock. Resulting maps provide probabilistic estimates of hydrogen occurrence, solubility, and purity. This framework provides a pathway for ranking prospective plays and guiding exploration decisions.

Topic: Modeling

Steam–Water Relative Permeability and Two-Phase Flow Properties in Geothermal Reservoirs: Measurements, Models, and Open Questions Reviewed Objectives/Scope: Flashing reservoirs run on steam-water relative permeability, a property measured by a handful of laboratories under conditions most core facilities cannot reach. We review steam-water kr and two-phase flow properties for geothermal systems: the measurement record, phase-transformation effects that make steam-water unlike gas-water, and the model curves simulators quietly default to. Enhanced-system forecasting inherits the gap most directly, its conditions furthest from the data. Methods, Procedures, Process: The steam-water measurement literature is compiled, sparse enough to review nearly completely; phase-change-coupled flow physics is reviewed to explain where steam-water departs from inert-gas analogy. Simulator default practices are examined against the data, and ML property estimation is surveyed briefly. Condition coverage is mapped against enhanced-system envelopes, sizing the extrapolation. Relative-permeability functional forms are compared for their behavior outside measured ranges. Results, Observations, Conclusions: Measured steam-water curves differed from nitrogen-water analogs on the same rocks, with phase transformation reducing apparent interference between phases; defaults borrowed from gas-water practice misstate that. Data cluster in a narrow condition range, and enhanced-system conditions extrapolate beyond all of it. Forecast sensitivity to the kr choice rivaled sensitivity to permeability itself in compiled studies. Functional-form choice dominated extrapolation behavior, curves agreeing where data exist and diverging beyond. Novel/Additive Information: A near-complete audit of a property everything in flashing-reservoir forecasting depends on and almost nobody measures, with measurement gaps ranked by forecast consequence and functional-form choice exposed as a hidden extrapolation policy. The consequence-ranked wish list is scoped for university and national-laboratory capability — a ready agenda for the geothermal research programs now expanding worldwide.

Topic: Enhanced Geothermal Systems

Geothermal Resource Capacity Estimation: Heat-in-Place, Power-Density, and Stochastic Methods Compared, with AI Applications Objectives/Scope: Resource estimates decide drilling budgets, and the estimation methods disagree by design. We compare geothermal capacity-estimation approaches: volumetric heat-in-place with recovery factors, power-density analogs, and stochastic implementations of both, asking what each actually predicts and why early-stage estimates keep running optimistic. Portfolio-level resource reporting inherits the bias, making calibration an industry credibility issue. Methods, Procedures, Process: Method foundations are reviewed with their buried assumptions surfaced, recovery factor above all; estimate-versus-outcome comparisons are compiled wherever developed fields allow the reckoning. Stochastic practice is examined for whether distributions reflect knowledge or convention, and ML-assisted estimation from sparse exploration data is surveyed. Recovery-factor evidence is assembled by system type from outcome comparisons. Cutoff-choice effects are compiled, temperature and depth cutoffs moving estimates quietly. Results, Observations, Conclusions: Recovery factor carried most of the uncertainty and least of the scrutiny; outcome comparisons showed early volumetric estimates biased high, with the bias traceable to recovery-factor and cutoff choices. Power-density methods performed respectably where analog quality was honest. Stochastic ranges published to date have been too narrow more often than too wide. Cutoff choices moved estimates as much as recovery factors in the compiled decompositions. Novel/Additive Information: Capacity-estimation methods are judged against realized drilling outcomes rather than internal logic, with the documented optimism traced to recovery-factor and cutoff choices — two levers of equal weight and unequal scrutiny. The system-type recovery-factor ranges are formatted for direct citation in resource disclosures, giving developers calibrated credibility in a sector whose inflated estimates keep spending it

Topic: Enhanced Geothermal Systems

Borehole thermal energy storage (BTES) can capture cooling from winter ambient air and retain it for data center cooling during warmer periods. Thermal interactions among boreholes, seasonal temperature evolution, and ground freezing and thawing influence storage capacity and cooling recovery. This study proposes a reduced-order modeling framework to evaluate and optimize cold BTES under ambient conditions in Urbana-Champaign, Illinois. The application considers a 10 MW AI-training data center with design loads of 9 MW for information technology (IT) and 1 MW for ancillary equipment, constant IT utilization of 95%, and equal fixed and utilization-dependent ancillary-load components. We will use GeoWISE, developed by Oak Ridge National Laboratory, to establish the ground heat exchanger configuration for meeting the data center cooling demand. The current GeoWISE formulation represents ground heat exchange for cooling and heating applications but does not explicitly model seasonal thermal storage operation. Its borehole-field and pipe heat-transfer calculations, together with pygfunction-based g-functions, will provide the initial field configuration and thermal-response basis. Prescribed charging temperatures will then support calculations of outlet-fluid temperatures and charging loads. We will extend this modeling basis to represent cold charging, idle periods, and cooling recovery. Calibration against a transient CFD model will incorporate the effects of ground freezing and thawing into the reduced-order formulation. The storage-aware model will then be optimized for borehole number, depth, spacing, and arrangement while meeting cooling-load and delivery-temperature requirements. The analysis will quantify stored cold energy, useful recovered cooling, thermal recovery efficiency, interactions among boreholes, and storage uniformity. The framework aims to translate detailed borehole-scale physics into efficient field-scale design calculations for seasonal BTES integration with data center cooling.

Topic: Emerging Technology

Identifying enhanced geothermal system (EGS) targets requires understanding how thermal, geological, and geophysical signatures vary with depth and differ from those of naturally permeable hydrothermal reservoirs. Individual proxy thresholds can overlook the multivariate relationships that define prospective subsurface volumes. This study presents an explainable framework that combines self-organizing maps (SOM) with depth-resolved statistical analysis to characterize these relationships at Utah FORGE, an EGS reference site, and at Roosevelt Hot Springs (RHS), a conventional hydrothermal system. An integrated subsurface model comprises approximately 4.5 million blocks at 100 m resolution and incorporates 11 attributes: P- and S-wave velocities, Vp/Vs ratio, gravity-derived density, temperature, fault density, and five lithology indicators. A 3 × 3 SOM, trained without geothermal system labels, identifies nine spatially coherent clusters. Component planes, cluster attribute profiles, and spatial back-projection reveal the proxy combinations associated with each subsurface domain. System contrasts are then evaluated within 500 m depth intervals using Welch’s t-test and Mann–Whitney U tests, complemented by Hedges’ g and Cliff’s δ effect sizes. The modeled temperature distributions show strong separation at depths shallower than 2,500 m, with RHS exhibiting higher temperatures and large effect sizes consistent with its convective hydrothermal regime. Thermal contrasts diminish near 3,000–3,500 m and locally reverse between 3,500 and 4,500 m, where FORGE exhibits higher median temperatures. Vp/Vs maintains comparatively persistent contrasts across the analyzed depth range, with Hedges’ g of approximately 0.4–0.6, while density and fault-density contrasts vary with depth. These findings demonstrate that the discriminatory value of individual proxies varies vertically, thereby supporting the joint interpretation of multiple attributes. Spatial dependence among blocks and sparse thermal constraints below approximately 5 km warrant caution in interpreting statistical significance and deeper model contrasts. The framework makes the basis for subsurface differentiation explicit by linking three-dimensional cluster geometry to proxy combinations and depth-specific effect sizes. It supports the delineation of candidate EGS target volumes and the prioritization of follow-up characterization, while distinguishing modeled geoscientific signatures from direct evidence of permeability or stimulation performance.

Topic: Enhanced Geothermal Systems

Cement sheaths in geothermal wells are subjected to elevated temperatures and casing deformation, which can cause cracking and compromise well integrity. This study investigates a ductile well cement to improve deformation capacity and crack control to enhance the structural integrity of geothermal wellbore. The experimental program combines material-level mechanical characterization with cylindrical casing–cement prototype testing to evaluate the ductile cement relative to conventional well cement. Material-level results demonstrate tensile strain capacity exceeding 10% after ten thermal cycles of elevated temperature. A compressive strength of 30 MPa was also measured, highlighting the potential to combine substantial tensile ductility with adequate compressive strength. The results from cylindrical specimens indicate enhanced cracking resistance compared with conventional well cement. To further evaluate this behavior, cylindrical prototypes consisting of an inner steel casing surrounded by an annular cement sheath is subjected to internal pressurization at multiple elevated temperatures in an environmental chamber, inducing casing expansion and loading the surrounding cement. Digital image correlation will be used to monitor surface deformation, strain localization, and crack development to compare the response of conventional and ductile cement sheaths. The prototype tests are expected to show that the excellent crack control at the material level translates into improved structural integrity under combined internal pressurization and elevated temperatures, showcasing a novel ductile well cement for next-generation of geothermal wells.

Topic: Emerging Technology

Geothermal energy is a low-emission, weather-independent power source. Expanding its deployment beyond hydrothermal resources requires enhanced geothermal systems (EGS), where stimulation is needed to establish a fracture network for fluid circulation and heat extraction. Packers provide the zonal isolation required for these operations. However, sustained high temperatures and differential pressures can compromise seal integrity. When a packer leaks, it is difficult to determine the gap along the sealing interface downhole, and the extent of leakage depends on the operating conditions. At the University of Oklahoma’s Well Construction Technology Center (WCTC), a dedicated setup evaluates metallic packers under high-pressure, high-temperature (HPHT) conditions. As part of this work, a method is proposed to estimate an equivalent hydraulic gap from measured leak rates. Here, single-phase liquid leakage is simulated using data from Utah FORGE well 58-32 as a reference. Reported temperatures were used to build a static temperature profile, which was combined with completion fluid data to calculate hydrostatic pressure above the packer. Injection and mini-frac tests documented in the literature, including a reported packer failure, provided zone depths and operating conditions. Leakage is modeled by unrolling the annular flow path into a slot between parallel plates. The governing equation is derived from Navier–Stokes for steady laminar flow and solved along the seal using local water properties. The equivalent hydraulic gap is obtained through an iterative process. Simulations covered 120–260°C, differential pressures of 2,000–6,000 psi, and prescribed leak rates of 0.05–0.5 gpm. Higher temperatures and differential pressures require smaller gaps for the same leak rate. For the reference case (189.5°C, 4,140 psi), 0.5 gpm corresponds to a gap of 12.61 µm. Holding this gap and differential pressure fixed, leakage rose from 0.31 to 0.69 gpm across the temperature range. All cases remained liquid and laminar. These results help assess how operating conditions affect leakage and provide a basis for comparison with acceptable operating limits.

Topic: FORGE

The Rio Grande Rift, which extends from central Colorado through New Mexico into West Texas, hosts one of the strongest heat-flow anomalies in the western United States. Despite this thermal potential, there are a few operating geothermal power plants in the region. Enhanced geothermal systems (EGS) in hot, dry rocks can overcome this limitation by creating permeability through reservoir stimulation, but no integrated, region-wide assessment of EGS favorability exists today. We built the 1st regional-scale 2D EGS play fairway map for the entire Rio Grande Rift by integrating four EGS evidence layers (thermal, reservoir, geomechanical, and operational). The framework combines depth to the 150 °C isotherm (thermal), lithology along that isotherm from integrated core, log, 3D gravity and magnetic data analysis (reservoir), Quaternary fault density, earthquake information, along with gravity and magnetic gradients (geomechanical), and brackish water depth (operational) to be used in EGS development. Lithology at 150°C isotherm varies spatially, including sediment, granite, gabbro, metabasalt, amphibolite, phyllite, schist, etc., with various implications for hydraulic fracturing and fluid loss. The highest EGS favorability occurs mainly on the rift flanks. On the east side, it is concentrated around the Estancia, Tularosa-Jornada del Muerto, and eastern Hueco basins, and on the west side, adjacent to the San Luis, Española, and Albuquerque basins. In contrast, the densely faulted axial basins show low favorability. Sensitivity analysis shows that lithology and brackish water components exert the strongest influence on the results. The approach provides a rapid screen of regional-scale EGS prospects and can help reduce exploration risk in rift settings.

Topic: Enhanced Geothermal Systems

The analysis compares medium-enthalpy geothermal systems in a direct configuration with hybrid geothermal-solar thermal systems for cooling applications in data centers. Direct geothermal uses resources with temperatures between 90 and 150 °C, providing a steady, stable, and continuously available energy source, although it is limited by the reservoir's own thermal conditions. The integration of solar thermal energy through concentration collectors allows the fluid temperature to rise to values close to 120–200 °C, significantly improving the available energy quality. This thermal boost makes it possible to use double-effect absorption chillers, increases the coefficient of performance (COP), and expands the system's operational flexibility. From an operational standpoint, geothermal provides the base thermal load while the solar resource adds extra energy during periods of high irradiation. Although the solar contribution introduces some daily and seasonal variability, the system's reliability is maintained thanks to the continuous backup from the geothermal source. Plus, using advanced control and automation systems allows optimizing the mix of both energy sources and adapting to changes in demand. In terms of performance, direct geothermal systems usually operate with thermal COPs between 0.6 and 0.75, while hybrid systems can reach values between 0.8 and 1.3. Although adding a solar field increases the initial investment (CAPEX), operating costs stay low because only renewable resources are used and there’s no need for fossil fuels. In conclusion, geothermal–solar hybrid systems represent a highly promising technological alternative for the sustainable cooling of data centers in Costa Rica. Combining the reliability of geothermal energy with the thermal boost capability of solar power allows for increased energy efficiency, reduced levelized cost of cooling (LCOC), and optimized use of available renewable resources, especially in regions with high geothermal and solar potential like Guanacaste.

Topic: Direct Use

This study investigates whether routine oilfield data, including measurement-while-drilling (MWD) temperature logs, reported bottom-hole temperature (BHT), and monthly production records, can support a defensible first-pass ranking of produced-water geothermal co-production potential in a mature field, without new geothermal-specific measurements. Applied to 25 horizontal Bakken wells in the Oliver Field, Williston Basin, North Dakota, the answer is a qualified yes: produced-water rate, not reservoir temperature, controls co-production ranking. Reservoir temperature varies only 35°F across the field (230-265°F, from 11 wells with a reported value), while produced-water rate varies four-fold. This finding is stress-tested rather than assumed: Monte Carlo resampling of the temperature assigned to wells lacking a reported BHT leaves the rank correlation with a water-rate-only ranking at 0.90-0.95 across 2,000 draws. Gross thermal power potential ranks 28-108 kWth per well (1,624 kWth field aggregate, brine-corrected), but converts to only 130-195 kWe field-aggregate net electric potential under conservative low-temperature binary-cycle assumptions, below typical minimum generation-unit thresholds per well, and at or below the 130°C threshold generally cited for economical geothermal power generation, favoring direct-use applications over electricity generation. Two common shortcuts were tested and found to fail under this field's data constraints: an empirical correction from MWD circulating temperature to static BHT (no meaningful relationship, r = -0.17, seven-well calibration set), and Arps hyperbolic decline forecasting of produced-water rate, which systematically underpredicts current observed rates in 21 of 25 wells. Both are reported as methodological findings. The result is a practical screening hierarchy for thermally narrow mature oilfields: rank first by sustained water rate, use temperature to refine, and validate any water-rate forecast against recent production before relying on it.

Topic: Field Studies

Geothermal power plays a pivotal role in global renewable energy goals by providing sustainable, low-carbon electricity. Nevertheless, a project's overall feasibility relies heavily on executing drilling programs that are both technically robust and economically sound. To attract financial institutions and secure necessary capital, developers must craft bankable drilling plans that satisfy strict operational, financial, and regulatory s. This paper investigates the essential methodologies and components required to formulate these plans, emphasizing the due diligence requirements from both technical and economic standpoints. Crafting a bankable geothermal drilling program demands a holistic methodology that merges industry best practices with site-specific environmental challenges. A resilient plan validates that well targeting, risk management, well design, and drilling sequencing are achievable and financially viable. Critical elements—such as architectural well design, testing protocols, and mitigation strategies—must be meticulously evaluated to circumvent expensive operational failures. From a due diligence perspective, this paper highlights the necessity of preemptively resolving common red flags that could diminish a project's credibility to lenders. Frequent areas of concern during evaluations include aligning the drilling strategy with the resource's conceptual model, establishing consistent success criteria for sequence decisions, and conducting precise rig sizing assessments. Selecting a drilling rig with the appropriate technical capacity is vital for maintaining economic efficiency and preventing budget overruns or schedule delays. Additionally, the paper details the incorporation of comprehensive cost modeling, ensuring that expenditures related to civil infrastructure, access roads, water supply, well testing, and remediation are accurately projected to satisfy investor scrutiny. Ultimately, this study delivers a structured framework designed to assist developers in generating comprehensive, financially appealing geothermal drilling proposals. By proactively tackling these fundamental due diligence criteria, the research aids in refining the project development lifecycle, bolstering lender confidence, and paving the way for successful project execution.

Topic: Drilling

Indonesia’s geothermal sector is pivoting from mature high-temperature fields toward vast low-to-medium temperature resources. However, the bankability of these future projects is threatened by power plant construction costs that remain well above global averages. This paper tackles this critical bottleneck by introducing a novel techno-commercial ing methodology that dissects the most capital-heavy project phases: Power Plant Engineering, Procurement, and Construction (EPC) and Steam Above Ground Systems (SAGS). Utilizing escalated global cost datasets from the U.S. and Türkiye, our analysis demonstrates that Indonesia’s CAPEX disparity stems from monopolized turbine markets, high import dependency, and inflated financial "safety margins" in early EPC bids. To counter these drivers, we propose a compounding cost-reduction framework featuring four actionable strategies: (1) utilizing Organic Rankine Cycle (ORC) technologies to break manufacturer monopolies; (2) de-risking execution through pilot testing; (3) replacing EPC safety margins with rigorous, geospatial data-driven Front-End Engineering Design (FEED); and (4) restructuring procurement into Split-EPC contracts to optimize tax exposure. The study translates these multidisciplinary findings into an Integrated Cost Optimization Roadmap, utilizing high-definition visual tools like waterfall charts and contracting visualizations. By bridging theoretical cost-saving measures with practical application, this blueprint offers Indonesian policymakers, operators, and financiers a replicable pathway to drastically reduce development costs and accelerate the nation's energy transition.

Topic: General

Cryogenic cold sources such as liquefied natural gas (LNG), liquid nitrogen, and liquid air contain substantial recoverable physical exergy that is commonly dissipated during regasification or vaporization processes. Thermoelectric generators (TEGs) have emerged as a promising solid-state technology for harvesting this cold energy because of their structural simplicity, scalability, reliability, and absence of moving parts. This review presents a comprehensive analysis of thermoelectric generator applications using cryogenic cold sources, with emphasis on LNG regasification systems, liquid nitrogen recovery systems, and liquid air energy storage applications. The review systematically examines thermoelectric operating principles under cryogenic conditions, material behavior at low temperatures, system configurations, heat transfer mechanisms, and reported electrical performance. Major technical challenges, including cold-side thermal resistance, frosting effects, contact resistance, and low conversion efficiency, are critically discussed. Reported experimental and numerical studies demonstrate that cryogenic thermoelectric systems can achieve electrical power outputs ranging from sub-watt laboratory-scale devices to hundreds-of-watts industrial-scale configurations, although practical efficiencies generally remain limited. The review further evaluates heat exchanger integration strategies, structural optimization approaches, and system-level performance limitations affecting large-scale implementation. Finally, current research gaps and future development directions are identified to support the advancement of efficient and reliable cryogenic thermoelectric energy harvesting technologies.

Topic: Emerging Technology

Project Ironside is a U.S. Department of Energy supported geothermal resource characterization effort led by Oriah Geothermal, LLC in collaboration with Idaho National Laboratory (INL) to evaluate a next generation Enhanced Geothermal System (EGS) prospect in Malheur County, Oregon. The project targets a volcanic–plutonic terrane where elevated geothermal gradients ( greater than 60 °C/km), potential radiogenic heat contributions, and proximity to major transmission corridors suggest strong commercial potential. Current project work and public-domain datasets, including groundwater well temperatures, geologic maps, magnetics, gravity, and the Stanford Thermal Model have been integrated into an updated conceptual model indicating that ≥200 °C reservoir temperatures may be reachable at depths of ~3,000–4,000 m. Project partners have supplemented these datasets with reconnaissance geologic mapping, thermal water sampling, and targeted geophysical surveys that refine structural interpretations and thermal regime uncertainty. Together, these datasets support preliminary resource capacity estimations that could support a 15 MWe pilot plant under conduction-dominated conditions and establish a strong technical foundation for confirmation drilling. Project Ironside will collect new subsurface data through a staged, decision guided drilling program. In the first phase, three thermal gradient (TG) wells (500–1,200 m) will be drilled, logged, and tested to constrain lithology, permeability indicators, stress regime, and the vertical evolution of temperature. These results will be integrated with preexisting data to update the 3D conceptual model and define drilling targets for the reservoir depth exploration well. During the 2nd phase, the project will drill and test a 3,000–4,000 m well to obtain direct measurements of reservoir temperature, geomechanics, mineralogy, and brine chemistry. INL will incorporate these datasets into reactive transport models and coupled thermal–hydraulic–mechanical–chemical (THMC) reservoir simulations to generate updated resource capacity estimates and evaluate feasibility of a ~15 MW pilot EGS system. Throughout the project, all non restricted datasets, metadata, conceptual model updates, and numerical modeling outputs will be synthesized and disseminated to support reproducibility, accelerate geothermal learning, and inform EGS deployment in volcanic–plutonic terranes across the western United States.

Topic: Field Studies

In deep U-shaped closed-loop geothermal systems, the ascending production leg passes through progressively cooler formation, and part of the recovered heat leaks back into the rock before it reaches the surface. In the present study, this loss is countered by insulating the upper production well. The system is a 4000 m deep, 200 °C loop with a 1500 m lateral, operated at 10 kg/s from a 20 °C injection state, modeled with a computationally efficient one-dimensional wellbore march coupled to transient radial formation conduction. 28 fluids that remain single-phase liquid throughout the loop are simulated over a 30-year life, with the insulated length swept from zero to the full well depth. Insulating the entire production well proves counterproductive, lowering cumulative exergy for only 22 of the 28 fluids, because the working fluid in the deeper production well sections remains colder than the surrounding formation, and blanketing it blocks heat gain. Each fluid instead has an interior optimum insulated depth; deepest for the halocarbon refrigerants and shallowest for water. Insulating to that optimum reorders the second-law ranking: R1234ze(E) rises to 42.2 GWh of exergy and displaces isobutane, which leads the bare-well case at 38.5 GWh, from the top position. The gain is largest for the fluids that lose the most heat during ascent, and it appears mainly as a hotter, higher-exergy-delivered stream. Production temperature reflects this most clearly: the refrigerants deliver the hottest streams and insulation widens their lead. The year-30 outlet temperature of R11 rises from 89 to 100 °C, while water barely changes, remaining near 54 °C. The temperature ranking is preserved even though the exergy ranking is altered. These results show that production-well insulation is a fluid-specific design variable and that working-fluid selection cannot be decoupled from completion design.

Topic: Modeling

Irregular discharge in multi-feed geothermal wells is commonly attributed to dynamic interactions between feedzones with contrasting pressures and enthalpies, yet quantitative constraints on coupled wellbore--reservoir responses remain limited. This study reconstructs a multi-feed wellbore model for well ULB-01 in the Ulumbu Geothermal Field using a pressure-drop approach calibrated against Pressure Temperature Spinner (PTS) surveys and production-test data. The inferred feedzone boundary conditions are then coupled to a TOUGH2 reservoir simulation to investigate feedzone interaction and wellbore--reservoir response under low wellhead-pressure operation. During a representative cycling interval (10.1--10.5 days), the observed wellhead pressure decreases from 6.8 to 2.7 barg and brine flow declines from 9.2 kg/s to zero, while steam flow remains relatively stable at 5.7--6.3 kg/s. The wellbore simulation captures the broader pressure and steam-rate response but does not fully reproduce the intermittent brine shut-off, indicating that additional transient wellbore or surface-flowline processes may contribute to the observed short-timescale cycling. The coupled TOUGH2--wellbore workflow provides a partial match to measured wellhead pressure, steam mass flow, and produced enthalpy during relatively stable periods, while larger discrepancies remain during transient brine surges and shut-off events. The results suggest that pressure drawdown promotes redistribution of feedzone contributions and cooler liquid inflow at intermediate depth, providing a plausible mechanism for discharge instability. The findings indicate that the coupled workflow is useful for diagnosing feedzone-controlled instability, although resolving intermittent brine shut-off would require a more transient representation of wellbore and surface-flowline processes.

Topic: Reservoir Engineering

Accurate delineation of the stimulated fracture volume (SFV) is essential for enhanced geothermal system (EGS) development. At Utah FORGE, the initial SFV was derived from the April 2022 microseismic catalog by clustering seismic sources. However, density-based methods such as HDBSCAN are highly sensitive to parameter choices, and the 2022 catalog alone provides insufficient constraints, leading to significant ambiguity in the resulting SFV geometry. To overcome this, we propose an iterative framework that uses subsequent seismic observations from the July 2023 circulation test to constrain the 2022 SFV. We processed distributed acoustic sensing (DAS) data from monitoring well 16B(78)-32 by denoising with a pre-trained Noise2Noise (N2N) network, band-pass and f-k filtering, and imaging via Kirchhoff migration to identify circulation-activated sources. In addition, microseismic events recorded by the near-surface seismic network during the circulation test were incorporated as an independent constraint. These sources are integrated into the clustering workflow to update the HDBSCAN–convex-hull SFV, and the update is repeated until the SFV converges. Preliminary results show that the later seismicity, primarily the near-surface microseismic events supplemented by DAS-detected sources, increases the SFV by 30% and reduces the variation in SFV across tested HDBSCAN parameters by 83%. This iterative approach leverages later seismicity to reduce uncertainty, moving beyond sole reliance on parameter selection. The resulting SFV offers a robust, observation-constrained representation of the fracture system stimulated in 2022 and reactivated during circulation. Future integration with pressure and temperature diagnostics will further evaluate the relationship among seismic activity, fracture connectivity, and reservoir stimulation.

Topic: Enhanced Geothermal Systems

Modern testing by reservoir engineers in petroleum practice originated at least 100 years ago. Traditional geothermal reservoir and well testing dates back more than 50 years. This long history provides a consensus standard for tasks such as reserves estimation. It also means that most test equipment is available from service companies, along with experienced field operators. By comparison, Engineered Geothermal System (EGS) development is in its infancy. The objectives for testing and the equipment required differ considerably from those of traditional petroleum and geothermal resources. The traditional petroleum or geothermal reservoir engineer, working in exploration and development, is typically trying to learn about the resource as it exists in its native state. Properties include permeability, porosity and geographic extent of the host rock. The fluids of interest are generally long in place and chemically equilibrated with the formation. The fluids produced in testing are examined to determine their in-situ composition and saturations. The situation facing the EGS reservoir engineer is nearly the opposite. The extent of hot rock is reasonably well known. Formation permeability and porosity are known to be nearly zero, as is the amount of connate water. The challenge of EGS testing is to determine what the reservoir will become, and how the rocks will react to a non-native working fluid in stimulated fractures. The contrasting situations lead to important differences in test design, ranging from mechanical test equipment, fluid handling, geochemical and surface chemical concerns, to regulatory compliance, operating costs, supervision and safety. These points are discussed in the context of the Utah FORGE long-term circulation test program.

Topic: Enhanced Geothermal Systems

Analytical heat transfer models are essential for feasibility studies and preliminary design of geothermal projects, yet several classical solutions rely on assumptions that systematically overestimate thermal performance, leading to optimistic engineering assessments. This study presents a practical analytical model for enhanced geothermal systems (EGS) that provides fast, consistent estimates for resource assessment. Based on Green's functions, the formulation explicitly captures thermal interactions between fractures, preserving analytical tractability, and it can be implemented directly in standard spreadsheets without Laplace-space transformations or numerical inversion algorithms. The model is validated in two stages: 1. Predictions are compared with classical approaches, including Gringarten et al. (1975) and Fox et al. (2016). The proposed model addresses the optimistic bias of these solutions and yields conservative predictions of production well outlet temperature. 2. Model predictions are compared with the numerical model of Zidane et al. (2025), showing close agreement in predicted temperature evolution (deviation ±4%). By bridging analytical solutions and numerical simulation, the model offers engineers an accessible tool for preliminary resource assessment, thermal performance forecasting, and early-stage EGS design.

Topic: Modeling

Characterization of the subsurface and understanding its uncertainty is a crucial part of de-risking an EGS project. We propose a framework that aims to infer subsurface fracture network information by combining an adaptive sampling statistical inference method with PFLOTRAN simulations of a doublet EGS system undergoing transient-injection tracer testing where heat is the tracer. Current state-of-the-art technologies aimed at characterizing subsurface flow paths include conventional hydraulic/pneumatic tomography and tracer testing [1,2], though long-term pump-test methods generally used to characterize fractures evaluate effective porous media parameters generally do not make use of transient injection [1]. With regard to inference of subsurface properties, there is a rich history of traditional inversion and statistical inference methods using borehole log or tracer test data [1]. Typically these methods either rely on an ECPM representation of the subsurface and produce a tomogram of fields such as permeability or hydraulic conductivity, which may lack relevant information about the fracture network topology and connectivity, or rely on DFN-representation based methods which may not accurately capture the heat exchange between the fluid and the rock matrix (a relevant mechanism in understanding a geothermal system's likelihood to thermally short-circuit). Meanwhile, geothermal heat has been utilized successfully as a tracer in methods such as travel time tomography [3], traditional inversion [4], or statistical inversion [4]; however, these methods either fail in regions that are not main flow channels, rely on assumptions about the connectivity of the fracture network, or utilize temperature borehole log data (passive) rather than active applied injection to the system. The work presented herein makes use of the PFLOTRAN simulator [6] and its geothermal fracture process model [7], which first generates or accepts a discrete fracture network (DFN) descrption of the subsurface, and then converts it into an equivalent continuum porous medium (ECPM) representation. This allows the inference strategy to be aware of the DFN descrptor of the fracture network (including fracture network position, size, orientation, and aperture) whilst still using ECPM simulations of each network which can better capture heat transfer between the fluid and surrounding rock matrix, and account for changes to fracture aperture size based on thermal contraction. The framework also aims to exploit transience in injection signals to improve inferability of fracture network properties and utilize active injection of cool fluid with the intent to use heat as the tracer. These aspects are combined with an adaptive sampling statistical inference method to infer properties about the fracture network based on an observed temperature and pressure response signal in response to transient injection of cool fluid. Herein, we present details on the proposed framework and explore preliminary results. Funding statement: This work was supported by the Laboratory Directed Research and Development program (project 240512) at Sandia National Laboratories, a multimission laboratory managed and operated by National Technology and Engineering Solutions of Sandia LLC, a wholly owned subsidiary of Honeywell International Inc. for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE NA0003525. References: 1. Ringel, L. M., Illman, W. A., & Bayer, P. (2024). Recent developments, challenges, and future research directions in tomographic characterization of fractured aquifers. Journal of Hydrology, 631, 130709. 2. Wu, H., Fu, P., Morris, J. P., Mattson, E. D., Neupane, G., Smith, M. M., ... & EGS Collab Team. (2021). Characterization of flow and transport in a fracture network at the EGS Collab field experiment through stochastic modeling of tracer recovery. Journal of Hydrology, 593, 125888. 3. Somogyvári, M., Bayer, P., & Brauchler, R. (2016). Travel-time-based thermal tracer tomography. Hydrology and Earth System Sciences, 20(5), 1885-1901. 4. Klepikova, M. V., Le Borgne, T., Bour, O., Gallagher, K., Hochreutener, R., & Lavenant, N. (2014). Passive temperature tomography experiments to characterize transmissivity and connectivity of preferential flow paths in fractured media. Journal of Hydrology, 512, 549-562. 5. Somogyvári, M., Kühn, M., & Reich, S. (2019). Reservoir-scale transdimensional fracture network inversion. Advances in Geosciences, 49, 207-214. 6. Hammond, Glenn, Noel, Michael, Leone, Rosemary, Park, Heeho, & Frederick, Jennifer (2020). PFLOTRAN. https://doi.org/10.11578/dc.20201103.3 7. Frederick, J. M., Gatz-Miller, H., & Lowry, T. S. (2025, February). Introducing PFLOTRAN’s New Geothermal Fracture Model for Simulation of Geothermal Systems. In PROCEEDINGS, 50th Workshop on Geothermal Reservoir Engineering. Presented at the Stanford Geothermal Workshop.

Topic: Modeling

In this study we history-match a single coupled stimulation-and-circulation ResFrac model against the 2024 stimulation and the 2024 to 2025 circulation tests at Utah FORGE, using treating pressure in both wells, fiber-optic fracture hits, microseismic extent, spinner-log flow allocation, tracer swept volume and water loss. We use two model families: one with prescribed fracture geometry and one with hydraulic fractures propagated from the field pumping schedules. Matching the observations requires (1) an effective fracture conductivity near 50 md-ft, (2) an effective matrix permeability of tens of µD, (3) stage-wise Shmin (0.72 to 0.92 psi/ft) and near-wellbore tortuosity, (4) prescribed pre-existing fractures that represent the mapped faults, and (5) enough open fracture volume to sustain the ~5,000 bbl tracer swept volume. The match is non-unique: the forecasted one-year decline in produced temperature ranges from ~5 to ~25 °C, depending mainly on how flow is distributed among the fractures.

Topic: FORGE

Multi-stage EGS completions rarely distribute flow evenly across the fracture network. Fracture conductivity can span orders of magnitude, so a few dominant paths take most of the flow, causing early thermal breakthrough and leaving much of the stimulated rock unswept. Flow control devices (FCDs) at the injection and/or production wells can enforce more uniform flow, but their selection for geothermal service is typically screened on flow uniformity alone rather than on lifetime economics. This study sizes FCDs against a levelized cost of energy (LCOE) objective over the full project life. An integrated reservoir and system simulation platform, coupled with fracture modeling software and a FCD characteristic model, resolves the per-fracture flow split, thermal drawdown, injector and producer wellbore hydraulics, and long-term thermo-poro-elastic conductivity evolution. An uncontrolled baseline, fixed orifices, and autonomous flow control devices are compared across heterogeneity scenarios. Results are reported as flow uniformity, breakthrough timing, recoverable energy, pumping load, and system LCOE.

Topic: Enhanced Geothermal Systems

Based on the rather trivial idea that cumulative tracer recovery values R(t) provide a more robust measure of fluid open-loop transport than any selection of data points, whatever their weighting, from a tracer breakthrough curve (TBC), and exploiting the fact that R(t) is a monotonous and thus invertible function of time, we propose using a normalized set of evolving R-indexed transit-time maps in parameter space as a more proficient tool for model calibration and thermal lifetime prediction, especially with almost-nil or just incipient TBCs recorded at production wells in seemingly very large reservoirs, with (seemingly!) long fluid residence times and/or extreme retardation of conservative tracer species in stagnant-fluid compartments; in such cases, a TBC-based prediction of thermal lifetime, and even a plain pore-volume estimation is further compounded by the inability to anticipate what the signal shape is going to look like, and what asymptotic value later tracer recoveries will tend toward. Though itself model-dependent, the transit-time mapping tool enables more robust tracer-based predictions of thermal lifetime for larger well-loop sizing options in various medium-enthalpy play types in Central Europe. It further yields a more plausible characterization, in terms of fluid transport parameters, of what had been perceived as a karstification 'black hole' within past studies on very low tracer recoveries from three hydrothermal well-loop tests in more or less karstified and tectonically overprinted geological formations in the Alpine foreland (Molasse) basin, enabling to correct past over-estimations of thermal lifetime.

Topic: Tracers

Wellbore quality is a critical factor in the successful construction of a horizontal geothermal wells. Its primary influence would be in directional efficiency, torque and drag, casing and liner deployment, and the risk of subsequent wellbore access problems. Wellbore quality is commonly evaluated using directional survey metrics such as dogleg severity (DLS) and qualitative observations including excessive torque, drag, reaming, or poor weight transfer. However, DLS describes changes in the surveyed wellbore trajectory and does not necessarily capture short-scale tortuosity, mechanical interaction between the drillstring and wellbore, or the cumulative operational consequences of wellbore geometry. Consequently, intervals with similar DLS may exhibit substantially different mechanical behavior. This study presents a data-driven framework for quantifying wellbore quality using routinely acquired drilling and directional data. Surface measurements such as hookload, torque, rotary speed, flow rate, standpipe pressure, rate of penetration, and drilling depth, are synchronized with directional surveys, bottomhole assembly assemblies, and physics-based torque-and-drag predictions. Differences between measured and modeled mechanical responses are used to characterize wellbore behavior. Candidate indicators include pickup and slackoff hookload residuals, rotating torque residuals, pickup-to-slackoff spread, calibrated friction factors, variability in mechanical response, reaming requirements, and trajectory-derived measures of curvature and tortuosity. These parameters are combined into a depth-dependent Wellbore Quality Index (WQI) designed to provide a quantitative measure of mechanical wellbore condition independent of DLS alone. Statistical methods are used to distinguish persistent changes in wellbore behavior from operational variability and measurement noise and to determine the relative contribution of individual indicators to the WQI. The framework is validated against subsequent operational observations, including reaming requirements, tight-hole behavior, abnormal torque and drag, weight-transfer limitations, tripping performance, and, where available, casing or liner deployment performance. The proposed methodology provides a continuous assessment of wellbore quality that can be compared across depth intervals, bottomhole assemblies, drilling practices, and wells. Particular emphasis is placed on determining whether the WQI identifies mechanically problematic intervals that are not apparent from conventional DLS analysis and whether deterioration in the index precedes operationally recognized wellbore problems. A quantitative measure of wellbore quality could provide geothermal drilling teams with an additional tool for evaluating directional-drilling performance, optimizing BHA and drilling-system selection, identifying deteriorating hole conditions, and assessing the operational value of smoother wellbore construction.

Topic: Enhanced Geothermal Systems

Geothermal wells use many of the same material classes across conventional geothermal systems (CGS), enhanced geothermal systems (EGS), and closed-loop geothermal systems (CLGS). However, the functions these materials must perform can differ substantially among applications. Carbon steel casing suitable for a cased closed-loop system may face different durability challenges than casing exposed to continuously flowing geothermal fluids, while a cement system designed primarily for zonal isolation may not provide the thermal performance required in a conductive closed-loop annulus. Consequently, materials design, development, performance requirements, and long-term integrity risks must be considered in the context of the geothermal technology and well architecture in which a material is deployed. This review examines long-term performance and degradation of geothermal well materials across major geothermal technologies and operating environments, including supercritical conditions where applicable. The review focuses on commonly used and advanced cementitious materials, metallic tubulars, cement-metal and cement-formation interfaces, elastomeric packers and sealers, and polymeric additives used to control slurry and fluid properties. Particular emphasis is placed on how material functions differ among geothermal applications and how those differences influence the properties required to maintain performance throughout the life of a well. Material behavior is considered under prolonged exposure to elevated temperature and pressure, geothermal fluids, thermal gradients, and thermal cycling. The review examines chemical, mineralogical, and physical evolution of cementitious materials; corrosion of metallic tubulars; and degradation of elastomeric and polymeric components, together with their effects on bonding and compatibility between adjacent materials. These changes are evaluated in terms of their impact on a material's ability to perform its intended function. The review integrates laboratory investigations, field observations, documented failures, modeling studies, and industry experience to compare material performance across geothermal technologies. Interactions among adjacent materials are also considered, recognizing that evolution of one material may alter the durability and performance of another. The resulting comparison identifies technology-specific design and qualification requirements and highlights areas where long-term field performance remains insufficiently understood.

Topic: Emerging Technology

Fluid injection in enhanced geothermal systems (EGS) aims to increase reservoir permeability to improve fluid circulation and heat extraction. Hydraulic stimulation can lead to seismic and aseismic slip on faults, both of which can enhance fault zone permeability. This permeability enhancement modifies fluid transport and the pore pressure distribution driving slip, which can modulate the spatiotemporal distribution of subsequent seismic and aseismic deformation. However, earthquake catalogs constrain only the seismic component of this deformation, and the aseismic component is difficult to constrain directly in field settings. Understanding these coupled processes is important for interpreting the hydraulic response of stimulated EGS reservoirs and managing the operational challenges posed by induced seismicity and aseismic slip. We develop a modeling framework to analyze the role of permeability enhancement, seismic slip, and aseismic slip during EGS stimulation of the Habanero Fault in Cooper Basin, Australia. We use 3D earthquake sequence simulations of slip on a velocity-strengthening rate-and-state fault, as well as a semi-analytical fracture mechanics model of constant-rate injection into a constant-friction circular shear crack, to model fluid injection into a single well. Both models account for permeability enhancement. Published logging and well testing, as well as geomechanical models of the Habanero Fault, provide initial estimates of the fault’s hydraulic properties and initial stress state. In our numerical simulations, we impose the measured variable-rate injection history, and we account for wellbore storage and turbulent pressure losses to relate simulated fault pore pressure to the wellhead pressure history. Our results show how wellhead pressure measurements and earthquake catalogs provide complementary constraints on fault zone permeability and deformation caused by EGS stimulation, including the aseismic deformation component that microseismic monitoring does not directly capture. Similar spatiotemporal patterns of seismicity can be reproduced by models with a range of permeability and total slip values. However, models with lower permeability produce more total slip and require greater pressurization to sustain injection, resulting in higher modeled wellhead pressures. The difference between modeled total slip and seismic slip inferred from the earthquake catalog provides an estimate of the aseismic contribution to deformation, allowing for a more complete assessment of fault deformation during EGS stimulation beyond what can be captured by earthquake catalogs alone.

Topic: Geophysics

Wellbore cement is a critical construction material in geothermal wells because it provides structural support, zonal isolation, and hydraulic separation throughout the service life of the well. As supercritical CO₂ (scCO₂) continues to be considered as a working fluid in enhanced geothermal systems, understanding its interaction with cement becomes increasingly important. CO₂ exposure can alter cement through carbonation, calcium redistribution, and decalcification of calcium-silicate-hydrate phases. These changes may affect barrier performance before externally visible damage develops. This study evaluates whether graphene-modified API Class H cement can provide a measurable electrical indication of such chemical alteration and thereby support periodic assessment of wellbore condition. Neat and 0.5 wt% graphene nanoplatelet (GNP)-modified cement samples were prepared using field-representative Class H cement formulations. After curing, baseline electrical resistance and resistivity were measured over 1 Hz–10 kHz. Samples were then exposed to scCO₂ for 30 days at 65 °C and 17.24 MPa (2,500 psi), followed by post-exposure electrical and materials characterization. Complete frequency sweeps were averaged by frequency, and pre- and post-exposure responses were compared over the full measured range and within the ≥100 Hz region, where bulk conduction is more dominant than low-frequency electrode polarization. Both cement systems exhibited a clear increase in resistance and resistivity after scCO₂ exposure. At 1 kHz, neat cement resistivity increased from approximately 25.0 to 53.8 Ω·m, while the 0.5 wt% GNP cement increased from approximately 31.1 to 95.2 Ω·m. Across the ≥100 Hz bulk-response region, mean resistivity increased by approximately 114% for neat cement and 203% for the GNP-modified cement. The larger relative shift in the GNP system indicates greater electrical sensitivity to changes induced by scCO₂ exposure. Post-exposure SEM-EDS provides supporting evidence that the electrical changes occurred alongside carbonation-related mineralogical and chemical transformation. Carbonate-rich domains were identified together with altered and decalcified C–S–H, silica-rich residues, and residual clinker-derived phases. In the GNP-containing cement, carbonation was spatially heterogeneous rather than uniform, with localized calcite occurrence and regions of strongly reduced Ca/Si ratios consistent with C–S–H decalcification. Our results show that scCO₂ exposure produces both measurable chemical alteration and a substantial shift in the electrical properties of wellbore cement. Rather than treating resistance as a direct measure of carbonation, the electrical response is interpreted as a practical indicator of changes in the cement matrix that can be evaluated during periodic well-integrity inspection. The stronger response observed in the GNP-modified cement suggests that graphene incorporation may increase the sensitivity of wellbore construction materials to scCO₂-induced alteration, providing a basis for future integrity-assessment methods in geothermal wells.

Topic: Geochemistry

Proppant packs keep hydraulic fractures conductive in enhanced geothermal systems (EGS), but at the Utah FORGE site they must carry closure stresses of 35-70 MPa at temperatures of 250 °C or more. Under these conditions, proppant grains may crush, generating fragments that alter the packing structure and pore geometry, potentially causing irreversible reductions in fracture conductivity. Stress and temperature damage the pack in different ways: higher stress breaks grains directly, whereas exposure to elevated temperatures and hydrothermal conditions can alter grain strength and increase susceptibility to crushing under a given stress. Laboratory conductivity tests measure how much conductivity remains, but they provide limited insight into the relative contributions of compaction of intact grains, grain crushing, and fragment redistribution. Discrete element models of proppant packs can follow grain breakage, but they often estimate permeability from average void ratio with the Kozeny-Carman relationship, whose predictive accuracy may be limited as crushing changes particle-size distributions and pore geometry. The extent to which different representations of grain breakage capture these microstructural changes and their consequences for hydraulic conductivity, particularly under elevated-temperature conditions, remains insufficiently understood. This study investigates how closure stress and temperature-dependent grain strength influence particle crushing, pack deformation, and hydraulic conductivity in propped fractures. Three proppant types are considered: conventional low-density ceramic particles, petroleum coke (petcoke), and magnetically tagged polymeric particles developed for enhanced durability under high-temperature and high-stress conditions. We hypothesize that grain crushing alters hydraulic conductivity beyond the effects of intact-grain compaction by modifying particle-size distributions and pore geometry, and that temperature-induced changes in grain strength influence the extent of crushing under a given closure stress. Our preliminary three-dimensional discrete element method (DEM) simulations focusing on low-density ceramic proppant packs using the fragment replacement method (FRM), calibrated against laboratory oedometer tests, indicate that incorporating particle breakage improves predictions of high-stress pack compressibility. Increasing the number of fragments generated per breakage event from two to four and six increased the mass fraction of particles smaller than the initial diameter from approximately 8% to 13% and 16%, respectively, while producing comparatively smaller differences in macroscopic compression and Kozeny-Carman-based permeability estimates. The simulations further indicate that fragment resolution influences contact-force redistribution, even when the mean coordination number remains nearly unchanged. Building on these preliminary findings, the ongoing study extends the DEM framework to incorporate the bonded particle method (BPM), which represents progressive intragranular fracture, and coupled computational fluid dynamics (CFD) to evaluate the hydraulic consequences of evolving particle-size distributions and packing structures. The effects of temperature are incorporated indirectly through mechanical parameters calibrated and validated against high-temperature single-particle crushing and/or oedometer compression experiments. The extended simulations will investigate how closure stress and temperature-dependent particle properties influence grain crushing, fragment generation, pack deformation, and hydraulic conductivity across the three proppant types. The preliminary results highlight the importance of particle-breakage representation in predicting pack deformation and contact-force redistribution. By extending the analysis to BPM and CFD, this study seeks to establish how these microscale mechanisms influence hydraulic conductivity degradation under EGS-relevant stress and temperature conditions, providing a mechanistic basis for evaluating proppant performance.

Topic: FORGE

As part of campus decarbonization efforts, district energy networks that incorporate ground sourced heat pumps and thermal storage are gaining popularity in the US universities. In this study, we examine the emission reduction potential of such systems on MIT campus. Our approach involves co-simulating the radial heat distribution in boreholes with optimal heat flows between the elements of the district energy network such as buildings, borefield, thermal batteries and central utility plant (CUP) to satisfy building demand. This approach demonstrates a simplified coupling of these two components that were extensively studied in the literature while providing reasonable estimates of emissions reduction potential compared to similar campuses. It also helps with straightforward interpretation of the drivers of the system such as coefficient of performance (COP) and carbon intensity of the electricity. Our results based on 2024 data demonstrated that the use of ground sourced heat and thermal storage provided around 20% emission reduction over the default CUP-only scenario. We also show the sensitivity of our results to the source of electricity used in heat pumps, optimization horizon, geological characteristics and the properties of thermal battery and ground heat exchangers.

Topic: Modeling

Geothermal development requires balancing early exploration uncertainty with the goal of minimizing environmental impact. Because subsurface conditions are poorly constrained at the beginning of a project, developers often permit and plan for a larger disturbance footprint than is ultimately required. As new geological, geophysical, and drilling data become available, development plans are refined, uncertainty is reduced, and surface impacts can often be substantially decreased. This study evaluates planned versus actual land disturbance across geothermal projects developed by Ormat in the western United States. Disturbance footprints were compared through the exploration, drilling, and development phases to quantify how project maturation influences land-use requirements. The analysis examines the relationship between subsurface uncertainty, permitting strategies, and drilling results, with particular emphasis on how technical learning translates into reduced surface impact. The findings highlight the importance of adaptive decision-making and staged project advancement in geothermal development. This approach allows developers to responsibly balance exploration flexibility with environmental stewardship. The results demonstrate that technical evaluation and iterative decision-making can reduce actual surface disturbance while maintaining exploration and development success.

Topic: Field Studies

Scaling and corrosion remain the most significant operational challenges affecting the efficiency, reliability, and sustainability of geothermal power generation. These processes reduce heat transfer efficiency, impair steam quality, accelerate equipment degradation, increase maintenance costs, and shorten the operational lifespan of geothermal facilities (Nogara & Zarrouk, 2018; Tassew, 2001). Although both phenomena have been widely investigated, they are frequently treated as independent operational problems despite being governed by the same physicochemical evolution of geothermal fluids during production and utilization (Mundhenk et al., 2013). Consequently, limited understanding exists regarding how changes in fluid chemistry within individual process units influence the coupled development of scaling and corrosion throughout a geothermal power plant. This research addresses this knowledge gap by developing an integrated geochemical framework for evaluating the mechanisms controlling scaling and corrosion within the Olkaria I Additional Units (IAU) geothermal power plant, Kenya. Unlike previous studies that primarily emphasize mineral characterization or corrosion assessment independently, this study integrates operational fluid chemistry, steam purity, deposit geochemistry, mineralogical characterization, and process operating conditions to evaluate the evolution of scaling and corrosion across multiple process environments. The framework is designed to identify process-unit-specific geochemical conditions that promote mineral deposition and material degradation, providing a scientific basis for targeted mitigation strategies. Representative deposits were collected from selected process units, including steam scrubbers, condensers, cooling tower basins, and gas extraction systems, where geothermal fluids experience significant physicochemical changes associated with flashing, steam separation, condensation, cooling, and non-condensable gas interactions. Deposit chemistry is being characterized using X-ray fluorescence (XRF), while mineralogical identification is being undertaken using X-ray diffraction (XRD). These datasets are interpreted alongside geothermal fluid chemistry and operational parameters to investigate the relationship between fluid evolution, mineral precipitation, and corrosion processes. Emphasis is placed on steam purity because impurity transport through liquid carryover and vapor-phase contaminants significantly influences deposition and localized corrosion within steam turbines and downstream equipment (Addison & Richardson, 2020; International Association for the Properties of Water and Steam [IAPWS], 2013). Rather than presenting a conventional characterization of geothermal deposits, this research proposes a process-unit-specific methodology for integrating geochemical and operational datasets to improve understanding of the spatial variability of scaling and corrosion within high-enthalpy geothermal systems. The proposed framework is expected to support evidence-based chemical treatment strategies by identifying locations requiring targeted intervention instead of uniform plant-wide chemical dosing. The outcomes are anticipated to contribute toward improved plant reliability, reduced maintenance requirements, enhanced steam quality, and more sustainable operation of geothermal power plants in Kenya and other high-enthalpy geothermal fields.

Topic: Geochemistry

The commercial deployment of superhot rock (SHR) geothermal resources requires more than reaching greater depths and temperatures; it demands drilling technologies and execution strategies capable of improving well productivity, increasing capital efficiency, and reducing the levelized cost of geothermal energy. Higher-temperature resources offer the potential to increase power generation per well, but realizing this advantage at scale depends on reliable, repeatable, and cost-effective well construction. Mazama Energy’s Athena well at Newberry Volcano, Oregon, represents a significant advancement toward these objectives, demonstrating improvements in drilling performance, equipment reliability, and operational efficiency while expanding the technical understanding of deep geothermal resource development. Reaching a measured depth of 15,110 ft (4,606 m), Athena builds upon lessons from Mazama’s preceding Newberry drilling campaign and provides an opportunity to evaluate engineering innovations under increasingly challenging downhole conditions. This paper presents Athena’s drilling performance, engineering design, operational optimization, and implementation of a multi-sleeve completion strategy intended to support thermal management, cementing operations, and future reservoir characterization. During the initial drilling phase, Athena reached 10,350 ft in just 15 drilling days, representing an operator-reported reduction of approximately 80% relative to the preceding well's drilling . Instantaneous rates of penetration exceeded 350 ft/hr, with sustained rates above 130 ft/hr across extended intervals. A single bit drilled 3,578 ft before reaching the planned casing point, while no downhole motor or directional drilling system failures were reported during this phase. These achievements demonstrate the potential of integrated drilling design, equipment selection, parameter optimization, and operational learning to reduce drilling time and improve well construction efficiency. The paper examines bit and bottomhole assembly performance, drilling parameter optimization, thermal management strategies, and operational practices across successive hole sections. Particular emphasis is placed on translating improved rates of penetration, extended bit runs, and downhole equipment reliability into reductions in drilling time, operational exposure, and ultimately well construction costs. Additionally, the paper discusses the engineering, planning, and implementation of a multi-sleeve program at predetermined depths, designed to address thermal requirements during cementing operations while enabling additional reservoir insights to inform subsequent horizontal well development. Comparative analysis with the preceding Newberry campaign evaluates how technological improvements and operational learning have translated into measurable drilling efficiencies, while identifying remaining technical and economic challenges associated with deeper SHR development. Athena's contribution extends beyond individual drilling records. By demonstrating improved drilling efficiency, advancing well construction practices, and expanding the accessible geothermal resource base, the project provides critical engineering insights into the technical and economic feasibility of SHR development. These findings establish an important foundation for Project Ceres, supporting Mazama Energy's broader objective of unlocking higher-capacity geothermal wells and enabling repeatable, cost-competitive SHR development at commercial scale.

Topic: Drilling

Superhot rock (SHR) resources ( greater than 374°C) hold the promise of efficient, commercially scalable extraction of geothermal energy. However, a critical challenge in drilling superhot formations is managing the temperature of the downhole measurement and directional tools, which, at most, are limited to 200°C. Several techniques have been proposed to overcome this challenge. Of these, Supercritical carbon dioxide (sCO₂) offers a promising approach to managing thermal exposure and extending the application of existing drilling technologies. By taking advantage of the phase behavior of supercritical and trans-critical CO2 through appropriate engineering and placement of downhole chokes, it is possible to maintain the temperature of downhole tools within existing tool limits at SHR conditions. This paper examines the engineering, operational planning, and implementation considerations required to translate this concept into a field-ready drilling system. The engineering scope integrates thermal and hydraulic modeling with well design, bottomhole assembly configuration, and surface equipment selection. Particular attention is given to CO₂ phase behavior throughout the circulation system, pressure-dependent cooling potential, transient temperature response, and compatibility with downhole tools and materials. Cooling objectives are evaluated alongside cuttings transport, pressure control, and surface handling requirements to define an operating envelope that supports both temperature management and drilling performance. The implementation framework covers system commissioning, staged testing, real-time monitoring, and predefined criteria for advancing to drilling operations. Performance assessment focuses on achieved cooling, stability of pressure and flow, tool exposure temperatures, hole-cleaning effectiveness, CO₂ consumption, and operational reliability. These measures establish a basis for comparing modeled behavior with field observations and identifying limitations to wider deployment. By connecting engineering design with practical execution requirements, this paper presents a structured pathway for using supercritical CO2 as a drilling-enabling technology. The central objective is to establish a controllable thermal operating window that supports reliable access to superhot geothermal resources.

Topic: Field Studies

An EGS project across most of Alberta, Canada, must drill deep into the Precambrian basement to find commercially viable reservoir temperatures exceeding 200°C. Despite hundreds of thousands of hydrocarbon wells drilled across Alberta, basement penetrations are sparse. Based on moderate-to-low surface heat flow estimates in the region, thermal conductivity samples, and radiogenic heat production from the granitoid/gneiss of the Rimbey Precambrian domain, we model a subsurface temperature of 209°C at 6,500 m depth at a proposed project location southwest of Edmonton. Published literature from the region, structural interpretations derived from airborne gravity and magnetic surveys, and 2D seismic data shot for research purposes indicate a compressional stress setting. Geomechanical data suggest a consistent stress-field trend, with maximum horizontal stress (SHmax) oriented NE/SW. Using simulation modelling and stress gradient estimates extrapolated from the sedimentary section to crystalline rock, hydraulic stimulation is expected to require bottomhole pressures exceeding 100 MPa and operating wellhead pressures of 37-44 MPa to sustain fracture aperture and commercial flow rates exceeding 60 kg/s per well. While these findings support the technical feasibility of EGS in Alberta's cratonic basement, direct measurements of deep basement temperature, stress, and fracture characteristics remain essential to reducing uncertainties associated with deep EGS projects. Resolving these uncertainties would significantly expand EGS deployment potential in stable cratonic settings worldwide.

Topic: Enhanced Geothermal Systems

Indonesia geothermal development has historically been dominated by high-temperature (≥230°C) volcanic systems using dry-steam and flash technologies. This pathway remains important for existing field optimization and near-term exploration projects. However, the future project pipeline is becoming more diverse, with growing attention to the exploration and development of low- to medium-temperature resources ( less than 230°C) and bottoming binary plant development. This raises two practical questions: “how significant could binary geothermal power plants become in Indonesia’s future pipeline, and under what project conditions are they technically and commercially viable?” This paper reviews global trends in binary geothermal power plant development and evaluates their relevance to Indonesia. A literature review and qualitative screening approach are applied, supported by project pipeline mapping using Indonesia’s planned geothermal projects. Benchmark cases from the United States, Türkiye, New Zealand, and the Philippines identify binary deployment practices, including standalone binary plants, bottoming binary units, and combined flash-binary cycle developments. The analysis suggests that binary geothermal power plants are unlikely to replace flash systems as Indonesia’s dominant geothermal power plant technology in the near term. Many operating fields and future priority projects remain high-temperature developments where flash technologies are more suitable. However, binary plants may become increasingly important through three pathways: brownfield optimization, low- to medium-temperature greenfield development, and emerging unconventional technology applications. High-temperature brownfield optimization through bottoming unit development is the most immediate opportunity and can serve as a practical learning platform for Indonesian developers before shifting into more challenging greenfield projects with lower-temperature resources. The paper argues that binary plants should not be viewed only as a niche add-on technology because, under viable project conditions, they can broaden Indonesia’s geothermal development beyond high-temperature fields. Nevertheless, the future role of binary geothermal plants will depend on resource deliverability, project economics, electricity tariff support, and local developers’ experience in developing and operating binary plants.

Topic: Field Studies

The decarbonisation of the energy sector and the need for greater energy security are key drivers of research into low-risk renewable energy resources. Unlike weather-dependent renewable resources such as wind and solar, geothermal energy can provide a constant, weather-independent baseload supply, with the potential to incorporate energy storage to help balance supply and demand. In recent years, alternative approaches that move beyond conventional geothermal development have been investigated. These approaches involve circulating a heat transfer fluid through closed-loop systems which have a range of geometrical configurations. Referred to as advanced geothermal systems (AGSs), these technologies offer potential advantages due to their comparatively low geological risk, particularly because they do not rely on the permeability or fluid availability in naturally occurring geothermal reservoirs. Typically, the focus has been on heat extraction potential, however in this work the focus is on seasonal storage. This study investigates the potential for thermal energy storage within AGSs through numerical modelling in OpenGeoSys (OGS) software, which utilises the finite-element method. Renewable energy resources, such as solar and wind, as well as waste heat from industrial processes, can provide heat for spatial heating; however, the availability of these sources often does not coincide with demand. Thermal energy storage enables the temporal shifting of energy supply, allowing heat generated during periods of excess availability to be stored and subsequently recovered when demand is higher. This study therefore examines the potential for thermal energy storage in deep closed-loop geothermal systems and evaluates the efficiency of such storage. Initial results show the impact of geometrical design, geological and engineering parameters.

Topic: Modeling

Hydraulic stimulation design in enhanced geothermal systems (EGS) requires exploration of large parameter spaces while accounting for complex interactions among fracture networks, fluid flow, heat transport, and reservoir properties. High-fidelity numerical simulations can capture these processes but are computationally expensive, limiting their direct use for rapid optimization and decision-making. As part of the U.S. Department of Energy’s Genesis Mission, this work develops an Agentic AI framework for hydraulic stimulation design and optimization and demonstrates the approach using the Utah FORGE geothermal site. The framework integrates fracture network generation, physics-based simulation, AI foundation model development, and optimization within an agent-managed workflow. Los Alamos National Laboratory’s dfnWorks is used to generate discrete fracture networks representing alternative fracture configurations, while the MOOSE/FALCON framework simulates coupled reservoir behavior under different fracture properties and stimulation scenarios. The resulting simulation ensemble provides training data for foundation models (FMs) that rapidly approximate reservoir responses. These FMs are coupled with multi-objective optimization to identify stimulation strategies that improve fracture connectivity, stimulated reservoir volume, injectivity, and energy-recovery-related performance while considering operational constraints. The agentic framework is implemented using the OpenAI Agents SDK, with scientific models and optimization algorithms exposed as tools that the AI agent can se1ect and execute. The AI agent serves as the orchestration and reasoning layer across these components, translating engineering objectives into computational tasks, invoking appropriate simulation and AI tools, evaluating results, performing iterative “what-if” analyses, and refining candidate stimulation designs with human-in-the-loop oversight. By connecting high-fidelity fracture and reservoir simulations with rapid AI prediction and optimization, the proposed framework seeks to move stimulation design from a computationally intensive, manually coordinated process toward an adaptive Agentic AI decision-support workflow. The approach provides a pathway for rapidly evaluating alternative stimulation strategies and improving fracture-system design for enhanced geothermal energy recovery.

Topic: FORGE

Conventional geothermal exploration has historically focused on areas with prominent surface manifestations such as hot springs, fumaroles, and hydrothermal alteration. However, many high-temperature geothermal resources remain concealed beneath sedimentary cover or volcanic sequences where surface evidence is limited or absent. These "blind" geothermal systems require integrated multidisciplinary exploration approaches capable of reducing drilling uncertainty. This study presents an integrated exploration workflow combining regional geological interpretation, structural analysis, two-dimensional seismic reflection data, magnetotelluric surveys, gravity and magnetic data, geochemical observations, and three-dimensional subsurface interpretation. Rather than relying on individual datasets, the workflow emphasizes the integration of complementary geoscientific information to identify structurally controlled permeability zones, reservoir compartments, potential heat sources, and optimal exploration drilling targets.

Topic: General

A new method for estimating power potential of electric-grade hydrothermal systems has been developed as an alternative to the volume method. The new method is used to estimate power potential for electric-grade systems in the Great Basin conventional hydrothermal assessment area. The results from this method are then compared to estimates from the volume method summarized by DeAngelo and Williams (2010). As part of their work, DeAngelo and Williams used Monte Carlo analysis to evaluate the estimated power capacity of 253 electric-grade hydrothermal systems across the Western US, 100 of which are within the current extent of the northern Great Basin. We have taken a similar approach using a set of parameters selected for Great Basin conventional hydrothermal systems, including distributions of reservoir volume, initial reservoir temperature, and effective fracture spacing. The results from the new method are shown to be comparable to those of DeAngelo and Williams. This new method has advantages over the volume method, including an explicit dependence of power potential on effective fracture spacing, and making the heat-to-electricity conversion a function of reservoir temperature. The data release accompanying this manuscrpt contains publicly available Python code for the estimation of likely power potential for any electric-grade hydrothermal systems (with uncertainty estimates) for systems with estimated distributions of the necessary parameters.

Topic: General

As global demand for renewable energy continues to increase, enhanced geothermal systems (EGS) have begun to emerge as a promising source of sustainable energy. A crucial component of EGS is the working fluid, which is initially injected into thermal reservoirs to increase permeability and then subsequently act as a circulating fluid for thermal energy extraction. However, the introduction of a working fluid can also promote unfavorable dissolution and precipitation reactions that could have potential implications on the mechanical behavior of EGS reservoirs. Despite this, few experimental studies investigate the effect geochemical reactions have on reservoir geomechanics within an EGS. In this study, the goal is to evaluate how dissolution and secondary mineralization may impact reservoir geomechanics through experiments simulating conditions of basaltic and rhyolitic reservoirs of the Snake River Plain (SRP), ID, which has been identified as an ideal candidate for EGS due to its high subsurface heat flow. Six hydrothermal experiments will be conducted in rocking autoclaves (rocker bombs) equipped with a flexible Au-Ti reaction cell (gold bag) at experimental conditions of 230 °C and 30 MPa for a minimum of 30 days. A synthetic Na-Cl-SO4 brackish water (pH = 6.70, I ≈ 0.092 mol/kg) was selected for all the experiments. This composition is based on the chemistry of waters that have been used as working fluids in successful EGS. The experiments are designed to address three important aspects of fluid-rock interactions in SRP reservoirs: (1) the mineral scaling that may occur; (2) the effects on mineral scaling if the brackish water is acidified with HCl and maintained at a pH of 5.5 or lower throughout the experimental duration; and (3) the effects of mineral scaling on geomechanical properties. Four experiments will address the first and second research objectives and will consist of either slightly altered basalt (30% plagioclase, 21% olivine, 7% pyroxene, 5% K-feldspar, 10% glass, and 9% smectite/vermiculite) or rhyolite (28% plagioclase, 26% K-feldspar, 25% quartz, 7% glass, and 4% smectite/vermiculite) rock chips (0.2–0.8 cm) collected from wells drilled in the SRP. These four experiments also will include a basalt or rhyolite rock cube (1 to 1.5 cm x 1 to 1.5 cm) that will undergo nanoindentation testing to measure local hardness and elastic response both pre- and post-experiment to address the third research objective. The remaining two experiments will contain a basalt or rhyolite core plug (1-inch x 2-inch) collected from outcrops in the SRP. These cores will be used to test how bulk rock compressibility changes under hydrostatic pressure and dry conditions both pre- and post- experiment using a rock compressibility system and will also address the third research objective. We hypothesize that initial dissolution of basaltic and rhyolitic glass will occur, causing increased concentrations of dissolved ions within the water. Basalt dissolution will likely enrich the water in Ca, Na, Mg, Al, and Fe, potentially promoting secondary mineralization of wairakite and mixed layer chlorite-smectites. Consequently, rhyolite dissolution will likely enrich the water in SiO2, Na, K, and Al, potentially promoting secondary mineralization of silica polymorphs and mixed layer illite-smectites. If dissolution and secondary mineralization occur, geomechanical properties will likely be affected as well, causing potential reservoir issues that could be critical for long-term performance of EGS.

Topic: Geochemistry

While aquifer thermal energy storage (ATES) development in low ambient groundwater flow systems has benefits that include higher recoverable stored thermal energy (heating/cooling potential) and lower likelihood to affect downgradient water users, groundwater flow rates in many natural systems may be sufficiently high to reduce efficiency due to advective heat leakage. The U.S. Geological Survey is tasked with developing maps that reflect resource potential under dense development scenarios. Herein, we consider alternative well placement configurations, demonstrating system performance under sparse and dense scenarios. We demonstrate that repeating five-spot and cross-gradient dense development scenarios can be used to simulate dense ATES resource development under a range of ambient groundwater flow conditions from low-flow to high-flow, respectively. Single system configurations show how these systems will perform during early adoption of technologies.

Topic: Low Temperature

Rapid expansion of U.S. enhanced geothermal systems (EGSs) requires scaling production of specialized, high-temperature materials and components. Supply chains for these materials and components are often constrained at key nodes to a limited number of locations and manufacturers. Understanding the dynamics of these supply chains is critical for identifying supply risks and solutions to support EGS growth. To address this, we are developing a supply chain risk assessment framework customized to EGS technologies through material flow analysis (MFA). Guided by industry engagement, we identified two key specialty materials and components for applying the framework: high-temperature cement and downhole tool semiconductors (SiC, GaN, silicon-on-insulators). In this research, we track material flows throughout the global supply chain, identifying key supply risks (e.g., high concentration of manufacturing in a single country or entity) and opportunities for mitigating risks (process improvement, alternative supplies, recycling potential).

Topic: Enhanced Geothermal Systems

Thermal short-circuiting in enhanced geothermal systems (EGS) occurs when injected cold fluids flow preferentially through high-permeability pathways toward the production well, causing localized heat depletion and premature thermal breakthrough. Thermoresponsive nanocomposite poly(N-isopropylacrylamide) (pNIPA) microgels have been proposed as a reservoir treatment for this problem: the microgels swell when cooled, so they preferentially clog cooled flow channels and divert flow toward hotter regions of the fracture where they shrink at high temperature. The microgels have previously been tested in Taylor-Couette and Hele-Shaw geometries, but neither setup was made of geologic materials. Here, a proof-of-concept study is presented using a meter-scale granite fracture analog heated from below to impose a temperature gradient. The microgels are used as an active tracer to clog an artificial preferential flow channel within the analog and redirect circulating fluids to hotter regions with higher effective heat transfer surface area. Initial results indicate that thermal breakthrough is reversed, as observed in wellhead fluid temperatures, inter-well pressure drop, and thermal imaging of the granite surface.

Topic: Emerging Technology

Interpreting chemical changes in enhanced geothermal systems requires connecting reservoir reactions to the conditions encountered during injection, production, and surface processing. This study develops an integrated thermodynamic descrption of the Utah FORGE circulation loop to test explanations for observed fluid chemistry, separating equilibrium effects from other processes. Geochemical equilibrium calculations using Reaktoro follow measured injection water compositions through the injection well, representative fracture environments, production well, and surface equipment. At each stage, the model evaluates how temperature and pressure affect mineral dissolution, precipitation, and the distribution of gases. Reservoir calculations at 210°C and 48 MPa were compared with measured production chemistry to assess how mineral composition, accessibility, dissolved gases, and finite salt inventories influence the predicted chemical response. Reactions with primary and fracture minerals reproduced the observed silica enrichment and calcium–magnesium depletion for all four 2024 sample pairs, matching magnesium removal within about 5% and capturing 70–80% of the observed calcium removal. Finite soluble salt inventories produced sodium, potassium, and chloride enrichment absent from granite equilibrium alone. Mineral transfers revealed competing sources and sinks for dissolved constituents, with some minerals switching between dissolution and precipitation depending on the surrounding assemblage and injection water. Granite and diorite assemblages containing the same minerals produced the same fluid despite different mineral proportions, whereas plagioclase composition and the availability of alteration minerals changed it substantially. These findings indicate that the minerals accessible to circulating water and the available chemical inventories are essential to interpreting production chemistry. Calculations are being extended through the wells and surface equipment, where planned phase boundary maps will examine how heating, cooling, pressure changes, and mixing affect mineral saturation and gas exsolution. These results will support operational assessment and provide thermodynamic constraints for future reactive transport modeling.

Topic: Geochemistry

Indonesia's growing demand for space cooling highlights the need for energy efficient alternatives, including ground source heat pump (GSHP) systems that utilize the relatively stable shallow subsurface as a heat sink. However, limited site specific thermal data and practical assessment methodologies remain significant challenges for early GSHP development in Indonesia. This study presents an analogue based desktop assessment methodology applied to a representative cooling application at a university campus in Yogyakarta, Indonesia. Published geological and hydrogeological information, thermal properties derived from comparable tropical GSHP installations, cooling load calculations, and preliminary techno economic assessments are integrated to evaluate subsurface suitability and system performance without requiring initial field investigations. The assessment adopts a representative ground thermal conductivity of 2.0 W/m·K, with a sensitivity range of 1.8 to 2.2 W/m·K, and estimates a peak cooling load of 22.72 kW for a 200 m² academic space. At an assumed cooling coefficient of performance (COP) of 5.0 and 5,760 annual operating hours, the proposed vertical closed loop GSHP is estimated to consume 26,173 kWh annually, compared with 37,391 kWh for conventional split air conditioning. The preliminary assessment indicates approximately 31% lower annual operating costs, equivalent to USD 1,322 in savings, and an estimated reduction of 8.13 tonnes of indirect CO₂ emissions per year. With an estimated installed investment of USD 20,000 to 30,000, the base case yields a simple payback period of approximately 12.9 years. The findings demonstrate how analogue based subsurface characterization can support preliminary GSHP screening, inform investment decisions, and identify the additional field investigations required before detailed engineering design.

Topic: Low Temperature

The importance of geothermal wells is steadily rising with the increased focus on renewable energy and the recent successes in Enhanced Geothermal Systems (EGS). Since higher arrival temperatures yield higher thermal and electric power, there has been a surge of interest in exploiting the potential of Super-Hot Rock (SHR) formations (resource temperature ≥ 374°C). Drilling wells in SHR resources involves a formidable host of challenges, primary amongst which is drilling near-horizontal wells, while maintaining sufficiently low downhole circulating bottom-hole temperatures in order to stay within the thermal limits of the downhole instrumentation to steer and maintain laterals parallel to each other. This, in turn, requires models and tools that can accurately predict circulating downhole temperatures. Many models are available to predict circulating temperatures. However, most appear to overpredict the temperature downhole, and thus underpredict the maximum resource temperatures that can be accessed with the limits of existing measurement while drilling (MWD) tools. A possible reason for this is that these tools do not account for the drilling history of the well. Accurate estimation of circulating temperatures requires a proper accounting of the entire operational sequence during drilling, so as to capture the thermal legacy of the well due to the cooling of the wellbore near-field during circulation and the subsequent recovery during a quiescent (Shut-in) period. This thermal legacy must account for the interaction of the wellbore with the formation, creating a near-field “thermal blanket” that acts as a heat sink during circulation and a heat source during thermal recovery when the wellbore is shut-in. In other words, the initial thermal condition of the well can no longer be represented by a single “undisturbed” geothermal temperature profile, but by temperature field which is a function of time. This study presents a semi-analytical approach to predict wellbore temperatures during an operational sequence at a level of detail not considered before in the literature. The approach considers the progression of the drill string down each hole section along with the wellbore geometry changes not considered in previous studies. The methodology involves a semi-analytical solution of the transport equations in a moving domain formulated in terms of enthalpy, and thermal coupling with the formation by means of Duhamel Convolution Integrals. The granularity of the analysis allows consideration all operations, such as Drilling, Running in Hole (of both drill pipe and casing), Pulling out of Hole, Fluids Displacement, Circulation on Bottom, and Waiting on Cement, in addition to what drillers call the Rebound, i.e., the downhole temperature recovery during the making up of a stand. The semi-analytical approach, and the numerical methods used, allow us the luxury of such granularity without suffering the resulting computational time and expense. Examples are shown to illustrate the methods described in this work. It is shown that proper modeling of all the stages of drilling (from spud) results in more accurate prediction of the MWD temperatures, and that these temperatures can be significantly lower than predicted by most commercial thermal simulators. Two key insights emerge from this work: that accounting for thermal legacy is necessary to accurately predict the circulating temperature at any time and depth of interest; and that circulating MWD temperatures can be better controlled by managing the legacy through proper operational planning.

Topic: Drilling

Enhanced Geothermal Systems (EGS) offer vast potential for baseload clean energy, but their long-term viability heavily depends on managing subsurface geofluid and heat sustainability. While fluid loss is a recognized operational challenge, its financial consequences are rarely ed across different project sizes and reservoir thermal quality. This paper presents a generalized techno-economic assessment of the impacts of water loss across various reservoir temperature and EGS project scales, ranging from small-scale modules (50MWe) to large-scale commercial plants (200MWe). Using a coupled reservoir-economic modeling framework, we evaluate how varying fluid loss rates influence key operational and financial metrics, including Net Power generated (MWe), Levelized Cost of Electricity (LCOE), Net Present Value (NPV), and Internal Rate of Return (IRR). Our findings reveal a distinct, scale-dependent sensitivity to fluid loss that is intertwined with reservoir thermal quality. Lower-temperatures reservoirs (150°C–200°C) amplify the economic penalties of water loss, rendering small-scale projects highly vulnerable due to the fixed overhead costs of makeup water infrastructure. Conversely, high-temperature reservoirs (200°C–300°C) with project scale more than 100 MWe significantly buffer these economic losses by yielding higher power outputs per unit of circulating fluid, though large-scale developments remain exposed to severe absolute volumetric makeup demands over a 30-year plant lifespan. Ultimately, this study demonstrates that mitigating water loss is a critical threshold constraint for low-temperature, small-scale viability, whereas for high-temperature, large-scale deployments, it dictates long-term reservoir thermal breakthrough dynamics and macroeconomic resilience. These results provide developers and policymakers with a multi-dimensional framework to optimize EGS development plan and site selection relative to local faults, natural fractures, water sources and thermal quality.

Topic: General

Large-scale closed-loop geothermal simulation is computationally expensive because field-scale models must resolve long-term heat transfer across large subsurface domains while capturing geological heterogeneity and thermal interactions among multiple wells. This cost also limits machine-learning surrogates, which are commonly trained on simulation domains comparable in scale to their intended deployment. Here, we investigate whether geothermal surrogates can instead be trained on small local domains and generalized directly to much larger multi-well systems without retraining. We develop a domain-extending neural-operator framework trained exclusively on one- and two-well domains. The model learns reusable single-well behavior and local well–well interactions using spatial representations that remain independent of overall system size, enabling prediction of transient subsurface temperature fields and well performance across larger configurations. Preliminary results for homogeneous closed-loop geothermal systems show that the surrogate generalizes zero-shot to previously unseen systems containing up to 25 wells, achieving a median relative error of 3.8% in three-dimensional rock-temperature change over a 20-year simulation horizon. We plan to further investigate the physical limits of this local-to-global generalization by varying well spacing, geological heterogeneity, thermal-property correlation length, and the characteristic length scales of thermal interference. By removing the need to generate new field-scale training libraries as development size increases, this approach could enable rapid evaluation and optimization of large closed-loop geothermal developments, including well placement, spacing, system scale, and long-term thermal performance.

Topic: Modeling

Long-term operation of geothermal systems is strongly affected by the propagation of the cold plume away from the injection well. A dedicated monitoring well provides a unique opportunity to detect plume arrival and evolution before it affects the production well, enabling early intervention, operational optimization, and reduction of geological uncertainty. However, the timing and spatial distribution of cooling at the monitoring well depend strongly on heterogeneous reservoir architecture and preferential flow pathways. This study investigates the use of a planned vertical monitoring well to characterize cold-plume evolution in the Delft Campus geothermal system. An ensemble of high-resolution geological realizations of the fluvial reservoir is simulated using the GPU-accelerated open-DARTS simulator to quantify uncertainty in plume propagation. Temperature distributions between the injector and monitoring well reveal substantial variability in both the lateral and vertical development of the cold plume. Connected high-permeability intervals generate preferential flow pathways, resulting in strongly asymmetric plume geometries and significant differences in the timing and magnitude of cooling at the monitoring well. The results provide a basis for future evaluation of monitoring strategies and their ability to detect and characterize the evolving cold plume. Once monitoring data become available, the measured temperature response at the monitoring well can be compared with the ensemble predictions to identify the most plausible geological realizations, constrain preferential flow pathways, and reduce uncertainty in plume evolution. This establishes the monitoring well as a key component of a future closed-loop geothermal management strategy, in which field observations can be assimilated to improve predictions and support optimization of injection and production over the lifetime of the system.

Topic: Reservoir Engineering

Silica scaling is a persistent challenge in geothermal energy production, where precipitation and particle aggregation can progressively reduce pore and fracture permeability, impair injectivity, and compromise long-term reservoir performance. These processes span a wide range of spatial and temporal scales, from molecular-scale silica polymerization and colloidal aggregation to near-wellbore pore clogging and reservoir-scale fracture evolution. Here, we develop a multiscale, multiphysics modeling framework to quantify silica-induced clogging and evaluate potential mitigation strategies in geothermal systems. At the near-wellbore scale, a population balance equation is coupled with Darcy–Brinkman–Forchheimer flow, multispecies reactive transport, silica polycondensation kinetics, and particle–pore-throat interactions. The framework explicitly resolves particle nucleation, surface growth, Brownian coagulation, and multi-particle bridging, enabling dynamic prediction of particle-size distributions, porosity loss, and permeability impairment. The model is validated against both the analytical Smoluchowski coagulation solution and laboratory column experiments. At the reservoir scale, a fully coupled thermal–hydraulic–chemical model represents flow and heat transport through discrete fracture networks while dynamically updating fracture apertures according to silica dissolution and precipitation. Simulations reveal that severe precipitation and permeability reduction are strongly localized near the injection well, whereas long-term thermal perturbations propagate hundreds of metres into the reservoir. Scale-inhibitor simulations further demonstrate that suppressing particle aggregation can reduce condensed silica concentrations by approximately one order of magnitude and substantially delay pore-throat blockage. The proposed framework connects particle-scale clogging mechanisms with reservoir-scale geothermal performance and provides a physics-based tool for predicting injectivity decline, evaluating anti-scaling strategies, and optimizing geothermal injection operations toward more sustainable heat extraction.

Topic: Geochemistry

Accurate prediction of fluid flow and heat transport in fractured reservoirs is essential for geothermal energy development. However, complex fracture connectivity and the computational cost of numerical simulation limit ensemble forecasting, uncertainty assessment, and reservoir characterization. We present FracCGM, a conditional denoising diffusion framework for rapid probabilistic prediction of spatiotemporal pressure and temperature fields in fractured geothermal reservoirs. The framework learns from stochastic discrete fracture networks and their corresponding physics-based simulations. Conditioned on fracture geometry and temporal information, it generates ensembles of reservoir states through iterative denoising, providing both predictive estimates and spatially resolved measures of ensemble variability. We evaluate FracCGM using three synthetic two-dimensional geothermal case studies with progressively increasing fracture complexity and its performance against a deterministic U-Net surrogate. FracCGM achieves higher predictive accuracy than the U-Net baseline and reproduces thermal-front propagation, preferential flow pathways, and production-well temperature evolution. Ensemble variability is concentrated near advancing thermal fronts and fracture intersections, highlighting regions of greater predictive uncertainty. The model also demonstrates generalization to previously unseen fracture configurations. Furthermore, integrating the pretrained model into an ensemble smoother enables fracture network inversion from sparse well pressure and temperature observations, improving data matching and reducing uncertainty in fracture locations. These results demonstrate a practical approach to combining rapid forward prediction, probabilistic forecasting, and observation-constrained fracture characterization. The framework supports computationally efficient scenario analysis for geothermal reservoir management and offers potential for extension to other fractured subsurface energy systems.

Topic: Reservoir Engineering

Cold underground thermal energy storage (UTES) offers a promising approach to reducing electric demand for data center cooling by storing winter cold in subsurface reservoirs for summer use. However, its performance remains difficult to predict because weather-dependent charging opportunities, cooling demand, and subsurface flow processes interact over seasonal timescales. In this study, we develop a three-dimensional coupled groundwater flow and heat transport model to investigate seasonal reservoir thermal energy storage (RTES) in the St. Peter Sandstone in Illinois, USA. The model incorporates realistic geological heterogeneity and reservoir geometry, with charging and discharging operations driven by hourly ambient air temperatures. Preliminary simulations indicate that seasonal cold recovery depends on the timing and duration of charging and discharging, with temperature-based and energy-based recovery metrics providing different assessments of system performance. Geological heterogeneity influences the spatial distribution and migration of stored cooling energy, while well configuration and hydraulic pressure constraints affect operational feasibility. The simulations also highlight the importance of numerical resolution in representing near-well thermal transport and predicting thermal breakthrough. These findings illustrate the interconnected roles of meteorological conditions, reservoir characteristics, and operating strategies in seasonal thermal energy storage. The modeling framework provides a basis for further evaluating cold RTES performance and exploring design and operational strategies in heterogeneous sedimentary reservoirs.

Topic: Modeling

As power systems evolve through the expansion of renewable generation, electrification, emerging large-load customers, and increasing reliability requirements, interest is growing in energy resources capable of providing reliable, dispatchable, low-carbon power. Geothermal energy is well positioned to contribute to future energy systems through its ability to provide firm generation, capacity value, operational flexibility, and long asset lifetimes. However, existing geothermal assessment tools typically evaluate subsurface performance, power plant operation, economics, and market participation independently, limiting the ability to quantify the full system value of geothermal resources and understand how alternative operating strategies affect long-term resource performance. This paper presents the development of an integrated geothermal value assessment framework that couples dynamic reservoir simulation, dynamic power plant modeling, power system operations, electricity market simulation, reliability assessment, and project economics within a unified workflow. The current effort focuses on developing the modeling architecture, data flows, and coupling methodologies needed to link reservoir behavior with power system outcomes. The framework is designed to support future evaluation of geothermal operating modes ranging from conventional baseload generation to flexible dispatch and hybrid energy system applications. The framework captures interactions among reservoir pressure, temperature, injectivity, thermal depletion, plant operating constraints, and evolving power system needs over the life of a geothermal asset. By integrating modeling domains that are traditionally evaluated separately, it provides a consistent approach for analyzing how operational decisions influence energy production, market participation, ancillary service provision, resource adequacy contributions, and long-term asset performance while maintaining fidelity to subsurface and plant constraints. The resulting capability provides a foundation for investigating key questions facing utilities, developers, policymakers, and investors, including the system value of geothermal energy, the degree of operational flexibility that can be achieved without compromising reservoir sustainability, and the conditions under which geothermal resources deliver the greatest reliability and economic benefits. By linking geoscience, power generation, electricity market operations, and reliability assessment, the framework establishes a foundation for evaluating the full system value of geothermal resources under a wide range of future power system conditions.

Topic: Modeling

[Choate]

Crossing and Recrossing the Brittle-Ductile Transition Paul CHOATE, John EICHELBERGER, Scott BROOKS, Cathleen LU [Alaska Division of Geological and Geophysical Surveys, USA]

The ductile regime, beginning at temperatures just above supercritical fluid conditions, requires different materials for drilling and involves different mechanisms by which permeability is generated. Much attention in superhot geothermal development has focused on producing fluid at or above its critical point (374°C, 22.1 MPa for pure water) because the high enthalpy of such fluids offers the potential for greatly increased electrical output per well. The critical point, however, is not a phase boundary, its pressure-temperature conditions vary with fluid composition, and pressure at a given depth depends on assumptions about the overlying fluid pathway. From an engineering standpoint, the simpler objective is to produce the highest-enthalpy fluid that can practicably be delivered to the surface. This requires confronting a more consequential frontier: the brittle–ductile transition (BDT) in reservoir rock and the corresponding high-temperature limitations of well materials. Moreover, heat transfer across a fluid boundary layer causes production temperature to be substantially lower than undisturbed formation temperature. Producing superhot fluid may therefore require drilling and completing wells in rock far hotter than the desired production temperature, potentially approaching magmatic conditions. Indeed, magma has been drilled as brittle rock that had been magma minutes before the drill bit reached it and quickly returned to being magma when fluid circulation stopped. Reaching superhot conditions therefore involves more than extending existing geothermal technology to higher temperatures. Crossing the brittle–ductile transition introduces three fundamental changes: 1) well construction materials and designs capable of operating at temperatures beyond those encountered by the oil and gas industry, 2) reservoir permeability generated by thermal contraction rather than hydraulic pressure, and 3) rapid physical and chemical dynamics, with processes that are reversible. These material, rock-mechanical, and hydrodynamic challenges, rather than the critical point of the fluid alone, may ultimately define the frontier of superhot geothermal development. Extensive experimentation near magma will be necessary.

Topic: Emerging Technology

Repurposing existing wells can accelerate geothermal development by reducing drilling requirements and leveraging legacy subsurface infrastructure. However, neither an existing well nor accessible subsurface heat alone constitutes a viable geothermal resource. This study develops an integrated framework linking well-scale feasibility with reservoir and regional resource assessment through geological characterization, well integrity and configuration, techno-economic performance, infrastructure, heat demand, and regulatory constraints. The framework was applied to regional assessments and representative sites in Berlin-Brandenburg, Lusatia, and Groß Schönebeck, Germany. Utilization pathways included Enhanced Geothermal Systems (EGS), Borehole Heat Exchangers (BHE), and Hydrothermal Energy (HE), with Aquifer Thermal Energy Storage (ATES) and Borehole Thermal Energy Storage (BTES) evaluated as screening-level alternatives. Results demonstrate that reservoir characteristics and conversion strategies are site-specific and that technical suitability alone does not ensure deployment viability. Successful development requires favorable reservoir performance, well configuration, infrastructure, and energy demand, with HE emerging as a promising regional pathway where naturally permeable reservoirs and suitable infrastructure coincide. Building on these findings, the framework can be extended from individual well assessment toward reservoir and regional geothermal play evaluation, with particular emphasis on Hot Sedimentary Aquifers (HSA). At the reservoir scale, temperature, transmissibility, thickness, continuity of reservoir distribution, and fluid properties determine resource quality, while basin architecture, stratigraphic and structural controls, legacy subsurface data, and regional play fairways define prospective development areas at the basin scale. Legacy hydrocarbon wells can serve both as potential geothermal assets and as sources of subsurface information for reservoir characterization and geothermal fairway identification. The resulting asset-reservoir-basin framework provides a systematic approach for identifying deployment-ready opportunities in mature hydrocarbon provinces and emerging sedimentary geothermal regions.

Topic: Field Studies

Distributed Acoustic Sensing (DAS) is valued as a microseismic monitoring tool for EGS due to its ability to endure high temperatures and its ability to resolve small magnitude events that would otherwise be undetected by surface networks, due to its borehole location. During stimulation at Fervo’s Cape Station Project, we detected over 30,000 small repeating events on DAS over the course of a ~2 week stimulation using a matched-filter technique. The repeating events (hereafter referred to as “burst repeaters”) are characterized by highly similar waveforms and identical moveouts across DAS channels, and often occur in bursts very close in time and space to the reference event that detected them. Due to their small magnitude, many burst repeaters are only detected on a single fiber. We use a novel technique to relocate single fiber detections and find that the detected repeater families fall along faults previously delineated by microseismic data. We show that a constant stress drop assumption fits the repeater families well and provides reasonable magnitude scaling. We also align the occurrence of repeating families to low-frequency DAS (LFDAS) signals observed on a horizontal fiber, which yields insight into the partitioning of aseismic and seismic behavior along the fault. Taken together, our results illuminate the spatiotemporal evolution of faults during EGS stimulation, with applications to hazards monitoring and reservoir development.

Topic: Enhanced Geothermal Systems

The study area is located near Diyarbakır city center, along the northern margin of the Arabian Plate in southeastern Türkiye. In 2025, Türkiye produced 47.9 million barrels of oil, corresponding to an annual average of approximately 131,200 barrels per day. More than 95% of the country’s oil production comes from southeastern Türkiye, where Cretaceous carbonate rocks host major petroleum reservoirs. This study assesses the geothermal potential of the Cretaceous Mardin Group carbonates in the Diyarbakır area. The assessment integrates geological and geophysical surveys conducted by the General Directorate of Mineral Research and Exploration (MTA) during 2010-2025 with records from numerous existing deep petroleum and water wells and data from three newly drilled geothermal exploration wells ranging in depth from 648.15 to 2,105 m. The results identify two principal geothermal reservoir intervals. The shallow interval comprises the Eocene-Oligocene Hoya-Gaziantep carbonates of the Midyat Group and exhibits temperatures of approximately 20-35 °C. The deeper interval comprises the Cretaceous Mardin Group carbonates, which also host important petroleum reservoirs, and exhibits temperatures of approximately 40-75 °C within the study area. The first well is 648.15 m deep and yielded geothermal fluid at 13 L/s and 32 °C during compressor-assisted production, while its artesian discharge was measured at 4.68 L/s. The second well is 2,105 m deep and yielded approximately 40 L/s at 64 °C during compressor-assisted production. The third well is 2,069 m deep and yielded 16-21 L/s at 66.3 °C during an electric submersible pump (ESP) test. Hydrochemical analyses indicate Ca-Mg-HCO₃-SO₄ water with low mineralization in the 648.15-m-deep first well and Na-Cl-HCO₃ water with higher mineralization in the 2,105-m-deep second well. The third well, which is 2,069 m deep, contains Na-Ca-Cl water with subordinate bicarbonate and a measured total dissolved solids concentration of 1,821 mg/L. These compositions distinguish the sampled shallow reservoir fluid from the deeper fluids and demonstrate hydrochemical variability between the two deep wells. The estimated geothermal gradient for the main target interval is approximately 27-28 °C/km. Together, the geological, thermal, and hydrochemical findings support the interpretation of a conduction-dominated foreland-basin geothermal system in the study area.

Topic: Low Temperature

Scientific advancements in enhanced geothermal systems (EGS) are essential for maximizing the potential to exploit low-permeability geothermal reservoirs. Analysis of geochemical interactions between working fluids and subsurface rock reveals risks to the long-term viability of an EGS. Basalts and rhyolites of the Snake River Plain (SRP), ID have been identified as prime EGS targets. Previous studies have characterized the rhyolite and basalt from the SRP. However, an understanding of how geothermal working fluids drive secondary mineralization in these reservoirs remains limited. In this study, we employ hydrothermal laboratory experiments to evaluate potential secondary mineralization in basalt and rhyolite EGS on the SRP. Two hydrothermal experiments simulate the highest reservoir temperature (230 °C) that an EGS is predicted to reach in the SRP. Two additional experiments simulate the lower temperature (80 °C) expected near the injection well. All four experiments react altered basalt (30% plagioclase, 21% olivine, 10% glass, 9% smectite/vermiculite, 7% pyroxene, 5% K-feldspar) or rhyolite (4% smectite/vermiculite, 7% glass) with synthetic brackish water (pH = 6.70, Ionic Strength = 0.092 m) at 30 MPa for at least 35 days. Rock samples used in the experiments were collected from the USGS Denver Core Library. The working hypothesis is that the thermodynamic instability of volcanic glass in both rocks will drive rapid initial dissolution of the glass promoting secondary clay and/or zeolite mineralization in the reservoir. For the 80 °C experiment with the rhyolite, the dissolution of glass is hypothesized to yield beidellite ± clinoptilolite whereas in the 230 °C experiments mixed layer clay + silica polymorphs are hypothesized to form. Conversely, for the 80 °C experiment with the basalt, the dissolution of glass is hypothesized to yield saponite whereas in the 230 °C experiments corrensite + wairakite are hypothesized to form. To conduct these experiments, rocking autoclaves (rocker bombs), Au-Ti reaction cells (Dickson cells), and previously established methods are being utilized to replicate EGS reservoir conditions. Thus far, six water samples were collected from the 80 °C basalt experiment. The first sample was collected at the start of the experiment and the sixth sample was collected after 840 hours of reaction. Water samples were analyzed using ion chromatography (IC) for anions, Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES) for major cations, Inductively Coupled Plasma Mass Spectrometry (ICP-MS) for trace metals, a pH electrode to measure pH, and carbon coulometry to measure dissolved inorganic carbon. Aqueous silica concentrations increased from 0.31 mmol/kg to 2.27 mmol/kg between the first and the sixth sample and aqueous magnesium concentrations decreased from 2.29 mmol/kg to 1.42 mmol/kg. Aluminum concentrations remained relatively constant, varying from 27.3 ppb to 24.6 ppb between the first and the sixth sample. pH decreased from an initial pH of 7.34 to 6.76 and DIC decreased from 39.84 ppm to 25.71 ppm between the first and the sixth sample. These results indicate that volcanic glass is dissolving to produce permeability-limiting secondary clays and zeolites. Identifying these mineralogical pathways will provide insight to mitigate reservoir degradation and enhance the long-term viability of an EGS.

Topic: Geochemistry

Neural operator surrogates have shown promise for accelerating computationally intensive geothermal reservoir workflows like rapid optimization, history matching, and uncertainty quantification. However, their reliance on regular-grid discretizations limits fidelity in fault- and fracture-dominated systems, where fault surfaces must be interpolated onto a uniform grid, discarding geometric detail precisely at fault interfaces, where local heterogeneity and thermal fronts are most pronounced. A prior study evaluated convolutional (U-Net) and spectral neural-operator architectures (FNO, U-FNO, Fourier-MIONet) on 3D temperature-field prediction across four fault-conceptualization scenarios of an anonymized geothermal field site, finding that FNO offered the strongest accuracy–robustness trade-off among grid-based surrogates. Here, we extend that to evaluate the Adaptive Physics Transformer (APT), a mesh-agnostic neural operator that can operate directly on unstructured geologic meshes without grid interpolation. APT combines a global attention encoder, which captures long-range correlations in reservoir properties, with a local graph-neural-operator encoder that resolves sharp, fault-localized heterogeneity; the resulting latent representation drives the prediction of thermal-front evolution in time.. Using the identical fault scenarios, train/test protocols, and error metrics established by the previous work, we perform a head-to-head comparison of APT and FNO on same-scenario, cross-scenario (leave-one-scenario-out), and combined-scenario prediction of 3D temperature evolution. This directly tests whether APT's native handling of fault geometry — a key limitation of grid-based surrogates — translates into improved predictive accuracy and generalization for geothermal decision workflows, while quantifying the associated trade-offs in training data requirements and inference cost.

Topic: Reservoir Engineering

In Massachusetts, where nearly 60% of household energy use is for space heating and cooling (mostly heating), Thermal Energy Networks (TENs) are an efficient and scalable method for reducing energy costs and public health burdens while reducing the need for extensive build-out of electric grid infrastructure. Built using well-understood engineering principles and mature technology, TENs are increasingly appealing to communities and are expanding from their historical prevalence in university campus energy systems to utility scale installations. However, there continues to be little research data on energy storage; performance projections; or other network considerations including site-specific feasibility, costing and upscaling. Through the multi-year Learning from the Ground Up (LeGUp) research program funded by the Massachusetts Clean Energy Center, this study synthesizes multiple data sets from discrete measurements of geological properties to thermal resources, building density, and energy use profiles to develop a data-driven mapping tool that can enable scaling TENs throughout the state. This mapping tool for assessing TENs potential throughout the state can provide decision-makers with a clear picture of the potential for this technology in Massachusetts and areas where networked geothermal is likely to be best suited. Built using the National Lab of the Rockies Geothermal Play Fairway Analysis (geoPFA) open source library, this TENs siting tool can be readily adapted for other regions as well as dynamically incorporate additional information as new data and information emerges in the thermal energy space.

Topic: Low Temperature

The U.S. Geological Survey (USGS) is currently assessing conventional geothermal energy resources in the northern Great Basin region of the western U.S. We use similar approaches to the previous 2008 conventional assessment (Williams and others, 2008b) but employ updated techniques and data to estimate discovered, undiscovered, and total resources in terms of electric power production potential. In this paper, we develop conditional statistics for use in estimating the resource potential. In companion papers, we describe the use of Monte Carlo analysis to estimate power production potential (Mordensky and others, 2027) which uses a method of estimating hydrothermal system power potential (Burns and other, 2027) that deviates from the volume method of Williams and others (2008a). We use a recently updated list of known/discovered conventional hydrothermal systems in the region (DeAngelo and others, 2026) to estimate the density of known systems in explored areas as a function of hydrothermal system favorability (Mordensky and others, 2026a, b). We develop a compilation of points where hydrothermal resources have been confirmed or verified not to exist, identifying these as locations that have been explored. We then apply a fixed-radius (2, 4, or 6-km) geographic buffer, assuming that this area has been explored. An estimate of the area allows us to use the known/discovered systems to estimate the density of systems as a function of hydrothermal favorability, from which the number of undiscovered systems can be estimated. A clear pattern of higher density of electric-grade hydrothermal systems is seen in more favorable areas. We also evaluated the estimated power capacity as a function of favorability but found no strong correlation between system size and favorability. In summary for conditional statistics, there are more electric-grade hydrothermal systems in more favorable areas, but in the absence of strong correlation between system size and favorability, all systems are assumed to belong to the same population. DeAngelo and others (2026) provide ranks for newly discovered resources, representing uncertainty that some locations will be viable resources. To evaluate this uncertainty in new resources, different confidence groups (e.g., high to low) were developed, and densities were developed for each group. For each group, we developed functional relationships of density as a function of favorability that accounts for exploration bias. In a companion paper, the relationships between system density and favorability described herein are used for a Monte Carlo analysis of discovered and undiscovered power production potential for the study area (Mordensky and others, 2027).

Topic: General

Torsional vibrations reduce drilling performance and damage tools in hard-rock geothermal wells, yet most rigs detect them only from surface torque and rotary speed. We tested the limits of surface-based detection on Utah FORGE wells 16A(78)-32, 78B-32 and 16B(78)-32, using downhole memory-tool data as ground truth. Detectors trained only on simulations of the Aarsnes-Shor torsional model were compared with a machine-learning model trained on the Volve field and with a simple torque-variability index. In a blind test on six FORGE bit runs, the physics-based detector reached an area under the ROC curve (AUC) of 0.76, above the transferred machine-learning model (0.71) but not above the simple index (0.79). Per-second downhole spectra from four bottom-hole assemblies explain this ceiling: most of what downhole tools record as stick-slip is high-frequency torsional oscillation (HFTO) at fixed BHA modes of 35-140 Hz, which persists even with the bit off bottom. A wave model of the drill string shows that tool joints form a torsional stop band that traps the dominant 140 Hz mode in the BHA; at the top drive its torque amplitude is about 25 times smaller than at the bottom of the drill pipe and below the noise of 1-s surface torque. When slow stick-slip is separated from HFTO, all detectors identify it well (AUC above 0.9 in the two runs where it could be defined), while none detects HFTO better than chance. Surface data are therefore adequate to monitor slow stick-slip in geothermal drilling, but HFTO requires downhole sensing. As an alarm, the physics-based detector raised 0.74 false alarms per drilling hour at FORGE while detecting 78% of stick-slip episodes, and simulation suggests that raising rotary speed while reducing weight on bit is the most effective response.

Topic: Drilling

Thermal short-circuiting is one of the key challenges that limit the long-term performance of Enhanced Geothermal Systems (EGS). When a dominant flow path develops by a highly conductive fracture, it reduces the reservoir volume available for heat exchange and causes premature thermal breakthrough. Although near-wellbore flow management has been used to avoid preferential flow path occurrence, controlling permeability away from the wellbore is still challenging, because injected materials need to remain mobile during transport and form a solid flow plug only after it reaches the target location. A laboratory investigation is underway to solve this problem by using microparticles encapsulating reactants within a thin polymer shell. This shell is designed to withstand the lower injection temperatures for a limited time before thermally degrading and releasing diverter-forming reactants at the higher reservoir temperatures, enabling delayed, targeted plug formation away from the wellbore. To deploy this technology in the subsurface at a reservoir scale, it is critical to understand and predict how these microparticles are transported by a viscous carrier fluid in a fracture system before forming diverter plugs and altering the fracture permeability. In this study, we used a Eulerian-Lagrangian approach to numerically simulate multiphase flow within a fracture and the formation of a gel-type diverter plug at a target location. The impact of the factors affecting plugging performance is examined, including carrier fluid properties (viscosity and density), particle properties (particle size, density, and concentration in the mixture), gelation time and stability, fracture geometry, flow rate and boundary condition. Each is varied to evaluate its effect on permeability reduction and plugging effectiveness within the fracture, to identify and optimize the injection conditions that improve the plugging performance. The results of the study will provide the basis for later upscaling to a reservoir model that can assess the field applicability of the technology and the long-term effect on short-circuiting and heat extraction in EGS reservoirs.

Topic: Enhanced Geothermal Systems

Closed-loop borefields are designed as if every bore is straight and lands where it was drawn. Deviation surveys the author has reviewed across New England show otherwise: in more than 95% of surveyed bores, drift averages 10–15° from vertical and reaches 35°. At 10–15°, the bottom of an 800 ft (244 m) bore can sit roughly 50–100 ft (15–30 m) from its design position, two to five times the 20–25 ft (6–7.6 m) spacing. The consequences are operational and legal as well as thermal. The author recorded collision rates of 2–6% on vertical projects, a rate expected to rise as designs reach 800–1,000 ft at unchanged spacing. In a 22-bore urban case study, every bore exceeded its 3%-of-depth verticality limit, 14 of 22 ended closer than the 20 ft minimum spacing, and bottom-holes on the drift side may lie beyond the property line. Characterizing this scatter statistically tells designers how much margin to carry; controlling it lets them carry less. This paper transfers oil-and-gas wellbore-positioning practice to closed-loop borefields. Each bore’s surveyed position is assigned an ellipsoid of uncertainty from an instrument error model. Anti-collision is managed with a separation factor between neighboring bores rather than nominal spacing. Property lines and utilities are treated as hard lines against the uncertainty envelope rather than the planned path. Each survey updates the as-built model, so the next bore’s trajectory and drilling parameters can be corrected. This closes the loop between measurement and control. On about 150 bores on the author’s projects, surveying and adjusting every bore kept position error at depth under 5% of depth, and under 2% on repeat bores once the plan and tools were tuned. Using g-function simulations of uncontrolled and controlled as-drilled geometry, we quantify how much spacing and bore length that accuracy returns for the same load, and we add the expected cost of collisions avoided. We close with acceptance criteria ready to write into specifications: position tolerance at depth at a stated confidence, minimum separation factor, boundary clearance and a standard survey report. With them, owners can require and verify that the field was built as designed before it is covered.

Topic: Low Temperature

The Upper Assam Basin, a Category-I petroliferous basin in Northeast India, contains substantial oil and gas resources and several depleted hydrocarbon wells with potential for geothermal utilisation. This study investigates the key drilling and wellbore stability challenges associated with retrofitting depleted wells within the Lakadong+Therria Formation for geothermal energy production. The investigated wells have recorded bottomhole temperatures ranging from 80 to 120°C at depths of approximately 3500–4500 m, indicating their potential for geothermal applications. A one-dimensional geomechanical model was developed to evaluate wellbore stability and assess the influence of well trajectory and temperature-induced stresses on drilling conditions. At a depth of 4051.5 m, the minimum mud weight required to prevent shear failure increases from approximately 10.6 to 14.1 ppg with increasing deviation at an azimuth of 10.72°, while the breakdown mud weight varies between approximately 16.8 and 18.5 ppg. At a fixed deviation of 26.3°, the shear-failure limit varies between 11.6 and 12.2 ppg with azimuth, whereas the breakdown limit ranges from 16.7 to 18.5 ppg, demonstrating a strong dependence of the safe mud-weight window on wellbore orientation. Comparison of thermal and non-thermal conditions further shows substantial changes in the near-wellbore tangential and axial stress distributions. The results demonstrate that well trajectory optimization and thermo-mechanical wellbore stability assessment are essential for defining safe drilling conditions and mitigating wellbore failure risks. The study provides a geomechanical framework for evaluating the feasibility of converting depleted hydrocarbon wells in the Upper Assam Basin into geothermal wells and supports the development of safer and more efficient strategies for subsurface heat recovery. The emphasis on trajectory-dependent mud-weight windows and thermal effects is consistent with established wellbore-stability research, where borehole orientation, in-situ stresses, mud pressure, and temperature are recognised as important controls on well integrity.

Topic: Drilling

Geothermal plants are increasingly digitized, with distributed sensors, SCADA/PLC control networks, remote operations, and third-party data interfaces. This improves efficiency and reservoir management, but it also widens the cyber-physical attack surface. A compromised operational technology (OT) system could alter injection setpoints or falsify sensor readings, with consequences ranging from equipment damage to unintended pressure and thermal transients in the reservoir, including induced seismicity. Current practice relies mainly on perimeter defenses and compliance frameworks. Few tools check whether the data reported to operators are consistent with subsurface physics. Our premise is that the reservoir cannot be spoofed. If an attacker changes what the reservoir experiences, the change should appear in observables the attacker does not control. We test this using THPoro, an in-house fully coupled thermo-hydro-mechanical simulator. We build a three-dimensional model of a representative horizontal-doublet EGS in a crystalline reservoir. The model is parameterized from published literature and digitized public field data, calibrated to nominal closed-loop operation, and used to define baseline operating envelopes. We then simulate attack scenarios in which operator-facing data appear nominal while the true state departs from it. The scenarios include spoofed wellhead pressure, a manipulated pump-rate setpoint, falsified flowmeter readings, and suppressed monitoring-well pressure. We compare the simulated reservoir and surface responses against the baseline to assess how well independent observables (monitoring-well pressure, fiber-optic strain, surface deformation from InSAR and tiltmeters, and forecast seismicity rates) can cross-check compromised downhole sensors. We present preliminary estimates of detection timescales and signature magnitudes for each scenario. The synthetic datasets also feed a lightweight physics-informed digital twin with AI-based anomaly detection (GeoSecure-RTS). We discuss implications for physics-informed monitoring of geothermal operations.

Topic: Modeling

Recent work has demonstrated that a unique, thermoresponsive hydrogel consisting of Poly(N-isopropylacrylamide) (pNIPA) can act as an “active” tracer that selectively clogs cold regions while leaving hot regions unaffected. Proof-of-concept has already been established at low pressures and temperatures up to roughly 60℃. Here, we study the volume phase transition behavior of these active tracers at pressures and temperatures consistent with commercial-scale geothermal reservoirs (e.g., 70 MPa and 150 ℃). The results demonstrate that the volume phase transition temperature favorably scales with pressure such that transition temperatures greater than 100℃ emerge at hydrostatic pressures as low as one MPa. The results suggest that the volume phase transition temperatures exhibited by our existing synthesis recipes are suitable for commercial-scale deployment.

Topic: Emerging Technology

While EGS systems currently use brine as the circulating fluid, supercritical CO2 (sCO2) introduces advantages including its minimal chemical reactivity (in the absence of any water) and favorable flow characteristics. However, its lower specific heat induces the need to circulate at a significantly higher rate. This paper offers a straightforward method to estimate the sCO2 rate that will deliver the same power generation as brine for the same downhole formation conditions. With the Fervo Energy Project Red as a base case, we estimated the sCO2 circulation rate required to deliver the same power generation for the same duration of time as the model simulation indicates for brine. Using density and specific heat data for brine and sCO2 at the conditions of the surface heat exchangers used to supply energy to the binary power plant results in a reasonable estimate for the sCO2 circulation rate that is easily refined as required in the simulation model. Having demonstrated the concept for the Fervo Energy Project Red case, we found the same approach works for other cases with different formation temperature, depth, and heat capacity. This simple approach enables a quick estimate for the sCO2 flow rate required for this alternative EGS approach.

Topic: Enhanced Geothermal Systems

A reason for reported lost EGS working fluid could be that a pump in the injection well does not supply sufficient energy to support production to the surface at a mass flux equal to the injection well mass flux. The volumetric rate of the circulating fluid flow increases as heat transfer from the formation lowers its density. While increased fluid temperature also reduces the fluid viscosity, the increase in volumetric rate may induce friction-related pressure losses both in propped fractures and in the horizontal and vertical segments of the production well. Overcoming the production well pressure losses may require a pump in the production well to ensure equal injection and production mass flux. While volumetric rate balance between injection and production is appropriate for usual miscible or immiscible processes common in oil and gas operations, EGS requires mass flux balanced fluid circulation. This study couples the simulation model for the downhole engineered heat changer with a process model simulator that addresses pump operation. Results suggest that a production well pump may be necessary to avoid lost fluid when brine is the circulating fluid, but the need for a pump in the production well is not expected when CO2 is the circulating fluid. Although current pump technology may not support pumping high temperature production fluid, this work underscores the potential need for a pump in the production well for EGS using brine as the circulating fluid.

Topic: Enhanced Geothermal Systems

[Eichelberger]

Augustine Volcano to Be Revealed by Geothermal Drilling John EICHELBERGER, Ally STEINLEITNER, Robert GILLIS, Marwan WARTES, Nina HARUN, Trystan HERRIOTT [Alaska Division of Geological and Geophysical Surveys, USA]

The first drilling into Augustine Volcano, planned for 2027 with funding to GeoAlaska by DOE, will reveal much about the structure and thermal budget of an Aleutian Range volcano. Augustine is one the most active subduction zone volcanoes in the U.S., perhaps in the top 10% worldwide. But despite intensive study and monitoring through three eruptions, its internal structure remains poorly known. Although located where it should be on the volcanic front, it stands alone in southern Cook Inlet while its equidistant neighbors are built high on batholithic basement. It lies at the western end of the transverse Augustine-Seldovia Arch, a structural hinge dividing the upper and lower Inlet basins. Outward dipping Jurassic sedimentary rocks are exposed on the south flank of Augustine (a topographically protected drilling site). Reconnaissance suggests this is the upper part of the Naknek Formation and that kilometers of forearc strata may be present there above the Talkeetna Formation volcanic basement. New magnetotelluric data show a large conductive body at shallow depth beneath this area, which may indicate that uplift and tilting of the Jurassic rocks resulted from magmatic intrusions or exposed Nanek may be a pre-existing structural high. New LiDAR gives the volume of the island, assumed to be mostly volcano, as 15.1 km3 (if Augustine’s base is at the top of the Naknek exposure, its volume is 1/10 that). The oldest eruption date of 15 ka then yields an effusion rate of 1 km3/ka, consistent with recent history. Counterintuitively, this is comparable to large caldera systems, despite Augustine’s smaller scale. In terms of time-averaged thermal output, the effusion rate is like a geothermal powerplant, but that neglects what remains unerupted. It is also comparable in volume to magma that erupted in 60 h at Katmai/Novarupta in 1912, with a thermal power of 100X US electrical consumption. It is hard to imagine that youth accounts for the small size of Augustine. Zircon antecrysts of 300 ka suggest an earlier beginning. Much volume could be hidden as intrusions beneath the volcano, perhaps as yet unerupted like at Katmai, and/or removal of the edifice by the massive Cook Inlet glacial ice stream. In any case, accessible geothermal energy is some combination of longevity and volume of the accumulating heat source, magma. A central andesitic magma chamber is convincingly postulated, sustained by frequent mafic replenishment, at ~4 km depth and equilibrated at 840oC. Scaling from Pinatubo Volcano would put Augustine’s 300oC isotherm at ~1 km depth, better than adequate for natural or EGS energy and a compelling choice for a superhot geothermal testbed. Existing cross sections provide conflicting subsurface geometries of Augustine that will soon be tested through drilling. This exploratory drilling effort will also illuminate the potentially enormous scale of geothermal resources in the Aleutian arc, advancing both volcanology and subduction zone geothermal energy understanding.

Topic: Field Studies

[Epping]

Permeable Fracture Imaging: Mapping Subsurface Permeability Architecture and Cross Well Flow J. EPPING, P. MALIN, A. MATHEWS, H. ONITVEROS, A. STROUJKOVA, S. VALENZUELA, T. FLEURE, W. MCLAIN, C. SICKING and F. HOROWITZ [Duke University, USA]

Successful geothermal resource development requires heat, fluid, and permeability to move the fluids. Permeability requires fractures, and uncertainty in their subsurface characterization is one of the primary risks in the sector. Associated problems with induced seismicity and water loss to the formation also are concerns for developers and investors, as these issues can occur when a well is unknowingly drilled into a fault or major natural fracture system. Data from the FORGE project suggests that advective flow, rather than conduction, dominates heat extraction by a wide margin. A Peclet number (PN), which is the ratio of advective to conductive heat flow, has been estimated at 7 at FORGE (Leary & Malin, 2025), indicating the system is dominated by advective flow. Assuming the ~100 m spacing between wells at FORGE with a steady 30 L/s stream of 185° C water, a direct fluid pathway between wells would have resulted in cooling over their 30-day flow test, dropping to less than 100° C in 100 days (Leary & Malin, 2025). This finding suggests that natural fracture advective fluid flow is critical for sustained heat extraction in enhanced geothermal systems (EGS) projects over the decades that a project is expected to last. Most modern seismic methods (e.g. microearthquakes) fall short of characterizing fracture flow systems prior to drilling. Instead, we have developed Permeable Fracture Imaging (PFITM), a passive seismic technique used to directly map the locations of fluid-filled fractures. PFI utilizes hundreds to thousands of seismic recorders on the surface to record passive seismic data, which is then processed to locate fractures and fluid activity. PFI has been used at Newberry Volcano in 2023 and 2025. In the 2025 survey, we mapped the fracture system surrounding an existing geothermal well doublet. Fracture orientations matched the tectonic trend (NW-SE) seen in geologic data. Changes in fracture density in 2023 data were also correlated with resistivity data and well logs, and results between the two surveys were reproduced. In 2025, we also mapped the location of near-wellbore fluids during a live injection test between the two wells at 3 km depth and achieved imagery down to 4 km. We show how fluid losses might be avoided and provide a preliminary technoeconomic analysis of how PFI-directed drilling can save the cost of 1 out of 3 geothermal wells.

Topic: Emerging Technology

Temperature gradient surveys are essential for modeling deep geothermal systems. In borehole temperature logging, resistance-sensitive temperature transducers (RTD or thermistor) are generally used. In such systems, use of 4-wire logging cables is required to eliminate resistance changes on the logging cable and obtain reliable ambient temperatures. This rises electrical integrity issues on the logging cable after certain times of usage in the field. We designed a portable temperature-depth tool that requires 2-wire logging cable only using AD590 current-sensitive temperature transducer. The sensor produces a current that is proportional to the ambient temperature independent of applied voltage, as a result the voltage drops due to cable length or voltage supply variability do not affect the measurements. Although the rated temperature sensitivity of the current sensor is lower than the resistance-sensitive sensors, use of 2-wire cable brings some advantages in terms of overall cable integrity including increased life time and easier replacement. Further advantages are smaller drum sizes and lower overall weight of the system. The surface read-out unit uses a low TCR resistor and a millivolt-reading precision voltmeter, and returns the real-time temperature-depth information. We also added an optional tilt switch to the downhole probe to obtain information when the probe gets stuck and tilted during logging. This helps preventing tangling issues on the cable during logging, especially for damaged boreholes. Overall, the portable logging system uses a simple analog circuitry which makes it easier to operate in the field and low-cost for temperature gradient surveys.

Topic: Geophysics

The current excitement around superhot geothermal ( greater than 375°C) is understandable. Higher temperatures increase enthalpy, improve conversion efficiency, and potentially enable much larger power outputs per well. But at what risk, cost and resource? From both an engineering and economic perspective, the more compelling opportunity may be widespread deployment of EGS in the 250-300°C range. The key insight is simple: The transition from today's EGS at the 200°C mark to reliable 250-300°C EGS is likely worth far more than the subsequent leap from 300°C EGS to 400°C superhot EGS. The reason is that technical feasibility, deployment speed, capital efficiency and accessible resource size are all dramatically better in the intermediate temperature range, compared to superhot temperatures. The risk vs. reward equation goes out of balance rapidly as we ponder temperatures above 300°C. In this talk, I will argue the case that industry and government need to put more emphasis on developing “incremental” technologies to deploy 250-300°C EGS at scale. There is no magic that happens at or around 375°C to dramatically improve brine efficiency, yet we need multiple “breakthrough” technologies (miracles) and very expensive materials to make it work, with relatively little incremental commercial resource base to unlock. The strongest case for geothermal decarbonization is likely not that the industry can someday reach 400-500°C resources. The strongest case is that reliable EGS in the 250-300°C range dramatically enlarges the economically accessible resource while remaining much closer to existing drilling, completion, reservoir, and power-generation technology.

Topic: Enhanced Geothermal Systems

Underground thermal energy storage (UTES) can provide valuable flexibility to integrated energy systems, yet its potential is difficult to represent in energy-system planning models. The nonlinear thermal and hydraulic behaviour of the subsurface is particularly challenging for mixed-integer linear programming (MILP) frameworks. GeoNinja addresses this gap by translating subsurface characteristics into simplified, MILP-compatible representations of Borehole Thermal Energy Storage (BTES) and Aquifer Thermal Energy Storage (ATES). The tool combines global geological and environmental datasets with the availability of energy technologies to assess UTES potential and derive system-relevant indicators such as storage capacity and energy density. A top-down approach provides consistent estimates across large regions, while a bottom-up approach enables refinement using site-specific data and subsurface simulations. This creates a pathway from global geothermal potential assessment to integration within energy-system models. A key element of GeoNinja's future development is its openness to contributions from the geothermal community. New measurements, site-specific studies, and improved subsurface data can be incorporated to progressively refine and expand the assessment. By combining global data, subsurface expertise, and energy-system modelling, GeoNinja aims to provide an evolving, community-driven platform for identifying and integrating UTES potential into future energy systems.

Topic: Modeling

The Upper Jurassic Smackover Formation in southwest Arkansas contains high-salinity brines with elevated temperatures and significant lithium concentrations, creating potential for combined lithium extraction and geothermal power co-generation. Using an extensive compilation of bottom-hole temperature (BHT) measurements from regional oil and gas wells, we estimate a lower-bound geothermal gradient of 36 C/km for Lafayette County. This gradient yields a Smackover reservoir temperature of approximately 116 C at a midpoint depth of 2725 m. With a total planned brine flow rate of 0.4166 m^3/s, deterministic flow-power calculations predict net 6.1 MW of geothermal co-generation, while Monte Carlo simulation (10K realizations) yields P5=5.94 MW, P50=8.01 MW, and P95=10.63 MW. Co-generation, using P50, could offset roughly 64.1 GWh/year (14.8%) of electrical demand for the planned lithium extraction facility. These results highlight the importance of robust gradient estimation using BHT-depth envelopes and demonstrate that geothermal co-production can provide a meaningful fraction of operational power for lithium brine facilities in southern Arkansas and adjacent areas.

Topic: Field Studies

Enhanced Geothermal Systems (EGS) require significant water for drilling, reservoir stimulation, and operation, a crucial concern where water is scarce. Drilling can use large volumes, but this water may be treated and reused in later wells. Reservoir stimulation, especially hydraulic fracturing, generally has the greatest demand. The induced fracture network must contact the largest practical volume of hot rock, which supplies heat over the project’s operating life. Fracturing fluid fills the reservoir and pressurizes the fractures, allowing wells to flow back during testing. During operation, losses to the rock into the existing fractures or due to continued fracture growth, must be made up from outside. These losses can be reduced by balancing injection and production pressures to maximize recovery of injected water while keeping injection pressure below the least principal stress, thereby avoiding additional fracturing. Long-term reservoir management therefore requires a clear understanding of the existing fracture network and the role of its fluids in recharging the engineered reservoir. This study proposes generating type curves for water use during stimulation and reservoir fill up and for consumptive water use during reservoir operation—water requiring external replacement per unit of geothermal energy recovered—as a practical metric linking reservoir performance to project development. Results are presented as dimensionless type curves relating reservoir pressure state due to injection and production and host-rock permeability to stimulation fluid volume, fluid loss, thermal performance and hydraulic performance of generated fractures.. These type curves can help to estimate the water needs for a project for both reservoir creation by fracturing and fluid loss to the rock during operations. The long term results at Soultz-sous-Forêts are reviewed with a focus on water consumption and field operating conditions. This example provides real world data for understanding EGS reservoir response and comparison to dimensionless type curves. A generic ResFrac EGS model was used to investigate the relationships between stimulated surface area to volume requirements, operating pressure, intrinsic host-rock permeability, and fluid exchange across the stimulated reservoir boundary. The numerical results are interpreted in the context of published circulation experiments at Soultz-sous-Forêts and other early HDR/EGS projects. These field observations provide an important physical precedent: an engineered reservoir need not behave as a closed underground heat exchanger, and the surrounding formation may act either as a sink for injected water or as a source of native reservoir fluid depending upon the imposed pressure field.

Topic: Enhanced Geothermal Systems

The Heber geothermal field in California’s Imperial Valley provides a well-documented case study for examining geothermal resource assessment and project-specific recoverability. This context is presented within the Geothermal Resources Management System (GRMS) recently sponsored by the Society of Petroleum Engineers. Initial exploration drilling, temperature measurements, pressure-transient testing, and numerical reservoir modeling during the 1970s and early 1980s established a substantial hydrothermal resource, with contemporaneous generating-capacity estimates ranging from approximately 150 to 500 MWe. By 1985, approximately $319 million had been committed to two commercial-scale developments employing dual-flash (47 MW) and supercritical binary conversion technology (45 MW). Upon startup, simultaneous field-scale production and injection revealed reservoir interference, aggregate deliverability constraints, and resource–conversion technology mismatches that were not adequately considered during appraisal. This study reconstructs the evolution of project uncertainty across three effective dates—pre-startup, startup (1985), and post-startup—using contemporaneous technical and operating evidence. An order-of-magnitude estimate of thermal energy initially in place (TEIIP), is compared with conceptual Low, Best, and High recoverable generating-capacity estimates and their implied 30-year thermal recovery factors. The analysis illustrates how estimates of recoverable quantities can contract as operating evidence replaces development expectations and conversion technologies used, although the underlying stored thermal energy remains substantially unchanged. The subsequent successful deployment of modular binary technology further demonstrates that commercial recoverability depends on the defined development project that is matched against reservoir performance, not simply the presence of a geothermal resource. Heber thus provides a historical example of the importance of effective-date reassessment, explicit treatment of reservoir and technology uncertainty, and distinguishing heat in place from technically and commercially recoverable geothermal energy under the GRMS.

Topic: Field Studies

Geothermal fluids are known to contain critical minerals at concentrations high enough to facilitate economical mineral recovery in certain locations. For example, the Salton Sea geothermal field is currently being developed for commercial scale lithium and manganese recovery. Despite previous work characterizing the mineral concentrations of hydrothermal fluids, gaps remain in our ability to identify and prioritize geographic areas for research and investment in geothermal-mineral co-extraction projects. To better understand which geothermal areas in the U.S. offer promise for mineral co-extraction, we analyze thousands of previously reported geothermal fluid chemistry measurements from DOE’s National Energy Water Treatment and Speciation (NEWTS) Version 2 data repository. We assess the concentrations and geographic distribution of over 20 elements in geothermal fluids and applied optimized Getis-Ord (Gi*) cluster analysis to identify statistically significant spatial clusters of high mineral concentrations. Our analysis confirms previously known critical mineral hot spots such as areas near the Great Salt Lake, Utah. We also identify previously underappreciated geothermal areas with elevated mineral concentrations. For selected areas, we assess the geological context that could be contributing to the mineral concentrations within the geothermal fluids. The project provides a quantitative basis for identifying and prioritizing investment opportunities in critical mineral co-extraction with geothermal energy resources.

Topic: Geochemistry

Enhanced geothermal systems (EGS) rely on engineered fracture networks between horizontal injection and production well pairs to transfer heat from hot rock to circulating water. Circulation rate per well pair affects both initial development costs and the subsequent costs of maintaining generating capacity. This paper quantifies the relationship between circulation rate and levelized cost of electricity (LCOE) using the Geothermal Insight Model (GIM), which couples hydraulic losses, reservoir heat transfer, organic Rankine cycle (ORC) turbine performance, and development costs to evaluate project economics. Maximizing rate per well pair does not necessarily minimize LCOE. An illustrative development with a reservoir temperature of 240 °C, a true vertical depth of 11,500 ft, and 7,500 ft laterals is modeled over a 30-year project life. Circulation rate is varied from 30,000 to 120,000 barrels per day per well pair. Each case is sized to the same net generating capacity (500 MW), with make-up wells added to maintain capacity as production temperatures decline. The selected circulation rate remains constant over each well pair’s operating life. The minimum modeled LCOE occurs at 60,000 barrels per day per well pair and is approximately 20% below the LCOE at 120,000 barrels per day. Contributions to LCOE are separated into initial costs (wells, ORC plant, and other facilities) and subsequent costs (water, make-up wells, ORC expansion, and operations). Above the optimum rate, lower initial costs are more than offset by higher subsequent costs. Pumping power increases nonlinearly with circulation rate, increasing parasitic losses. Earlier thermal breakthrough means that less heat is recovered before production temperatures decline, so a greater share of recovered heat is converted to electricity at lower ORC efficiency. Higher parasitic losses and reduced ORC efficiency result in lower cumulative net electricity generation per well pair than at the optimum rate. As the field’s average production temperature declines, maintaining net generating capacity requires greater total water throughput and additional ORC capacity to process that flow. Together, these effects increase required make-up wells and operating costs. Below the optimum rate, more initial wells are required to reach design capacity, increasing initial capital costs. Production temperatures remain above the 150 °C abandonment threshold at the end of the 30-year project life, leaving useful heat unrecovered. Savings in make-up wells and operating costs do not offset the higher initial investment and reduced heat recovery. Within the ranges modeled, the LCOE-minimizing circulation rate increases with more closely spaced fracture clusters, larger diameter casing, or longer laterals, and decreases with lower reservoir temperature or smaller injector–producer spacing. These results show that the optimum circulation rate depends on reservoir conditions, completion design, and well geometry. Circulation rate per well pair should therefore be optimized for the selected reservoir and development configuration using integrated techno-economic modeling.

Topic: Enhanced Geothermal Systems

Enhanced geothermal systems (EGS) rely on engineered fracture networks to extract heat from the subsurface at economically viable rates. Hydraulically conductive faults that connect injection and production wells pose an economic risk by diverting flow away from engineered fractures, accelerating thermal breakthrough, and increasing the cost of maintaining net power capacity. We quantify these impacts using the Geothermal Insight Model (GIM), which couples hydraulic losses, reservoir heat transfer, organic Rankine cycle (ORC) performance, and development costs to calculate project economics. GIM allocates flow between faults and the engineered fracture network based on their relative total transmissibility, calculated as conductivity divided by the respective fault and fracture fluid viscosities. It calculates power-plant inlet temperatures using heat-transfer type curves derived from simulations for engineered fractures and Gringarten’s analytical heat-transfer solutions for faults. For every modeled case, circulation rate is reoptimized to minimize levelized cost of electricity (LCOE), make-up wells are added as needed to maintain net power capacity, and water-handling costs reflect the circulation volumes required to sustain that capacity. We examine three drivers of the economic impact of conductive faults: the number of connecting faults per injection–production well pair, the conductivity of each fault relative to one engineered fracture, and the fraction of project well pairs affected. For example, under the modeled conditions, LCOE remains less than 5% above the no-fault case when every well pair encounters one connecting fault with less than ten times the conductivity of one engineered fracture. However, if every well pair encounters two connecting faults, each with 100 times the conductivity of one engineered fracture, modeled LCOE triples relative to the no-fault case, if the faults remain unmitigated. We also evaluate fault avoidance and mitigation. Avoidance from the outset is assumed to achieve no-fault economics; we separately quantify the LCOE effects of avoidance after the initial development wells have been drilled. Mitigation options include skipping fracture stages where wells intersect faults and isolating fault-connected intervals that contribute to early thermal breakthrough. These interventions sacrifice effective engineered fractures, a trade-off represented in the model by reducing their number. Implementation options are discussed qualitatively. The results quantify the potential for hydraulically conductive faults to be a major source of downside risk in EGS developments. Their prevalence, detectability, and potential for mitigation warrant particular attention during appraisal.

Topic: Enhanced Geothermal Systems

Developing Enhanced Geothermal Systems (EGS) at national scale requires coordinating project decisions with portfolio objectives under uncertain resource performance, costs, electricity demand, and regulatory pathways. Building on previous analyses of EGS as resilient baseload capacity under geopolitical and policy stress, this study applies an established probabilistic risk and decision framework to evaluate pathways toward a 5 GW portfolio across the United States. The analysis considers candidate projects across states with differing regulatory environments, linking development alternatives and financing requirements through shared uncertainties and capital constraints. Alternative objectives, including financial return, downside protection, deployment pace, and meeting growing electricity demand from artificial intelligence and data centers, are evaluated under regulatory, market, and geopolitical scenarios. Grid interconnection availability, costs, and timelines are incorporated as constraints on project selection and delivery. At the tactical level, the framework compares project scale, appraisal and development sequencing, contracting, offtake arrangements, interconnection strategies, and financing structures. At the strategic level, it evaluates geographic diversification, capital allocation, and deployment scheduling relative to demand growth and transmission capacity, including opportunities for hybridization and shared infrastructure. Dependencies among projects are retained to assess diversification benefits and common exposures. Strategies are compared using distributions of value, capital requirements, debt-service coverage, and delivered capacity, alongside the probability of achieving the deployment target. The study examines how demand scenarios and interconnection constraints change preferred investments and which risk-management interventions improve financeability. The contribution connects project decisions with adaptive capital planning to support EGS deployment where and when additional firm power is economically valuable.

Topic: General

We have developed a laboratory based model of flow through a porous fracture surrounded by impermeable rock, simulating the flow between the injection and production well in an enhanced geothermal system. We can vary the flow rate, fracture thickness, fracture geometry, and the permeability of proppant. We measure the time evolution of the temperature distribution within the fracture and neighbouring impermeable rock. Comparison of the experimental results with a numerical simulation are leading to insights about the controls of the fracture geometry and flow properties on the long term evolution of these systems, and their resulting efficiency in terms of power production.

Topic: Enhanced Geothermal Systems

Near-wellbore fracture tortuosity can increase the injection pressure required for hydraulic stimulation and the pumping requirements during fluid circulation in enhanced geothermal systems (EGS). The resulting fracture pathways depend on the completion geometry and the interaction between local stress concentrations and the far-field stress conditions. Understanding these interactions is important for designing completions that establish effective hydraulic connections between the wellbore and the stimulated reservoir. In this paper, we present high-resolution, three-dimensional numerical modeling and quantitative characterization of near-wellbore fracture tortuosity in hard-rock EGS. The models use a discrete element approach to simulate fully coupled hydro-mechanical fracture initiation and propagation. Perforation tunnels, casing, cement sheath, and surrounding rock are explicitly represented for cased-hole completions, while open-hole configurations incorporate the corresponding borehole and notch geometries. The simulations reproduce borehole excavation, casing and cement installation, and creation of perforations or notches to account for the stress changes associated with well construction. This enables high-fidelity simulation of fracture initiation, interaction, merging, and reorientation under the combined influence of completion geometry and wellbore–stress alignment. Fracture area, volume, aperture distribution, and their evolution are extracted from the simulations to develop quantitative measures of near-wellbore tortuosity. These measures enable systematic evaluation of how perforation arrangements, notching configurations, and wellbore orientations influence fracture development. By connecting completion design choices to the resulting tortuosity, this study provides an engineering basis for identifying configurations that limit near-wellbore hydraulic restrictions during stimulation and subsequent geothermal circulation.

Topic: Enhanced Geothermal Systems

Geothermal field-development evaluation requires repeated forecasts of reservoir pressure, temperature, and heat production across uncertain subsurface conditions and operating scenarios. High-fidelity numerical simulation can make this evaluation computationally demanding. This study presents a Control-Aware Physics-Informed Surrogate (CAPIS) framework integrating three-dimensional reservoir-state compression, control-conditioned latent dynamics, physics-based regularization, and output-specific ensemble prediction. A field-derived static reservoir model constructed in Seequent Leapfrog was implemented in CMG STARS to generate simulation datasets comprising 750 training cases, 250 validation cases, and 128 reserved test cases. A three-dimensional convolutional encoder–decoder compresses pressure and temperature fields on a retained 34 × 16 × 28 grid into 64- or 128-dimensional latent representations. The transition architecture includes a six-input pathway designed for injection-rate and producer bottom-hole-pressure controls, while an observation head predicts injector bottom-hole pressures, producer water rates, and thermal-energy rates. Training combines field reconstruction, latent-transition, and well-response losses with finite-volume mass- and energy-residual matching and Darcy-flux penalties. State-based re-encoding and direct latent propagation were evaluated over 28 recursive forecast steps without supplying subsequent simulator states. Physics-informed configurations provided the strongest pressure predictions, whereas the data-driven latent model achieved the lowest temperature error, motivating output-specific combinations of complementary predictions. Ensemble weights were fitted on the validation cases and frozen before testing. Across the 128 reserved cases, the weighted ensemble achieved pooled pressure and temperature root-mean-square errors (RMSEs) of 49.99 psi and 3.25 °F, respectively, reducing errors by 12.45% and 23.56% relative to the strongest individual predictor for each field. A dedicated useful-heat ensemble achieved an RMSE of 1.899 MWth for case-level mean useful thermal production, 48.2% below the best individual predictor. Complete 28-step ensemble forecasts required approximately 5.3 seconds on a four-thread CPU, while individual latent observation-only forecasts required 0.019–0.030 seconds. These results demonstrate rapid, field-scale reproduction of reservoir and well responses within the evaluated simulation family and highlight the value of output-specific model selection. The physics terms encourage agreement with approximate reference physics rather than exact conservation. Prospective application to geothermal scenario screening requires verified operating-input extraction, consistent time alignment, and high-fidelity confirmation of selected scenarios.

Topic: Reservoir Engineering

Flow channeling associated with changes in fracture aperture risks thermal short circuiting, with implications for power production over the lifetime of a geothermal plant. Previous work using subsurface numerical models suggests that inclusion of thermo-hydro-mechanical (THM) processes when simulating Enhanced Geothermal Systems (EGS) is crucial to capturing the risk of flow channeling for a range of fracture-network types, with results varying widely between different, stochastically generated fracture networks. Previous modeling work focused on individual fractures also suggests that geochemical processes like mineral precipitation on the fracture face has the potential to mitigate fracture expansion from thermal changes. However, mineral volume changes may also clog fractures, limiting network connectivity and worsening flow channeling, and these combined effects have yet to be examined in the context of larger, 3-dimensional (3D) fracture networks. Thermo-hydro-mechanical-chemical (THMC) inclusive simulations with complex 3D fracture networks are needed to improve understanding of these processes. We used PFLOTRAN’s fracture process module in combination with reactive transport to simulate a limited silica geochemical reaction network in a 3D EGS fracture network, and compared thermo-hydro (TH), thermo-hydro-mechanical (THM), thermo-hydro-chemical (THC), and thermo-hydro-mechanical-chemical (THMC) outcomes. Results address the effect of geochemical processes in combination with thermo-mechanical processes on fracture flow in an EGS context, with implications for costs of and risks to power production for EGS. Sandia National Laboratories is a multimission laboratory operated by National Technology and Engineering Solutions of Sandia LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy's National Nuclear Security Administration. SAND2026-27336A

Topic: Modeling

Drilling hard, abrasive, and confined formations subjects polycrystalline diamond compact (PDC) cutters to high loads, accelerated wear, and inefficient rock grinding. This study quantifies whether abrasive-water-jet (AWJ) preconditioning immediately ahead of a PDC cutter can reduce cutter loads and mechanical energy demand while increasing rock failure. The scope is limited to completed single-cutter simulations in granite and compares mechanical-only, AWJ-only, and simultaneous hybrid cutting under matched traverse velocities and confinement conditions. The work targets hard-rock drilling applications utilizing current bit architecture. A three-phase LS-DYNA smoothed-particle-hydrodynamics (SPH)/finite-element (FE) framework used a 100×50×20mm granite domain represented by a calibrated Riedel-Hiermaier-Thoma (RHT) material model. Phase 1 simulated mechanical cutting with a rigid single PDC cutter. Phase 2 represented water and 5 wt% iron abrasive using SPH. Phase 3 combined them, positioning a 9.525 mm jet 5 mm ahead of the cutter. Eighteen accepted simulations covered traverse velocities of 1,000, 5,000, and 10,000 mm/s under unconfined and 10 MPa biaxially confined conditions. Matched comparisons used stabilized cutting and axial forces, failed volume per unit travel, and cutter mechanical specific energy (MSE). Mechanical only failed volume remained nearly unchanged across the velocity matrix. In contrast, AWJ only failure was strongly exposure time dependent: increasing traverse velocity from 1,000 to 10,000 mm/s reduced failed volume by 99.76% in unconfined granite and 99.33% under confinement. In all six matched comparisons, hybrid cutting reduced stabilized cutting force by 3.38-18.79% and axial force magnitude by 1.09-17.64%, while increasing failed volume per unit travel by 3.84-65.31%. Cutter MSE decreased by 5.83-51.54% in every hybrid case. The greatest benefits occurred at 1,000 mm/s: failed volume per unit travel increased by 65.31% and 59.46%, and cutter MSE decreased by 51.54% and 49.08%, for unconfined and confined granite, respectively. Confinement had its clearest restrictive effect at this velocity, where AWJ exposure was longest. The consistent force reduction, larger failed zone, and lower cutter MSE demonstrate effective preconditioning, although the benefit diminished rapidly as traverse velocity increased. This study provides a traceable, staged SPH-FEM comparison that separates mechanical cutting and abrasive jet erosion before combining them under identical rock geometry, velocity, and stress conditions. Unlike assessments based on a single force or crater metric, the hybrid response is evaluated using two cutter force components, normalized failed volume, and mechanical energy per failed volume. The resulting dataset quantifies the exposure time dependence of AWJ assistance and demonstrates that a leading abrasive jet can reduce local PDC cutter demand in confined hard rock while enlarging the failed zone.

Topic: Drilling

As part of ongoing Utah FORGE research into novel multi-fracture stimulation concepts for EGS, abrasive jet perforation is being evaluated to create controlled perforation tunnels in hard-rock completions, with the objective of improving near-wellbore pressure losses and supporting closely spaced fracture development. However, published data on abrasive perforation in cased-and-cemented high-strength crystalline rock remain limited. A proof-of-concept yard test was conducted using a 24 × 24 × 48 in. granite block with an unconfined compressive strength of approximately 165 MPa (23,880 psi). An 8.5-in. bore was prepared through the block, and 7-in., 38-lb/ft T95 casing was cemented concentrically to represent the Utah FORGE EGS completion geometry. Testing was performed utilizing a 4-in. OD abrasive-jet perforating gun with nitrified abrasive slurry at approximately 2,600 psi nozzle differential pressure. The test matrix varied abrasive media, solid loading, and carrier-fluid gel concentration using sand and garnet systems. The results demonstrate the feasibility of abrasive jet perforation through casing, cement, and high-strength granite, while indicating that both penetration depth and tunnel geometry vary with abrasive selection, solid loading, and slurry formulation. The work provides a basis for further optimization of these parameters and for future testing under representative downhole pressure and temperature conditions prior to hydraulic-fracture initiation studies.

Topic: Enhanced Geothermal Systems

Superhot rock, reservoirs at temperatures below the critical point of water but with high fluid enthalpy, is emerging as a target for meeting growing energy demand, yet the thermo-hydro-mechanical (THM) response of these reservoirs during production remains poorly understood. In these systems, temperature evolution and phase behavior are governed by the interaction between the rock matrix and the fracture continuum. Because flow through the high-permeability fractures is rapid, advection removes heat quickly, and the thermal response depends on the balance between latent heat absorbed during flash vaporization and heat supplied from the matrix to the fracture. While such behavior has been studied in conventional vapor-dominated systems, we investigate these interactions under superhot conditions (340 °C) and additionally account for the coupled mechanical response. Using a dual-permeability THM model, we examine the reservoir under natural production and compare the effects of oriented natural fractures with those of hydraulically stimulated fractures. The results show that advection rapidly removes heat from the fracture continuum, while latent heat absorbed by flash vaporization accounts for only a small fraction of the energy balance; the remainder is supplied by continued heat flow from the matrix, which sustains the enthalpy of the produced fluid. In the stimulated case, the larger temperature decline causes the effective stress near the well to decrease at late times, even as pore pressure continues to fall, a signature of the transition from poroelastic to thermoelastic dominance in the near-well mechanical response.

Topic: Modeling

In the Eavor-Loop™ design, multilateral well trajectories are designed to maximize heat transfer surface area between the circulating fluid and the surrounding rock. During transient events such as initial circulation, mass flow rates in the laterals can evolve rapidly and deviate greatly from a uniform distribution as frictional and gravitational pressure differences change with fluid temperature and density. This study presents a coupled transient fluid-flow and heat transfer simulation of water circulation in an Eavor-Loop™. Pressure continuity is evaluated at lateral junctions, and the magnitude and direction of mass flow in each lateral are determined as the thermal and hydraulic conditions evolve. The impacts of multilateral trajectory design, operating conditions, and initial thermal state on flow distribution and thermal duty are analyzed from circulation startup to stable operation. Results are cross validated across multiple simulation tools to assess model consistency. Simulation results indicate significant flow imbalances can arise during transient operation due to the competing effects of friction and thermally induced gravitational pressure differences. As the fluid and surrounding rock temperature fields develop, the lateral flow distribution and the thermal output approach a stable condition. Operating and design measures are also evaluated for their ability to control or mitigate the initial flow instability.

Topic: Modeling

Following its successful launch at the 2025 Geothermal Rising Conference, the Geothermal Energy Atlas (GEA) has become an essential, publicly available resource for the geothermal community. This presentation will offer a one-year review of the GEA, a platform developed by the National Laboratory of the Rockies (NLR) and the Department of Energy’s Office of Geothermal (OG) to make complex geothermal geospatial data more discoverable and actionable. We will review platform usage metrics, user feedback, and the incremental data enhancements and refinements implemented since its debut. The GEA continues to address the needs of research scientists, geologists, and energy analysts by providing a centralized platform for evaluating geothermal opportunities. In this second year, our focus shifts to demonstrating the platform's value and charting its future. This session will highlight key capabilities and showcase how the GEA supports a wide spectrum of geothermal applications, from power to mineral extraction and energy storage. Looking ahead, we will outline the strategic roadmap for the GEA. This includes a discussion of potential integrations and new capabilities designed to further support informed decision-making across the geothermal community. We will explore the planned data set additions and modeling results that aim to accelerate innovation and deployment within the industry, ensuring the GEA evolves to meet the dynamic needs of its users. This forward-looking review will provide stakeholders with a comprehensive understanding of the GEA's progress and its planned trajectory for the coming year.

Topic: Modeling

The Utah FORGE (Frontier Observatory for Research in Geothermal Energy) is a field-scale laboratory with the goal to enhance our knowledge on how to utilize hot, dry rock for geothermal energy production, which may act as one of the base-load capable components contributing towards the clean energy transition. To transport heat from the underground to the surface, fluids circulate though the reservoir. Considering the low permeability of granite at the targeted depth of ~3km, it is necessary to create fluid pathways through hydraulic stimulation. To better understand where and when stimulation induces seismicity and what role natural seismicity plays in this geological setting, it is essential to have a comprehensive overview of the seismic activity and potential structures in the larger area. Utah FORGE provides a seismic dataset that includes recordings from a permanent network first emplaced in 2016 and continuously expanded since, multiple temporary nodal arrays, and downhole geophone chains. Using this extensive database, we utilize machine learning models to create a continuous enhanced event catalogue ( greater than 130,000 events). With multiple geothermal projects active in this area and the vast amount of seismicity produced by others, this comprehensive catalogue for the Milford valley can help to identify fractures that have been created or activated by the stimulation/circulation activities, or if there are pre-existing structures that are either acting as hydraulic pathways or barriers. In addition to the anthropogenic events, there is also a record of natural seismic swarms mostly east of the Utah FORGE associated with the Mineral Mountains. This comprehensive catalogue can be used in the future as a basis for imaging or more in-depth studies on the evolution of the seismicity in the Milford valley.

Topic: FORGE

Closed-loop geothermal systems offer a potential pathway to expand geothermal electricity generation beyond conventional hydrothermal resources, but their performance and cost competitiveness depend strongly on subsurface conditions and engineering design. Building on the nationwide modeling framework previously presented at the 2026 Geothermal Rising Conference, this study presents completed maps and comparative techno-economic results for the conterminous United States. The assessment evaluates three down-selected representative configurations under consistent modeling assumptions: a two-lateral U-loop, an inclined multilateral system, and an L-shaped coaxial system in crystalline rock. Spatially varying subsurface temperatures and thermal conductivities from the Stanford Thermal Earth Model are extracted by the Python wrapper and integrated with slender-body-theory heat-transfer modeling and techno-economic calculations in GEOPHIRES through an automated high-performance computing workflow. These nationwide simulation outputs are post-processed into maps of produced fluid temperature, average annual electricity generation over a 30-year operating lifetime, and levelized cost of electricity. The assessment compares resource uncertainty cases and examines sensitivity to target depth, circulation flow rate, and horizontal drilling cost. Land exclusion layers are applied to the mapped results to distinguish potentially developable locations. The maps show substantial geographic variation in modeled performance. Under the evaluated full-cost assumptions, inclined multilateral systems had the strongest cost performance, with the most favorable results at 6 km and some locations below $300/MWh threshold, reflecting its greater formation-contact area and heat-extraction capability. The evaluated coaxial and U-loop systems did not reach this threshold. Lower horizontal drilling costs improve cost competitiveness, while access to deeper, hotter resources introduces a tradeoff between increased electricity generation and drilling expenditure. These maps provide a consistent national-scale screening basis for comparing CLG configurations and identifying priorities for further evaluation. Results remain conditional on conduction-only heat transfer, prescribed geometries, and standardized economic assumptions, and should not be interpreted as site-optimized project estimates or economically recoverable capacity.

Topic: Modeling

As hydraulic fracturing sees wider use in enhanced geothermal systems (EGS), it’s worth asking whether proppant testing methods borrowed from oil and gas actually hold up under geothermal conditions. High temperatures, high effective stresses, harsh fluid chemistry, and long-term exposure to reservoir conditions all present challenges that standard lab procedures may not fully capture. In static testing, proppants exposed to elevated temperatures have consistently shown no measurable degradation in the properties we looked at, a result echoed in independent testing programs as well. That’s a reassuring finding, but it also raises a real question: does static testing capture the mechanisms that actually matter once a well is operating? To dig into this, we adapted existing API 19D equipment and procedures to introduce dynamic conditions that better reflect what happens in an EGS well. The approach stays grounded in established oil and gas testing practice but extends it to account for the fluid movement and exposure patterns that set geothermal operation apart from static lab conditions. This talk walks through what we observed, the reasoning behind the modified test method, and how we’re applying it to assess proppant durability under geothermal conditions. The broader point is simple: reliable proppant selection for EGS depends on testing that actually represents what the material experiences downhole.

Topic: Enhanced Geothermal Systems

Enhanced geothermal system (EGS) projects require relatively evenly distributed flow across the fluid pathways that connect injection and production wells to prevent short-circuiting and early thermal breakthrough. Characterizing flow distribution requires accurate knowledge of the hydraulic properties of individual fractures in the reservoir, but quantifying these properties at such resolution is difficult. Traditional techniques can resolve bulk properties of the reservoir, but not those of individual features. Low-frequency distributed acoustic sensing (LF-DAS), which measures dynamic strain rate at high spatial resolution along a fiber cemented behind casing, provides a unique way to record fracture-related deformation along the length of a well. This study presents a novel method to estimate the hydraulic diffusivity of individual faults or fractures using LF-DAS. The method is applied at Fervo’s Project Cape in Milford Valley, Utah. After drilling through a known fault in the subsurface, LF-DAS strain data in an offset monitoring well ~150 meters away showed extension and compression at the fault zone intersection in response to pressure changes in the drilling well. Building on research from both hydrology (periodic pumping tests) and seismology (attenuation estimation), and assuming a linear pressure-strain relationship, we use the diffusion equation to model the strain response to the drilling-well pressure signal and match it to the measured LF-DAS strain at the offset well, thereby estimating the hydraulic diffusivity of the fault. We also investigate and compare several different ways to estimate diffusivity using this approach, such as 1D hydraulic vs. 2D coupled hydro-mechanical numerical simulations, 1D analytical solutions, linear vs. radial flow assumptions, time-domain vs. frequency-domain data-fitting, and other data processing choices. We find that across all choices, the best-fit estimated diffusivity of the fault lies between 13 and 21 m2/s. The range of values is largely due to which flow regime is assumed (linear or radial). For a given choice of flow regime, uncertainty can be constrained using the data misfit. Estimates are less sensitive to time- or frequency-domain data fitting and the use of analytical solutions or numerical simulation. By taking advantage of the analytical solutions to the diffusion equation, future work can apply this technique to many individual fractures and perform real-time monitoring for changes in hydraulic diffusivity. Our study has implications for reservoir modeling, mitigating operational risk, and predicting reservoir performance.

Topic: Enhanced Geothermal Systems

Closed-loop geothermal systems extract heat through the wall of a sealed well. This enables them to operate without a permeable reservoir and without the injection or production of formation fluid. Any location with sufficiently hot rock becomes a candidate for implementation. However, there is a trade-off in terms of heat output. Heat mainly reaches the well by conduction, which limits how much a single well can deliver. Therefore, the cost of drilling weighs heavily against the energy the well can extract. One way to offset this cost is to reuse existing deep wells, such as abandoned oil and gas wells, although these are often low-enthalpy and too cool for direct use. An established closed-loop option is the deep borehole heat exchanger (DBHE), which circulates a fluid down and back up a single well. The gravity heat pipe is an alternative closed-loop option that works as a two-phase thermosyphon. The working fluid evaporates at the bottom, rises as vapor, condenses at the top, and returns by gravity. Consequently, no circulation pump is needed. Only a handful of kilometer-scale installations, known as super-long gravity heat pipes (SLGHPs), have been built, almost all in China. There are too few in operation, and their field data cover too short a period, to show how the technology performs over a service life and how it compares with other closed-loop systems. We present a model that couples the heat pipe with the surrounding rock, validated against field measurements and published simulations. The paper covers two heat pipe configurations. The conventional one is open to the rock along its full length. The second, developed by our group, is designed to deliver higher temperatures, particularly from low-enthalpy wells. Its packer limits heat exchange to the section beneath it, where the rock is hotter than the saturation temperature of the working fluid. We study both configurations as standalone wells and as the heat source of integrated systems. These systems supply heating, cooling, or both simultaneously, through heat pumps or through sorption chillers, with intermediate thermal storage for waste heat recovery. Our ongoing work evaluates the behavior and long-term performance of the standalone SLGHP and its integrated systems in single wells and multi-well fields across a range of site conditions, and compares them with DBHEs. The aim is to place the gravity heat pipe among established closed-loop options.

Topic: Modeling

As geothermal fields are exploited for power production, the wellhead pressure of some production wells falls below the plant's operating pressure. In addition, some newly drilled wells never reach that pressure. These wells may then be left unused and maintaining plant capacity typically requires costly make-up drilling. The GeoEjector project was launched in 2022, and the aim of the project has been to study if ejectors can be used to bring such low-pressure wells back into production by connecting them with high pressure wells through the ejectors. In an ejector, high-pressure primary flow entrains low-pressure secondary flow, and the combined flow is discharged at an intermediate pressure. This can be realized for geothermal wells, where the low-pressure flow (from secondary wells) is connected with high pressure flow (from primary wells) in ejectors and returning the mixed flow at intermediate pressure, which is high enough for the operational conditions of the plant. The purpose of connecting them is to bring wells with insufficient operational pressure back in operation. This concept has been developed in this study through rudimentary initial field setups, laboratory-scale experiments, pilot- and field-scale installations and operation, and analytical and computational fluid dynamics (CFD) modelling of the ejector flow and validation of the models. The results from this work are promising for geothermal operations. They have shown that through an ejector, the primary flow can entrain secondary flow and increase the overall flow and power potential of the combined flow, compared to the primary flow only. A full-scale ejector deployment at the Theistareykir power plant in Iceland has the potential to generate an additional 2.3 MWe of electrical power, elevating the total mixed-flow output to 16.8 MWe, which is a significant gain over the primary well's 14.5 MWe standalone output.

Topic: Production Engineering

[Gudmundsdottir]

Operational Strategies to Mitigate Cooling and Enhance Peak Load Capacity in the Reykir Geothermal Field Halldora GUDMUNDSDOTTIR, Thrainn FRIDRIKSSON, Bjarni Reyr KRISTJANSSON [Stanford University, USA]

The Reykir low temperature geothermal field has experienced localized cooling associated with long term production and wellbore crossflow. Wells that experience crossflow often cool while idle, reducing their effectiveness during periods of peak demand. To address these challenges, Reykjavík Energy has initiated field scale experiments focused on two operational strategies: (1) maintaining selected wells in continuous low flow production to prevent crossflow and preserve well temperatures, and (2) controlled injection of hot water into cooled wells during low demand periods to store thermal energy for future peak-load utilization. The study evaluates thermal and operational responses through well monitoring and historical performance analysis. These experiments aim to assess whether relatively simple operational changes can slow thermal decline, improve well performance, and increase available production capacity during winter demand peaks.

Topic: Low Temperature

Geothermal wells are increasingly being considered for CO2 service in two distinct applications: geological CO2 storage, including basalt-hosted systems such as CarbFix in Iceland, and supercritical CO2 (sCO2) circulation for Enhanced Geothermal Systems (EGS). While published research has focused extensively on subsurface processes—including CO2 transport, mineralization, storage capacity, and heat extraction—less attention has been given to the integrity of the wellbore itself. This creates a critical technical gap when existing geothermal wells are repurposed or new wells are designed for CO2 service. This paper presents a well-integrity screening framework for evaluating geothermal wells exposed to CO2 injection or circulation. Geothermal wells originally designed for hot-brine service can experience substantially different mechanical, thermal, and chemical loading when subjected to CO2. The framework considers rapid thermal transients and cyclic thermal loading, trapped annular pressure buildup (APB), wellhead movement, casing and cement response, and corrosion associated with CO2 and, where present, H2S exposure. It addresses dissolved-phase and dense-phase CO2 injection as well as sCO2 circulation, including load paths associated with Joule-Thomson cooling and phase behavior. The proposed methodology adapts established deepwater and HPHT well-integrity practices to geothermal applications. It defines critical load cases, required diagnostic inputs—including the original design basis, casing and cement condition, pressure and temperature history, and transient thermal profiles—and a structured risk-ranking protocol. Wells are classified as suitable as-is, suitable with remediation, or not suitable, with explicit confidence flags to account for incomplete or uncertain legacy data. Using a representative well built from published geothermal data, we demonstrate the framework's application, illustrating how the load-case checks and risk ranking sort wells into the three suitability categories. The resulting workflow provides asset managers and well construction teams with a practical screening tool to identify integrity vulnerabilities, quantify data uncertainty, and define potential remediation requirements before placing geothermal wells into CO2 service.

Topic: Injection

Climate change poses a major challenge to human civilization and ecosystems. The over reliance on fossil fuels for energy production has been a primary contributor to rising global temperatures, sea levels, and extreme weather events and many others. Geothermal energy, a clean, sustainable and abundant resource within the Williston basin plays a critical role in addressing this crisis. This study conducts reservoir characterization on the Deadwood Formation, a numerical analysis study on the Beaver Lodge field within the Williston Basin, a sensitivity analysis of key design and reservoir parameters and, finally, an evaluation of sensitivity parameters on energy generation. A numerical simulation study is conducted for three years and to account for individual key design and reservoir parameters, each parameter is analysed independently. A comparative sensitivity study is performed to highlight the independent design and reservoir parameters effects on thermal recovery. From the results, the insulator and casing number, play a key role in reducing the overall heat loss coefficient. Also, the production rate is the main factor that contributes the most to heat loss within the formation. A further comparative sensitivity study is performed to highlight the independent design and reservoir parameters effects on energy generation. From the results, production rate has the potential to generate the most electricity due to a high fluid flow rate and substantial amount of heat conserved from reservoir to surface. Production rate is the only parameter that has a major effect on the surface equipment for electricity generation, compared to other design and reservoir parameters for electricity generation. In conclusion, significant quantities of heat in the form of geothermal energy lie within the Williston basin, and the role of key design and reservoir parameters are highlighted in this study for thermal recovery and energy generation.

Topic: Modeling

Lumut Balai, one of Pertamina Geothermal Energy’s own operated areas, which in 2025 doubled the installed capacity from 55 MW to 110 MW. Lumut Balai Unit 1, with 55 MW installed capacity and commercial operation date (COD) in 2019, is operating with 9 production wells and 3 injection wells (in November 2024, the injection wells are being relocated as these wells are planned to be used as production wells for Lumut Balai Unit 2). Before the COD of Unit 2, Lumut Balai reservoir monitoring shows a good sign with average steam production decline less than 2% per year, which this number is lower than the typical decline rate of conventional two-phase geothermal fields, generally ranges from 3% to 5% per year. Lumut Balai Unit 2 steam is supplied from 4 production wells with 2 injection wells. As Lumut Balai doubled the installed capacity, it is greatly anticipated the increasing of the production decline. Surprisingly, one year after the COD of Lumut Balai Unit 2 in June 2025, reservoir monitoring shows the production decline remains at around 2% per year. The number shows Lumut Balai’s great reservoir potential as it aligns with PGE’s plan to increase the installed capacity into 4 x 55 MW.

Topic: Reservoir Engineering

A well reaching 400°C can produce up to ten times the electric power of a 200°C well. However, long-term performance remains uncertain because of emergent phenomena such as dissolution, precipitation, proppant performance and embedment, and transmissivity localization. In particular, fracture closure is partly governed by pressure-solution at proppant-rock contacts. Mineral solubility in supercritical water is strongly dependent on fluid density, which is highly sensitive to pressure in this regime. For example, quartz solubility increases from 263 ppm at 200°C to 757 ppm at 350°C, but decreases to ~120 ppm at 400°C and 25 MPa, as the density of supercritical water drops to approximately 0.17 g/cm³. If solubility controls the rate of proppant embedment, fractures in low-density superhot reservoirs may close more slowly than fractures in 200-250°C reservoirs. Reservoir pressure would then become an important design variable for fracture lifetime. We test proppant performance and embedment into granite under liquid-water and supercritical-water conditions. Steel and alumina spheres are pressed into granite to impose an equivalent boundary effective stress of sigma'=70 MPa for seven days. Under supercritical conditions, fluid pressure is varied from u=25-to-30 MPa at constant T=400°C, corresponding to water densities of ~0.17-to-0.40 g/cm³; these conditions produce up to a sixfold variation in quartz solubility. For reference, we run equivalent tests under liquid-water conditions (T=250°C and u= 4 MPa), as well as a dry test (T=400°C). All tests are conducted at the same equivalent effective stress (sigma'=70 MPa). Pre- and post-test characterization includes detailed SEM/EDS mapping, embedment-depth measurements, and assessment of proppant performance. Results confirm the sensitivity of proppant–matrix interactions to supercritical conditions, provide evidence of dissolution-precipitation near proppant-rock contacts, and demonstrate the excellent performance of alumina proppants under all tested conditions.

Topic: Enhanced Geothermal Systems

In situ combustion generates thermal energy directly within oil bearing reservoirs through oxidation of a fraction of the in place hydrocarbons. While this energy is conventionally used for enhanced oil recovery, the high temperatures generated within and behind the combustion front create an opportunity for simultaneous or subsequent geothermal heat extraction. This study experimentally investigates closed loop heat recovery from an in situ combustion heated porous medium using a dedicated heat extraction wellbore embedded within a laboratory combustion system. The geothermal working fluid remains hydraulically isolated from reservoir oil, brine, and combustion gases. The objectives are to quantify recoverable thermal power and energy, compare simultaneous and post combustion heat extraction, and determine how active heat removal influences combustion front propagation and stability. A laboratory scale combustion tube is packed with oil and brine saturated porous material. A separate metal tube extending through the heated porous medium serves as a closed loop heat extraction wellbore. Compressed air is injected to initiate and sustain combustion, while thermocouples distributed along the combustion tube monitor spatial temperature evolution and combustion front propagation. Water is circulated through the embedded wellbore at controlled rates, and inlet and outlet temperatures, water flow rate, and pressure differential are continuously measured. Recovered thermal power is determined from water mass flow and enthalpy gain and integrated over time to quantify cumulative recovered energy. Experiments compare simultaneous heat extraction during active combustion, delayed extraction following combustion, and baseline combustion without active heat removal. Water circulation rate and extraction start time are varied to determine their effects on thermal drawdown and combustion behavior. The experiments evaluate the tradeoff between maximizing geothermal heat extraction and maintaining a stable self sustained combustion front. Increasing circulation rate can increase heat removal but may decrease outlet temperature and locally reduce reservoir temperature. During simultaneous recovery, excessive heat extraction may also decrease combustion front temperature, alter propagation velocity, or reduce oxygen utilization. Comparison with baseline combustion therefore provides a direct measure of the thermal energy that can be extracted without compromising the primary enhanced oil recovery process. The study introduces a closed loop approach in which combustion generated reservoir heat can be recovered without direct circulation of the geothermal working fluid through the formation and provides an experimental basis for integrating enhanced oil recovery and geothermal heat production within the same subsurface system.

Topic: Emerging Technology

Centralized, high-fidelity subsurface data is essential to reducing exploration risk and accelerating natural resource development. To support these efforts, the Utah Geological Survey (UGS) upgraded its digital infrastructure by transforming fragmented geological and geophysical datasets into high-performance, interactive web portals funded in part by Utah’s Operation Gigawatt initiative and the Department of Energy. The focus of this paper is the UGS Geophysical & Geothermal Data Portal, a modern web application which aggregates critical geothermal data including well records, thermal springs, temperature logs, and heat flow measurements alongside comprehensive geophysical datasets such as gravity, magnetic, and electromagnetic surveys. By standardizing and integrating diverse geologic data,, the Geophysical & Geothermal Data Portal presents data in a machine-learning-ready format for rapid mapping and subsurface characterization. This web app is relevant to all geothermal systems, from direct-use and conventional hydrothermal to Enhanced Geothermal Systems (EGS). Together with complementary subsurface frameworks, such as the UGS Carbon Storage Portal, this ecosystem provides stakeholders with accessible subsurface data to identify high-probability resource zones. The volume and complexity of geophysical data available in the Geothermal web application warranted upgrades in how UGS serves data. The data populating the Geophysical & Geothermal Data Portal moves through three connected open-source systems: (1) an automated pipeline that checks and loads new data into a central database, (2) a data warehouse that publishes the data as ready-to-use files, and (3) a web map that displays the data layers. Every dataset is listed in one searchable catalog (STAC—SpatioTemporal Asset Catalog), making it accessible, viewable, and retrievable from one place in a variety of formats for download or via standard web service endpoints (OGC—Open Geospatial Consortium—API Features and ArcGIS-ready map tiles). In addition to the data warehouse, utilization of open-source file types (Parquet, GeoParquet, PMTiles), and cloud-optimized bucket storage enables fast-loading web applications, smaller download files, and lower storage costs. The Geophysical & Geothermal Data Portal showcases a curated view of over 75 years of data collection where each layer can be explored, queried, and downloaded on its own. Custom tools built for the portal let users search geophysical and geothermal records by keywords, read geophysical anomaly values anywhere on a statewide grid, view data in a tabular format, and follow links from each data acquisition site to complete survey files hosted on the UGS GeoDataArchive and, when applicable, other public data repositories. User-defined subsets of layers can be downloaded with graphic tools and the app URL preserves the exact map view and filters for sharing with colleagues, researchers, industry professionals, and members of the public. By delivering pre-competitive, public-facing data and tools, the UGS provides the technical foundation needed to streamline decision-making, foster research, and advance energy deployment across Utah.

Topic: General

Distributed acoustic sensing (DAS) is producing growing waveform archives across geothermal, earthquake, urban, marine, and infrastructure environments. DAS-focused models, including PhaseNet-DAS, DASFormer, and DAS-MAE, have demonstrated spatially aware, semi-supervised, and self-supervised learning. However, comparison remains difficult because studies use different datasets, preprocessing, label sources, splits, and metrics. It remains unclear which representations generalize across sites, acquisition systems, tasks, and label regimes. We present DASFM-Bench, a versioned, multi-domain for evaluating existing DAS models and guiding future foundation-model development. It provides standardized waveform and metadata contracts, reproducible preprocessing, task-aware manifests, source-balanced sampling, and dataset-size tiers. Group-aware splits and complete-source holdouts reduce leakage between correlated windows and support in-domain, cross-site, and cross-task evaluation. Because DAS collections range from unlabeled recordings and automatic labels to limited expert annotations, the records label provenance and evaluates label efficiency and weak-label utility. The will compare architectures with complementary inductive biases: vision-based patch models and masked autoencoders; graph neural networks representing fiber geometry and physics-derived channel relationships; and audio-inspired waveform and token models. Task-specific DAS and conventional seismic models provide supervised and transfer-learning baselines for phase picking, detection, classification, denoising, and interpolation. We demonstrate these capabilities through an initial geothermal phase-picking study using 264 expert-labeled events, divided into 185 training, 26 validation, and 53 held-out test events, together with 1,127 events containing automatic P picks. On the fixed test set, PhaseNet mean P/S F1 at ±20 ms increases from 0.884 without adaptation to 0.926 with 10 expert-labeled events, 0.944 with 50, and 0.947 with all 185. Adding the larger automatically labeled pool does not improve performance. This result shows why ing must account for label provenance and quality rather than treating all labeled samples as equivalent, while establishing a workflow for evaluating transferability across future DAS architectures, sites, and tasks.

Topic: Enhanced Geothermal Systems

Geothermal research and operations requires information from publications, data repositories, monitoring networks, and specialized scientific software. Reliable scientific agents must therefore do more than generate plausible answers: they must identify evidence and data, coordinate heterogeneous tools, retain intermediate products, and expose uncertainty and quality-control decisions. We present an optimized Geothermal Analytics and Intelligent Agent (GAIA) as the supervisory layer of a multi-agent scientific workflow system. GAIA interprets and grounds a user request, identifies available and missing data, constructs a bounded plan, and delegates domain-specific tasks. Execution is provided by GAIA Digital Twin (GAIA-DT), an in-development domain-subagent runtime and scientific execution substrate. GAIA and GAIA-DT communicate through the Model Context Protocol. GAIA sends structured tasks and artifact references to role-bounded subagents, monitors their status, and receives results containing scientific outputs, quality-control information, provenance, failures, and new artifacts. GAIA then coordinates tasks and reports the combined evidence to the user. An evaluation-driven redesign also introduces hybrid retrieval and reranking over a geothermal knowledge base, with live discovery across the Geothermal Data Repository (GDR), USGS ComCat catalogs, web resources, and OSDU-compatible sources. On 117 retrieval questions, first-result recall increased from 0.62 to 0.79 and nDCG@10 from 0.78 to 0.87 compared to the initial implementation. The context engine abstained on all evaluated off-topic questions, with a 2% false-abstention rate. We demonstrate the integrated system through a seismicity-update case at The Geysers. Starting from waveform recordings and station metadata, GAIA retrieves scientific context and delegates event characterization, historical-catalog retrieval, catalog updating, and forecasting to GAIA-DT subagents. Artifacts move between tasks by reference and remain linked to execution records and scientific caveats. The resulting update is provisional and subject to human review. This example demonstrates auditable multi-agent automation and showcasing the framework as a foundation toward a digital twin system.

Topic: Emerging Technology

Fluid injection can promote fracture slip by increasing fluid pressure, thereby reducing effective normal stress and frictional resistance. The resulting shearing response depends in part on how pressure develops and dissipates along the fracture, making hydraulic boundary conditions relevant to shear stimulation in enhanced geothermal systems (EGS). This study investigates these effects using a custom biaxial direct shear system and smooth and rough polymethyl methacrylate (PMMA) fracture interfaces under sealed and unsealed hydraulic boundary conditions. Transparent specimens enable high-speed imaging of dyed-fluid migration, while synchronized measurements of injection pressure, shear stress, and shear and normal displacements characterize the coupled hydraulic and mechanical response. Under the tested conditions, hydraulic boundary conditions strongly influence pressure evolution, slip magnitude, and displacement patterns at the monitored locations. Sealed configurations restrict fluid escape and exhibit larger slip, whereas unsealed configurations exhibit smaller slip and more asymmetric deformation. These observations suggest that differences in hydraulic confinement modify pressure redistribution and effective normal stress along the interface, influencing the mechanical response to injection. The findings highlight the importance of representing hydraulic boundary conditions when interpreting laboratory experiments and modeling injection-induced fracture reactivation. They provide experimental insights into hydraulic controls on fracture slip relevant to shear stimulation and fluid circulation in EGS, while informing the broader understanding of processes that may contribute to injection-induced seismicity.

Topic: Enhanced Geothermal Systems

Closed-loop geothermal systems circulate a working fluid through a closed subsurface flow path and recover heat transferred from the surrounding formation. Because these systems do not directly produce subsurface fluids, they have the potential to utilize geothermal resources without relying on reservoir permeability. In addition, the thermosiphon effect, driven by temperature and density differences in the circulating fluid, may enable natural circulation to continue after the circulation pump is stopped. In conventional simulations, the injection and production sides are treated as independent boundaries, and the feasibility of natural circulation is evaluated based on the potential difference between the two sides. However, in an actual closed-loop system, the injection and production sides are connected at the surface, and the circulation rate and flow direction in each branch are determined by the pressure and temperature conditions of the entire system. Therefore, conventional approaches cannot directly predict the circulation state after a fully closed loop is established. In this study, a numerical simulator was developed to model a fully closed-loop system, including the surface connection, and to directly calculate the circulation rate and flow direction in each branch induced by the thermosiphon effect. Furthermore, the influence of forced-circulation conditions prior to the transition to natural circulation on the subsequent steady-state flow was investigated. Two cases with initial flow rates of 100 m³/day and 10,000 m³/day were examined. After one day of forced circulation, the injection and production sides were connected to form a fully closed loop. In the 100 m³/day case, reverse flow developed in one branch during forced circulation and persisted after the transition to natural circulation, eventually reaching a steady-state circulation rate of approximately 8,000 m³/day. In contrast, no reverse flow occurred in the 10,000 m³/day case, and the system reached a steady-state circulation rate of approximately 10,800 m³/day. These results indicate that even for the same closed-loop system, different steady-state circulation patterns may develop depending on the flow conditions prior to the transition to natural circulation. Under the conditions investigated in this study, the steady state with reverse flow also exhibited a higher total heat extraction rate than the state without reverse flow. Future work will systematically vary the initial circulation rate to identify the conditions under which different steady-state flow patterns develop and to evaluate their effects on heat recovery performance.

Topic: Reservoir Engineering

Reliable assessment of geothermal potential remains challenging because subsurface thermal behavior reflects nonlinear interactions among geological architecture, structural permeability, geophysical signatures, and well conditions, while available observations are spatially heterogeneous and often unevenly sampled. This study develops a generalizable physics informed machine learning framework for predicting geothermal thermal potential, quantifying predictive uncertainty, interpreting the physical controls learned by data driven models, and translating complex model behavior into an analytical screening criterion. A large nationwide geothermal well database from the United States is used as a heterogeneous development and validation test bed, while the framework is formulated using physically defined predictors that can be reconstructed in other geothermal provinces. Exploratory data analysis, spatial statistics, feature engineering, nonlinear relevance analysis, redundancy screening, and spatially stable feature selection are first used to construct a physically consistent predictor space while preventing target leakage and geographic memorization. Strong tabular and neural baselines are compared with a physics informed cross attention architecture that represents geological, structural, geophysical, and well context as interacting information domains. Model selection and hyperparameter optimization are performed using nested spatial cross validation, with geographically separated holdouts used to assess transferability beyond randomly partitioned observations. Predictive reliability is characterized through heteroscedastic uncertainty modeling, deep ensembles, and conformal calibration to distinguish high predicted geothermal potential from high confidence geothermal potential. Model reasoning is investigated using SHAP attribution, SHAP interaction analysis, accumulated local effects, attention based interpretation, and physics consistency tests to identify both individual controls and coupled mechanisms governing predicted thermal conditions. Finally, symbolic regression is applied to stable physical predictors and out of fold model responses to derive a compact geothermal screening index and a locked decision threshold for distinguishing elevated thermal potential. The proposed framework therefore moves beyond black box geothermal prediction toward an uncertainty aware and physically interpretable scientific machine learning approach capable of producing continuous thermal estimates, prospect probabilities, confidence bounds, mechanistic insights, and a transferable analytical screening relationship for regional geothermal exploration.

Topic: Emerging Technology

While active seismic exploration methods are a staple of oil-and-gas resource estimation, they are rarely used in geothermal exploration in magmatic rift settings. In these settings the lithological units are composed of highly scattering material and interleaved in complex ways. This is in contrast to uniform layers observed in sedimentary settings, with sharp, reflective boundaries between them, making them ideally suited for controlled seismic reflection methods. Furthermore, in magmatic rift geothermal systems the resource does not necessarily follow lithological units, and geothermal fluid-rock interaction may have altered the rock, affecting both porosity and/or permeability. Although active seismic methods may not be considered suitable for geothermal settings, other seismological methods and sensing techniques may generate better results. In recent years, several projects (DEEPEN, Carbfix2, S4CE, COSEISMIQ, SUCCEED) have focused on applying various passive seismological methods to the Hengill volcano, and its associated geothermal and carbon mineralization fields. The Nesjavellir field, at the northern edge of Hengill volcano, is the location of IDDP-3, the proposed superhot well planned to be drilled winter 2026-2027. In this study, we will review the results from the seismological studies, compare them with other datasets from Hengill, and comment on the value of information provided by each method. The main focus is on understanding the potential superhot (above 380°C) resource which is the target for IDDP-3.

Topic: Field Studies

Geothermal exploration programs acquire data in stages to characterize undrilled opportunities and improve the probability of a successful exploration campaign. However, acquisition plans often follow play-oriented best practices rather than adaptive policies sensitive to information value, uncertain outcomes, and budget constraints. This study reframes acquisition choices as a decision-centered tradespace constructed as a graph of belief states. Graph nodes represent the play hypothesis, evidence already acquired, and remaining uncertainty in resource potential. Edges define acquisition choices, including geological mapping, geochemistry, gravity and magnetics, magnetotellurics (MT), seismic, temperature gradient drilling, and well testing. Each acquisition has a cost and delay, and its possible outcomes connect to successor nodes with updated belief states. Cumulative acquisition cost defines one tradespace axis such that a project budget sets the ceiling for feasible routes without changing the underlying graph. Terminal choices include drill (information supports action over further acquisition), exit (opportunity is no longer viable), or defer (pause until additional funding or evidence becomes available). The expected value of sample information links observations to prospect value distributions, and the framework identifies Pareto-efficient adaptive policies for a program instead of a single route to complete subsurface characterization. The work demonstrates profiles for six overlapping geothermal archetypes: conventional hydrothermal, enhanced geothermal systems (EGS), closed-loop, deep sedimentary, ultra hot/deep, and direct use. In addition, a synthetic comparison of blind hydrothermal and EGS prospects illustrates how alternative existing datasets and budget ceilings affect expected net decision value, acquisition cost, and downside exposure under fixed survey sequences and greedy VOI-per-dollar selection. The resulting tradespace identifies when the preferred next acquisition changes, and when further acquisition is less valuable than acting or deferring on a geothermal opportunity.

Topic: Modeling

[Holt]

Thermal Exclusion Zones: What Does a Gradient Well Actually Exclude? Richard HOLT [RESPEC, USA]

Temperature-gradient wells are widely used to guide drilling decisions in both conventional geothermal and enhanced geothermal system (EGS) exploration. Conventional projects typically seek convective hydrothermal systems, whereas EGS projects often seek conductive hot rock with limited natural fluid circulation. Despite these opposing objectives, both communities face the same question following an unexpected drilling result: How far must the next well be moved to encounter a different thermal regime? This study introduces the concept of a thermal exclusion zone, defined as the region surrounding a borehole within which the occurrence of a contrasting thermal regime is unlikely. An ensemble of anonymized three-dimensional geothermal natural-state models is analyzed to quantify relationships between observed borehole temperature signatures and proximity to conductive, convective, and outflow-dominated thermal structures. Temperature fields are normalized to allow comparison across diverse geothermal systems while preserving confidentiality. Millions of synthetic temperature profiles extracted from the model ensemble are classified according to gradient character, thermal curvature, and temperature-reversal behavior. Statistical relationships are then developed between observed thermal signatures and the distance to contrasting thermal regimes. Results demonstrate that borehole temperature profiles contain information not only about local thermal conditions but also about the spatial distribution of nearby hydrothermal features. The resulting framework provides a quantitative basis for estimating thermal exclusion zones and evaluating step-out distances following unsuccessful exploration drilling. The methodology applies equally to conventional geothermal projects seeking convective systems and EGS projects seeking conductive hot rock.

Topic: Modeling

Strabo Engineering is developing a heat mining system designed to operate in single boreholes optimally at superhot conditions. The Enhanced Geothermal System (EGS) paradigm of creating fracture networks between injection and production wells is currently working well for intermediate temperature conditions where industry growth is fastest. As reservoir targets progress towards higher temperatures and greater depths, the difficulty of reliably connecting multiple boreholes will likely increase. Single-borehole operation will become a lower risk option. Operating from a single-borehole requires either alternating injection and production or a closed loop. Coaxial heat exchangers, while elegant, are relatively inefficient and enable little reservoir control. Our approach incorporates down- and up-flow pipes within the borehole, connecting a high-surface area heat exchanger between the borehole fluid and the inner working fluid, emplaced above the bottom of the borehole. The design ensures that the working fluid never touches the rock, reaches a supercritical state in the heat exchanger, then flashes to steam near or above ground, and flows straight into the turbine with minimal energy loss. Producing pure supercritical water or steam is a primary energetic advantage of this system. Another is that because primarily heat, not fluid, is being extracted from the reservoir, fluid re-injection is not necessary. We will present thermodynamic models of the deep heat exchanger, coupled to simple reservoir heat extraction models. The system can be optimized for heat extraction from single boreholes for long duration, high output thermal power, and multiple boreholes can be networked at the surface for large scale power generation. As heat delivery to the heat exchanger will limit production, the thermodynamic model includes convective/advective fluid flow through the surrounding rock to the borehole. Performance will be assessed for systems ranging from rock thermal conduction-limited- to high permeability cases, e.g. tapping into natural hydrothermal convection systems, such as those in Iceland and Japan.

Topic: Emerging Technology

[Horne]

Introduction to the 52nd Stanford Geothermal Workshop Roland HORNE [Stanford University, USA]

An introduction to the 52nd Stanford Geothermal Workshop.

Topic: Introduction

[Horne3]

Implications of Wellbore Heat Transfer for Conductive Closed-Loop Geothermal Systems Roland HORNE [Stanford University, USA]

Closed-loop, or “advanced”, geothermal systems circulate a working fluid through a sealed conduit in hot rock, taking up heat by conduction alone and never producing formation fluid. The concept appears to be attractive because it removes the requirement for permeability, and with it the dry-hole risk, the scaling problems of produced fluid, and the pressure disturbance that accompanies injection. This paper examines what such a conduction-only system can deliver, using the standard single-phase wellbore heat transmission equations. In a coaxial well the two legs form a countercurrent heat exchanger, and the governing equations become a two-point boundary value problem whose linearized form has one growing eigenmode. A closed-form solution is given for uniform properties, together with an affine backward sweep that accommodates arbitrary variation of the conductances along the flow path and avoids the ill-conditioning that defeats a shooting method. With a perfectly insulated return leg the delivered heat is exactly the heat gain that an injection well experiences on its way down, and its ceiling is the formation conductance multiplied by the temperature difference available along the path, independent of circulation rate. This defines the “conduction ceiling”, which is the maximum heat output attainable, and which is easily calculable in advance. The conduction ceiling is almost exclusively a function of the formation properties, and the design of the completion only determines how much of the ceiling can be recovered. The consequences are restrictive. A 3000 m vertical well continued 3000 m as a lateral in rock at 150ºC delivers 765 kW of heat at 131ºC with perfect insulation and 568 kW at 108ºC with good vacuum-insulated tubing, against a conduction ceiling of 1660 kW. Thermal energy in the rock more than about 20 m from the hole is barely disturbed after ten years, and the system recovers under three percent of the heat lying within 50 m of the wellbore over its first decade. The ratio of leg-to-leg to leg-to-formation conductance determines how much of the ceiling survives. The U-loop configuration evades the short circuit by separating the legs with rock, at the cost of a second well and of an ordinary production heat loss on the return. Closed-loop designs suit direct use, heat storage and shallow borehole heat exchangers with heat pumps; for power generation they would require tens of kilometers of hole.

Topic: Reservoir Engineering

Supercritical geothermal energy has long been regarded as one of the most promising pathways to transformational renewable power generation; however, despite more than four decades of international drilling efforts, sustained commercial production from confirmed supercritical reservoirs has not yet been achieved. This paper presents the first comprehensive engineering-framework post-mortem analysis of supercritical geothermal wells (T greater than 374°C and/or P greater than 221 bar) drilled worldwide between 1981 and 2026. The dataset comprises 20 entries across six countries-Iceland, Japan, Italy, the United States, Mexico, and Kenya-evaluated using a consistent seven-category failure mode taxonomy (FM-1 to FM-7), including one aggregated entry representing five Los Humeros wellbores with an identical failure profile. For each well, failure mechanisms are systematically examined through quantified operational parameters, root-cause determination, and cross-well pattern analysis to identify recurring engineering limitations. Moving beyond retrospective assessment, the paper proposes four original candidate engineering frameworks intended for independent review and validation by the wider geothermal community: (1) a Deterministic Failure Chain (DFC) describing progressive failure evolution across all basin types; (2) an Axial Compliance Framework (ACF) integrating five coupled thermomechanical mechanisms with the Norton Power Law and Larson-Miller creep rupture criterion to quantify structural survivability; (3) a dual-constraint Thermo-Mechanical and Chemical Survivability Envelope (DCSE) defining simultaneous operating limits for long-term well integrity; and (4) a Basin Risk Index (BRI) providing a structured predictive methodology for assessing technical risk during future supercritical geothermal well planning.

Topic: General

Storing heat underground at scale, in aquifers and depleted reservoirs, converts intermittent thermal supply into dispatchable heating, and the high-temperature variants now being piloted push into geochemical and design territory the low-temperature ATES record does not cover. We review high-temperature underground thermal storage: round-trip efficiency evidence and its controls, the geochemical management burden heat imposes, well and pattern design, and the system economics against alternatives. Dispatchable heat from intermittent supply is the energy transition's least glamorous requirement, and the subsurface answer has an operating record worth consolidating.The first temperature-resolved synthesis of underground thermal storage with the early-cycle conditioning effect and geochemical management burden given design-stage status. Heat-network planners and geothermal operators gain screening and design guidance for the storage layer that makes intermittent thermal supply dispatchable. Heat-network planners gain screening and design guidance with the geochemical management burden priced in from the start rather than discovered in operations. The pilot and operating record was assembled across temperature classes with recovery efficiencies normalized, efficiency-control evidence reviewed, buoyancy-driven tilting, conduction losses, ambient flow, geochemical experience compiled, scaling and clogging dominate the operational record, and design responses examined, with machine-learning operational optimization surveyed. Recovery efficiencies are normalized across temperature classes and cycle counts before comparison, because early-cycle numbers systematically understate mature performance. Recovery efficiency climbed over the first cycles everywhere as the storage volume thermally conditioned, making early-cycle performance a poor basis for project judgment, and buoyancy tilting set the efficiency ceiling in permeable aquifers. Geochemistry, carbonate scaling on heating, iron clogging on mixing, caused most operational interventions and was manageable where water treatment was designed in rather than retrofitted. The economics favored large, seasonal, high-utilization applications with district-scale heat sinks. The conditioning finding protects good projects from premature judgment: efficiency climbs for cycles as the ground warms to its job.

Topic: General

Geothermal prediction problems, stimulation, seismicity, long-term circulation, thermal drawdown, are coupled thermo-hydro-mechanical-chemical problems, and a growing family of THMC codes answers them with confidence their validation record has not caught up to. We review coupled THMC modeling for geothermal: the and code-comparison record, which couplings each application genuinely requires, and the predictive-confidence evidence from history-matched field cases. Coupled codes are answering billion-dollar geothermal questions with confidence their validation record has not yet purchased, and the gap deserves mapping.The first application-mapped assessment of THMC modeling that tells practitioners which couplings their question requires and what confidence the validation record supports. Geothermal modelers gain cost-rational coupling guidance; reviewers gain the skepticism calibration the field's confident predictions have needed. Code-comparison exercises were assembled with their agreement statistics by coupling class, problems catalogued against the applications they actually constrain, field validation cases reviewed where THMC predictions met subsequent measurements, and machine-learning surrogates for coupled simulation surveyed with fidelity evidence. Agreement statistics are extracted by coupling class rather than headline problem, which localizes the divergence to the physics responsible. Codes agreed well on TH coupling, diverged on mechanical response wherever fracture representation differed, and diverged most on chemistry, where kinetic parameterization dominated everything. Full four-way coupling was rarely necessary: application-specific coupling subsets reproduced full-physics results at large cost savings, and the review maps which subsets suffice where. Field-validated predictive skill existed for thermal breakthrough and remained thin for stimulation outcomes. The subset finding is the budget finding: most applications need two or three couplings, not four, and the map of which suffices where is provided.

Topic: Modeling

Effective cement placement is critical to the long-term integrity of geothermal wells, where incomplete displacement of drilling fluids can lead to poor zonal isolation, weak cement–formation bonding, and potential pathways for fluid migration. This study develops a computational fluid dynamics framework to investigate the displacement of drilling fluid by cement slurry in the annular space between a well casing and the surrounding formation. The volume-of-fluid method implemented in ANSYS Fluent is used to track the interfaces among drilling fluid, spacer fluid, and cement slurry, while the non-Newtonian behavior of the cement slurry is represented using the Herschel–Bulkley rheological model. Three-dimensional wellbore models with both smooth and irregular formation surfaces are examined to evaluate the influence of annular geometry and cement injection velocity on displacement behavior. The simulations reveal that wellbore geometry strongly affects flow patterns, recirculation, and cement placement efficiency. For the smooth-wall geometry, pronounced recirculation and stagnation regions develop near the bottom of the annulus, resulting in less efficient drilling-fluid displacement. In contrast, the irregular-wall configuration with a larger annular cross-section reduces the recirculation region and produces more uniform cement placement, leading to higher replacement efficiency. The influence of cement injection velocity is further evaluated over inlet velocities ranging from 0.05 to 0.5 m/s. Among the conditions investigated, an inlet velocity of 0.5 m/s provides the highest filling rate and most rapid displacement of drilling fluid, whereas an intermediate velocity of 0.2 m/s is more susceptible to recirculation and two-phase flow instabilities. Overall, the results demonstrate that both wellbore geometry and slurry injection conditions play important roles in controlling cement placement and residual drilling-fluid distribution. The proposed multiphase CFD framework provides a physics-based approach for evaluating well cementing operations and can support future optimization of cement placement strategies for geothermal wells with complex borehole geometries and operating conditions.

Topic: Modeling

The mid-to-deep single-well closed-loop geothermal system (CLGS) is considered an important pathway for efficient geothermal energy exploitation. However, under long-distance circulation conditions, significant hydraulic pressure losses in the wellbore result in excessive pumping power consumption, which limits the overall efficiency of the system. Existing studies mainly focus on wellbore structure and heat transfer parameter optimization, while the coupling between working-fluid rheology and deep-well hydraulic losses has not been systematically quantified. This study proposes a non-Newtonian fluid-based drag reduction strategy for geothermal wellbores. The objective is to investigate whether polymer additives can effectively reduce circulation resistance and identify optimal operating conditions for non-Newtonian fluids through multi-parameter optimization, thereby reducing pumping energy consumption. An integrated framework combining experimental characterization, numerical simulation, and optimization is developed. Polymer materials including Hydrolyzed Polyacrylamide (HPAM), Xanthan Gum (XG), Polyethylene Oxide (PEO), and Welan Gum (WG) are used to modify conventional water-based working fluids. Rheological properties are characterized through power-law parameters measured under different concentrations and temperatures. A thermo-hydraulic coupled model of a single-well CLGS is established to simulate heat transfer and pressure loss under deep geothermal conditions. Furthermore, a genetic algorithm (GA) is applied to optimize operating parameters, including circulation flow rate, well depth, and geothermal gradient. Compared with conventional water, non-Newtonian polymer fluids at specific concentrations demonstrate significant drag reduction performance in deep well circulation, with wellbore pressure losses reduced by 63%–75%. The GA optimization identifies optimal operating parameter combinations under different geothermal gradients, further improving the drag reduction performance of various non-Newtonian fluids. The study provides a new fluid regulation strategy for CLGS design and offers theoretical insights and engineering guidance for achieving low-pumping-power and high-efficiency development of mid-to-deep geothermal resources.

Topic: Modeling

Stimulated geologic hydrogen (SGH) aims to generate hydrogen in the subsurface by enhancing water–rock reactions, including serpentinization in ultramafic rocks. SGH shares major engineering challenges with enhanced geothermal systems (EGS): identifying suitable reservoirs, drilling, creating and managing flow pathways, maintaining fluid circulation, and mitigating induced seismicity. The main differences involve the water–rock reactions and the types of rocks, which may fracture, react, and transport fluids differently from rocks in geothermal reservoirs. This work examines how advances in EGS can inform SGH development and where these differences require new approaches. We assess EGS experience in reservoir characterization, drilling, hydraulic fracturing and hydroshearing, zonal isolation, injection and production strategies, and seismic monitoring and mitigation under conditions relevant to SGH. The assessment is based on published research, including papers from previous Stanford Geothermal Workshops, and geothermal R&D at Berkeley Lab spanning fundamental studies to field demonstrations. We then identify research questions and technology needs specific to SGH, including the coupling of fluid transport, mineral reactions, and hydrogen production in fractured ultramafic reservoirs. Building on EGS successes, we define a pathway to accelerate SGH development through lower costs, safer operations, and more efficient production.

Topic: General

In low porosity geothermal systems, fracture properties, including contacts, aperture, and roughness, and the connections between fractures control network permeability and heat transfer. However, these fracture properties can evolve during stress changes from reservoir operations. To quantify this dynamic behavior, we perform in-situ synchrotron microtomography experiments on a fractured rock at reservoir conditions. We create a fracture network by deforming a horizontal, mismatched brittle fracture at 400°C. As the main fracture closes, secondary fractures develop from contacts and form an increasingly connected network. We acquire time-lapse microtomography scans, segment each network, and use the lattice Boltzmann method (LBM) to simulate fluid flow and heat transfer. We compare two representations: the main fracture alone and the complete fracture network. For each, we examine two flow directions, one aligned with the dominant secondary fracture strike and one approximately perpendicular to it. The results show that main fracture closure expands contacts and reduces aperture. The resulting lower flow rate gives fluid more time to exchange heat with the rock, producing a higher outlet temperature. Secondary fractures improve thermal efficiency by providing additional flow channels and exposing flowing fluid to more hot-rock surface. The enhancement is controlled by network layout. When flow aligns with the dominant secondary fracture strike, the branches provide additional inlet-to-outlet pathways. When flow is approximately perpendicular, secondary fractures connect main fracture pathways separated by contacts and redirect fluid along longer routes for heat exchange. Despite increases in total porosity and outlet temperature, thermal power tends to decrease because the reduction in flow rate outweighs the greater heat gain per unit of fluid. Thus, a higher outlet temperature does not necessarily indicate greater heat extraction.

Topic: General

Energy systems are becoming increasingly complex as multiple generation, conversion, and storage technologies must be coordinated, typically to minimize total cost while satisfying technical and operational constraints. Underground thermal energy storage (UTES) can provide significant economic and operational benefits in such systems, but its cost-effectiveness depends strongly on system design, geological properties, operating conditions, and interactions with other technologies. Suitable, easy-to-handle integration methods are therefore needed to identify promising applications and to determine the role of UTES within the overall energy system. In this work, we develop methods for the optimal system integration of UTES and apply them to two real-world installations: a high-temperature borehole thermal energy storage (HT-BTES) system on the Empa campus in Dübendorf, Switzerland, and a 12-well aquifer thermal energy storage (ATES) system at Bispebjerg Hospital in Copenhagen, Denmark. High-fidelity storage models, operational data, and lessons learned from both installations inform simplified design models that abstract the physical storage behavior to be compatible with linear energy system models. We present the resulting modeling and design approach and demonstrate the integration of the HT-BTES and ATES models into mixed-integer linear programming-based energy system planning tools. For both sites, we compare cost-optimal system configurations with UTES to alternative technology choices. This identifies cost-effective system configurations and suitable applications for UTES.

Topic: Modeling

Regional groundwater flow changes both the apparent thermal conductivity measured around a geothermal borehole and the operating life of the surrounding borefield for ground source heat pump applications. Conventional thermal-response-test interpretations assume conduction-dominated heat transfer. When a borehole intersects an active aquifer, groundwater advects heat away from the borehole and produces a directional, depth-dependent temperature response. A conduction-only interpretation attributes this additional heat transfer to an elevated thermal conductivity, although the intrinsic conductivity of the formation has not changed. This study develops a three-dimensional coupled groundwater-flow and heat-transport model for geothermal boreholes partially penetrating an active aquifer. The simulations quantify the effects of groundwater velocity, aquifer thickness, penetration ratio, formation thermal properties, borehole spacing, and operating time. Synthetic temperature responses are interpreted using a conventional conduction-based model to determine how groundwater advection affects the inferred apparent thermal conductivity. The results are organized using the thermal Péclet number, dimensionless aquifer thickness and penetration ratios, conductivity and diffusivity ratios, and the borehole Biot number. The Nusselt number measures the resulting heat-transfer enhancement. These dimensionless groups define conduction-dominated, transitional, and advection-dominated regimes and allow the results to be transferred among different aquifer and borehole configurations. Groundwater flow increases apparent thermal conductivity, displaces the thermal plume downstream, and directly controls borefield service life. It supplies thermally undisturbed water to upstream boreholes and can delay temperature decline, while transporting cooled or heated water toward downstream boreholes and accelerating thermal interference. Consequently, the commonly adopted 20-year design horizon is not a universal limit. Depending on groundwater velocity, flow direction, and borefield configuration, it can either underestimate or overestimate the useful operating life. The resulting dimensionless response maps support correction of thermal-response-test interpretations and performance-based prediction of borefield service life. They also provide a practical basis for selecting borehole depth, spacing, orientation, and sustainable thermal load in direct-use geothermal, ground-source heat-pump, and borehole thermal-energy-storage systems.

Topic: Direct Use

Optimizing stimulated fracture spacing design is essential for improving energy extraction in enhanced geothermal systems. However, resolving it requires a forward model fast enough to sit inside an optimization loop, while still capturing the physics that governs long-term heat extraction. Current physics-based models are computationally expensive. We propose a workflow based on dfnWorks (https://dfnworks.lanl.gov) that builds a stochastic discrete fracture network (DFN) calibrated from field data and embeds a set of stimulated fractures between an injection and production well at a candidate spacing. Graph-based flow and particle tracking in dfnWorks then generates a breakthrough curve at a fraction of the cost of a fully meshed solve, grounding the optimization in a fracture network that reflects real site conditions rather than a purely synthetic one. That breakthrough curve is converted to a heat extraction estimate using a time-domain random-walk formulation with a finite matrix-block (slab) return-time distribution, rather than the semi-infinite conduction assumption used in standard tracer-based thermal breakthrough methods. The predicted spacing-dependent heat extraction from this fast surrogate workflow is validated against high-fidelity simulations using a newly developed handshake between dfnWorks and MOOSE, allowing direct comparison to fully resolved coupled thermal-hydraulic-mechanical simulation. We demonstrate the workflow using the Utah FORGE background fracture population, calibrated to field data from the FORGE 16A(78)-32 well and the associated 2025 v1 reference DFN parameterization. Anchoring both the fast surrogate and the high-fidelity validation runs to the same real-site DFN parameterization keeps the comparison meaningful for an actual engineered reservoir rather than an idealized one and creates a natural path to further validation against FORGE's own production and thermal monitoring data as it becomes available. This validation step is designed to test whether the fast surrogate preserves the correct relative ranking of candidate spacings, the property an optimizer actually depends on, rather than absolute accuracy, which the reduced-order workflow is not expected to match. Establishing that correspondence would support a practical, field-data-informed design workflow for stimulated fracture spacing in fractured-rock geothermal reservoirs, with the dfnWorks-MOOSE handshake itself standing as a reusable capability for validating other reduced-order or surrogate models against full multiphysics simulation.

Topic: Enhanced Geothermal Systems

Several approaches implementing supervised machine learning have been proposed for mapping hydrothermal favorability in the Great Basin over the last two decades. Models are typically pointwise (each cell predicted from its own feature values alone) and evaluated with random train-test splits that can place training and testing wells spatially very near to one another. Because both labels and evidence layers are spatially autocorrelated, random pointwise splits can reward memorizing local conditions over learning transferable relationships, inflating reported predictive skill. We quantify this inflated predictive skill using three architecturally distinct models trained on the same 53 evidence layers and 100 Monte Carlo realizations: a pointwise linear regression, XGBoost, and an image-based convolutional neural network (U-Net) that predicts favorability from 12.8 km patches, adding spatial context. Models are evaluated under two spatial split protocols: 1) random pointwise splits used by prior work; and 2) spatially buffered cross-validation (SB-CV), which withholds entire image patches rather than individual wells. The SB-CV protocol also includes a buffer to guarantee no training point is within 4 km of a test point or a known power-producing system. Each model produces a favorability map from the median prediction over the 100 realizations; we score the area under the curve (AUC) of power-producing systems versus prediction percentile. All models lose skill under the SB-CV versus random pointwise splits: AUC falls from 0.916 to 0.887 for linear regression, from 0.959 to 0.894 for XGBoost, and from 0.955 to 0.856 for the U-Net, the largest drop belonging to the model with the most spatial context, which under SB-CV ranks below both pointwise models. Skill and the footprint of the map reported as highly favorable are decoupled; at comparable AUC, the fraction of the basin flagged as highly favorable ranges from ~1% (U-Net) to ~20–37% (pointwise models). Model architecture also changes which known systems rank as having the highest favorability. The two pointwise models rank systems nearly identically (Spearman correlation ρ≈0.81), whereas the U-Net favors a materially different set (ρ≈0.47). We recommend reporting spatially buffered metrics, map footprint, and system-ranking agreement alongside random-split AUC.

Topic: General

Foam-based drilling fluids offer significant advantages for geothermal well construction, particularly in underbalanced drilling and in formations prone to lost circulation. However, maintaining foam stability at elevated temperatures remains a major challenge because accelerated liquid drainage can compromise foam rheology, cuttings-transport capacity, and wellbore pressure control. Accurate prediction of foam drainage under geothermal drilling conditions is therefore essential to optimize fluid formulations and ensure reliable drilling performance. This study proposes an improved temperature-dependent foam stability model to predict liquid drainage in polymer-based foams under elevated-temperature and elevated-pressure conditions. The model incorporates the coupled effects of temperature, foam quality, liquid-phase rheology, gravitational drainage, capillary resistance, and time-dependent changes in foam structure and permeability. Particular emphasis is placed on capturing the three characteristic drainage regimes: early drainage, steady drainage, and drying. The proposed model is evaluated against previously reported experimental measurements at 6.8 MPa, with temperatures from 25 to 90°C and foam qualities from 45% to 65%. Experimental drainage profiles, derived from differential pressure measurements along a vertical stability cell, form the basis for model calibration and validation. Model predictions are compared with measured fractional drainage, drainage rates, and foam half-life to assess accuracy across different operating conditions. The analysis focuses on predicting temperature-induced foam destabilization, the stabilizing effects of polymer additives, and the influence of foam quality on liquid retention. For geothermal drilling applications, the model provides a predictive framework for estimating foam stability during circulation, assessing the effects of elevated downhole temperatures, and identifying suitable foam formulations and operating conditions to maintain adequate rheological properties and cuttings transport performance. The proposed approach is intended to support improved foam drilling fluid design, reduce the risk of premature foam degradation, and enhance operational reliability in high-temperature geothermal wells.

Topic: Drilling

Non-condensable gas (NCG) reinjection in geothermal fields is an emerging strategy to reduce greenhouse gas emissions from geothermal power plants, primarily carbon dioxide (CO2). However, this promising strategy introduces potential subsurface risks, including injectate breakthrough, external well casing corrosion, and permeability alterations due to mineral dissolution and precipitation. To rigorously assess these risks, reactive transport modelling (RTM) serves as an essential method, as it couples non-isothermal multiphase fluid flow, heat transfer, and geochemical reactions within porous and fractured subsurface media. Among available reactive transport simulators like PHREEQC and MRST, TOUGHREACT stands out as the industry-standard code for geothermal NCG reinjection due to its robust handling of multi-dimensional, complex thermodynamic conditions and multi-component reactive geochemical systems. This paper aims to establish a comprehensive methodological framework for conducting RTM in geothermal NCG reinjection, addressing critical model design choices, including spatial dimensionality (1D radial for near-wellbore effects, 2D cross-sectional for fluid migration, and full 3D for field-wide dynamics). Furthermore, we evaluate the fundamental necessity of natural-state geochemical modelling to define initial mineralogy and fluid chemistry. We systematically examine major sources of uncertainty in RTM, specifically focusing on the mineral’s reactive surface area (RSA) of minerals, initial rock mineralogy, and baseline fluid chemistry, emphasizing the critical need for sensitivity analyses. Drawing upon pivotal lessons extracted from previous TOUGHREACT studies on NCG reinjection, we highlight best practices for mitigating numerical instabilities and ensuring representative geochemical constraints. To demonstrate the practical application of this framework, we present a simplified 1D radial RTM case study modelling the feed zone permeability and near-wellbore porosity changes typical of the Ngā Tamariki geothermal field in New Zealand, which has successfully operated active NCG reinjection since 2021. The simulation results quantitatively indicate that the influx of lower-pH, NCG-enriched brine initially promotes primary mineral dissolution, temporarily enhancing near-wellbore permeability, followed by secondary carbonate precipitation further into the formation, which may gradually impede fluid flow. In conclusion, deploying structured RTM workflows using TOUGHREACT significantly improves the predictability of subsurface geochemical interactions, enabling operators to proactively design reinjection strategies that maximize carbon mineralization while minimizing operational risks such as injectivity decline and wellbore scaling.

Topic: Injection

Conventional geothermal power plants are a mature technology, with well-established flash, dry steam and binary designs for fluids typically below 300 °C. Supercritical geothermal resources, where fluids exceed the critical point of water (374 °C, 22.1 MPa), are claimed to offer higher power production potential per well, yet no power plant has been designed or operated for such fluids. In contrast, the first commercial supercritical coal-fired unit entered service in 1957, and modern ultra-supercritical units operate at main steam temperatures of about 600–630 °C. This paper reviews the literature on both technologies and identifies which elements of coal-fired supercritical practice can be transferred to geothermal power plant design and which cannot. Six areas of transferable experience are examined: cycle configuration, high-temperature materials, steam purity and cycle chemistry, turbine solid particle erosion and phase transition zone damage, high-pressure valves and components, and operational practice. Fundamental differences between the open geothermal cycle and the closed coal-fired cycle are discussed, candidate plant configurations are compared, and the main challenges and opportunities are summarised.

Topic: Emerging Technology

Production casing design for superhot rock wells must accommodate two opposing thermal load regimes: cold, high pressure hydraulic fracturing loads that generate high tensile stresses in the deeper sections of the well; and hot production loads that generate compressive yielding and cyclic thermal strains, particularly in the shallower sections. Reliability Based Design (RBD) may be required to qualify casing for hydraulic fracturing (HF) operations, while strain-based acceptance and Low Cycle Fatigue (LCF) methods are needed where yielding occurs during production. A material grade crossover balances these demands by placing higher strength material in the lower section and higher ductility material in the upper section. However, as wells get deeper and the pressure demand on hydraulic fracturing increases, increasingly higher strength—and therefore lower ductility—materials in the lower section of the well will be needed. Further, the material crossover point must simultaneously shift to a shallower depth. As a result, low ductility materials may become subjected to high temperature cycles. These temperatures also threaten cement integrity. This paper presents an approach for thermally managing the cement set temperature to establish a favorable initial condition for balancing the loading requirements while preserving cement integrity. The results show that controlling the set temperature can deepen the material crossover, enabling a production casing design that maintains integrity under both HF and production loads. Two case studies are presented to illustrate this approach. The authors conclude that thermal management of production casing installation conditions is critical to maintaining well integrity throughout the intended service life.

Topic: Modeling

Previous oil-and-gas studies have evaluated surface-to-borehole, borehole-to-surface, crosswell, and single-well electromagnetic (EM) configurations for appraising fractures containing electrically conductive proppant. For deep enhanced geothermal systems (EGS), a single-well architecture is operationally attractive but electromagnetically challenging: both transmitted and fracture-induced secondary fields must traverse steel casing, which attenuates and redistributes them. Resistive crystalline host rocks improve field propagation and target contrast outside the well but do not eliminate casing-related attenuation of EM fields. This study develops a synthetic 2D axisymmetric COMSOL forward-modeling and inversion workflow to quantify the detectability and parameter recoverability of a single circular fracture containing electrically conductive proppant under representative cased-hole geothermal conditions found in the Utah FORGE site. A transition boundary condition represents the thin steel casing. The analysis identifies spacing-specific operating frequencies and compares residual-weighting strategies, background-response transformations, and receiver configurations. Our optimized workflow was evaluated using a noise model comprising zero-mean Gaussian perturbations with an 8% relative standard deviation and a 10-9 V absolute noise floor combined in quadrature. Under these conditions, the workflow yields a mean 95th-percentile absolute percentage error of 23.47% across the evaluated parameters and target cases. For the 92 m target fracture radius, the P95 absolute radius error is 41.43% (38.11 m), whereas for the 14 m target radius, the error decreases to 17.30% (2.42 m). Reducing the absolute noise floor to 10-10 V decreases these errors to 26.12% (24.03 m) and 7.86% (1.10 m), respectively, indicating that long-spacing measurements are strongly limited by the assumed voltage floor. Under the corresponding fiberglass or electrically nonconductive casing model, errors decrease further to 7.44% (6.84 m) and 1.59% (0.22 m), respectively. These results establish a quantitative synthetic basis for designing multi-frequency, multi-spacing, single-well EM measurements through steel casing to constrain conductive-proppant extent. Coupling this workflow with short-spacing 3D inversion could additionally enable fracture dip and asymmetric geometry to be represented within a unified inversion result. The assumed noise model represents a conservative stress test rather than a calibrated instrument specification.

Topic: Reservoir Engineering

Hot dry rock (HDR) geothermal resources are a vast energy source for China’s carbon neutrality goals. This review synthesizes the challenges and future perspectives of HDR exploitation through Enhanced Geothermal Systems (EGS) in China from a view of fluid-reservoir interactions. The diverse utilization pathways of geothermal resources, including geothermal-assisted oil recovery, geothermal-driven natural gas hydrate dissociation, and co-mining of mineral and geothermal resources, indicate a great potential of exploiting geothermal energy. The challenges specifically summarize reservoir damages induced by working fluids that contact with reservoir formations. Further, the perspectives highlight closed-loop heat extraction systems and reservoir protection working fluid technologies for open-loop heat extraction, as promising solutions to enhance economic viability and sustainability. Key technological advances in reservoir working fluid design assisted with artificial intelligence techniques and muti-field coupling numerical simulation analysis of fluid-reservoir interactions, are critically reviewed. This review concludes that promoting EGS development should integrate traditional empirical and experimental methods with coordinated and interdisciplinary approaches. These insights provide a strategic road map for sustainable EGS development in China and beyond.

Topic: Enhanced Geothermal Systems

This study develops a data-driven framework to forecast subsurface temperature fields using distributed temperature sensing (DTS) measurements from one of North America's largest ground-source heat exchange (GHX) fields. Accurate subsurface temperature forecasting is critical for the sustainable operation of large-scale GHX systems, where thermal response depends on long-term heat storage, reservoir heat flux, and seasonal variations in the thermal load profile. A neural network uses two years of historical ground heat flux and ambient air temperature, along with the temperature response of a 2596-borehole, 150-m-deep geothermal field in the Upper Midwest of the United States, to predict the temperature field over the next 90 days at 150-m depth. Current results show a Root Mean Square Deviation (RMSD) of 0.11-0.13 °C during temporal extrapolation. To further investigate the influence of model complexity, we systematically vary the hidden-layer width while holding all other training settings fixed. Our results show that neural-network architecture substantially influences both fitting behavior and extrapolation performance, with test RMSD varying from 0.11 °C to 0.21 °C. These findings emphasize the importance of model capacity, optimization, and temporal generalization in subsurface thermal forecasting. This work provides a basis for future development of more robust data-driven approaches for geothermal temperature prediction and for the rational operation of district-scale, low-enthalpy geothermal facilities.

Topic: Low Temperature

Wells receiving flow from two or more discharge zones can exhibit cycling behavior at the wellhead. This instability is driven by interactions between feed zones with contrasting permeability and storage characteristics. Consequently, one or more zones may display geyser-like behavior, characterized by a sudden surface surge in fluid flow followed by a gradual decline until flow ceases. This phenomenon can choke neighboring wells connected to the same gathering system and destabilize power generation. Mitigation strategies include surface venting or executing a targeted workover to identify and stimulate the problematic feed zone. In the Berlín Geothermal Field, Well TR-17B induced steam flow oscillations of up to 5 kg/s across the system. Several pressure-temperature (PT) log including a deep TFT were conducted to pinpoint the unstable feed zones. Subsequently, a matrix acidization workover using a mud acid preflush of 15%HCl and a mainflush of 12%HCl/5%HF was executed to restore permeability and remove near-wellbore scaling. Following thermal recovery, the cycling amplitude dropped to 1.5 kg/s, demonstrating that comprehensive well evaluation significantly increases the likelihood of a successful workover.

Topic: Reservoir Engineering

We present the Advanced Reactive Transport Simulator (ARTS) for estimating thermal breakthrough and water loss in fracture networks of enhanced geothermal systems (EGS). ARTS, an in-house EGS simulator developed by the Energy & Geoscience Institute at the University of Utah, can generate discrete fracture networks (DFNs) stochastically or based on in-situ stress, import existing DFNs such as the Utah FORGE DFN, and account for fracture opening as the rock cools. ARTS reproduces the Gringarten analytical solution with a root-mean-square error of 0.62 °C over 30 years. In a case study using 125 fractures from the Utah FORGE DFN and an injection rate of 26.5 kg/s, the produced water temperature begins to decline after approximately 0.25 years (3 months) and decreases by 7 °C within 1.6 years. In a stochastic network containing 8 hydraulic fractures and approximately 1,200 natural fractures, increasing the injection rate from 3 to 20 kg/s raises the average thermal power over 30 years from 1.7 to 5.2 MWth but reduces the time to a 7 °C temperature decline from 13.2 to 0.45 years. Faults and matrix permeability can also be incorporated to represent geologically plausible explanations of water loss ranging from low (1%) to high (40%). ARTS can further evaluate the heat recovery from a specified rock volume with multiple wells in any configuration.

Topic: Modeling

Bulk X-ray diffraction (XRD) analyses were performed on powdered cuttings collected throughout the deviated sections of Utah FORGE injection well 16A(78)-32 and production well 16B(78)-32. Samples were collected at 10-ft intervals, resulting in 366 analyses from 16A and 395 analyses from 16B, representing 3,655 ft and 3,945 ft, respectively, of excavated rock. Pulsed neutron logging measures neutron-induced gamma rays to characterize the elemental composition of the formation. The total gamma-ray energy spectrum is decomposed to estimate concentrations of Si, Al, Fe, Mg, Ca, Na, K, Mn, Ti, C, and S. High-resolution data was obtained in intervals of interest over 1,000 ft in each well. The tool was deployed inside drill pipe, within cemented and cased sections of the wells. The XRD mineralogic results will also be compared with published X-ray fluorescence (XRF) elemental data from the same 16A cuttings samples between 7,000 and 9,500 ft measured depth. Several intervals in 16A contain overlapping XRD, XRF, and pulsed neutron datasets, providing a direct basis for comparison among mineralogical and elemental measurements. XRD and XRF data are used to ground truth the pulsed neutron results and other geophysical well logs. This integrated dataset is expected to identify alteration associated with paleo-hydrothermal water–rock interaction and potential planes of weakness, including faults and mineralized fractures to better resolve structures that may influence fracture propagation during stimulation. Mineralogical and chemical variations will be used to correlate lithologies with broadly similar compositions between the wells, and estimate lithology-dependent thermal conductivity, heat capacity, and thermal diffusivity. Ultimately, the work will refine the geologic model of the EGS reservoir, improve understanding of lithologic controls on fracturing, including whether fractures preferentially occur within particular units or at lithologic contacts.

Topic: FORGE

Optimizing geothermal systems requires accurate predictions of coupled fluid flow and heat transport across a wide range of operating scenarios and time scales. Spatial variability in permeability and porosity further complicates these predictions by controlling hydraulic connectivity, fluid residence times, and thermal depletion. Although physics-based reservoir models provide a robust framework for evaluating operating strategies, their repeated use in optimization and uncertainty analysis can require substantial computational resources. We developed a surrogate model based on a Universal Physics Transformer (UPT) to predict reservoir pressure and temperature evolution and resulting energy production. Unlike grid-dependent surrogate approaches, UPTs can represent spatially heterogeneous reservoir properties while operating independently of a fixed computational grid, providing greater flexibility across reservoir geometries and discretization. The model is trained on a comprehensive dataset of 1,500 thermal-hydraulic reservoir simulations performed with the MOOSE/FALCON framework. Across the simulation ensemble, reservoir pressure, temperature, permeability, porosity, well separation, and operating parameters are systematically varied to capture a broad range of geothermal reservoir conditions. The resulting surrogate reduces the computational cost of physics-based geothermal reservoir prediction by more than two orders of magnitude. We further develop an optimization workflow that couples the surrogate model with well-response and energy-production calculations, while retaining numerical reservoir simulation for verification of candidate operating strategies. Energy extraction is optimized by adjusting reservoir operating parameters over a 30-year production period. We present examples demonstrating the predictive performance and computational efficiency of the proposed data-driven modeling and optimization framework. Finally, UPT-based predictions and optimized operating strategies are compared with numerical simulations representative of real geothermal reservoir conditions.

Topic: Modeling

Silica scaling can constrain geothermal energy production by reducing fracture transmissivity and fouling surface equipment. Predicting this behavior requires resolving the coupled effects of fluid flow, heat transport, mineral reactions, and operating conditions. HeatFlow is developing a Trilateral Flash Cycle (TFC) technology designed to extract more thermal energy from produced geothermal brine than conventional power-conversion approaches. The increased heat extraction results in substantially lower brine temperatures prior to reinjection, raising questions about how cooler reinjection may affect silica precipitation and scaling within geothermal systems. In this study, we investigate the potential tradeoff between enhanced energy recovery and silica-scaling risk. We used the MOOSE/FALCON framework to conduct three-dimensional thermal–hydrological–chemical (THC) simulations of a fractured enhanced geothermal system (EGS) and evaluate the tradeoff between thermal extraction and silica production. The modeled reservoir based on the legacy Fenton Hill EGS site was 3 km × 3 km × 1 km and contained approximately 72,000 hexahedral matrix elements with 50 m resolution. Fractures were represented explicitly as lower-dimensional quadrilateral elements embedded within the three-dimensional matrix. Fluid and heat transport were coupled with spatially distributed geochemical reactions. A kinetic quartz-like mineral was used to represent silica–rock interaction within the reservoir, while equilibrium precipitation from produced fluid was calculated at the reinjection temperature. Production temperature and cumulative silica precipitation were tracked over 30 years. The results demonstrate nonlinear interactions among reservoir properties, fracture geometry, and operating conditions. At constant flow rate, concentrating flow within a single hydraulic fracture accelerated cold-plume migration and thermal breakthrough, whereas distributing flow among multiple fractures increased rock–fluid heat-transfer area. Injection temperature had comparatively little effect, while injection rate produced the largest response: higher rates increased total throughput but accelerated reservoir cooling and reduced silica recovery per kilogram of produced fluid. These results show that operating conditions that maximize short-term energy extraction may also accelerate thermal depletion or increase total silica transport. Coupled THC simulations can therefore help define operating strategies that balance energy production, reservoir longevity, and silica-scaling risk, providing insights for the deployment of high-efficiency geothermal conversion technologies such as TFC.

Topic: Enhanced Geothermal Systems

High enthalpy geothermal systems producing liquid-rich and steam-rich geothermal fluid have been identified in several regions. Although silica scaling risk is more prevalent in liquid-dominated fluid, the steam-rich fluid is also prone to scaling as seen from the case of Olkaria geothermal powerplant where massive silica scale formed in a steam-dominated, high enthalpy production well. The cause of scale formation in steam-dominated environment remains unclear due to the extremely low solubility of silica in steam-phase, and no detailed characterization of the scale from such system. Therefore, we have collected scale samples from a steam-dominated production well in Olkaria geothermal power plant, Kenya to infer their deposition condition through characterization of major and trace chemical components. The scale is composed of 97% SiO2 with minor CaO (0.16%), MgO (0.10%), and Fe2O3 (0.08%). The silica component occurs mostly as amorphous silica as indicated by XRD analysis. Observation under SEM shows a predominantly dense texture with rare and dispersed occurrence of microparticles of clay minerals. Although bulk concentration of Al is low (undetected by XRF), EDX mapping reveals its occurrence along the scale-pipe interface, suggesting it might play an important role in the early formation of the scale. LA-ICP-MS analysis shows Al concentration of ~6,700 ppm (max. 21,000 ppm) at this zone, substantially higher than the edge of the scale (~400 ppm). The pipe interface zone also shows elevated Ga and Ti concentrations, indicating higher temperature, whereas the edge zone shows significantly lower Ga and Ti content suggesting lower temperature formation. Particles within the scale show elevated Ga and Ti content than the surrounding scale. They are likely formed at the reservoir and transported as carried particles. Our study shows that the scale at the pipe initially formed as Al-rich silica precipitated at higher temperature. The scale is mostly composed of silica that are precipitated directly from the geothermal fluid with few carried particles from the reservoir.

Topic: Geochemistry

Geophysical data, including time-lapse gravity and self-potential, have been periodically observed before and during ongoing operations around the Wasabizawa geothermal power plant in Akita Prefecture, Japan (Horikawa et al., 2025). A three-dimensional numerical reservoir model developed prior to the beginning of the operation in May 2019 (Nakanishi et al., 2017, 2021) has been improved using logging data from newly drilled wells, operational data from the power plant, and geophysical survey data before and during operations (Kano et al., 2025). In this study, we report the latest results from repeated geophysical monitoring at the site, a production-reinjection history-matching simulation that incorporates the operational history for comparison with the monitoring data, and efforts to further improve the reservoir model based on the history-matching results. The observed gravity that has reached a quasi-steady state a few years after the beginning of the power plant operation suggests a recharge effect, which is not included in the present model for conservative resource estimates. We incorporate recharge effects into the model trying to account reasonably and comprehensively for obtained monitoring data, and conduct long-term simulations to evaluate their effects on reservoir behavior, power plant operation, and its sustainability for decades. Acknowledgement: The authors are grateful to Yuzawa Geothermal Power Corporation for their generous support on developing the model, conducting geophysical monitoring, and the history matching simulations. References: Kano et al. (2025) SGP-TR-229.; Horikawa et al. (2025) SGP-TR-229.; Nakanishi et al. (2017) GRC Transactions 41, 2811-2825.; Nakanishi et al. (2021) Proceedings, World Geothermal Congress 2020+1, 06006.

Topic: Field Studies

Geothermal power plants, particularly those based on Organic Rankine Cycle (ORC) technology, present operational challenges that conventional monitoring and SCADA systems do not fully address. Important performance indicators, including individual turbine power output, the progression of heat exchanger fouling, and cycle-level thermodynamic efficiencies, are often not directly monitored and may require separate engineering calculations. In addition, many ORC plants do not maintain a centralized data historian. Operational records may instead remain in paper-based shift logs or be distributed among several spreadsheets, which limits long-term performance assessment and makes operational comparisons more difficult. This paper presents Thermosoft, a software platform developed specifically for geothermal power plants to combine long-term operational data storage with real-time thermodynamic and heat transfer calculations, regression- and statistics-based performance analysis, and machine-learning-based production estimation. Thermosoft communicates directly with plant control systems through PLC gateways and continuously acquires and archives operational data from turbines, heat exchangers, condensers, and auxiliary equipment. Its thermodynamic engine calculates ORC cycle efficiency, turbine isentropic efficiency, heat exchanger thermal performance, fouling ratios, individual turbine power contribution, and the loss in electricity production associated with heat exchanger fouling. Historical operating data can also be combined with weather forecasts to estimate short-term electricity production. In parallel, the contribution of each production well to gross and net plant output is evaluated individually. A financial module converts the resulting operational indicators into economic performance information that can be used by plant managers and owners. These functions are accessible through a cross-platform web interface that includes configurable dashboards, automated alarms, digital shift records, and flexible tools for exporting and reviewing plant data. A pilot deployment at a binary geothermal power plant in western Turkiye showed that Thermosoft could identify thermodynamic inefficiencies that had not previously been recognized through routine plant monitoring. Corrective actions based on these findings resulted in a measurable improvement in power output, demonstrating the practical value of combining real-time thermodynamic calculations with historical operating data. Future development will focus on extending Thermosoft beyond monitoring, diagnostics, and performance analysis toward an AI-assisted autonomous operational control system capable of supporting real-time optimization of geothermal power plants.

Topic: Emerging Technology

Reservoir and power-plant models of enhanced geothermal systems (EGS) are usually run apart from each other. Reservoir simulations take the reinjection temperature as a fixed input, and plant studies take the produced brine as a fixed resource, although the two are tied together. The plant sets the temperature at which the brine returns to the reservoir, and the reservoir sets the temperature and flow the plant has to work with, both of which change over the life of the field. This paper presents a coupled digital model that closes this loop and uses it to examine how the interaction between the two sides affects the long-term behavior of an EGS project. The subsurface side is a semi-analytical model of a stimulated reservoir with many parallel fractures connected by wellbore segments. It tracks the thermal drawdown of each fracture, the redistribution of flow between fractures as their temperatures and apertures change, the pressure loss across the network and the heat remaining in the rock. The surface side is a recuperated organic Rankine cycle with an air-cooled condenser, modeled with fixed exchanger areas so that the plant responds to a changing brine temperature the way a built plant would. In every interval the reservoir hands the production temperature and pressure loss to the plant, and the plant hands the resulting reinjection temperature back. The tool also reports operating limits and failures, bare-module equipment costs and a levelized cost of electricity. Preliminary cases indicate that the reinjection temperature delivered by the plant can differ considerably from the fixed value normally assumed, that it depends on the working fluid and on the plant design and operating strategy, and that these differences carry through to the reservoir thermal profile, the heat extracted and the number of years the plant can operate before its turndown limits are reached. The paper will describe the modeling framework, illustrate it with selected cases, and discuss which aspects of the coupling matter most for the assessment of EGS projects.

Topic: Enhanced Geothermal Systems

Wellbore integrity problem faces distinct challenges in Conventional Geothermal Systems (CGS), Enhanced Geothermal Systems (EGS), and Closed Loop Geothermal Systems (CLGS). In CGS, the wellbore casing and cement experience large pressure and temperature fluctuations leading to mechanical deformation, cement debonding, and loss of zonal isolation. In EGS, additional integrity risks arise from the high temperature stimulation, proppant placement, elevated injection pressure, and repeated thermo-mechanical cycling. In contrast, CLGS tubulars generally experience smaller pressure and temperature cycle amplitudes compared to CGS and EGS. Nevertheless, CLGS wells still require robust integrity due to sustained high temperature exposure over long operation periods. Wellbore integrity issues in CLGS wells such as cement debonding and micro-annulus formation are of critical importance as they can significantly reduce heat transfer between the host rock and the circulating fluid inside the wellbore. Furthermore, the type of cement and cement formulations used in well completions across these geothermal systems are different resulting in different rheological properties, placement behavior, mechanical properties, and associated wellbore integrity risks. These challenges are further amplified in horizontal well geometries, where eccentricity and nonuniform placement can compromise cement-sheath quality. In this paper, we review the existing literature to identify and analyze the major wellbore integrity issues encountered in these geothermal systems. In addition, we compare both short-term and long-term wellbore integrity issues across these geothermal systems under different operational scenarios, well geometries, fluid rheology, and cement formulations using advanced wellbore integrity modeling software. This software simulates initial stresses during cement placement and hydration along with long-term thermo-mechanical interaction between the casing, cement, and surrounding formation allowing us to capture integrity issues throughout the wellbore’s lifetime. This integrated approach enables the assessment of wellbore integrity throughout the entire lifecycle of geothermal wells and provides insight into the key factors governing integrity performance across different geothermal technologies.

Topic: Emerging Technology

This study uses statistical and machine learning models to estimate radiogenic heat production (RHP), thermal conductivity (TC), and heat flow (q) on global and regional scales. Predicting subsurface properties is necessary for assessing geothermal resources and siting power projects. A total of 220,527 data points compiled from global open-source geothermal and geochemical databases (the IHFC Global Heat Flow Database, GRANITE2017, Gard et al., 2019, and EarthChem) were used to train regression methods: 2nd- and 3rd-degree polynomials, random forest, and XGBoost. Gravity, magnetics, the P-wave-to-S-wave velocity ratio, depth to the Moho, and proximity to active faults and earthquakes were covariates for all three targets. Model performance was evaluated via k-fold cross-validation across each of CRUST1.0's crustal layers, from unconsolidated sediments to the upper mantle, to assess how predictive performance varies with depth. Global prediction models yielded cross-validated r and R-squared values as high as 0.84 and 0.71, respectively, for thermal conductivity predictions, indicating a moderately strong fit given the heterogeneity of the underlying data. After separating the dataset across individual continents and an ocean/land split, r and R-squared values as high as 0.81 and 0.65 were still calculated, underscoring the importance of regional data density and geologic setting. A regional model was also developed to predict RHP, TC, and q with increased spatial resolution. Results from the global model suggest that machine learning methods can predict RHP, TC, and q with moderate to high accuracy when sufficient training data are available. Results from the regional model can directly inform geothermal prospecting.

Topic: Geophysics

North Dakota produces over 500 million bushels of corn annually (USDA NASS, 2024), the majority of which requires mechanical drying, predominantly with propane-fired systems at high cost and carbon intensity. Ground-source heat pumps (GSHPs), which draw on shallow ground loops at 10–15 °C and deliver drying air near 43 °C, offer a renewable alternative if their thermodynamic limits are properly characterized. This study characterizes thin-layer drying kinetics of one feed-corn lot rewetted to two moisture levels (approximately 17% and 25% wet basis) over a 50–103 °C temperature range using a gravimetric protocol based on ASABE S352. Dry-matter content was established by taking the 50 °C samples to bone-dry at 103 °C and applying the resulting dry-matter fractions to all runs of the same corn lot. Moisture content was tracked at intervals until samples approached equilibrium. The Page model (MR = exp[−k tⁿ]) was fitted to eight temperature–container combinations using a pooled Arrhenius regression, yielding an apparent activation energy Eₐ = 21 kJ/mol with a pooled Page exponent n = 0.60. At 60 °C, the higher-moisture sample (starting near 25% wb) reached safe-storage moisture (15% wb) in approximately 112 min; at 80 °C this shortened to approximately 55 min. Extrapolating to a 43 °C drying-air condition, consistent with a GSHP drawing from a 10–15 °C ground loop, predicts a drying time of approximately 4 h for corn starting near 25% moisture. These results are integrated into a preliminary systems analysis for a 10,000-bushel farm-scale GSHP batch-bin drying system. The kinetic model provides the throughput constraint for dryer sizing, while the geothermal resource assessment defines the recoverable heat supply. Because rewetted feed corn may not fully replicate the drying behavior of freshly harvested grain, validation with fresh field corn at harvest moisture is planned for the 2026 season. Preliminary analysis indicates geothermal-assisted drying is technically feasible and warrants a full economic and lifecycle assessment.

Topic: Low Temperature

Geothermal wells frequently experience declining injectivity and productivity due to stress-induced alteration of flow paths, mineral scaling, and loss of connectivity within natural fracture networks. Stimulation through Hydroshearing has historically been employed to enhance permeability, where elevated pressures promote slip along pre-existing fractures. However, extensive field evidence demonstrates that the effects are largely transient. Recent field diagnostics show that sustained enhancement in geothermal systems requires maintaining mechanical aperture, achievable only through opening-mode (tensile) fractures stabilized by various proppants. This article builds a technical and economic case for proppant-based hydraulic fracturing as a sustainable stimulation strategy in hydrothermal geothermal reservoirs. Drawing on historical EGS examples, three principal paradigms where proppant placement provides distinctive value, have been proposed: (1) extending permeability away from major faults to connect low-flow zones into active reservoir corridors; (2) strategically fracturing within injector–producer patterns to improve field-scale connectivity and thermal sweep; and (3) directly propping fault and feed-zone structures to prevent stress- or thermally-induced closure. Collectively, these strategies can maintain or restore plant capacity without costly make-up wells. The article addresses overcoming the existing operational challenges, including deviated well trajectories, slotted liners, high formation temperatures, and complex fracture geometries. Modern completions and fracturing technologies including advances in stimulation chemistry portfolio are discussed. High-viscosity, non-damaging fluids, zonal isolation packers, reliable degradable diverters, and temperature-resistant ceramic proppants can enable reliable placement even in geothermal wells at 250–300°C. Laboratory tests and modeling confirm that such proppants maintain conductivity under sustained stress and thermal cycling, resisting creep and chemical degradation that limit traditional sand systems. An economic analysis contrasts the cost-benefit analysis of proppant stimulation ($1 – 2 million per well) with new, makeup well drilling ($6 – 10 million) and shows that modest injectivity improvements (5 – 10%) can preserve several megawatts of capacity in a 30 - 50 MW power plant, translating to a significant reduction in levelized cost of electricity. Beyond economics, proppant-based fracturing offers improved predictability and design control relative to hydroshearing, supported by decades of oil-and-gas modeling tools adaptable to geothermal stress regimes. In conclusion, proppant-assisted hydraulic fracturing provides a durable, mechanically supported permeability enhancement mechanism for hydrothermal reservoirs where shear and thermal stimulation alone are inadequate. By integrating EGS learnings, high-temperature materials, and modern stimulation design, this approach can extend the productive life of hydrothermal fields, reduce the frequency of make-up drilling, and help close the gap between hydrothermal and engineered geothermal resources. The study advocates for pilot demonstrations to quantify performance gains and refine techno-economic parameters, marking a critical step toward sustainable, high-efficiency geothermal power generation.

Topic: Enhanced Geothermal Systems

Efficient fluid transport is essential for maximizing production and reducing operational costs in geothermal systems. However, liquid-dominated geothermal wells often experience significant frictional pressure losses due to high rates and turbulence, especially in tubing and surface pipelines. This increases the parasitic load due to the required pumping pressure. Drag-reducing agents (DRAs) are one of the solutions used to decrease frictional pressure losses by decreasing the turbulent flow in long pipelines. This study investigates the potential of DRAs to decrease the frictional pressure losses in multiphase water-air flow under controlled laboratory conditions. The tests aim to mimic flow in geothermal production systems. Experiments were conducted using a 30-ft horizontal flow section with air/water and air/water+DRA mixtures across a range of liquid superficial velocities (0.1-2 ft/s) and gas superficial velocities (2-20 ft/s). Pressure gradients and multiphase flow behavior were evaluated using differential pressure measurements and high-speed video recordings. Detailed video analysis was performed to characterize the liquid-gas flow with stratified, pseudo-slug, or slug flow pattern. The changes in flow pattern, pressure gradient, and DRA efficiency were mapped across the test matrix. Flow pattern dependent variables like wave celerity, slug propagation, slug frequency, and slug translational velocity were also evaluated to track changes in gas-liquid flow dynamics associated with DRA addition. Results demonstrate that the addition of DRA decreases frictional pressure losses by up to 60%. Substantial reductions are observed at higher liquid and gas superficial velocities and under slug flow conditions. High-speed flow visualization and quantitative video analysis provide further insight into how DRA addition affects the liquid-gas flow pattern, promoting stratified flow. It also impacts slug behavior and gas-liquid flow characteristics. These findings demonstrate the potential of DRA application to reduce frictional pressure losses and associated pumping requirements in geothermal fluid transport systems.

Topic: Production Engineering

Hydrogeochemistry of thermal and non-thermal waters at Rukwa Rift Basin (RRB), Southwestern Tanzania. ARIPH Kimani, Kotaro YONEZU, SAEFUDIN Juhri [Kyushu University, Japan] [Tanzania Geothermal Development Company Ltd,TGDC] The Rukwa Rift Basin (RRB), located within the western branch of the East African Rift System (EARS), hosts numerous thermal springs that indicate the presence of an active geothermal system associated with rift-related tectonics and elevated heat flow. However, the hydrogeochemical processes governing fluid evolution, reservoir characteristics, and geothermal potential within the basin remain poorly constrained. This study investigates the hydrogeochemistry of geothermal waters from the southern and southeastern parts of the RRB to determine fluid origin, evaluate water–rock interaction processes, assess mixing and cooling mechanisms, estimate reservoir temperatures, and develop a conceptual geothermal model. Hydrogeochemistry of major, trace elements, and geothermometric analyses were performed on thermal spring waters and associated groundwater samples. The geothermal fluids are predominantly characterized by the Na–HCO₃ hydrochemical facies, indicating extensive interaction between meteoric recharge waters and the reservoir rocks during deep circulation within the rift system. The hydrogeochemical signatures further indicate that fluid migration is strongly controlled by fault systems, which facilitate deep circulation, heating, and prolonged water–rock interaction. Geothermometric evaluations based on cation geothermometers (Na/K Truesdell,1976) indicate reservoir temperatures ranging from 140 to 150 °C, suggesting a moderate-enthalpy geothermal resource. Magnesium enrichment in groundwater suggests that during ascending, geothermal fluids undergo varying degrees of mixing with shallow groundwater and conductive cooling before discharge at the surface. Based on the integrated hydrogeochemical evidence, a conceptual geothermal model is proposed in which meteoric waters recharge the system through elevated rift shoulders and surrounding highlands with minimum inputs from Lake Rukwa. Infiltration occurs along major extensional faults and circulates to depth where they are heated by anomalous geothermal gradients associated with lithospheric extension and tectono-magmatic activity. The heated fluids subsequently ascend through permeable fault networks, undergo further water–rock interaction, partial mixing, and cooling, and finally emerge as thermal springs at the surface.

Topic: Geochemistry

Active tectonic, volcanic, and geothermal environments are often associated with high background seismic activity. Therefore, conducting seismic monitoring is essential to enhance a better understanding of the Earth’s dynamic processes occurring in these environments. The Greater Olkaria Geothermal Area (GOGA) is one of Kenya’s most developed geothermal environments, located in the Kenya Rift System. Passive seismic monitoring is being employed in GOGA to enhance exploration of geothermal resources and monitor seismic hazards. Several temporal seismic networks of varying sizes have been deployed in GOGA since 2019 and have acquired continuous seismic data for several months. Analysis of these data reveals GOGA is characterized by a high seismic activity that originates from both tectonic and geothermal activities. Most tectonic earthquakes are fault-controlled and occur at both shallow and deep depths greater than 5 km below the surface. For geothermal activity-related earthquakes, most occur in clusters near production wells and at shallow depths. The distribution of these earthquakes in GOGA suggests the area is characterized by deep structures (faults), which allow deep circulation of fluids. The occurrence of earthquakes near the geothermal production (fluid production and injection) wells shows a clear correlation of geothermal production and induced seismicity. Seismic monitoring has helped to show that GOGA has potential for deep geothermal resources, which can be explored through deep drilling. Access to these deep geothermal resources will increase geothermal production in GOGA for power generation, help the county improve access to clean energy resources, and accelerate the transition to green energy sources.

Topic: Geophysics

Tracer flow testing is the routine measurement of well output in terms of mass flow and enthalpy in two phase pipelines and is of utmost importance in understanding reservoir performance. This is carried out by injection of exclusively non-radioactive and non-toxic tracers of high chemical and thermal stability- greater than 330oC. High precision multi-phase tracer metering systems with rapid-pulse tracer injection is used as the injection equipment to maximize concentrations at high flow and pressure. Multi-phase sampling separators are used for water-brine and gas-steam sampling. Analysis of samples is carried out by use ultra-specific and sensitive methods which include Gas chromatography and UV-Vis/fluorescence. Interpretation of data is carried out by numerical reservoir simulation services and not just qualitative and quantitative interpretation of results. This kind of tracers are injected in the flow line and sampling is downstream of the injection point. Liquid tracers injected must be conservative- cannot partition significantly to the vapor phase, decay chemically or thermally while in the reservoir for the measurement period. They must also be very detectable at ultra low concentrations without interference from high concentrations of dissolved minerals in the produced water. Naphthalene sulphonates (NSA) are used for brine phase and Sulphur Hexafluoride (SF6) for gas phase. Interpretation of tracer flow test data is only used quantitatively.

Topic: Tracers

Geothermal investment is constrained by high upfront capital outlay, subsurface uncertainties, lengthy project development periods, and risks that change throughout project cycles, all of which limit the bankability of geothermal investments. The bankability of a geothermal development project and its ability to attract financing from commercial sources increases gradually as each successive development phase delivers better results and reduces uncertainty. The objective of this paper is to develop a quantitative Geothermal Bankability Index (GBI) for geothermal projects by considering technical, financial, commercial, environmental, and institutional dimensions in a transparent 0-100 investment-readiness metric. The index combines eight weighted dimensions: resource confidence, drilling and well performance, technical and infrastructure readiness, project economics, financing readiness, commercial and off-take arrangements, environmental and social readiness, and institutional and regulatory readiness. The paper applies a five-level maturity rubric to four Kenyan geothermal cases representing operating, resource-confirmed, field-development, and exploration stages. Prototype scores differentiate the cases in the expected order, while a ±20% weight-sensitivity test preserves the ranking in all simulations. The analysis indicates that resource proof is necessary but not a sufficient parameter of bankability: as projects advance, commercial arrangements, financing structure, infrastructure, and environmental-social readiness become increasingly important. The proposed GBI provides a stage-gate tool for project prioritization, de-risking, financing discussions, and portfolio monitoring. Although developed for Kenya, the architecture is designed for recalibration across other geothermal markets.

Topic: General

The development of the Menengai Geothermal Field in Kenya constitutes a big step toward improving reliable and sustainable renewable electricity supply. The initial phase of the project, targeting 105 MW of power output, is predicted to impose considerable changes in reservoir pressure, temperature, fluid flow, and production characteristics. The first phase of the project targets 105 MWe; however, the field has been subjected to progressive rather than full-scale exploitation at this level. Commercial power generation increased from approximately 38–40 MWe in mid-2023 to about 70 MWe by December 2025 and approximately 92 Mwe currently, with further drilling, well testing and field development ongoing towards the 105 MWe target. This study evaluates the reservoir response to this staged development over a two-year period, distinguishing observed reservoir behaviour under the actual production levels from model-based predictions of reservoir performance at the planned 105 MWe exploitation level. To create baseline reservoir conditions and model the response during continuous production and reinjection, numerical reservoir modeling is integrated with historical production and well-test data. Key indicators of reservoir sustainability, including pressure decline, temperature evolution, production-well deliverability and the contribution of reinjection to pressure support, are assessed in relation to the progressive increase in power generation. Particular attention is given to the reservoir response under the prevailing production levels and the effects associated with partial reinjection during the assessment period. It is anticipated that the findings will offer vital information for controlling reservoir pressure, reducing thermal breakthrough, maximizing production and reinjection strategies, and preserving long-term steam supply. Observed data show current reservoir behaviour; models predict performance under planned 105 MWe exploitation. The study offers a technical foundation for future power generation growth at the Menengai Geothermal Field as well as adaptive reservoir management.

Topic: Reservoir Engineering

Maps of favorable structural settings (FSS), such as stepovers, fault terminations, intersections, and accommodation zones, are among the strongest predictors of geothermal favorability in the Great Basin. Training a machine learning model for FSS prediction could therefore lead to improved geothermal resource estimation. However, FSS labels are interpreted polygons, meaning that a mapped polygon indicates favorable structure somewhere within it, not that every enclosed cell is productive. Training an ordinary pixelwise loss against such labels therefore forces the model to treat entire polygon interiors as uniformly favorable, an assumption the labels do not support. We treat FSS prediction as a weakly supervised segmentation problem, training a U-Net convolutional neural network and comparing a pixelwise classification loss (Dice/BCE, a combination of Dice and binary cross entropy) against three polygon-aware losses (top-k pooling, multiple-instance learning, and generalized-mean pooling) that require only that each polygon contain confident predictions somewhere in its interior. Models are trained and evaluated across 120 spatial cross-validation splits, each holding out whole FSS polygons. We judge the losses primarily on two detection metrics that mirror the exploration task of locating favorable structure: 1) detection of held-out FSS polygons at a fixed background false-positive rate; and 2) recovery of 28 operating geothermal power plants. Both s are scored out-of-fold. For the known-system , we additionally remove the 17 FSS polygons within 4 km of any plant from supervision and build the favorability map as an out-of-fold ensemble mean. On polygon detection, top-k pooling recovers 97% of held-out polygons at a 5% background false-positive rate on every split, whereas the Dice/BCE recovers only 36%. On the known-system , generalized-mean pooling obtains the strongest performance, detecting 12 of 28 plants at a 10% false-positive budget, including 5 of 9 systems with no surface expression (Tungsten Mountain, Blue Mountain, Stillwater, Don Campbell, and Cove Fort, in order of predicted favorability). We recommend that weak-supervision losses be judged by detection metrics matched to the exploration question, namely regional polygon detection and known-system recovery, on which the polygon-aware losses consistently outperform Dice/BCE.

Topic: General

Geothermal energy is increasingly recognized as a significant contributor to the United States' clean energy portfolio, drawing on a diverse array of subsurface heat sources. Beyond traditional hydrothermal resources, Enhanced Geothermal Systems (EGS) offer vast potential for primary energy extraction through advanced heat-mining techniques. However, the development of EGS is still greatly limited by issues such as thermal short-circuiting. The problem of thermal short-circuiting in geothermal reservoirs is a direct consequence of the presence of heterogeneous natural and enhanced fractures. This preferential channeling is a direct consequence of fracture network heterogeneity in natural and engineered reservoirs. Mitigating thermal short-circuiting is therefore critical to maximizing EGS performance. This study investigates the gelation kinetics of silica-based nanofluids under high-temperature, high-pressure (HTHP) reservoir conditions as a potential strategy for dynamic flow management and fracture permeability control. In this study, A series of silica-based nanofluids were formulated using fumed silica (SiO₂) nanoparticles dispersed in a base fluid of deionized water and polyethylene glycol of molecular weight 6,000 (PEG 6000) at a weight ratio of 90:10. The nanoparticle weight fraction was systematically varied from 2 to 6 wt% to assess concentration-dependent gelation behavior. Gelation kinetics were characterized using a high-pressure, high-temperature (HPHT) rheometer under dynamic oscillatory conditions. Experiments were conducted at temperatures of 100°C, 150°C, and 200°C and pressures of 100, 150, 250, and 500 psi, spanning conditions representative of EGS reservoirs. The gel point was identified from the crossover of the storage modulus (G') and loss modulus (G″), which marks the transition from a viscoelastic liquid to a viscoelastic solid. Key gelation parameters, including gelation time, gel strength, and normalized activation energy, were extracted to quantify the flow control potential of each formulation under dynamic reservoir conditions. The gelation kinetics of the formulated silica-based nanofluids were governed by nanoparticle concentration, reservoir temperature, and confining pressure. Increasing silica content accelerated gelation onset and enhanced the resulting gel strength, consistent with greater particle network density at higher loadings. Elevated temperature and pressure both promoted nanoparticle aggregation and accelerated gelation, with pressure exerting a particularly notable effect: the activation energy required to initiate gelation decreased systematically with increasing pressure, indicating that high-pressure reservoir conditions are inherently favorable for gelation-based flow control. Collectively, these findings demonstrate that silica-based nanofluids can be engineered to gel in situ under EGS reservoir conditions, enabling dynamic regulation of fracture permeability and targeted redirection of working fluid toward high-temperature, low-permeability zones, thereby improving overall heat extraction efficiency.

Topic: Enhanced Geothermal Systems

The Neal geothermal system is a deep-circulation-type geothermal field in Malheur County, Oregon owned and operated by Ormat. Electricity production began in 2012, with the commissioning of a 32 MW (gross) power plant. The resource is hosted in Miocene basaltic rocks of the Malheur Gorge, overlain by Miocene-Pliocene volcanic and sedimentary rocks of the Idaho-Graben and western Snake River Plain. Surface manifestations include the Neal Hot Spring that discharged low volumes of ~90°C fluids forming an opaline sinter along the silicified fault trace of the west-dipping Neal Fault. The hot spring is geographically situated proximal to the upflow zone, without any notable chemical indications of mixing, and chalcedony geothermometry does a good job of predicting the resource temperatures. Permeability pathways are inferred within the damage zone of the Neal Fault and associated splays in a normal fault stepover (or relay ramp) structural setting. Recent helicopter-borne time-domain electromagnetic (HTEM) indicates possible shallow clay-alteration overlying the geothermal system as indicated by a low-resistivity anomaly. Since 2008, 27 wells have been drilled at Neal. There are currently four producing wells that intersect feedzones in the Neal Fault or its splays between 690 and 1098 mMD. The Na-HCO3-Cl-HCO3 reservoir fluid has a TDS of ~1000 mg/L, a pH of 6.5-6.8, and Cl concentrations of less than 15 mg/kg, reflective of equilibration within Cl-poor basaltic lithologies. Injection is spread across five wells with feedzones from 224-1934 mMD. Production temperatures initiated at ~142 °C in September 2012, before stabilizing at ~141 °C between 2012 and 2017. Starting in 2018, production temperatures have declined ~0.8 °C per year, corresponding with higher total production flow. The current 2025 field averaged temperature is 134 °C: a modest decline of about 8 °C over 13 years of operations. The modest decline rate is attributed in part to the favorable vertical separation of injection-production feedzones. Here we review the history of the field and present an updated resource conceptual model for Neal that integrates all available geologic, geophysical, geochemical, and reservoir engineering datasets.

Topic: Geology

Superhot rock geothermal development requires drilling technologies capable of operating in extreme downhole environments where circulating temperatures exceed the limits of conventional directional drilling systems. Measurement-While-Drilling (MWD) tools are critical for wellbore placement and drilling optimization, however, their electronics and associated thermal management systems degrade rapidly at elevated temperatures. Most commercial directional drilling tools are currently rated for circulating temperatures between 150°C and 175°C, with only limited systems capable of approaching 200°C. Adjacent technologies including coated or insulated drill pipe and mud chillers can mitigate circulating temperatures, but are of limited use when making connections or when running into a hot hole. As a result, reliable directional drilling in higher temperature geothermal reservoirs remains a key technical barrier to accessing deeper and hotter resources. Chillers, low heat coefficient coatings, insulated pipe and diverter subs are all expensive rental items and in the case of chillers they consume a lot of diesel to run the associated generators. The geothermal operators want to build the most cost-effective high temperature drilling system and running high temperature tools can reduce cost in operational time and in the overall drilling system. This paper presents the design and field validation of an ultra high temperature MWD system rated for drilling mud temperatures up to 230°C. The system incorporates a novel circular stacked circuit architecture developed to improve thermal management and component reliability under high temperature conditions. Unlike conventional rectangular circuit board configurations, the circular boards are integrated with thermally conductive rings that form a continuous axial heat transfer pathway along the tool body. This structure acts as a thermal superhighway, rapidly transporting heat away from sensitive components toward cooling interfaces and ultimately to be removed by the drilling fluid through advection, and thereby minimizing the temperature difference between the working parts of electronic components and the surrounding environment. Field deployments demonstrate that the system can operate reliably under sustained high temperature drilling conditions, validating both the thermal management strategy and electronics design in real-world geothermal environments. The system has been deployed in next-generation geothermal wells, providing field validation under increasingly demanding drilling conditions, and has enabled drilling the hottest EGS well to date. Despite the extreme downhole environment, the system remained operational with no equipment failure, providing a significant validation of its thermal architecture and high temperature electronics. Field deployments in different temperature and frictional environments are crucial to understanding the nuances of drilling deeper and hotter wells. It is critical to test the high temperature tools and to understand their durability in high temperatures and in rocks with high friction coefficients that tend to be prone to vibration. These results represent a critical step forward in extending the operating envelope of directional drilling systems beyond the traditional 200°C threshold. This novel thermal management technology provides a pathway toward drilling and measurement equipment capable of operating at significantly higher temperatures, with development targeting circulating temperatures up to 300°C in the coming years. In turn, this capability will support the development of deeper and hotter geothermal systems, including superhot rock resources, where higher reservoir temperatures have the potential to significantly increase energy output per well and improve project economics.

Topic: Drilling

Thermal cycling in geothermal wells can generate repeated plastic strain that drives low-cycle fatigue damage in tubulars and connections. While strain-based design methodologies have been proposed to address these conditions, experimental validation of the underlying material assumptions has been limited. This paper presents a multi-year experimental program conducted to validate a previously developed low-cycle fatigue methodology based on Critical Strain and a Ductile Fracture Damage Indicator (DFDI). The study focused on three questions relevant to geothermal material qualification: (1) does cyclic plastic deformation degrade Critical Strain, (2) are DFDI-based life predictions representative of actual thermal fatigue behavior, and (3) can Critical Strain be used as a material parameter for screening alternative geothermal alloys? Coupon-scale testing including monotonic true stress-true strain characterization, strain-controlled cyclic loading, thermal cycling, and post-cycling pull-to-failure testing on K55 and L80 materials was performed to validate this approach. Results showed that Critical Strain remained essentially unchanged following cyclic loading, under the conditions tested, even after stabilization of cyclic response and after exposure to significant fractions of the predicted design life. We conducted several additional experiments by cycling specimens to failure under both controlled strain amplitudes and direct thermomechanical loading representative of geothermal service. In the tests performed, experimental lives consistently exceeded DFDI predictions, in many cases, by substantial margins. Materials subjected to ±0.35% to ±0.50% strain amplitudes survived thousands of cycles before failure, while thermomechanical testing produced hundreds to thousands of thermal cycles to failure at severe temperature excursions. The results indicate that, for the materials and conditions tested, the DFDI framework provides conservative life predictions and suggest that Critical Strain is a stable material property that may be measured independently and applied to fatigue-life assessment. The work further discusses how the methodology could potentially be extended to qualify advanced geothermal tubular materials, using a common physically based parameter tied directly to ductility and crack-initiation resistance. These results support the use of Critical Strain as a practical metric for geothermal material screening, qualification, and strain-based well design.

Topic: General

The traffic-light protocols employed in enhanced geothermal systems generally consist of modifying the injection process, but it is rarely directly verified whether the largest induced events occur when the injection process is still capable of being altered. The public records from the April 2024 stimulation of the Utah FORGE wells 16A(78)-32 and 16B(78)-32 have been brought into alignment on a common timeline; this has included one-second treatment data supplied by the fracturing contractor, two independent microseismic catalogs from the surface and downhole networks, the well survey, and 76 daily operations reports which were interpreted by a language model. The stage start and end times extracted correspond with the treatment record within 15 minutes 85% of the time, and the reports provide the perforation depths so that each newly perforated stage can be located. Upon cross-referencing the various sources, four silent errors were found, namely the mistaking of rig pump pressure for treating pressure and a six-hour time-zone offset in the seismic catalog, both of which would have made the analysis invalid. Seismic activity occurs 30 minutes after changes in the injection rate, with a correlation coefficient of r = 0.68 and a family-wise p-value of 0.00006; however, the largest earthquakes occurred after injection had ceased. In the surface catalog, 8 out of the 9 events of magnitude 1.0 or higher, that is, the Amber magnitude defined in the protocol, took place after the shut-in had been carried out, even though 61% of all the events occurred while pumping was underway (p = 0.003; p = 0.013 after applying the Holm correction). The conclusion is even more compelling when the first hour after each shutdown is considered to still be part of the pumping phase (p = 0.0003). The downhole catalog, which employs independent moment magnitudes and contains only those events detected by both networks, also shows this same trend (p = 0.005 for events at Mw 0.5 and above; p = 0.09 for events at Mw 1.0 and above). Although the two catalogs differ in regard to which event was the largest and whether the Amber count was reached, in both cases the largest event happened 8 to 11 hours after shut-in and 62 to 89% of the events of magnitude 1.0 or greater occurred when no injection was taking place. After shut-in, larger events were found to lie outside the cluster of seismicity that had developed during pumping more often than smaller events (56% compared with 37% for events at Mw 0.5 and above, p = 0.0002), a contrast that was not observed during pumping and is consistent with pressure continuing to spread outwards. Accurately timed operator events improved a boosted-tree seismicity forecaster in 7 out of 9 comparisons, although no learned model consistently outperformed a persistence model. Since most of the events which contribute to protocol thresholds cannot be affected by responses that are triggered by injection, the monitoring and response rules should last for many hours after shut-in. Also, before carrying out the modelling, multimodal geothermal analyses need cross-source verification.

Topic: FORGE

Dixie Valley, Nevada, is a fault-controlled geothermal field with reservoir temperatures of roughly 240-250°C. Its operating history points to geochemical risks that should be quantified before enhanced geothermal system (EGS) development, where stimulation and circulation further accelerate fluid-rock chemistry. Production wells have shown amorphous silica scale with minor calcite. The reservoir fluid is a moderately saline brine carrying mostly CO2 with a small amount of H2S. Geothermal production drives degassing and large changes in pH. Reinjected fluid, a mixture of evaporated brine and Ca- and Mg-rich shallow water, adds further potential for scaling and alteration. This study uses batch geochemical modeling to evaluate mineralogical alteration, scaling, and corrosion potential in a multiphase system. The simulations use lopolith and basalt mineralogy and measured reservoir brine and gas compositions to evaluate geochemical changes under operational scenarios such as boiling, cooling, and mixing with injectate. The modeling identifies which minerals dissolve or precipitate, and under what temperature, pressure, and gas-loss conditions silica, calcite, and aluminosilicate scales form. It also predicts how pH and dissolved CO2, H2S, and chloride vary with temperature and pressure, which indicates corrosion risk. The results will support operating strategies and mitigation measures for EGS development at Dixie Valley and similar Basin and Range systems.

Topic: Geochemistry

Injection flow in a multistage enhanced geothermal system is rarely distributed uniformly, yet the long-term thermal consequences of that heterogeneity are not well constrained. This study uses a three-dimensional, field-informed Utah FORGE discrete-fracture model to examine how alternative stage-level injection allocations affect production temperature, thermal breakthrough, instantaneous power, and cumulative heat recovery. Decadal thermo-hydraulic simulations compare the field-observed allocation with equal, toe-biased, mid-well-concentrated, and heel-and-toe-biased profiles over a range of total circulation rates. The results show a consistent performance ranking across the tested rates. Toe-biased injection preserves the highest production temperature, delays thermal decline, and delivers the greatest cumulative energy, whereas the field-observed allocation cools most rapidly. The toe-side advantage reflects both favorable circulation pathways and access to deeper, hotter reservoir zones along the inclined Utah FORGE wells. Equal and mid-well allocations provide intermediate improvements, while concentrating flow at both ends offers only a modest benefit. The advantage of favorable allocation becomes more pronounced as total rate increases, demonstrating that injection rate and flow placement cannot be optimized independently. These findings identify stage-level flow allocation as a practical reservoir-management control. Redistributing flow toward pathways with greater effective heat-transfer volume and higher formation temperature can improve thermal sweep and sustain heat recovery without changing the stimulated reservoir volume.

Topic: FORGE

Average pore pressure is an important parameter in a two-phase fractured reservoir. A few applications of the parameter include reservoir management, history match for reservoir simulation, estimation of effective stress, and calculation of water influx. The in-situ water and gas saturations are normally not known in a two-phase fractured reservoir; therefore, the estimates of average pore pressure are difficult in field studies. With the help of a case study prior to the 2003 Mw 6.8 Chengkung earthquake, groundwater radon was used as a tracer to determine the water and gas saturations in a small naturally fractured aquifer. By taking the volume average of the pressure in the two phases (gas and water), we can estimate average pore pressure. Using the radon-concentration data prior to the 2003 Mw 6.8 Chengkung earthquake, we gain an improved understanding of precursory behavior of fluid saturations, average pore pressure, and effective stress in a small fractured aquifer.

Topic: Reservoir Engineering

Comparative Hydrothermal Alteration and Clay Mineralogy of Intracaldera and Extracaldera Wells in the Olkaria Geothermal Field: Within Caldera OW-736 and outside the ring OW-922 Abstract Clay minerals serve as critical indicators of hydrothermal alteration, providing valuable insights into temperature gradients, fluid chemistry, permeability patterns, and the evolution of geothermal reservoirs. This study compares the subsurface clay mineralogy between intracaldera and extracaldera environments within the Olkaria geothermal field to evaluate structural controls on hydrothermal alteration. Drill cuttings and core samples from well OW-736 (drilled within the caldera boundary) and well OW-922 (drilled outside the caldera ring structure) were analyzed using X-ray Diffraction (XRD), the Methylene Blue spot test, and Scanning Electron Microscopy (SEM). The investigation evaluated the vertical distribution, transition zones, and relative abundance of key clay assemblages, including smectite, mixed-layer illite-smectite (I/S), illite, chlorite, and chlorite-smectite (C/S). Results indicate that well OW-736 displays a shallower onset of high-temperature alteration assemblages (illite and chlorite) and more intense mineral replacement, reflecting higher heat flow, enhanced permeability, and proximity to primary upflow zones driven by intracaldera magmatic activity. Conversely, well OW-922 is characterized by a deeper smectite-to-illite transition zone and a dominance of low-to-moderate temperature clays, consistent with lower thermal gradients and lateral fluid outflow towards the extracaldera margins. These mineralogical distinctions elucidate the subsurface hydrological and thermal framework of Olkaria, offering critical criteria for refining the conceptual reservoir model, delineating clay-cap geometry, and optimizing future drilling targets.

Topic: Geology

Kalimantan hosts extensive sedimentary basins and crystalline basement complexes but lacks the widespread active volcanism that characterizes Indonesia's principal conventional geothermal provinces. Consequently, geothermal development in the region remains limited, despite increasing electricity demand associated with urban expansion and the development of Indonesia's prospective new capital, Nusantara. This study investigates whether unconventional geothermal systems, particularly Hot Dry Rock (HDR) and Hot Sedimentary Aquifer (HSA) resources, can provide additional geothermal development opportunities in Kalimantan. A play-based spatial screening framework is developed to identify areas where the geological architecture and thermal regime satisfy the fundamental requirements of each unconventional resource type. The assessment integrates geological and structural information, including lithology, stratigraphy, basement configuration, and fault systems, with measured temperature data and regional thermal indicators. Additional geological and crustal proxies are incorporated to characterize the subsurface thermal regime and estimate the distribution of temperatures at depth. For HDR systems, the analysis emphasizes heat-producing crystalline basement, thermal conditions, basement depth, and structural characteristics that may support engineered reservoir development. For HSA systems, the screening focuses on thermally favorable sedimentary basins, reservoir depth and stratigraphy, potential permeability pathways, and geological conditions favorable for fluid circulation and heat extraction. The resulting play-fairway maps distinguish areas of favorable HDR and HSA conditions and highlight their contrasting geological controls across Kalimantan. Preliminary results indicate that favorable unconventional geothermal settings extend beyond the currently recognized conventional geothermal occurrences, with HDR and HSA plays occupying distinct geological domains. The identified fairways are further evaluated in relation to existing and planned electricity infrastructure to assess their potential contribution to future geothermal development in Kalimantan, including areas surrounding the future Nusantara capital region.

Topic: Low Temperature

Enhanced geothermal systems (EGS) aim to harness high temperature formations enabling enhanced grid scale power production. The composition of geothermal working fluids vary substantially among different geothermal wells containing several corrosive chemical species such as oxygen, hydrogen ion, chloride ion, sulphate ion, sulfide, carbon dioxide, ammonia, etc. Steel, conventionally used as casing material, has limited resistance against high temperature harsh environments and experiences corrosion, loss of mechanical strength, scaling, and high thermal expansion, ultimately leading to early casing failure. Again, due to continuous high temperature exposure of EGS system above 350 °C, high cost engineering polymers are also unfavorable, attributing to the thermos-oxidative decomposition of organic molecules. This leaves ceramic matrix composites containing polymer derived ceramics such as siloxane derived silicon oxycarbide (SiOC) uniquely suited as both a high temperature stable and low cost material that can be formed into EGS casing. Furthermore, low temperature pryolysis required to convert polymer into amorphous ceramics allows the use of low-cost fiber reinforcement (such as glass or basalt fiber). The resulting amorphous ceramic matrix composites can maintain stability under geothermal temperature while suppresses the energy requirements for manufacturing and thus the production cost as compared to high temperature sintering process. Here, we investigate the effect of catalyst, crosslinker, and filler on key CMC properties including density (porosity), ceramic conversion efficiency, and corrosion in relevant EGS conditions.

Topic: Enhanced Geothermal Systems

Kilohertz frequency microseismic emissions recorded by km-deep sensors explicate Engineered/Enhanced Geothermal System (EGS) flow stimulation needed to commercially extract ambient crust heat. Deep-sensor data from the Finland 2016-2020 and UtahForge 2022-2026 EGS projects find: (i) emission spectral peaks at kHz frequencies; (ii) bidirectional emission first motions; (iii) lognormal emission amplitude distributions; (iv) two-point emission spatial correlation GR) ~ 1/R, R = pairwise emission source offsets. These findings disqualify smooth slip on fault-like planes in an elastic continuum. They instead posit radial rough slip within ambient crust poro-permeability structures attested worldwide by ambient crust well-log, well-core, and well-flow data. Further, and importantly, deep-sensor sub-millisecond emission arrival times open the way to map emission source locations with m-scale precision. This paper replicates deep sensor emissions to simulate EGS monitoring on the scale of the UtahForge 2024 EGS cross-well flow V ~ 30 L/s extracting Q ~ 20MWth heat energy. The model well scheme comprises a 200m diameter ring of 500m-reach horizontal producer wells centered on a parallel injector well. EGS stimulation occurs between the central injector and the bounding quartet of producers equipped with acoustic sensors. The bounding ring of sensors detect EGS emission travel-times by which to locate emission sites in time and space. At MWh electrical energy pricing, five-year power production of the UtahForge EGS 2024 model is comparable to the drilling costs.

Topic: Enhanced Geothermal Systems

The investigation of hydraulic stimulation and reservoir circulation at the Newberry Enhanced Geothermal System (EGS) is performed using field data from the multi-stage stimulation of wells 55-29 and 55A-29. A pressure-constrained hydraulic fracture growth model is used to estimate fracture radius and aperture through fluid mass balance, accounting for fracture volume and pressure-dependent leakoff. The measured surface treating pressure is corrected for hydrostatic pressure and flow-dependent wellbore, perforation, and near-wellbore pressure losses to calculate the fracture fluid pressure. Net pressure is then related to fracture aperture using an elastic opening relationship. Field observations are used to constrain key model parameters. Stage-specific fluid efficiencies are used to calibrate formation leakoff and matrix permeability, while pressure responses measured in Well 55-29 during stimulation of Well 55A-29 are used to characterize interwell hydraulic connectivity. The field responses show strong stage-dependent connectivity, with significantly stronger connectivity for Stages 4 and 5 than for the remaining stages. The calibrated hydraulic fractures are then coupled with a discrete fracture network model to analyze pressure distribution, shear dilation, and stimulation-induced permeability enhancement. The resulting permeability fields are used in reservoir circulation models to investigate the relationship between stimulation-created fracture connectivity and post-stimulation reservoir performance.

Topic: Enhanced Geothermal Systems

Steamboat 28A-32 was drilled in the Steamboat geothermal field, Nevada, to access a fracture-controlled geothermal reservoir identified from nearby production wells. During drilling, the well encountered substantial lost circulation exceeding 100,000 barrels, indicating extensive reservoir permeability and strong hydraulic connectivity within the target interval. Despite these encouraging drilling indicators, subsequent flow and injection testing revealed lower-than-expected well performance, suggesting significant near-wellbore impairment. The diagnostic program integrated drilling observations, flow testing, injection testing, pressure transient analysis (PTA), pressure-temperature-spinner (PTS) surveys, and feedzone interpretation. Initial PTA results indicated a skin factor as high as 20, while spinner surveys identified multiple permeable feedzones contributing to flow across the production interval. The combination of high drilling losses and elevated skin suggested that the well's limited productivity was primarily controlled by near-wellbore damage rather than reservoir quality. Based on these findings, an acid stimulation treatment was designed and executed to improve hydraulic communication between the wellbore and the surrounding fracture network. The results demonstrated a substantial improvement in injectivity following the stimulation. PTA analysis further indicated a significant reduction in skin, confirming that the stimulation successfully removed a significant portion of the near-wellbore restriction. Independent evaluation from step-rate analysis and transient response showed consistent improvement in well performance. This paper presents the integrated workflow used to diagnose reservoir impairment, design the stimulation program, and quantify post-treatment performance improvements.

Topic: Field Studies

The geothermal power industry’s focus in the western United States has shifted over the last half-decade from exploring for and developing hydrothermal resources to constructing enhanced geothermal systems (EGS). This shift has been driven primarily by two factors: (a) the enormous progress made toward improving EGS economics by utilizing hydraulically fractured horizontal wells and other oil industry technologies, and (b) the perception that commercially viable hydrothermal systems are relatively rare, hard to find, and usually small compared to enhanced geothermal systems. In this paper, we present evidence indicating that negative perceptions about the prevalence and size of economically viable hydrothermal systems are likely an artifact of the use of vertical to moderately deviated well designs to explore for and develop these systems rather than from inherent physical scarcity. Namely, we note that nearly all known higher temperature geothermal systems in the Great Basin of the western United States exist in structurally complex settings – settings containing multiple overlapping and/or intersecting critically stressed faults. In such settings, some faults may contain highly permeable damage zones that can act as conduits for hot fluids flowing upward from deeper, warmer levels within the crust. However, determining where such faults are located is challenging. Given this reality, deciding where to drill a vertical or modestly deviated well to intersect an economically viable permeable pathway is problematic and fraught with peril, since the permeable sweet spots could be easily missed and thereby mistakenly condemn a prospect that contains viable pathways nearby. To avoid such outcomes, horizontal or highly deviated wells should be routinely incorporated into geothermal drilling programs to explore for, appraise, and ultimately develop hydrothermal systems. This is an approach that has served the oil & gas industry extremely well in fractured reservoirs, with horizontal wells dramatically increasing both success rates and per well production rates compared to vertical and moderately deviated well designs. It is also an approach supported by results from the drilling campaigns at the Utah FORGE site and adjacent Cape Station project, where horizontal and high-angle wells have encountered numerous conductive faults while vertical wells did not. This suggests that conductive fault damage zones are far more prevalent in the Great Basin region and other parts of the western United States than is generally recognized. It also demonstrates the usefulness of horizontal wells for intersecting steeply dipping faults and fractures that are hard or impossible to image prior to drilling. Combining these observations leads one to conclude that the hydrothermal resource endowment of the western United States has likely been significantly underestimated. It is therefore hoped that the recent excitement surrounding EGS related technological advancements will not cause geothermal power developers to lose sight of the potential associated with hydrothermal systems, especially as: (a) horizontal well designs are likely to become a common element underpinning the economic optimization of both EGS and hydrothermal systems, and (b) many geothermal anomalies will likely be best developed using a combination of these two approaches.

Topic: Emerging Technology

Enhanced geothermal systems have attracted increasing attention for the development of deep hot dry rock geothermal resources. Water is generally considered the preferred working fluid for extracting heat from hot dry rock reservoirs. CO2 can serve as an alternative working fluid while providing a pathway for carbon capture and utilization. However, whether CO2 provides economic advantages over water remains uncertain. This study establishes a three-dimensional coupled thermo-hydraulic numerical model to simulate an enhanced geothermal system with dimensions of 300 m × 300 m × 300 m over a 30-year operational period. Discrete fracture networks generated through Monte Carlo simulations are coupled with an equivalent porous medium model to represent fractured reservoirs. For reservoir conditions where temperature exceeds 31 °C and pressure exceeds 7.38 MPa, CO2 exists in a supercritical state. A life-cycle techno-economic assessment framework is developed, covering capital expenditures such as well drilling, surface power plant construction, pressurization equipment, and working fluid procurement, as well as geothermal power generation revenue, to calculate the levelized cost of electricity and discounted payback period. The results indicate that under the investigated reservoir conditions and operating parameters, for the same working pressure, the specific heat capacity of water is 2.8 times that of CO2. However, CO2 has much lower viscosity and achieves a mass flow rate of 17.5 kg/s, compared with only 3.4 kg/s for water. Thermal breakthrough occurs at Year 4 for the CO2 system, earlier than Year 5.2 for the water system. Nevertheless, the CO2 system maintains a higher thermal extraction rate throughout the entire operation period. Economic calculation shows that the upfront capital investment of the CO2 system is approximately twice that of the water system. However, higher annual electricity generation revenues, resulting in a discounted payback period of only 15.8 years, compared to 18.6 years for the water system. The minimum levelized cost of electricity of the CO2 system reaches 0.32 CNY/(kW·h), less than half of the water system's value of 0.67 CNY/(kW·h). Overall, under the specific reservoir parameters and operating conditions investigated in this study, CO2 has better techno-economic performance than water as a heat extraction fluid in the studied hot dry rock enhanced geothermal system, indicating its potential as a cost-effective pathway for geothermal power generation. These findings provide quantitative guidance for working fluid selection in hot dry rock reservoir development.

Topic: Enhanced Geothermal Systems

The long-term performance of a multistage enhanced geothermal system (EGS) depends on the ability of individual fractures to extract heat from their surrounding rock volumes. Conventional one-dimensional half-space models neglect lateral heat recruitment beyond the active fracture footprint, whereas infinite-domain models do not represent depletion of the finite tributary volume assigned to each fracture. These assumptions can produce substantial and opposing errors as fracture geometry, stage spacing, and operating time change. This study develops a reduced-order framework for quantifying conductive heat transfer between a thermally active fracture and a finite anisotropic rock matrix, with direct application to multistage fractured geothermal reservoirs. Each fracture face is represented by a finite axisymmetric matrix cell containing a circular active contact. The model accounts for aligned thermal anisotropy, insulated outer boundaries, and a prescribed area-averaged contact temperature. The resulting boundary-value problem is reduced to a Laplace-domain thermal impedance. Responses to maintained temperature conditions are obtained by numerical inverse Laplace transformation, while a direct spectral-state formulation evaluates arbitrary temperature histories without repeated convolution. The behavior is described primarily by dimensionless variables including the cell-to-contact radius ratio, the effective thermal aspect ratio, and the normal Fourier number. These scaling variables separate the effects of fracture contact area, stage spacing, anisotropy, thermal storage, and operating duration, allowing results to be transferred across reservoir scales and used efficiently in system design and optimization. The illustrative example is designed to identify controlling heat-transfer regimes and quantify errors introduced by simplified matrix models, rather than reproduce a specific field history. The analysis is therefore presented first in dimensionless form. The resulting response maps show when lateral heat recruitment increases fracture heat supply and when finite-volume depletion becomes dominant. A dimensional baseline with a 24 m symmetric fracture spacing is then used to demonstrate the engineering interpretation. For an 80 K maintained mean temperature deficit, the model extracts 27.06 TJ after 30 years, corresponding to 50.9% of the assigned thermal inventory. The one-dimensional half-space approximation underpredicts this result by approximately 43% because it excludes lateral heat supply. Comparison with a finite-volume simulation gives differences of 0.033% in heat rate and 0.117% in cumulative energy for the finest tested mesh. The spectral solver also reproduces thermal memory and heat-flow reversal under a nonmonotonic temperature history. For multistage EGS applications, the finite-cell response provides a stage-level thermal kernel that can be coupled with fracture-flow and reservoir models. Dimensionless response maps enable rapid screening and optimization of stage spacing, thermally active area, conductivity anisotropy, and operating duration. Extension to hydraulically communicating fractures, unequal stage loading, and coupled production-temperature prediction provides a path toward practical reservoir-scale design.

Topic: Reservoir Engineering

Cold-water injection is essential for sustaining geothermal reservoir pressure and recovering heat, but the resulting thermal disturbance can substantially alter the mechanical and hydraulic behavior of the near-wellbore formation. Cooling induces thermal contraction of the rock matrix, reduces effective normal stress acting on fractures, and may increase fracture aperture and permeability. These changes can improve injectivity and are sometimes described as thermal or cold stimulation. At the same time, the reduction in compressive stress may lower the fracture-initiation or fracture-reactivation pressure, thereby reducing the allowable injection-pressure margin. This study develops a reduced-order thermo-hydro-mechanical framework to quantify these competing effects and determine how the safe injection window evolves during prolonged geothermal injection. A transient radial injection model is developed in which fluid and rock temperatures are treated separately to account for possible local thermal nonequilibrium at high injection rates. The thermal solution incorporates advective transport by the injected water, heat exchange between fluid and rock, and radial conduction through the formation. The calculated rock-temperature field is coupled to a thermo-poroelastic stress model to determine changes in effective normal stress. Fracture aperture and permeability are related to effective stress through a stress-sensitive constitutive relationship, allowing the permeability field to evolve as the cooled region expands. The resulting permeability distribution is then fed back into the radial pressure equation to calculate transient injectivity and required bottomhole injection pressure. In parallel, the thermoelastic reduction in minimum compressive stress is used to update the fracture-pressure limit. Semi-analytical and successive-coupling solutions are used to retain the dominant physical relationships without requiring full numerical reservoir simulation. The analysis is expected to identify three characteristic stages of cold-water injection. Initially, cooling is concentrated near the well and produces a rapid thermoelastic reduction in fracture-normal stress. As the cooled region expands, fracture aperture and permeability increase, reducing the pressure required to maintain a specified injection rate. At longer times, the stimulated permeability region grows while the fracture-pressure limit may continue to decline. The relative rates of these two processes determine whether the safe injection-pressure margin widens or narrows. Parametric results will quantify the effects of injection rate, temperature contrast, formation permeability, fracture stress sensitivity, thermal diffusivity, interphase heat transfer, Young's modulus, Poisson's ratio, and thermal expansion coefficient. Regime maps will identify conditions under which cooling produces negligible hydraulic change, beneficial thermal stimulation, or a fracture-pressure constraint.

Topic: Reservoir Engineering

In geothermal wells, CO₂-bearing, chloride-rich fluids and high temperature can lead to coupled cement carbonation and steel casing corrosion, threatening long-term well integrity. Cracks in conventional brittle well cement provide preferential pathways for these aggressive fluids (chloride-rich) to reach the casing, highlighting the need for cementitious barriers with improved crack-width control. This study developed a series of mix designs for ductile well cements using American Petroleum Institute (API) Class H cement, silica flour, and fly ash, reinforced with polyethylene (PE) fibers to address the inherent brittleness of traditional oil well cements. The mechanical behavior and microstructural evolution of the composites were evaluated under four conditions: ambient curing; carbonation curing (10% CO₂); chloride exposure (14 days in 5% NaCl solution); and elevated-temperature curing (120°C). The composites were characterized using uniaxial tensile testing, thermogravimetric analysis, and X-ray diffraction. Results indicated that PE fiber reinforcement transformed the failure mechanism from brittle fracture to ductile strain hardening. Carbonation curing stabilized the strain-hardening behavior across all test ages, effectively preventing the late-age embrittlement observed in ambient-cured specimens. The material remain ductile after chloride exposure. An interpretable supervised machine learning framework was further employed to establish quantitative links between hydration- and carbonation- and chloride exposure driven microstructural evolution and the resulting strength and ductility. These findings demonstrate that fiber-reinforced Class H composites offer strong potential to improve zonal isolation in CO₂ and chloride-rich geothermal environments.

Topic: General

Rock fracturing and fluid flow are central to enhanced geothermal systems and hydrocarbon recovery from tight formations. Current hydraulic fracturing theory predicts planar fracture propagation normal to the minimum principal stress in homogeneous media. Fracture complexity, when presents, is often attributed to either heterogeneous in-situ stresses or structural geologic discontinuities. However, highly instrumented laboratory experiments and many field tests often demonstrate that fractures are far from planar with a high level of complex curvature, under conditions expected to favor planar propagation. Here, we report repeated experimental evidence of curving and branching of fluid-driven fractures in homogeneous, porous rock-analogs subjected to anisotropic three-dimensional boundary stresses. The fracturing process was video-recorded, enabled by a custom-designed transparent true triaxial cell. To investigate the mechanisms underlying these observations, we performed numerical simulations using a material point method framework that supports mesh-independent fracture propagation and physics-based crack curving and branching. The results identify T-stress, the nonsingular in-plane stress term in the crack-tip stress field, as a key contributor to fracture complexity. Our findings suggest that fracture criterion may need to be revised to incorporate both fracture toughness and T-stress, in order to predict fracture propagation paths and explain complex hydraulic fracture geometries in the subsurface.

Topic: Enhanced Geothermal Systems

Safe development of enhanced geothermal systems (EGS) requires managing induced seismicity and aseismic fault slip, which depends on understanding hydraulic fracture–fault interactions and the evolution of fault-zone hydraulic properties during stimulation. We investigate these processes using a dataset from the multistage stimulation of horizontal wells at Cape Station project, Utah, integrating injection schedules, wellhead pressures, stress measurements, and microseismicity. Fault activation begins when hydraulic fractures intersect a fault and introduce localized fluid-pressure sources that drive pressure diffusion, weakening, and slip on the fault. A rapid pressure increase recorded in an offset observation well approximately 300 m from the stimulated well indicates that hydraulic communication with the pressurized fault was established due to fault slip. Negligible cumulative seismic slip at that time indicates that this slip was predominantly aseismic. We model fluid flow from hydraulic fracture–fault intersections through the fault zone into the offset well. Matching the observed offset-well pressure transient constrains fault-zone diffusivity, transmissibility (permeability multiplied by thickness), and overpressure. The pressure peak is dominated by diffusion from the nearest fracture–fault intersection. Its timing and amplitude jointly constrain diffusivity and fracture closure time. Inversion gives diffusivity of 0.01 m²/s. The fractures remain as effective pressure sources for about one day after shut-in, until leak-off closes them. Independently, the subsequent 10-day pressure decline is dominated by along-fault diffusion, rather than leak-off into the reservoir, and yields the same diffusivity. Migration of the seismicity triggering front suggests a consistent range of 0.004–0.032 m²/s. A series of discharge events were conducted at the wellhead, and the pressure recovery after each event constrains transmissibility to 10^-16–10^-15 m³. The inferred transmissibility increases by a factor of 2–3 over several days, suggesting slip-induced permeability enhancement. In summary, our interpretation indicates a low-permeability fault zone locally pressurized by 15–25 MPa around fracture–fault intersections. The pressurization drives slip and potentially fault opening or dilation. Three-dimensional simulations successfully reproduce both the offset-well pressure history and the observed seismicity migration. Offset-well pressure transients can therefore constrain fault-zone pressurization, hydraulic properties, and their evolution during EGS stimulation, even where downhole instrumentation is unavailable.

Topic: Enhanced Geothermal Systems

In geothermal well stimulation, where hydraulic fractures are intentionally created to develop high-permeability fluid pathways, LFDAS is a promising tool for observing both fracture opening and aseismic fault slip. To date, however, most inferences have been qualitative detecting, for example, whether a hydraulic fracture has reached a fiber-instrumented well, without recovering quantitative opening profiles or fracture orientations. Recent studies have performed time-lapse inversions of fracture opening in reservoir and magmatic-dike settings, but these rely on LFDAS data alone and provide no quantification of uncertainty. We present a physics-constrained, multi-modal inversion framework that jointly integrates LFDAS strain rates, fluid injection histories, and downhole pressure measurements to resolve spatiotemporal fracture opening and fault slip. We parameterize spatiotemporal front evolution using an arrival surface level set (ASLS) that represents the arrival time of the fracture front over the entire fracture plane. The inversion promotes physical plausibility by enforcing tip asymptotic solutions from fracture mechanics, positive apertures, monotonic front growth, and volume conservation directly within the model space rather than through ad hoc regularization. This mechanics-informed formulation reduces parameter dimensionality relative to conventional grid-based discretizations. The inversion framework allows quantification of posterior uncertainty through Monte-Carlo sampling. We validate the framework on physically consistent planar fracture simulations exhibiting non-radial front growth, and test it against field data from the Fervo Energy Enhanced Geothermal System (EGS) project in Utah. The field inversion reveals that severe non-uniqueness is driven by limited angular coverage of borehole fibers, high parameter-to-observation ratios, and single-component tangential strain rate recordings. The inversion framework demonstrates the capability to propose optimized fiber deployment geometries that better regularize the inverse problem, advancing seismic hazard mitigation. Our approach opens the door to the design and testing, with quantitative metrics, of optimal fracture dynamic imaging and monitoring needed for long-term energy-producing EGS projects.

Topic: Enhanced Geothermal Systems

Sustaining commercial circulation rates while minimizing pumping requirements poses a fundamental challenge for enhanced geothermal systems (EGS). During prolonged circulation at the Utah Frontier Observatory for Research in Geothermal Energy (FORGE), an injection rate of approximately 9.5 bbl/min into Well 16A(78)-32 necessitated an early-time surface injection pressure of around 2,780 psi. This significant pressure requirement constitutes a considerable parasitic load. Reducing this power requirement requires comprehending the causal mechanisms. However, the relative contributions of near-wellbore restrictions, fracture conductivity, and heterogeneous interwell flow remain unclear. This study employs numerical simulation and systematic sensitivity analysis to quantify the primary factors influencing parasitic hydraulic losses during circulation between a generic injector and producer, such as the system at Utah FORGE. Three mechanisms are examined. First, near-wellbore tortuosity and fracture complexity are systematically altered to assess their impact on hydraulic impedance, circulation pressure, and pumping requirements. Second, interwell fracture conductivity is analyzed using quantified proppant-concentration pumping schedules that range from low to high proppant loading. This establishes relationships among proppant placement, resulting fracture conductivity, and parasitic hydraulic loss. Third, flow-path localization is explored by comparing relatively uniform interwell flow networks with increasingly localized preferential flow channels across a spectrum of injection rates. These simulations demonstrate how the concentration of flow into a limited number of conductive pathways influences local flow velocity, pressure gradients, and system-level hydraulic losses. The resulting simulation ensemble undergoes statistical analysis to quantify the relative impacts of near-wellbore complexity, fracture conductivity, flow-path localization, and injection rate on parasitic hydraulic losses. Pearson correlation coefficients, along with complementary sensitivity metrics, are employed to rank the controlling parameters and to identify nonlinear or interacting effects. This analysis differentiates localized near-wellbore restrictions from distributed pressure losses within the interwell fracture network and assesses the conditions under which increasing fracture conductivity leads to diminishing reductions in circulation pressure. The study presents a quantitative framework that connects stimulation design, proppant placement, fracture-network architecture, and circulation rate to hydraulic impedance and parasitic pumping power. This framework offers guidance for enhancing circulation efficiency and optimizing net energy generation in future Utah FORGE operations and commercial-scale EGS developments.

Topic: Enhanced Geothermal Systems

Outflows are the regions of lateral convection/advection away from deep upflow zones of geothermal systems that are a function of buoyant hot fluid reaching the water table or shallower permeability features that initiate more horizontally-direct flow. These features are intrinsic components of liquid-dominated convective geothermal systems, and many initial indications of geothermal systems that can be detected at the surface, commonly including thermal manifestations and a significant fraction of low resistivity caps, are more closely related to outflows rather than to upflows. Outflow attributes are controlled by, and provide insight into, the subsurface permeability structure, fluid saturation state, interaction with adjacent marginal fluids, regional hydrology, and natural state mass flux, and their characterization provides valuable information that can be applied to vectoring upflow locations during exploration. Additionally, while outflows are common targets for injection, high-, medium-, and low-temperature geothermal outflows are also targets for production in many developed systems around the world. Some notable advantages of targeting outflows for either injection or production include the shallower associated target well depths as well as the elevated matrix or fracture-based permeability that is commonly encountered in outflows. Outflow parameters such as temperature, lateral temperature gradient, depth, thickness, area, directionality, chemistry, permeability, flux, quantity, and termination style all affect their development potential for production. Twenty well-characterized reservoir case histories from around the world are included to illustrate the parameters for outflows from producing and drilled but undeveloped geothermal fields and to support recommendations for terminology to assist with outflow characterization. The distribution of manifestations, trends in geochemistry, geophysical data, and temperature profiles are included to illustrate the importance of considering outflow properties into conceptual models and numerical simulation sensitivity analyses are performed to illustrate the influence of mass flux and permeability contrasts on outflow geometries.

Topic: Geology

Direct energy drilling using Millimeter Wave (MMW) technology is an ablation process which generates a dusty plume of rock fragments or vapor. From field measurements, the hot dust cloud has a mean particle diameter of 100 µm. Although scattering losses are limited for such fine particles, prolonged residence may increase local particle concentrations and promote plasma breakdown. Thus, the plume must be rapidly quenched and removed from the ablation cavity, both to maintain exposure of fresh rock and to limit particulate accumulation within the high-field region. While one-dimensional wellbore models are capable of describing particle transport through the annulus, gas–particle dynamics within the ablation cavity remain unresolved and must otherwise be represented through prescribed closure conditions. To address this, a compressible Eulerian–Eulerian multiphase model is applied to study the evolution and transport of a dusty plume within a representative MMW ablation cavity. The model captures viscous gas–particle interactions in a confined geometry across dilute to dense particle concentrations and has been validated across subsonic to hypersonic flow conditions. These simulations can then be used to characterize entrainment, recirculation, accumulation, and particle residence time within the cavity, and to quantify the particulate flux entering the return annulus.

Topic: Modeling

This paper presents a time-domain series approximation, herein termed the erfc solution, for the Gringarten et al.\cite{1975gwo} geothermal reservoir model involving multiple parallel vertical fractures. The original model requires numerical inversion from the Laplace domain, while this approximation offers an analytically transparent alternative derived through a linear expansion of the exponential function. The time-domain indicates clearly how the numerous parameters of the problem enter into the solution. Using parameters from the Fervo Blue Mountain project, we evaluate the accuracy of this approximation against the exact Laplace solution. Results indicate that the dimensionless Fourier number serves as a critical threshold for solution validity. The erfc solution is highly accurate for short-term simulations where thermal interaction is weak. However, it significantly overestimates produced fluid temperature and electrical power output as thermal interference becomes strong over extended production times. This work provides a practical tool for early-stage geothermal well design and reservoir cooling forecasts.

Topic: Modeling

Drilling long inclined/horizontal wells in hard, crystalline rock remains one of the principal economic barriers to next-generation geothermal systems. Downhole Drive Unit (DDU) addresses these chalenges directly – modular, stackable downhole units grip the formation and apply thrust and torque at the bit itself, independent of drillstring weight, under closed-loop control electronics capable of modulating WOB with high frequency. DDU has now completed three sequential field-test phases in 2026 at the NORCE Ullrigg Research and Test Centre in Stavanger, Norway, demonstrating that hydraulically gripping the borehole wall to react weight-on-bit (WOB) and torque directly into the formation is a viable route to overcoming the two dominant barriers to economic hard-rock geothermal drilling: insufficient, poorly controlled WOB delivered through a long, compliant drillstring, and torsional stick-slip dysfunction that caps achievable rate of penetration (ROP). Across the programme, the system delivered up to 32,000 lbf of controlled downhole thrust while the drillstring above continued to rotate, drilled hard Phyllite and operated also in oversized and irregular borehole geometry. The Ullrigg programme progressed from single-unit gripping and drive validation through synchronised dual-unit drilling towards drilling with downhole motor. A phase planned for October 2026, will run a direct side-by-side with and without the system in harder and deeper formation to isolate the performance uplift attributable to the technology, with results available ahead of the workshop. These results speak directly to the drilling economics that constrain deep and enhanced geothermal development. The paper presents the full 2026 Ullrigg field validation dataset, the engineering challenges encountered and resolved between phases, and the outlook for commercial deployment as an enabling technology for cost-effective drilling of deep, hot, hard-rock geothermal wells.

Topic: Drilling

Geothermal resource exploration along the eastern Castle Mountain Fault (ECMF) in Alaska requires methods for translating fault outcrop observations into subsurface models beneath Quaternary cover. This work is part of a U.S. Department of Energy-funded project awarded to the University of Alaska Anchorage, with university and state data collection supporting geothermal reservoir modeling and exploration by Teverra. A temperature anomaly occurs along the ECMF in an area covered by hundreds of meters of sediments, making it difficult to determine the underlying lithology and fault architecture. A bedrock exposure of the ECMF approximately 38 km along strike to the east provides an opportunity to examine a structural analog at the surface, but the relationship between this exposure and the covered structure and lithology remains uncertain. This study will compare existing published methods for using structural outcrop observations to produce geothermal reservoir model inputs. We will assess how measurements of fault architecture and fracture properties can be translated below the Quaternary cover and into model properties or conceptual scenarios. For example, this could involve constructing discrete fracture networks to produce a permeability tensor or developing alternative geological interpretations for modeling. We will develop a decision framework to determine which approaches are most appropriate for the ECMF, or how they could be integrated. The comparison will account for each approach’s data requirements, assumptions, uncertainties, practical feasibility when applied to the ECMF, and how useful they are for the final reservoir model. Special consideration will be given to fracture mineralization and aperture, lithological uncertainties, and field access. The intended outcomes for this study are an assessment of existing methods, a recommended workflow, and priorities for future fieldwork and subsurface data collection. The framework will determine what our outcrop observations can contribute to the reservoir model, what assumptions are needed, and what additional information is needed. The recommendations would provide a starting point for testing and refining a workflow as the project develops.

Topic: Geology

Time-dependent deformation of fractured crystalline rock can influence stress redistribution, fracture compliance, and the persistence of permeability in enhanced geothermal systems. However, the dissipative mechanical response of geothermal reservoir rocks remains poorly constrained across the temperatures and loading timescales relevant to reservoir operations. We investigate this response using low-frequency torsional forced oscillation measurements on granitoid core from the Utah Frontier Observatory for Research in Geothermal Energy (FORGE). Experiments were conducted under nominally dry conditions at temperatures from 25 to 300 °C and frequencies from 0.01 to 100 Hz. Sinusoidal shear loading at a nominal strain amplitude of 10⁻⁵ was applied after thermal equilibration at each temperature. The amplitude and phase relationship between applied torque and deformation were used to determine storage and loss shear moduli and shear attenuation. These measurements probe both recoverable stiffness and energy dissipation associated with deformation of the rock framework. We develop a fractional viscoelastic framework to interpret the measurements in terms of a distribution of relaxation times within a heterogeneous population of microcracks and grain contacts. By combining temperature-dependent forced oscillation measurements with an interpretable constitutive framework, this study examines how local crack and contact processes can produce an effective viscoelastic response in geothermal granitoids. The approach provides a basis for evaluating frequency-dependent stiffness and dissipation and for developing laboratory constraints on models of stress relaxation in enhanced geothermal reservoirs.

Topic: Geophysics

The thermal performance of enhanced geothermal systems (EGSs) can be substantially degraded by premature thermal breakthrough caused by preferential fluid pathways and hydraulic short-circuiting. This study provides a critical review of recent techniques for alleviating thermal breakthrough in EGS reservoirs and outlines emerging research opportunities. The review is intended to highlight advanced methods with demonstrated or potential effectiveness, clarify the principal technical limitations that remain unresolved, and establish perspectives for future investigation. Distinct from earlier reviews, this work synthesizes recent advances from multiple disciplines and systematically examines the relationships, differences, advantages, and limitations among major thermal breakthrough mitigation strategies. Five principal approaches that have been proposed or investigated are assessed: optimization of fluid-flow management, intermittent heat-production schemes, application of carbon dioxide (CO2) as a circulating fluid, modification of fracture conductivity, and implementation of temperature-responsive fluids. Results reported in the literature indicate that regulation of circulation rates together with real-time reservoir monitoring and operational control can significantly enhance EGS thermal performance; however, such measures may increase operational expenditures and potentially elevate induced-seismicity risks. Intermittent heat extraction, involving alternating production and shut-in periods, facilitates thermal recharge of the reservoir and consequently postpones the deterioration of production-fluid temperature. The use of CO2 as a heat-transfer medium may improve fluid distribution and expand effective heat exchange throughout fracture networks, thereby increasing recoverable thermal energy. Temperature-sensitive proppants provide another promising mechanism by adaptively modifying fracture flow capacity in response to local thermal conditions. Similarly, temperature-responsive fluids can increase flow resistance within cooled regions and preferentially divert circulation toward hotter fractures, resulting in a more spatially uniform thermal sweep. Promising future directions include the development of nanofluid-based heat-transfer systems and reactive-tracer technologies for improved reservoir characterization and control. Ultimately, combining complementary mitigation mechanisms within the proposed multi-level framework may enable an integrated strategy for managing thermal breakthrough and sustaining long-term EGS heat production.

Topic: Enhanced Geothermal Systems

Heat extraction in traditional closed-loop geothermal systems is limited by the small heat exchange area between the rock and wellbore. To overcome this limitation, an augmented closed-loop geothermal system (ACLGS) is proposed in this study to improve heat production. The ACLGS integrates a hydraulic fracture, segmented by a horizontal isolator to achieve zonal isolation of fluid flow within the fracture, into the closed-loop system's circulation. A three-dimensional numerical model of the ACLGS was developed and validated to simulate heat extraction under various conditions, including different fracture geometries, rock permeability, and temperature dependence of rock thermal properties. The performance of smart working fluid, with automatic response to temperature, in improving heat extraction is also explored. The effectiveness of complex fracture designs, including branched and multi-wing fractures, in boosting heat production was also investigated. Results show that incorporating a double-wing fracture significantly increased cumulative heat extraction by more than 150% over a 20-year period. Both fracture half-length and height were found to have a substantial impact on the system's heat production. A branched fracture improved cumulative heat extraction by above 320% over 20 years, and incorporating interlaced sub-branches can further enhance heat extraction. The system’s thermal performance is better when considering temperature-dependent rock thermal properties than when assuming constant properties. Notably, smart fluid with strong temperature responsiveness can noticeably mitigate thermal breakthrough of the ACLGS, thus improving long-term heat extraction performance.

Topic: Emerging Technology

Long-term geothermal well integrity depends critically on the durability of the cement-metal interface, where coupled hydrothermal reactions, corrosion, and thermomechanical mismatch can initiate debonding and compromise zonal isolation. Here, we systematically compare cement-metal interfacial degradation across multiple cement formulations under high-temperature hydrothermal and thermomechanical conditions. Synchrotron X-ray diffraction and three-dimensional X-ray tomography were combined with mechanical and microstructural characterization to relate phase evolution to pore, crack, and interfacial damage development. The results reveal pronounced formulation- and exposure-dependent differences in the extent and morphology of interfacial degradation. Across the investigated systems, degradation manifests through varying combinations of interfacial debonding, localized cracking, bulk matrix damage, and reaction-driven microstructural reorganization. Correlative phase and structural analyses show that phase evolution and three-dimensional damage morphology are associated with these contrasting failure behaviors. These results establish a comparative framework for linking cement formulation and exposure history to interfacial failure behavior, providing guidance for the selection and design of more durable cement-casing systems for high-temperature geothermal wells.

Topic: General

Matrix acidizing is a widely employed stimulation technique for enhancing productivity in carbonate geothermal wells, yet the fundamental mechanisms governing acid-rock interactions under high-temperature reservoir conditions remain unclear. Temperature exerts complex influences on reaction kinetics, mass transport, and fluid rheology through multiple pathways, posing significant challenges for optimizing treatment design. This study systematically investigates the temperature-dependent mechanisms that govern acid stimulation performance in geothermal reservoirs using an integrated experimental and numerical modeling approach. High-temperature rotating disk apparatus experiments were conducted to systematically determine acid-rock reaction kinetic parameters under different temperature conditions, obtaining key data including reaction rate constants, hydrogen ion diffusion coefficients, and apparent activation energies. Based on experimental results, a mathematical acidizing model incorporating temperature field evolution was established through parameter fitting. This model couples chemical reaction kinetics, hydrogen ion mass transport, fluid flow, and heat transfer within a unified multi-field coupling framework. Furthermore, numerical simulations of high-temperature matrix acidizing were performed to elucidate the dynamic evolution of acid-etched pore structures and stimulation mechanisms under multi-field synergistic effects. Results reveal that temperature elevation exerts typical nonlinear effects on the acidizing process. Both acid-rock reaction rates and hydrogen ion diffusion coefficients exhibit exponential growth with increasing temperature; however, the temperature sensitivity of reaction rates is higher than that of diffusion coefficients, and their competitive relationship undergoes dynamic transitions with temperature variations, thereby influencing the selection of optimal operational parameters. For non-Newtonian acid systems including gelled acids and VES-based acids, temperature exerts significant effects on fluid rheology. In summary, temperature acts as a typical and important nonlinear multi-factor regulatory variable in the multi-field coupled acidizing process of geothermal wells. This study provides theoretical foundations and numerical tools for optimizing acidizing treatment designs and predicting stimulation outcomes in high-temperature carbonate geothermal wells.

Topic: Enhanced Geothermal Systems

The success of geothermal energy production, particularly in Enhanced Geothermal Systems (EGS), relies heavily on a thorough understanding of subsurface fracture networks. However, leveraging multiple datasets from drilling, completion, monitoring, testing, and production to accurately characterize fractures remains challenging due to data scarcity, resolution limitations, and inherent geological complexity. This study presents an advanced machine learning-based framework to integrate and analyze various datasets from a geothermal field, addressing these challenges and enhancing fracture network characterization to support unleashing American energy. The proposed methodology applies multiple machine learning techniques to datasets obtained from the U.S. Department of Energy’s (DOE) Frontier Observatory for Research in Geothermal Energy (FORGE) initiative in Utah. Initially, up to five unsupervised machine learning algorithms are employed to analyze passive seismic data from wells 16A and 16B, quantifying b-values and categorizing diffusivity to delineate the fracture geometry, which consists of both hydraulic and natural fractures. Subsequently, supervised machine learning techniques are applied to well logs, core testing results, stimulation and pumping data, tracer records, and circulation tests, serving as validation measures for the fracture networks derived from seismic analysis. Further validation is conducted by integrating simulated completion and stimulation processes, incorporating treatment, injection, and testing pressure measurements. A key aspect of this approach is the implementation of data fusion techniques to systematically integrate fracture network interpretations across multiple datasets. This multi-level analytical framework enhances fracture characterization, providing a holistic understanding of subsurface fracture behavior. To facilitate visualization and user interaction, a web-based interactive platform has been developed. This platform enables users to explore input datasets, analytical processes, machine learning algorithms, and final fracture network models in an interactive manner, improving accessibility and interpretability. The insights derived from this study offer direct benefits to geothermal operations, including optimization of completion and stimulation designs, fracture interference analysis, and production performance forecasting. Additionally, the developed technology has been successfully tested in hydraulic fracture network characterization for unconventional oil and gas applications, demonstrating innovation in energy super intelligent technologies and broader applicability in subsurface energy extraction.

Topic: FORGE

As the Geothermal Development Company (GDC) expands its footprint across high-potential green energy zones spanning Nairobi headquarters, Central Rift operations, and remote exploration fields in Baringo (Paka, Korosi, Silali) and the South Rift (Suswa) securing critical infrastructure presents unique geographical and logistical challenges. This study outlines GDC's transition from traditional, reactive guarding toward an intelligence-led, technology-driven security paradigm. The framework integrates biometric access control, AI-enabled thermal PTZ CCTV monitoring, automated fuel telemetry, and real-time GPS fleet tracking to safeguard high-value assets such as drilling rigs, wellheads, and transmission networks. Coupled with robust incident response units, standardized operating procedures, and rigorous vulnerability assessments, this multi-layered approach ensures uninterrupted business continuity, regulatory compliance, and maximum safety for personnel operating in rugged environments (Fosner, 2024).

Topic: General

Using CO₂ as the working fluid in enhanced geothermal systems (EGS) offers potential advantages over water, including viability at lower reservoir temperatures, concurrent geologic carbon storage, and reduced pumping requirements through a natural thermosiphon effect. However, field experience with CO₂ as a geothermal working fluid remains limited. Co-located with industrial emitters, CO₂-EGS could create a valuable use for captured CO₂ as a productive working fluid, pairing clean firm power with permanent CO₂ storage and offering a scalable pathway to decarbonize hard-to-abate industry. Recent developments in commercial EGS, subsurface CO₂ management from enhanced oil recovery, supercritical CO₂ power cycles, and carbon storage policy now make a field pilot more feasible. Building on these advances, we present an integrated design framework for a CO₂-EGS pilot in fractured crystalline rock that couples well architecture, multistage completion design, and produced-water management. Using wellbore and reservoir models, we evaluate how design choices affect self-circulation, heat extraction, and operational risk. The goal is a set of conceptual design guidelines for a near-term pilot.

Topic: Enhanced Geothermal Systems

Fault-controlled fluid flow and heat transfer play an important role in heat transport within geothermal systems. Field physics and recent studies have demonstrated that fault geometry and its thermal properties can produce distinct subsurface temperature distributions and thermal signatures. Inspired by these studies and information from geothermal fields, this work considers the possible effects of spatial and temporal variations in fault hydraulic properties on transient geothermal heat transport. In particular, fracture aperture and permeability may vary in both space and time, affecting fluid flow and the resulting thermal response. As an initial step, a reduced model is proposed to examine heat transport along a fault or fracture with spatially and temporally varying hydraulic properties. The resulting transient heat-transfer problem is investigated using analytical and semi-analytical methods, with particular attention to the relationship between variations in fault properties and observable dynamic temperature responses. The model will be used to examine how these variations influence transient temperature profiles and the development of thermal anomalies across the domain under investigation. This work aims to develop a tractable framework for understanding the thermal responses associated with spatially and temporally varying fault hydraulic properties. The reduced formulation may also provide a basis for incorporating additional coupled physical processes and for comparison with numerical simulations and field temperature observations in future studies and field practice.

Topic: Modeling

Geothermal exploration integrates diverse geological, structural, thermal, geochemical, and geophysical datasets that differ substantially in schema, vocabulary, coordinate reference systems, spatial and temporal resolutions, units, and completeness. The INGENIOUS project has compiled extensive regional and prospect-scale datasets across the Great Basin region, Western US, and has used many of these datasets to further play fairway analysis and machine-learning applications[NW1.1]. Substantial processing is required in order to transform raw exploration observations into cell-based feature matrices used as quantitative analysis input. Such transformations are commonly developed ad-hoc for individual datasets or analyses, which can limit consistency, reproducibility, provenance tracking, and reuse across exploration workflows. This study develops a standardized, geometry-aware framework for transforming heterogeneous geothermal exploration data into analysis-ready spatial features. Source-data standardization first establishes dataset-specific schemas that define required fields, naming conventions, data types, units, controlled vocabularies, and missing-value representations, along with temporal information, persistent record identifiers, and source references to support traceability and reuse. Geometry-specific transformation methods are then applied according to source data type. Point observations can generate nearest-observation, neighborhood, density, interpolation, and spatial-support features. Linear datasets can generate distance, density, intersection, termination, orientation, and other structural-context features. Polygon datasets can be represented using categorical membership, proportional coverage, boundary relationships, and related spatial metrics. Existing raster surfaces are harmonized through reprojection, alignment, resampling, masking, unit standardization, and scientifically justified derivative calculations. The framework is demonstrated using representative Great Basin datasets developed and compiled through the INGENIOUS project, including known geothermal system and thermal observations, Quaternary fault attributes (e.g., location, slip rate, and recency),[NW2.1] favorable structural settings, and geophysical datasets (e.g., gravity and magnetics). The workflow produces aligned feature grids, a cell-based feature matrix, and associated metadata that record the source and transformation history of each analytical feature. Particular attention is given to distinguishing true zero values from missing observations and limited spatial data support. The framework is intended to provide a reproducible connection between geothermal data compilation, feature engineering,[EM3.1] play fairway analysis, and machine learning, while also supporting standardized data practices.

Topic: General

Geothermal reservoirs are commonly associated with elevated rates of earthquakes, motivating concerns about potential seismic hazards with new or expanded geothermal energy projects. While this seems to be true at high-enthalpy hydrothermal systems (Coso, The Geysers) and in enhanced geothermal systems (Basel, Cape Station), low-enthalpy hydrothermal systems in the Basin and Range are typically seismically quiet. We monitored one such system, Lightning Dock Geothermal in New Mexico, with four borehole geophones for over 17 months (05/2025 – 09/2026). We implemented a near-real time monitoring system for seismic detection and location using QuakeMigrate, then re-processed the full dataset using QSeek. We observed 18 microearthquakes total, with moment magnitudes ranging from Mw=-0.5 to Mw=0.75. The microearthquakes did not vary in response to major changes in the reservoir such as the initiation of a new production well, or to intermittent hiatuses in power operations (grid trips). We hypothesize that the low rate of microseismicity is related to a combination of low tectonic stress conditions and extensive permeability networks that rapidly distribute any localized changes in pore pressure.

Topic: Geophysics

Geothermal is an important component in the Kenyan energy portfolio and the preferred source of base load due to its reliability and cleanliness as an energy source. However, development of geothermal energy is capital intensive, complex and subject to uncertainties and risks. Financial modeling is an important component of strategic planning for geothermal projects and can help companies evaluate the financial feasibility, risk profile and returns of a geothermal project. Financial models consider financial input including costs, revenues, financing opportunities and operating variables to generate valuable financial output including net profit, cash flow forecasts, return on investment and payback period. Investors should be aware of the amount of investment required to complete the project and what they can expect to gain from their investment. Financiers, on the other hand, would like to verify that the project can pay off the loan principal and interest timely. The ones who must approve and deal with the utilities think in terms of the greatest money. Hence, the financial sustainability of the project needs to be measured with a sustainable financial model in the long run. Input data includes important decisions made in project implementation that will influence resources that will be allocated to the project, how they will be sourced, assumptions about the characteristics of the resources and project development costs. One of the most important problems in making a financial model for a geothermal project is coming up with representative data. Financial models generally give financial statements, referred to as pro forma, that follow various accounting standards as per requirement. Optimization using Ms Excel further fine tunes implementation strategy and if there are changes in the assumptions made, can be used to evaluate the impacts. So, models can also facilitate the tracking of the project at different development phases

Topic: General

High-pressure high-temperature (HPHT) oil and gas wells and geothermal wells share significant similarities in thermal and mechanical environments; nevertheless, their engineering challenges are commonly addressed separately. This paper presents an integrated literature review that looks into both well types through a shared engineering perspective, examining how well construction, drilling-fluid, and downhole tool performance manifest as some of the most important challenges for these types of wells. The review uses documented field and technology development case studies showing that, although the underlying failure mechanisms are often the same, geothermal wells are more severe environments than most HPHT well applications, with temperatures exceeding 300°C. Some of the challenges include temperature-dependent fluid degradation, thermal damage to downhole tools, and thermally induced cement degradation, all of which should be addressed not only by applying documented HPHT knowledge but also by developing new testing standards for these extreme conditions. The review finds that geothermal operations have largely adopted HPHT well protocols for cementing, tool qualification, and fluid selection, even when those protocols were not designed for geothermal well conditions. Progress requires both transferring existing HPHT knowledge and extending it: developing higher-temperature material standards, adapting downhole tools for higher thermal resistance, and recognizing that water-based fluids serve a different role in geothermal than oil-based mud does in HPHT wells. The findings support treating the operational challenges of both well types as related engineering problems, while identifying where technology can be transferred directly, where adaptation is needed, and where new standards are required.

Topic: Drilling

Relay ramps within a composite normal fault system control permeability and geothermal upflow at Thermo No. 1, Utah. This updated structural framework integrates geologic, geophysical, and geochemical data to refine resource delineation and constrain fluid migration. En-echelon segments of north-striking, west-dipping normal faults along the western edge of the legacy well field create relay ramps that concentrate fracture permeability, allowing geothermal fluids to discharge at the Thermo hot springs. In revised gravity survey interpretations, gradient lines stretch and step past one another to define the relay ramps between adjacent fault segments. These interpretations are synthesized with regional deformation trends, geologic mapping, and 3D geologic modeling, then tested against existing reservoir datasets. Borehole image logs from step-out well 73-32 confirm the mapped fault orientations and identify the principal feedzones on permeable north-striking faults. A secondary WNW-ESE trend, also mapped in outcrop, enhances permeability only locally. Geochemical data show the hot springs yield a less-mixed, more thermally mature fluid than the wellfield, and a tracer test confirms the two are hydraulically connected. These findings are consistent with previous studies identifying step-overs as the most common structural setting for geothermal systems in the Great Basin region. Most current production at Thermo No. 1 comes from a shallow, highly permeable mixing zone, whereas the deeper fault-controlled reservoir feeding the hot springs remains largely untested. Well 73-32 approached the interpreted upflow zone but intersected the main fault zone only at its margin. Weak, late tracer returns at 73-32 and the hot springs indicate an indirect flow path from the wellfield. Fault geometry and connectivity, rather than heat flow alone, control where geothermal fluids can be produced. The framework therefore guides targeting of deeper upflow in future exploration wells.

Topic: Geology

The scanning electron microscopy has been used for the microscopic work in the analysis of rock cuttings dependent on the electron emission and provides a more detailed field characterisation images It has been used worldwide for the analysis of both organic and inorganic materials on a nanometres and micrometre scale. The EDS, Energy dispersive x-ray spectroscopy (EDS) Works together with the SEM to provide qualitative and semi quantitative results, these two techniques can therefore provide the composition of the material scanned, which sometimes cannot be done with the other laboratory microscopic techniques. The SEMS are equipped with EDS and modern software to analyse the received data. The EDS adds the advantage of evaluating the composition of various elements in the sample with the aid of a computer programme. The EDS converts the intensity of the x-ray’s ratios to chemical composition in a short time and increases the effectiveness of the quantitative analysis.

Topic: Field Studies

Hydrothermal alteration mapping is considered as a widely adopted step in hydrothermal mineral exploration. In this works, the wavelength mapping and Spectral Angle Mapper (SAM) techniques were applied to map and identify hydrothermal alteration minerals in geothermal drill cuttings. Hydrothermal alteration minerals rich in Al-OH, Mg-OH and Fe-OH were targeted. Using wavelength mapping approach. As a result, wavelength position and the depth of the absorption features were generated. Wavelength stretches of between 1850-1950 nm, 2160-2228nm and 2300-2370 nm were generated to map Al-OH, Mg-OH and Fe-OH rich group of minerals respectively. For the wavelength stretch between 1350-2400 nm pixels of light green to yellowish occurring near 1900 nm represent deepest absorption features within that range. Equally, wavelength stretch between 2160-2228 nm differentiates various Al-OH minerals found in the sample with the deepest features occurring at 2214 nm. Results for the third stretch between 2300-2370 nm shows dominant yellow pixels at 2337-2342 nm which corresponds to epidote minerals sample while cyan color pixels between 2314-2320 nm correspond to actinolite minerals. SAM classification was performed on all images using various spectra subsets and image-derived endmembers and 0.1 thresholds. The most commonly hydrothermal mineral classified by this algorithm includes zeolite, smectite, illite, chlorite, calcite, epidote and amphiboles. Mineral mixtures were classified in separate classes to distinguish them from pure endmembers minerals. This study shows the combination of wavelength mapping and SAM techniques proves to be a powerful approach for accurately identifying and characterizing hydrothermal alteration features using specific wavelength ranges. The findings from this study can aid future mineral exploration endeavors in similar geological settings, providing guidance for locating important hydrothermal alteration minerals in a geothermal system.

Topic: Emerging Technology

Reservoir engineers forecast how enhanced geothermal systems (EGS) will perform using detailed physics simulators such as ResFrac, which models hydraulic fractures, wells, and the reservoir together. One 30-year run takes hours to days of computing, so most studies are limited to a few hundred runs, a full Monte Carlo uncertainty study is rarely affordable, and each new question usually needs a new batch of runs. ResFrac Copilot is a browser-based tool that learns from runs a team has already made and builds a fast stand-in model (a surrogate) that engineers can question directly, so one study can answer many questions. Given the input parameters and simulated output time series of a study (e.g., net power, cumulative net energy, produced fluid temperature, and flow rates), the workflow trains several types of surrogate (Gaussian process, polynomial, nearest-neighbor, and others), and keeps the surrogate with the lowest k-fold cross-validation error on held-out simulations. The trained model can be used for probabilistic forecasting, inverse estimation of injection wellhead pressure schedules, history-match ranking, and sensitivity analysis. A language-model assistant translates engineering questions into these computations. We tested the workflow on an existing Cape Station Phase 1 uncertainty study: 90 ResFrac runs that varied nine reservoir properties over 30 years. Using the selected Gaussian process surrogate, a 10,000-case Monte Carlo study ran in seconds, while in ResFrac it would have required about 10,000 more simulations. To validate the surrogate, six parameter sets corresponding to the P10, P30, P70, and P90 of the predicted year-10 cumulative net energy distribution were simulated in ResFrac. Compared with ResFrac, the relative error of the surrogate in year-10 cumulative net energy ranged from 0.4% to 3.3% (1.4% on average). A surrogate-based sensitivity analysis further identified the parameter that dominates forecast uncertainty, reducing the P10 to P90 range by 57% when constrained, providing a quantitative basis for prioritizing future data acquisition.

Topic: Modeling

Though the drilling crew has no exclusive control on the downhole formation fracture ecosystem, there are moments where bad drilling practices may lead to induced formation fractures. Induced formation fracture is an unintentional breakdown of an open hole formation that happens when bottom hole or hydrostatic pressure exceeds the rock's fracture gradient during drilling. This becomes a compounded problem in Olkaria geothermal fields which are highly characterized by natural fractures as well. The net effect to this scenario is the loss of circulation returns which is one of the most unwelcome occurrences that not only increases the drilling costs, but also diminishes the success rate to target depth. When lost circulation occurs, the possibility of getting a kick, stuck event, loss of drill string or the entire well is increased. It is therefore critically important that at any given moment in drilling process, the drilling engineer should be able to distinguish between natural fractures from those induced by the drilling process so that prompt corrective actions can be taken to minimize the impact of the induced fractures. The induced formation fracture is a phenomenon that has not been well understood for requisite attention to be granted. Therefore, the main objective of this work is to seek to understand how induced formation fracture occurs and bring to light with clarity the pathway to navigate through the same in order to ensure safe, efficient and economic drilling.

Topic: Drilling

Geothermal well drilling process is one of those activities that are replete with technical challenges and almost impossible to avoid entirely. Driving the bit through the multiple rock’s formation for several thousand of feets require as much logistical and mental capital as walking a blind path out of vision. The Inevitable incidents that befall the process has been persistent for as long as the history of geothermal drilling itself and leads to non-productive time and undesirable additional drilling costs. In well planning, the key to achieving objectives successfully is to design drilling programs on the basis of anticipated potential hole problems and plan for solutions in advance. This paper therefore seeks to bring to fore the viable mitigation measures that are commensurate to most common problems. The initial examination of the data points to the fact that the stuck situations, wait- on events and unplanned trip runs constitutes the most unwarranted nonproductive times. The research data was examined from the daily drilling reports on a wide range of drilled wells. OW-729A data was used as a representative sample of the conventional drilling obstacles.

Topic: Drilling

Aurora prospect in Aurora Crater, Nevada is proposed as a compelling greenfield Geothermal project, supported by strong financial metrics and critical market demand. The generated P50 base case (12.98MW) forecasts an Internal Rate of Return (IRR) of 12.51% and a Net Present Value (NPV) of $33.32M. Aurora is currently undergoing Phase 1: Resource Procurement (exploration stage). The main objective is to meet the energy demand target of 3MW to 15MW suitable for moderate-scale Aurora gold mining operations. This project acknowledges the target customers’ pain points and offers a viable solution, offering an LCOE of 10.82 cents/kWh equivalent to up to 57% electricity savings. The project is bankable based on an established long-term PPA with the gold mine authorities to guarantee revenue exceeding the projected LCOE. Overall good potential to explore synergies between Geothermal and Mining Industries for sustainable project development.

Topic: Field Studies

Prolonged exploitation of geothermal reservoirs may lead to gradual changes in production performance, including reductions in wellhead temperature, pressure, mass flow rate and power output. Detecting these changes at an early stage is important for sustainable reservoir management because delayed identification of declining performance can complicate operational decision-making and reservoir optimization. Conventional reservoir monitoring generally relies on individual parameters and numerical reservoir models, which can be computationally demanding and may not fully exploit the large volumes of spatial and temporal data generated during geothermal operations. This study proposes an integrated Geographic Information System (GIS), remote sensing and machine learning framework for early detection and spatial characterization of geothermal reservoir decline. The proposed approach integrates time-series well production data, wellhead temperature, pressure, mass flow rate, electricity generation, injection rate, reinjection temperature, reservoir pressure and temperature, together with satellite-derived land surface temperature (LST), normalized difference vegetation index (NDVI), digital elevation model (DEM) and structural information. Temporal features including moving averages, rates of change, seasonal anomalies and lagged variables are derived and used to train Extreme Gradient Boosting (XGBoost), Support Vector Machine (SVM) and Long Short-Term Memory (LSTM) models. The models classify wells or monitoring periods into stable, transitional, and declining conditions and predict the probability of future decline. GIS is subsequently used to spatially visualize predicted decline zones and their relationships with production and reinjection infrastructure. The model performance will be evaluated using accuracy, precision, recall, F1-score, area under the receiver operating characteristic curve and root mean square error for continuous predictions. The framework is intended to provide an early-warning system capable of identifying emerging reservoir-performance changes before substantial production losses occur. The study demonstrates how integrating operational, spatial and Earth-observation data with machine learning can support data-driven geothermal reservoir management and sustainable exploitation.

Topic: Reservoir Engineering

One key aspect of geothermal surface exploration studies is to determine permeable zones which act as primary fluids circulation pathways. The permeable zones are characterized by the presence of faults, fractures, and fissures. Conventional geothermal exploration studies are done through geological, geophysical, and geochemical field surveys. Remote sensing can also be used to map areas with high concentration of structures, hence assists in determining the permeable zones prior to carrying out detailed geoscientific studies. This study focuses on utilizing the Shuttle Radar Topography Mission (SRTM) Digital Elevation Model (DEM) obtained from the United States Geological Survey (USGS) and LiDAR DEM to delineate these zones within the Paka Volcano. Terrain attributes such as hillshade and slope were extracted from the two DEMs. Lineaments were extracted using multi-directional hill shades and Canny edge detection algorithm. The extracted features were converted into a binary raster and analyzed using a moving window (kernel) approach to generate lineament density maps of the study area. The results show that SRTM – derived lineament density zones capture the regional structural trends and major fault lines whereas the LiDAR – derived lineament density zones uncovered the detailed fractures and subtle fault scarps. High-density zones identified in both datasets correspond to areas of enhanced permeability, but LiDAR provides greater spatial precision for delineating potential geothermal fluid pathways. Despite its advantages, LiDAR is limited by data availability and processing requirements, while SRTM offers a cost-effective solution for regional assessments.

Topic: Field Studies

Building on American GeoPower`s 2026 Stanford Geothermal Workshop paper, this study models the physics of thermal drawdown to size well architecture and molten-salt flow for target electrical output. In dry rock, conduction replenishes the wellbore; conduction and radiation transfer heat into the closed loop. Below water’s critical point (373.95°C; 22.064 MPa), deviating sealed inserts and connected laterals expand exchange area. Superheated formation-fluid circulation reduces this need. Yara International’s calcium–potassium–sodium nitrate mix, demonstrated to 500°C, is the preferred thermal transfer fluid. A coaxial well in granite initially at 375°C is modeled for a 10 MWe target. Real-time AI reduces molten-salt flow near 370°C and increases it as the wellbore recovers toward 375°C; thermal storage supplies the power block during reduced flow. The analysis calculates area, flow schedule, rock recovery, storage capacity, and long-term heat balance. Although directional drilling guidance above 400°C is still being developed, this study models diagonal and lateral molten-salt closed loops at 400°C, 500°C, and beyond as future deployment cases. The analysis calculates well spacing, thermal drawdown, and rock recovery to demonstrate the conditions required for sustained gigawatt-scale power from a single formation through managed heat extraction and routine maintenance.

Topic: Modeling

Geothermal energy development faces significant challenges due to subsurface uncertainties, particularly in identifying economically viable plays where elevated temperatures occur at accessible depths. Geothermal Play Fairway Analysis (GPFA) emerges as a powerful risk-reduction framework, adapted from petroleum exploration, to systematically integrate geological, geophysical, and thermal data at basin and play scales. By mapping Common Risk Segment (CRS) and Composite Common Risk Segment (CCRS) elements, GPFA highlights prospects with anomalous heat mechanisms that elevate local geothermal gradients beyond regional norms, enabling high-temperature resources at shallower depths and thereby slashing drilling costs while boosting return on investment (ROI). Anomalous heat mechanisms often drive these localized thermal enhancements. For instance, salt diapirs act as thermal conductors, funneling heat upward and creating hotspots in sedimentary basins. Overpressured zones can advect deeper, hotter fluids toward shallower reservoirs, amplifying gradients through convective heat transfer. In the Basin and Range province, high-permeability faults linked to deep crystalline basement rocks facilitate upward circulation of hot fluids, yielding gradients far exceeding background levels. GPFA's data-driven approach identifies these features by quantifying their impact on heat flow, reservoir quality, and fluid pathways. This targeted identification allows operators to prioritize prospects where temperatures suitable for power generation (greater than 150°C) are achievable at depths under 6 km, reducing capital-intensive deep drilling and enhancing project economics. A critical pitfall in geothermal assessment is the misuse of geothermal gradients for temperature extrapolation. Linear gradients, derived from bottom-hole temperatures (BHT), cannot be reliably extended beyond well control, as they overestimate deeper temperatures. Gradients inherently decrease with depth due to rising thermal conductivity in compacted, less porous rocks—governed by the equation: gradient = heat flow / thermal conductivity. This leads to flawed temperature-depth profiles; for example, a projected 200°C at 3 km might actually yield only 150°C, inflating drilling budgets and jeopardizing projects amid high upfront costs. To mitigate this, GPFA incorporates 1D basin modeling, applying Adaptive Heat Flow/Temperature Modeling, simulating heat flow from the lithosphere-asthenosphere boundary (at ~1330°C) upward through a stratigraphic column parameterized with lithology-specific thermal conductivities. This physics-based method accurately predicts non-linear gradients, accounting for variable heat flow, thermal conductivity, radiogenic heat production, temperature dependent thermal conductivity and conductive/convective processes. Case studies from numerous geothermal exploration prospects demonstrate how GPFA, augmented by such modeling, has de-risked plays with anomalous gradients, improving success rates by 20-30% and optimizing ROI through shallower, cost-effective developments.

Topic: Enhanced Geothermal Systems

At the 51st Stanford Geothermal Workshop we proposed that the Brawley Seismic Zone (BSZ), the transtensional stepover between the San Andreas and Imperial faults, hosts a Naturally Enhanced Geothermal System (NEGS™): a hydrothermal reservoir in which tectonically induced, structurally governed fracture permeability supplies commercial transmissivity without hydraulic stimulation. This paper presents the field test designed to confirm or refute that concept. The DAVINCI INTACT project (Imperial NEGS™ Testing And Confirmation Trial) has been selected by the U.S. DOE Office of Geothermal for award negotiation under DE-FOA-0003472. The test pairs two new horizontal wells drilled from a common pad about 2 miles northeast of the producing Brawley field. Both wells kick off near 6,500 ft TVD and land at about 7,000 ft TVD in Pliocene–Recent fluvial-lacustrine sandstones, where pre-drill mapping indicates temperatures of roughly 230–240°C. The 1,100-ft producer lateral runs due west, perpendicular to SHmax (N5–12°E) and across the dominant NE-SW and subordinate NW-SE fracture sets mapped from high-resolution aeromagnetic data , 3D seismic and relocated seismicity; the 1,200-ft injector lateral diverges to the northwest. A vertical pilot hole precedes the producer lateral. The producer is completed open hole with distributed fiber-optic sensing; the injector is cased and selectively perforated at intervals about 550, 800 and 1,200 ft from the producer, so that isolated injection, pressure-pulse and tracer tests sample each spacing before full-lateral circulation. No stimulation is planned, and injection pressures remain below the fracture gradient under a seismic traffic-light protocol. We state the outcomes that would confirm or reject NEGS™: (1) unstimulated lateral productivity exceeding that of offset vertical wells; (2) measurable pressure and tracer response at the producer from discrete injector intervals; (3) permeability anisotropy aligned with the predicted fracture azimuths; and (4) sustained circulation minimal thermal decline over 120 days. We present the pre-drill structural model, the well and test design, and the interwell-spacing criteria that will guide horizontal development of tectonically fractured sedimentary basins.

Topic: Emerging Technology

Geothermal reservoir modeling requires effective integration between three-dimensional geological interpretation and physics-based numerical simulation. This paper documents a practical workflow for interoperability between Leapfrog and ResFrac, enabling geological models developed in Leapfrog to be incorporated into coupled hydraulic-fracturing multi-physics EGS simulations. The workflow covers transferring key model components, including well trajectories, geological surfaces, structural features, reservoir domains, and spatially varying rock and reservoir properties. These inputs are used to construct ResFrac models capable of simulating hydraulic stimulation and long-term EGS circulation, including fracture propagation, fluid flow, geomechanical and chemical responses, and heat transfer. The workflow is demonstrated through an EGS modeling example, illustrating the progression from geological interpretation in Leapfrog to simulation in ResFrac and visualization of simulation results within the geological context. Particular attention is given to maintaining consistency between the geological and simulation models and enabling efficient iteration as subsurface interpretations evolve. The resulting workflow provides a repeatable connection between geological characterization and coupled EGS simulation, improving integration between geoscience and reservoir engineering and supporting systematic evaluation of geological uncertainty in EGS development.

Topic: Modeling

The source of steam for sealing turbine-shaft gland areas affects geothermal plant safety as well as maintenance and reliability. The gland-seal steam systems are not necessarily familiar to technical and operating staff. Further, real-world examples in this paper demonstrate that there are important safety considerations that deserve wider recognition in the geothermal industry, which is the purpose of this paper. Like steam turbines in other industries, flash and dry-steam geothermal plants use steam to seal gaps between the rotating shaft and the stationary casing of the steam turbine. The steam source is usually geothermal steam. The steam is removed from the gland-seal assemblies by the gland-steam leak off (GSLO) system. Because of the typically high H2S content of geothermal steam, it is important to carefully consider where the GSLO is vented. It needs to be vented to prevent worker exposure to high H2S, yet it should not be vented where it will cause other problems. For example, venting the GSLO into the non-condensable gas (NCG) would eliminate the H2S concern, but may cause the NCG to become an explosive gas mixture with associated dangers. If clean steam (e.g., without H2S) is used, most of the potential dangers are mitigated. Further, clean steam offers potential maintenance and reliability benefits for the steam turbine. This paper will review the purpose of gland-steam systems and provide an explanation of how they work. Further, the paper will review the associated gland-steam leak off (GSLO) systems, options for safely venting GSLO, alternative steam supplies, and other information useful for techniucal staff at flash and dry-steam geothermal plants. Real world examples from operating geothermal power plants will be provided, including ones where the GSLO vent was a serious hazard that was subsequently mitigated. A goal of this pape is to support a safety-focused culture within the geothermal industry.

Topic: Production Engineering

The Utah Frontier Observatory for Research in Geothermal Energy (Utah FORGE) is the U.S. Department of Energy’s dedicated field laboratory for advancing enhanced geothermal systems (EGS) technology in low-permeability, high-temperature crystalline rock. Since receiving the FORGE designation in 2018, the project has advanced site characterization, drilling, stimulation, reservoir monitoring, and open-data activities intended to reduce the technical and economic risk of commercial EGS deployment. In 2026, Utah FORGE completed construction and commissioning of the Phase 4A circulation facility and initiated circulation testing in mid-August between the existing highly deviated injection and production wells. The facility enables controlled, field-scale testing of a hydraulicly stimulated fracture network under conditions scalable to commercial EGS operations. A comprehensive monitoring program acquired surface and subsurface operational data throughout the circulation campaign, including injection and production conditions, pressure response, flow behavior and distribution along the injector, temperature, seismicity, geochemistry, and distributed fiber-optic measurements. This presentation will provide an update on facility construction, commissioning, and the evolving interpretation of results of early circulation operations. Analyses evaluate water loss and reservoir connectivity, parasitic load and surface-system performance, induced seismicity, fracture response during injection, thermal behavior and evidence of thermal decline, and geochemical, microbial, scale, and corrosion responses across the circulation system. This integrated data set will be used to assess the operability and sustainability of the stimulated reservoir, identify constraints affecting longer-duration circulation, and inform the design of future monitoring, stimulation, and circulation experiments at Utah FORGE. The results offer a landmark comprehensive operational data set from an actively circulated, instrumented EGS reservoir in crystalline rock and will help define the technical pathways toward reliable, scalable geothermal energy production.

Topic: FORGE

Cold Underground Thermal Energy Storage (Cold UTES) is an ultra-long-duration grid energy storage technology. With Cold UTES, low-cost grid power is converted to cold thermal energy and stored in subsurface systems, including boreholes and reservoirs, at the point of use. Cold UTES is one approach within the general category of engineered geothermal systems. Previous articles at the Stanford Geothermal Workshop have explored how data center cooling systems could benefit from Cold UTES. Ref. [1] introduced the concept Cold-UTES to manage data center load shapes, while Ref. [2] quantified the value to the electrical grid of managing data center peak loads. In this paper, a system-level techno-economic model developed at the National Laboratory of the Rockies is described. The model is used to quantify how Cold-UTES reduces data center peak electrical loads, reduces electricity consumption and cost, and can potentially increase data center compute capacity for a given grid interconnection. A key research gap for deployment of this technology is demonstration of Cold UTES integrated with an end application. We provide a discussion on the objectives, steps, timeline, and metrics for a successful demonstration that advances the UTES industry. [1] J. Winick , G. Zhu , K. Kitz , and A. Adams, Reducing Data Center Peak Cooling Demand and Energy Costs With Underground Thermal Energy Storage (UTES), SGW, 2025, https://pangea.stanford.edu/ERE/pdf/IGAstandard/SGW/2025/Winick.pdf [2] C Halloran , Q. Luo , G. Schivley , A.T.D. Perera , L. Salmeron , J. Jenkins, W. Cole, Grid Impact of Reservoir Thermal Energy Storage for Data Center Cooling, SGW, 2026, https://pangea.stanford.edu/ERE/pdf/IGAstandard/SGW/2026/Halloran.pdf

Topic: Emerging Technology

Geological Thermal Energy Storage (GeoTES) is a method of storing thermal energy in permeable subsurface formations at temperatures up to 300°C. GeoTES systems provide daily and seasonal energy storage and can be used to deliver industrial process heating or cooling. GeoTES costs are predicted to be very low – once the wells have been drilled, additional energy capacity has very low marginal costs. GeoTES is a novel technology and clearly communicating and illustrating its benefits to prospective clients is a key step towards deployment. Therefore, we have developed the Rapid and Accessible Techno-economic Evaluation (RATE) webtool to assess GeoTES for industrial process heat sites. The objective of the RATE GeoTES webtool is to build customer confidence in GeoTES as a technology solution to process heat systems. The webtool uses information about the site’s location, thermal loads and temperatures, and energy costs to evaluate the benefits and costs of GeoTES, and thereby enable industrial organizations to assess the commerciality of their GeoTES projects. The RATE GeoTES webtool is based on previously developed GeoTES techno-economic models [1,2] and evaluation methods [3,4]. Different charging configurations can be explored such as grid electricity with heat pumps, solar thermal, and solar photovoltaics with electric heaters. The webtool is designed to be fast and simple to use, but also allows more detailed analysis to be conducted. In this article, the methods and usage of the RATE GeoTES webtool are described and an illustrative case study is presented. [1] J. McTigue, W. Cole, H. Oh, R. Barney, E. Witter, A. Gold, C. Halloran, S. Smith, M. Grahovac, M. Ganeshalingam, A. A. Chien, W. E. Gnibga, J. Jenkins, Q. Luo, D. Perera, G. Schivley, J. Winick, K. Kitz, L. Baker, G. Zhu, Project Phase 1 Report: Reducing Data Center Peak Cooling Demand and Energy Costs With Cold Underground Thermal Energy Storage (Cold UTES), NLR Technical Report, NLR/TP-5700-100780 June 2026, https://docs.nlr.gov/docs/fy26osti/100780.pdf [2] D. Akindipe, J. McTigue, P. Dobson, T. Atkinson, E. Witter, R. Kumar, E. Sonnenthal, M. Umbro, J. Lederhos, D. Adams, and G. Zhu, Techno-Economic Analysis and Market Potential of Geological Thermal Energy Storage (GeoTES) Charged With Solar Thermal and Heat Pumps, NLR Technical Report, NREL/TP-5700-91225, October 2024, https://www.nrel.gov/docs/fy25osti/91225.pdf. [3] J. Gifford, J. Kadlec, J. Martinek, W. Hamilton, A. Zolan, Z. Ma, J. Sment, K. Plewe, S. S. M. Alaqel, L. G. Maldonado, and B. Bean, First Solar Thermal Energy Planner (STEP 1) and Nationwide Industrial Heat and Power Analysis, NLR Technical Report, NLR/TP-5700-96614 February 2026, https://docs.nlr.gov/docs/fy26osti/96614.pdf [4] E. Witter, N. Taverna, R. Barney, T. Atkinson, R. Colldeweih, E. Sonnenthal, D. Dwivedi, G. Zhu, J. McTigue, Applying Play Fairway Analysis Methods to Geological Thermal Energy Storage Analysis, Transactions of the GRC, 2026

Topic: Emerging Technology

Enhanced Geothermal Systems (EGS) are emerging as a key source of dispatchable carbon-free energy. Unlike conventional geothermal developments, EGS wells commonly rely on hydraulic stimulation operations that subject casing strings to repeated cycles of high internal pressure, elevated axial loads, and severe thermal environments. These operating conditions create loading combinations that are not fully represented by traditional tubular connection qualification programs developed for conventional oil and gas wells. To address this gap, Vallourec developed a fit-for-purpose qualification protocol for tubular connections intended for EGS service. The protocol was derived from established API RP 5C5 and API 5SF methodologies while incorporating loading sequences representative of EGS well construction, stimulation, and production operations. Special emphasis was placed on cyclic pressure and tension loading associated with reservoir stimulation, elevated-temperature performance, thermal cycling, and post-stimulation production conditions. The protocol was developed in collaboration with EGS operators and applied to the qualification of connections used in large-diameter geothermal casing strings. Testing included combined-load sealability assessments, elevated-temperature exposure up to EGS temperatures (400°F-500°F), cyclic stimulation loading, sealability, thermal cycling, and limit-load evaluations. The qualified connections successfully completed the prescribed loading sequences and demonstrated robust sealing performance throughout the tested operating envelope. This paper presents the rationale behind the protocol development, the engineering basis for the selected loading conditions, key differences between conventional qualification approaches and EGS requirements. The results provide a framework for evaluating connections in EGS wells and support broader industry adoption of fit-for-purpose qualification methodologies for next-generation geothermal developments.

Topic: Enhanced Geothermal Systems

Explicit tortuosity is not included in broomstick plots and torque and drag models in the well planning process. This study investigates the borehole friction factor used in torque and drag modeling and examines how wellbore tortuosity affects it. Because the friction factor is a number that quantifies many different effects such as borehole condition, drillstring contact, and fluid behavior, this research evaluates the influence of tortuosity on predicted torque and drag forces. The study also investigates whether variations in tortuosity are associated with drilling parameters (ROP & MSE) and fracture characteristics observed in formation microimager (FMI) logs. This study uses data from the FORGE Utah geothermal project, including BHA configurations, wellbore survey data, measured hookload and torque readings, drilling parameters, and FMI logs. Using the torque and drag equations, a friction factor is calculated from measured data for each BHA run across wells of varying inclination. Planned trajectories are then generated and compared against actual tortuous wellbore paths to evaluate how tortuosity affects the calculated friction factor. FMI logs are also reviewed to classify conductive, partially conductive, and induced fractures and evaluate possible correlations between fracture occurrence, ROP, MSE, and observed tortuosity. Wellbore tortuosity has a significant effect on the calculated friction factor, and by explicitly including expected tortuosity, based on historic data and planned operations, the friction factor estimate reveals more detail about downhole conditions and removes guesswork from analysis. This study compares BHA modeling at different points along the wellbore to reveal friction factor sensitivity. Preliminary observations from the FORGE dataset suggest that friction factors calculated from measured hookload and torque data deviate meaningfully from planned values, particularly in higher inclination wellbore sections. After isolating the contribution of wellbore tortuosity from the friction coefficient, the remaining friction could be used to evaluate additional sources of drag, including cuttings loading and formation related effects. This separation also provides a basis for identifying operational conditions such as bit balling and cuttings loading. Furthermore, FMI analysis will also evaluate whether sections with higher tortuosity correspond with changes in ROP and MSE and whether these trends vary across conductive, partially conductive, and induced fracture classifications. This may help determine whether wellpath tortuosity is related to formation characteristics, drilling response, or both. While torque and drag modeling is well established, the explicit relationship between wellbore tortuosity and friction factor does not remain quantified. This study contributes by using FORGE geothermal field data to calculate and compare friction factors across varying wellbore paths and BHA configurations, while also evaluating relationships between tortuosity, drilling response, and FMI derived fracture classifications. The findings will give a more reliable basis for friction factor selection during well planning, reducing unexpected drag events and improving drilling efficiency.

Topic: Drilling

Fluid circulation in hydrothermal systems depends on connected pathways through fractured rock. Although slip and dilation tendency identify structures favorably oriented for reactivation or opening, hydraulic behavior also depends on fracture geometry, mechanical properties, fluid pressure, and deformation history. This work investigates whether incorporating mechanical information can improve predictions of potential fluid pathways in hydrothermal systems. I first propose rock-strength and stiffness adjustments to fault-orientation and stress data to identify areas where mechanical properties may enhance fracture opening or reactivation. These adjustments are intended as rapid exploration screening tools because rock type, strength, and stiffness can be readily estimated or measured. I then evaluate a more complete Barton–Bandis-style model of shear dilation and use Monte Carlo simulations to explore uncertainty in parameters including rock strength, stiffness, and joint roughness. The analysis highlights the substantial assumptions required for mechanics-based screening while identifying predictions that can be tested against observed fluid pathways in the field.

Topic: Field Studies

Geothermal wells experience large temperature and pressure variations during drilling, stimulation, injection, production, and shutdown. Although well integrity analyses commonly account for coupled effects of thermal loading and mechanical loading, the role of hydraulic effects such as pore pressure evolution within the porous cement sheath is often simplified or neglected. Cement pore pressure can directly influence the effective stress state of the cement and therefore affect yielding, tensile failure, casing-cement debonding, and the potential failure at sites of local defects during geothermal operation. These effects may become particularly important during rapid thermal transients, when fluid pressure and temperature within the cement cannot fully equilibrate. This study investigates these two effects using complementary numerical approaches. First, a fully coupled thermo-poroelastic model is used to quantify the evolution and redistribution of cement pore pressure during geothermal thermal cycling. The model evaluates how temperature changes, thermal expansion of the pore fluid and solid skeleton, hydraulic diffusion, and cement transport properties influence transient pore pressure and the corresponding effective stress state. Parametric analyses are performed to determine the sensitivity of cement pore-pressure response to permeability, porosity, thermal loading rate, initial pore pressure, and other relevant material properties. Second, a thermo-elasto-plastic wellbore model is used to investigate the mechanical consequences of local cement cavities or incompletely cemented regions. The analysis evaluates how cavity size, location, and geometry alter stress concentration, cement yielding, casing deformation, and interface integrity during thermal and pressure cycling. Particular attention is given to the interaction between the initial cement stress state and local geometric defects, because the absolute stress state determines the proximity of cement and casing materials to yielding. Together, the two analyses clarify the importance of cement pore-pressure evolution and local cement defects in geothermal well-integrity assessment. The results identify conditions under which pore-pressure effects should be explicitly considered and quantify how cement cavities can reduce the mechanical margin of the wellbore system. The study provides guidance for selecting appropriate modeling complexity and improving the assessment of geothermal well integrity under repeated thermal cycling.

Topic: Modeling

Optimal placement of exploration wells can reduce subsurface uncertainty and support geothermal field development. This research develops a framework for selecting sequential drilling locations to reduce temperature uncertainty under a fixed exploration budget. An ensemble of 50 three-dimensional temperature realizations generated using an LLM-based modeling approach that captures relevant information from reports and peer-reiviewed publications to represent prior uncertainty. A kernel-based conditioning approach to update the relative weights of the realizations is utilized. Conditional temperature distributions are estimated using a kernel-density formulation with Nadaraya–Watson weights, incorporating the temperature observations accumulated across previously drilled wells. Dynamic programming evaluates candidate drilling decisions and their possible observations to identify an adaptive policy that maximizes expected uncertainty reduction within the available budget. The framework integrates ensemble-based subsurface modeling, conditioning on continuous observations, and sequential experimental design to guide geothermal exploration wherein the optimal policy is compared with a naïve policy that drills locations based on standard deviation across realizations. 

Topic: Modeling

Geothermal and high-pressure, high-temperature (HPHT) drilling exposes the formation surrounding the wellbore to large temperature differences as relatively cool drilling fluid circulates through hot formations. The resulting thermal disturbance evolves continuously with circulation time, fluid temperature, wellbore geometry, and operational conditions and can generate thermally induced stresses that alter the mechanical state of the near-wellbore formation. Accurate prediction of this thermal history is therefore important for understanding thermal loading and its implications for wellbore stability in high-temperature wells. This study develops a hierarchical thermal–mechanical framework for predicting transient formation temperature and thermally induced stress during drilling. A transient analytical solution for radial heat conduction between the wellbore and surrounding formation was first derived to establish a for temperature evolution. A corresponding numerical finite-difference model was then developed for the same T(r,t) problem and verified against the analytical solution. The verified approach provides the foundation for the main numerical model currently being developed in ANSYS Fluent to represent realistic circulating-well conditions, including the coaxial drillstring–annulus configuration, evolving drilling-fluid temperature, axial heat transfer, wellbore geometry, formation heterogeneity, and temperature-dependent thermal properties. The resulting model captures transient formation cooling during circulation, connections, and trips. The predicted temperature histories are subsequently used to quantify the magnitude and spatial distribution of thermally induced stresses under varying fluid–formation temperature differences, circulation durations, cooling rates, and exposure histories. These thermal stresses are combined with the in-situ and drilling-induced stress state to evaluate changes in the near-wellbore mechanical environment and their implications for wellbore stability. Particular attention is given to cyclic cooling and reheating and the resulting time-dependent thermal contribution to the wellbore stress state. The completed analytical and numerical T(r,t)solutions establish a verified foundation for the development of the full thermo-mechanical framework. The study provides a physically based approach for replacing prescribed temperature changes with an operationally dependent thermal history and establishes a pathway for incorporating thermally induced stress into wellbore-stability analysis in geothermal and HPHT wells. The resulting temperature and stress framework will also provide a foundation for subsequent investigation of the effects of thermal loading on drilling performance.

Topic: Drilling

Geothermal Arrays -- networks of large numbers of standard geothermal wells. Networking creates scaling opportunities for clean, cheap, safe, reliable, configurable, and abundant geothermal energy. We test these opportunities with Julia Language simulations. Standard well types can include EGS, AGS, HGS, and others producing electricity and/or heat through networked standard boreholes. The scaling mindset is taken from the Internet, which scaled up successfully into the billions of users over 50 years. In case you are wondering, EGS wells are open, AGS wells are closed, and HGS wells are both -- hybrid -- closed with fracturing to enhance circulated heat harvesting. Our favorite digital twin is a 10x10 array of standard wells producing upwards of a gigawatt over 1,000 acres. Our longer term cost goal, at the next standards plateau, is LCOE less than 1 ¢/kWh.

Topic: Modeling

Risks to Enhanced Geothermal Systems (EGS) across the well life cycle are scattered among lessons-learned reports, drilling logs, conference papers, and expert opinion, which makes systematic comparison and prioritization difficult. This paper presents a Geothermal Modeling, Evaluation & Risk Assessment Tool (GeoMERiT), a spreadsheet-based tool that (1) organizes EGS risks in a phase-based risk register, (2) scores each risk on a 5×5 likelihood-by-severity scale, (3) displays results in automatically generated heat maps, and (4) links selected high-priority risks to input parameters of a techno-economic model (TEM) for levelized cost of electricity (LCOE) sensitivity analysis. The tool follows a four-step workflow: risk identification and categorization, scoring-scheme development, subject-matter-expert (SME) scoring and selection of high-priority risks, and verification of those risks against high-priority risks reported by SMEs with commercial-operations experience. The current register contains 128 risks spanning four project phases (Drilling, Stimulation, Production, and Abandonment) and four risk categories. Only risks rated High or Critical by the SMEs and also identified as high priority from commercial operations are treated as verified operational risks and carried into the TEM linkage. By combining literature evidence, expert judgment, and operational experience in one traceable workbook, GeoMERiT helps in identifying and prioritize geothermal risks at different well life cycle phases.

Topic: Enhanced Geothermal Systems

[Miyora]

Time-Based Performance Analysis of Drilled Wells in the Paka Geothermal Field: Development of Deterministic Drilling Benchmarks for Future Well Planning in the Kenyan North Rift MIYORA Thomas Ong'au [Geothermal Development Company Ltd., Kenya]

The Geothermal Development Company (GDC) commenced exploration drilling in the Kenyan North Rift in late 2018, and over thirty wells have been drilled to date. Geothermal development is capital-intensive and involves high risks, particularly during the initial stages of field development. About 50% of the total cost of geothermal development is spent on drilling wells. It is therefore of paramount importance to ensure that drilling operations are carried out as efficiently and effectively as possible. Drilling a geothermal well involves several activities, including drilling, connections, tripping, BHA and bit changes and other related operations. The time spent on these activities may be classified as productive time (PT) or non-productive time (NPT). However, some time losses are hidden within productive operations and are commonly referred to as invisible lost time. Although these losses may not be easily identified during routine drilling operations, their cumulative effect can significantly increase the total time taken to complete a well. All these analyses can only be possible if there exists a reference well that is field specific. There is currently no established, data-backed, Paka-specific drilling-time that can be systematically used for planning and costing future wells. This paper analyses time data from twenty (20) deep geothermal wells drilled in the Paka Geothermal Field. The time spent on individual activities is evaluated, and the results are compared with drilling performance in other geothermal fields. Each well section is also analysed independently to establish its contribution to the overall drilling time and performance. The historical data will be used to develop a Paka reference well devoid of invisible NPT and unplanned NPT, with representative drilling durations established for individual well sections and activities. These s will support the planning of future geothermal wells in the Kenyan North Rift. Excel macros developed for the analysis, together with a personal application currently under development, will facilitate drilling-time analysis, estimation of future well durations, and cost forecasting. Variations in drilling performance across the twenty wells will be evaluated to establish representative drilling durations for individual well sections and activities, forming a reliable for planning future wells.

Topic: Drilling

In geothermal systems, solid solution compositions serve as natural geothermometers and indicators of fluid flow paths, providing insights into diagenetic and metamorphic processes. Solid solutions also play a critical role as both hosts and regulators of trace element distribution. In the Salton Sea Geothermal Field (SSGF) Li is found in multiple minerals from near-surface clays (hectorite) to chlorite at reservoir temperatures, and biotite in the hottest zones. Understanding how critical elements (e.g. Li) incorporate into mineral solid solutions is essential for predicting mineral stability, fluid composition, and co-produced resource availability over the lifecycle of a reservoir. In addition, solid solution formation is an important factor in predicting scaling and reservoir permeability. The partitioning of elements into solid solutions is controlled by thermodynamics, kinetics, brine and host rock compositions, temperature, and hydrothermal fluxes. Reactive transport models can evaluate the relative contributions of these processes and test conceptual models for reservoir evolution and exploitation over large temporal and spatial scales. In complex mineral assemblages with multicomponent solid solutions, the implementation of solid solution models in reactive transport simulators is hampered by the lack of accurate thermodynamic data, making it difficult to simulate fluid-rock interactions. Furthermore, the thermodynamics of these solutions are often non-ideal, meaning that calculating properties by linearly interpolating between pure endmembers may not be warranted. Kinetics also play an important role when mineral formation is limited by the rate of host rock alteration. TOUGHREACT, a parallel simulator for non-isothermal multiphase reactive transport in porous and fractured rock that is widely used for geothermal simulations, currently represents kinetically controlled solid solutions as ideal mixtures. In this work, we evaluate the treatment of solid solutions in geochemical models for implementation in TOUGHREACT. Reactive transport codes usually represent solid solutions in one of three ways: at equilibrium (e.g., PHREEQC and Gibbs-energy-minimization couplings); kinetically with ideal mixing; or by discretizing composition into many fixed-stoichiometry minerals (Lichtner and Carey, 2006), which becomes impractical for complex mineral assemblages. We implement a treatment for regular non-ideal solutions in TOUGHREACT that builds on the current ideal model. The model is demonstrated on simple systems by comparing results with PHREEQC equilibrium predictions with equivalent simulations using the discrete-composition method of Lichtner and Carey (2006). For the SSGF application, motivated by Fe-Mg-Li partitioning in silicates at different temperatures, the approach requires solving coupled equations for precipitate composition and performing a multicomponent stability test. Because interaction parameters for Li-bearing endmembers have not yet been calibrated, we perform sensitivity analyses and an independent application based on the discrete-composition approach of Lichtner and Carey (2006). We analyze the differences among these approaches and discuss the implications for using kinetic non-ideal multicomponent models in geothermal applications, including the associated data needs.

Topic: Geochemistry

Proppants constitute an important aspect of the longevity of subsurface systems. Their evolution and performance can impact on the overall project economics and longevity. While fractures are the conduits for flow in unconventional reservoir subsurface fluid production and injection systems, proppants are the elements that keeps these conduits open for as long as desired or designed. Although the transport and deposition of these proppants have been widely studied, there is still a dearth in the number of research around the evolution of proppant conductivity in harsh environments such as in Enhanced Geothermal Systems (EGS). Thus, the objective of this study is to characterize the conductivity evolution of different proppant types under EGS relevant conditions of pressures and temperatures. In this work, we conduct preliminary numerical study on the stability and evolution of proppant pack conductivity under thermal, hydraulic and mechanical stresses using a coupled CFD-DEM approach. The geometry is modified after the standardized API 19D proppant conductivity test. The boundary conditions are as described in the standardization methods (ISO 13503-5/API RP-19D, 2008) to maintain a laminar and stable flow conditions. Particles are allowed to move, rotate, crush and break under stress following the Traverse breakage model in ANSYS Rocky. The tests are conducted under Utah FORGE relevant pressure and temperature conditions. Analysis was done for rough and smooth particles with sizes ranging from 200-400 microns (40/70 mesh) to 400-850 microns (20/40 microns), different injection rate (5 ml/min – 10 ml/min), temperatures, injection fluid (H2O and CO2) and particle types, including sand and ceramics (different mechanical properties, size distribution and sphericity), for 30 – 120 s in mono- to multi-layer packs. Preliminary hypothesis follows that particles with higher mechanical strength and tighter size distribution range (ceramics) are more mechanically resistant to stress and break into fewer number of particles. Conductivity evolution shows slower decay for these particles than for sand under the same conditions. These preliminary hypotheses are actively being researched for suitable confirmatory results. The findings of this study will help optimize proppant selection to maintain long-term fracture conductivity and ensure the longevity in EGS projects

Topic: Enhanced Geothermal Systems

Millimetre-wave drilling produces rapid, localized heating of the wellbore wall, generating thermal stresses that promote rock removal through thermal spallation. However, the resulting stress redistribution, crack propagation, and material degradation may also compromise wellbore stability. The drilling response therefore depends on the complex interaction between energy deposition, thermal expansion, in-situ stress, fracture growth, damage accumulation, and progressive rock failure. To investigate these processes, we have developed a coupled thermomechanical model within the Alamo multiphysics framework to simulate the response of crystalline rock subjected to millimetre-wave irradiation. Volumetric heating associated with millimetre-wave absorption is represented using Beer-Lambert attenuation within the rock, producing a transient heat source that drives thermal diffusion and thermally induced strain. Mechanical equilibrium is solved quasi-statically using a Newton multigrid method acting on a degraded elasticity operator. The model incorporates tensile and shear damage together with temperature-dependent rock-strength degradation. Failure is represented either by a temperature-dependent Mohr-Coulomb spallation model or by a Rock Failure Process Analysis (RFPA) formulation that captures progressive damage evolution, irreversible softening, and rate-dependent failure. Damaged regions remain attached to the rock mass until a fracture-energy criterion for buckling-driven delamination is satisfied, allowing episodic chip formation and release. Following detachment, the borehole geometry is updated, and both the stress field and millimetre-wave heat deposition are transferred to the newly exposed rock surface. The model is then used to investigate the influence of formation temperature, in-situ stress, wellbore pressure, irradiation intensity, and rock strength on near-wellbore damage evolution and material removal. Particular attention is given to the transition between shallow thermal spallation, which enhances drilling performance, and deeper progressive breakout, which threatens wellbore integrity. The resulting simulations provide new insight into the mechanisms governing thermally induced wellbore failure and establish operating envelopes for stable millimetre-wave drilling in high-temperature crystalline geothermal reservoirs.

Topic: Drilling

Direct-energy ablation utilizes a high-power gyrotron beam to remove material through a combination of spallation, melting, and vaporization. At moderate beam intensities, however, granite fails predominantly by thermal spallation. The heated surface layer, constrained by the colder rock beneath, develops compressive stresses that fragment the layer and eject millimetre-scale spalls. We present a simplified model of this process at 100 GHz and evaluate drilling performance as a function of depth and rock properties. An axisymmetric bidirectional beam-propagation method is used to compute the electric field within both the rock and the wellbore cavity, including reflections from the evolving rough surface. The model uses temperature-dependent permittivity measurements for granite and basalt up to 1273 K. The absorbed power drives transient heat conduction and generates thermal stresses, which are corrected for borehole curvature and superposed on the in-situ stress field. When the stress exceeds the rock yield strength, damage accumulates according to a temperature-dependent Hoek-Brown criterion. The damaged layer is subsequently released only when buckling-driven delamination satisfies an energy-based criterion. As a result, spalls detach episodically and the borehole geometry evolves self-consistently. The effects of pore water on thermal properties, dielectric permittivity, and pore pressure are also incorporated into the model. In the results section, predicted rates of penetration are compared with measurements from field drilling tests. Several beam profiles and intensities are investigated for granite under an 80 °C/km geothermal gradient. The simulations predict that the rate of penetration more than doubles between the surface and 3 km depth, while the specific energy decreases by approximately 40%, owing to higher formation temperatures and bottom-hole stress concentrations that promote spallation. The effects of rock strength and pore-water content are also examined.

Topic: Drilling

In enhanced geothermal system (EGS) wells that are pushing toward superhot conditions, directional and MWD electronics often run close to their rated limits. The highest exposure comes during operations that conventional practice treats as routine: tripping into a hot hole, pump-off periods and re-establishing circulation. On the Athena well at Newberry, Oregon, Mazama Energy set out to understand how surface equipment, downhole measurements and rig practices each affect bottomhole temperature, so that downhole conditions could be adjusted and managed to create the environment drilling operations require. Memory data from the MWD tools was aligned with surface drilling records, and each part of the system was assessed for what it contributes and what limits it. Mud inlet temperature and surface cooling equipment set the baseline for heat removal. Pump capacity and circulation time control how much of that cooling reaches the bottomhole assembly. Flowline (outlet) temperature was tested as a real-time proxy for downhole conditions. The MWD tool acts as both a sensor and the component being protected, and its real-time and memory readings behave differently. Rig crews and tripping procedures decide whether a plan is carried out consistently from crew to crew and tour to tour. Fiber-optic distributed temperature sensing (DTS) reads continuously whether the well is circulating, tripping or static. It was used to estimate downhole temperatures, to extrapolate static conditions below the deepest fiber reading as the well recovers thermally, and to estimate dynamic temperatures for each operation. It was also used to monitor cement setting behind casing, so design temperatures could be compared directly with downhole conditions during and after placement. The results showed the trade-offs between these levers and where each one stops being effective, and that relying on any single surface indicator gives an incomplete picture. The findings were turned into a depth-staged trip-in procedure and simple field instructions with escalation criteria, supported by a calibrated planning model. For the Ceres wells, this paper describes how the system is being optimized as a package: sizing surface cooling and pumping capacity together, defining which surface signals the crew should watch and how to read them, using continuous fiber-optic monitoring in both pre-job planning and real-time decisions, and preparing crews before the hot sections rather than reacting during them. The fiber-derived static and estimated dynamic temperature profiles, together with the lessons from cement set monitoring, will be applied to the cement design for the deep casing strings. The aim is a repeatable temperature management practice that improves with each well.

Topic: Enhanced Geothermal Systems

The U.S. Geological Survey (USGS) is currently updating the conventional hydrothermal resource assessment of the Great Basin. As part of the update, the methods from the 2008 USGS assessment are being revised to use the best available strategies and to incorporate new data. The new assessment workflow estimates total electric power generation potential from both discovered and undiscovered resources for conventional hydrothermal systems in the Great Basin. Estimates of electric power generation potential for identified systems with high confidence in electric-grade viability (i.e., discovered systems) are stochastically estimated using an updated Monte Carlo analysis from the 2008 method that substitutes a fractured and/or porous reservoir model for the previously used volume method (Burns et al., 2027). The 2008 list of discovered systems is stochastically amended with additional discovered systems identified as potentially being electric-grade in DeAngelo et al. (2026) and then used to estimate the density of electric-grade systems per unit area (DeAngelo et al., 2027). Combining the Monte Carlo approach with the system-per-area density enables stochastic estimation of the undiscovered resources. Estimates of the frequency of undiscovered systems, and therefore the total electric power generation potential, depend upon how much of the region is considered explored. Area explored is estimated using the masking method of DeAngelo et al. (2027), where masks of different exploration radii are constructed around the set of points where electric-grade geothermal resources are known to exist or known not to exist. Exploration bias, from past exploration having potentially focused on more favorable areas based on best professional judgment, is addressed in the Monte Carlo analysis by stochastically allowing for the possibility that system-per-area density may be less in unexplored areas than explored areas. Masking radii of 2 km, 4 km, and 6 km are considered, and the range of undiscovered resources is estimated for each radius. The 6-km mask produces estimates of total hydrothermal power potential similar to the 2008 estimate. The 4-km and 2-km masks produce sequentially larger resource estimates with the mean prediction being, respectively, 43% and 309% larger than that of the 2008 assessment, demonstrating the dependence on the selected estimate of explored area.

Topic: General

Reliable prediction and operation of geothermal systems requires a model that captures the coupled dynamics of the reservoir and wellbore. This is particularly important because most operational and monitoring data are acquired in the wells, while a substantial fraction of the observed pressure and temperature response is transient and reflects changes occurring in the reservoir. Consequently, interpreting field observations and predicting system behavior requires a tightly coupled well–reservoir model, rather than separate wellbore and reservoir simulations. In this study, the fully coupled well–reservoir simulator DARTS-well is validated at field scale using measurements from the TU Delft geothermal doublet. The doublet targets the Lower Cretaceous Delft Sandstone at approximately 2.2 km depth. A field-scale model is constructed using the as-built well trajectories and completion geometry together with reservoir properties obtained from reservoir characterization. The actual operational conditions of the doublet are used, and simulated pressure and temperature responses are compared against field observations acquired in the wells. The field comparison provides a test of the ability of DARTS-well to reproduce the coupled hydraulic and thermal response of the well–reservoir system under realistic operating conditions. Beyond model validation, the study establishes the foundation for a digital twin of the TU Delft geothermal system, in which transient and steady-state observations are continuously assimilated into the coupled model to update its state and parameters. Such a closed-loop data-assimilation framework provides a pathway from field-scale simulation and validation toward predictive monitoring, operational optimization, and reliable long-term management of geothermal systems.

Topic: Modeling

The present paper is a comprehensive review on induced seismicity in Enhanced Geothermal Systems (EGS) as well as on the related processes. It furthermore describes the monitoring of an injection area for seismic activities and outlines different options for seismic hazard prediction and mitigation in order to handle the risks. An integrated descrption of the mechanisms of induced seismicity (pore pressure increase and thermal stress) and of the criteria for fault reactivation. The monitoring of microearthquakes in the area of injection. The last advances of machine-learning methods for the improvement of the short-term seismic hazard forecast and for the real-time seismic-risk management (so-called “traffic-light control”) of geothermal operations. Results, Observations, Conclusions. Increasing both pore pressure and thermal stress can cause reactivation of faults leading to earthquakes. The related increase in seismic hazard can be managed with an appropriate injection strategy. Short-term seismic hazard forecasts can be improved with the aid of machine learning. However, their accuracy strongly depends on the corresponding monitoring density. This Review provides a fully integrated view of induced seismicity and of associated seismic hazard, for EGS, including management. It also briefly outlines the recent approaches, using machine-learning, of short-term and of real-time seismic hazard and of seismic risk monitoring and management for geothermal projects, and for their on-line real-time on-line seismic risk management (traffic-light control).

Topic: Enhanced Geothermal Systems

This review is intended to give a status report to the field of tracer testing for reservoir characterization of geothermal wells. The main issues are identification of the flow paths, determination of swept-volume and of thermal-breakthrough and the use of machine-learning for improved interpretation of the obtained tracer profiles. We review conservative and reactive tracers as well as flow-path and swept-volume assessment, thermal breakthrough prediction and most recently machine-learning enhanced interpretation of a tracer return to improve our analysis of the connectivity and heat-life of a geothermal reservoir. Tracer testing as a dynamic characterization tool going beyond static models in order to better constrain inter-well-connections as well as thermal breakthrough. In addition, machine learning can be used to improve the interpretation of complex return signals of tracers. But also tracer testing has its limitations. Non-uniqueness of results cannot be solved by tracer testing itself. Additional data is required. the review is of the view that tracer testing is a form of dynamic characterization that can be added to static models, and Machine Learning can be used to interpret the tracer returns in terms of reservoir connectivity and heat lives.

Topic: Tracers

In this review we critically review recent advances of machine-learning in geothermal resource exploration and reservoir management, and assess their current stage of development for favorability, property estimation as well as for process optimization. The review covers supervised learning for favorability and property prediction as well as physics-constrained and hybrid modeling for optimization. The review discriminates between points regarding data-scarcity, validation and transferability to other applications. Data-driven exploration targeting and property prediction for geothermal fields can be supported with ML approaches. However, due to data-scarcity, results are sensitive to the degree of site-specific data incorporation and the corresponding disciplinary validation scope. Results also have limited transferability between sites. A Maturity map for tasks within geothermal applications is presented and it is argued that, in a data-scarce domain such as geothermal resources, the key to durable value from ML is the extent of physics constraints and disciplined validation, rather than the specific ML model used. A maturity map is developed for the application of machine learning in geothermal exploration and reservoir management. The value of machine learning models in geothermal is largely determined by the degree to which they are embedded in physical models and are subject to rigorous validation as opposed to specific model choices.

Topic: Enhanced Geothermal Systems

Successful development of naturally fractured hydrothermal systems requires drilling strategies that target connected, permeable fault and fracture networks while accounting for uncertainty in geology, temperature, hydraulic properties, and drilling response. We are developing the G2-Steering Agent, an initial proof-of-concept agentic AI framework that integrates geologic models, thermo-hydrological simulations, machine-learning surrogates, drilling measurements, uncertainty quantification, and trajectory optimization into a closed-loop workflow for geothermal geosteering. The AI-agent coordinates specialized tools and sub-agents to evaluate pre-drill trajectories using predicted injectivity, productivity, thermal breakthrough, heat extraction, and operational constraints, and to update steering decisions as measurement-while-drilling and logging data provide new information on fault and fracture probability, intensity, and orientation. As an initial test environment, we use data from the Lightning Dock geothermal system that incorporate faults, stratigraphy, observed well temperatures, anisotropic permeability, and coupled flow and heat transport. These data-to-decision models also expose important uncertainties, including spatially variable temperature misfit and imperfect representation of hydrothermal upflow, providing a realistic setting for testing uncertainty-based decision making. Rather than replacing physics-based modeling or expert judgment, the G2-Steering Agent acts as an orchestrator that connects and coordinates these capabilities. It evaluates multi-physics and geomodel reliability, compares alternative subsurface interpretations, and invokes physics-based simulations or machine-learning surrogates as needed. Based on the available evidence and quantified uncertainty, the agent recommends candidate well trajectories with traceable physical justification and confidence bounds. This provides a pathway toward adaptive, performance-based geothermal geosteering.

Topic: Modeling

Enhanced geothermal systems (EGS) require predictive tools that can integrate field observations, multiphysics reservoir simulations, uncertainty, and rapidly evolving operational data to support reservoir characterization, forecasting, and management. High-fidelity numerical models provide the physics needed to understand coupled thermal and hydrological processes, but their computational cost can limit the large ensembles required for calibration, sensitivity analysis, uncertainty quantification, and near-real-time forecasting. To address this challenge, in Year-1, we are developing EGS-Twin-v0.1, a proof-of-concept AI-enabled digital-twin framework that integrates field data, physics-based simulation, machine-learning surrogates, uncertainty quantification, advanced visualization, and ultimately agentic workflow orchestration. The initial demonstration focuses on Fervo Energy’s Project Red in Nevada. Field injection and production datasets are curated and transformed into inputs and observational targets for PFLOTRAN-based multi-physics simulations. The modeling strategy deliberately progresses from simple, interpretable conceptual models toward field-scale representations. A baseline conceptual model consisting of a single hydraulic-fracture stage was first developed to evaluate reservoir connectivity, pressure response, heat transport, numerical robustness, and computational requirements. The same framework was subsequently scaled from one stage to 2-, 4-, 8-, and 16-stage representations of the full lateral, reaching approximately 6.55 million computational cells. An important outcome of this initial work is demonstrating that field-scale simulations can be performed at computational costs compatible with systematic ensemble generation. Through numerical and timestep optimization, the one-stage 41.7-day simulation was reduced from approximately two hours to about 20 minutes, while the 16-stage, 6.55-million-cell model can be simulated in approximately three hours using 256 CPU processes. These results provide the computational foundation for generating PFLOTRAN ensembles that vary uncertain matrix, fracture, and well properties; identifying controlling parameters through sensitivity analysis; and calibrating reservoir models against measured production and pressure responses. EGS-Twin-v0.1 is being designed as a modular workflow in which these physics-based ensembles form the training and validation foundation for fast AI surrogate models. Candidate approaches include convolutional architectures, Fourier Neural Operators, and physics-informed models implemented using NVIDIA GPU-enabled tools such as PhysicsNeMo. The resulting surrogates are intended to accelerate parameter exploration, forecasting, history matching, and uncertainty quantification while retaining connection to the underlying high-fidelity simulations. In parallel, PFLOTRAN outputs are being translated into OpenUSD-compatible representations that separate reservoir, fracture, and well components and support visualization of time-dependent temperature, pressure, and tracer fields. Initial PFLOTRAN-to-OpenUSD prototypes have already demonstrated interactive representation of wells, fractures, and reservoir state variables needed for NVIDIA Omniverse platform. In Year-2, our vision is to connect these capabilities through an agentic EGS-Twin orchestration layer that can coordinate data processing, physics-based simulation, ensemble design, surrogate prediction, uncertainty assessment, and visualization via NVIDIA Omniverse for field sites in Nevada and Utah. This progressive workflow provides a pathway from high-fidelity reservoir simulation toward computationally efficient digital twins capable of incorporating new observations, updating forecasts, and supporting adaptive management of geothermal reservoirs. The Project Red demonstration establishes the initial foundation for extending the framework to increasingly complex EGS representations, including pre-existing fractures, additional field datasets, and coupled thermal-hydrological-mechanical-chemical processes.

Topic: Enhanced Geothermal Systems

Financing is one of the main barriers to a faster global buildout of geothermal energy, and the cost of capital is among the largest cost items of most projects. The difficulty is concentrated in the phases before the resource is confirmed: exploration, confirmation drilling and flow testing absorb a substantial share of total capital while well productivity, temperature and long-term reservoir performance are still uncertain. Investors in these phases wait many years for returns and can rarely sell a project that is not yet fully developed. Other industries handle similar problems: oil and gas companies spread subsurface risk across large portfolios of wells, mining has an active market for pre-production projects, and wind and solar are financed at low cost despite high upfront investment. Comparable solutions are not widely available in geothermal, and the existing evidence on geothermal financing is mostly grey literature covering few regions and periods. We study deep geothermal financing across the project lifecycle in two steps. First, semi-structured interviews with about 40 developers, investors, insurers and advisors identify the most widespread financing problems, from which we derive hypotheses about their causes and possible solutions. Second, we assess these solutions with a model of how much capital is spent in each phase, from exploration and confirmation drilling to construction and operation, and how long each phase takes. Model inputs come from the literature and desk research and are verified in the interviews. Using levelized cost of electricity (LCOE) modelling, we show how policies and new technologies change development times, risk levels and the cost of capital, and thereby generation costs. The interviews so far point to three hypotheses. First, the bottleneck lies in financing projects, not technologies: investors are willing to back technology companies, but few are prepared to carry the risk of individual projects. Second, capable project developers are scarce, so that actors from other parts of the value chain step in as developers, such as the equipment supplier Ormat or the technology firm Fervo. Third, a portfolio approach to risk is hard to achieve, because it would require many projects and a large capital base, while little experience carries over between projects and markets. We also ask whether enhanced geothermal systems (EGS) and closed-loop systems substantially change this picture. Both aim to replace site-specific exploration risk with engineered and repeatable well designs, but shift risk towards drilling cost, stimulation performance and long-term thermal output. We conclude with approaches to overcome the barriers identified and implications for public policy.

Topic: General

During exploration drilling campaigns, early predictions of reservoir temperature can improve targeting decisions and help to optimize dynamic campaign decisions. Fully stabilized downhole temperature measurements require months of waiting, but data while drilling, such as from maximum registering thermometers (MRTs), and from early-heat up surveys can be used mid-campaign. Corrections which are proposed based on a systematic review of dozens of drilling examples which can use early data to reliably predict the true reservoir temperature. These methods are particularly critical in small-diameter wells which cannot be reliably flowed. If a flow test is possible, temperatures from flow tests can also provide a highly reliable predicton of reservoir temperature in permeable zones, with appropriate correction.

Topic: Reservoir Engineering

The upstream phase of geothermal energy development—drilling and wellfield construction—is among the most resource- and emission-intensive stages in the geothermal value chain, yet it remains poorly quantified and documented in the African context. This study presents a cradle-to-gate Life Cycle Assessment (LCA) of geothermal wellfield development in the Olkaria geothermal field within the Kenyan Rift, focusing on embodied carbon and material flows for drilling, casing, cementing, and surface infrastructure. Using a process-based LCA framework in openLCA, the analysis estimates greenhouse gas emissions (kg CO₂e per MWh), cumulative energy demand, and key environmental indicators for the Olkaria 1AU 4&5 power plants. The study compares upstream emissions from geothermal with renewable s—wind and photovoltaic. Results show that upstream activities account for 15% to 30% of geothermal electricity's total life-cycle emissions, driven mainly by steel, diesel fuel, and cement consumption. Despite these contributions, geothermal systems exhibit a markedly lower environmental footprint—over 60% lower than the EU energy mix—and outperform comparable renewables such as PV and wind in overall environmental performance. Opportunities to reduce emissions include improving drilling efficiency, adopting low-carbon materials, and optimising well design. The findings provide quantitative evidence to support sustainable drilling practices and more accurate carbon accounting in geothermal development, contributing to ongoing discussions on geothermal energy’s environmental competitiveness within global decarbonisation pathways.

Topic: Modeling

As the global energy transition accelerates, Project Financing Institutions are increasingly tasked with funding a wider spectrum of geothermal resources. The established Probability of Discovery (POD) framework provides a quantitative assessment of exploration risk. However, its current grading parameters implicitly favor high-enthalpy, vapor-dominated systems associated with classical volcanic heat sources and extensive surface manifestations. Consequently, structurally controlled, low-to-medium temperature systems are often penalized in the scoring matrix, resulting in low POD grades that could impact the project became unfavorable and have high risk profiles. This study aims to identify and adjust specific parameters within the existing POD framework to accurately reflect the commercial viability of low-to-medium temperature geothermal fields. By recalibrating the weighting of geological, geochemical, and permeability indicators, this research seeks to provide lenders with a modified assessment tool that maintains rigorous risk mitigation while ensuring lower-enthalpy projects remain favorable for project financing. The study utilizes pre-exploration drilling data from two low-to-medium temperature geothermal prospect areas. The original POD framework which evaluates temperature expectation, temperature probability, permeability, chemistry, and drilling probability is applied to both fields to establish a baseline score. Subsequently, a sensitivity analysis is conducted to test parameter adjustments. Key modifications include down-weighting the reliance on high-temperature surface manifestations, such as fumaroles, and up-weighting indicators critical to binary plant success, including deep fault permeability and the presence of neutral, benign reservoir fluids. The application of high-enthalpy POD parameters increase the risk of low-to-medium temperature geothermal resources. By selectively adjusting permeability and chemistry weightings, lenders can utilize a more applicable POD framework for low-medium temperature system. This recalibrated tool not only prevents the premature abandonment of viable projects but actively expands the pipeline of bankable geothermal investments, accelerating the deployment of renewable energy generation.

Topic: Low Temperature

Indonesia is entering a period of increasing geothermal development, with new working areas and exploration programs being pursued to support the country’s growing geothermal capacity. At the same time, socio-environmental issues are becoming an increasingly important consideration for geothermal projects. Community concerns, land-use conflicts, environmental sensitivities, cultural considerations, and regulatory requirements can become project bottlenecks even before a geothermal working area reaches the development stage. Such issues may affect exploration access, permitting, stakeholder acceptance, and the implementation of planned project activities, highlighting the need to identify potential constraints at an early stage. This study presents a GIS-based approach for screening socio-environmental constraints across geothermal working areas in Indonesia. The framework integrates spatial analysis with publicly available environmental, regulatory, and social information to provide an early understanding of surface conditions that may influence exploration and subsequent development. Four principal dimensions are evaluated: forestry status, biodiversity and critical habitat, community and social issues, and cultural heritage and Indigenous Peoples. The community and social assessment also incorporates spatial proxies for potential land-use and land-access complexity, including settlement distribution and built-up areas, without requiring predefined wellpad or infrastructure locations. Spatial intersections are used to identify areas requiring different levels of attention, while a semi-quantitative multi-criteria assessment is applied to compare the overall socio-environmental profiles of several geothermal working areas. The resulting assessment differentiates geothermal working areas with multiple overlapping constraints from those where constraints are more localized or comparatively limited. A Composite ESS Index provides a consistent basis for comparing the relative level of socio-environmental attention required during early project planning. The results demonstrate how a common screening framework can reveal substantially different surface constraints among geothermal working areas and provide an initial basis for adapting exploration and project preparation strategies accordingly. The framework is not intended to replace detailed environmental and social assessments, field investigations, stakeholder consultation, or project-specific land acquisition studies. Instead, it provides an early decision-support layer that complements conventional geological, geochemical, and geophysical resource assessment. By identifying potential socio-environmental bottlenecks before major exploration or development commitments are made, the approach can support exploration-area optimization, permitting strategy, stakeholder engagement, survey sequencing, and early infrastructure planning. The study demonstrates the potential of GIS-based socio-environmental screening to help Indonesian geothermal developers anticipate surface constraints and incorporate them into project planning before they become significant obstacles to project implementation.

Topic: General

Maintaining conductive fractures under elevated temperature and closure stress is essential for sustained fluid circulation in enhanced geothermal systems (EGS). Polymer-nanoparticle coatings offer a potential approach for modifying proppant surface and mechanical behavior; however, their influence on propped-fracture hydraulic performance under geothermal conditions remains insufficiently understood. This study investigated the performance of a graphene-epoxy-coated ceramic proppant using a combined experimental and modeling framework consisting of long-term fracture conductivity testing, parallel proppant-pack permeability experiments, and propped-fracture flow modeling. Long-term conductivity testing was performed following API RP 19D at 200 °C (392 °F), a proppant loading of 2 lb/ft², and closure stresses of 2,000, 4,500, and 6,000 psi. Uncoated 40/70 CarboLite provided fracture conductivities of 2,170, 1,638, and 1,338 md-ft, respectively. Graphene-epoxy-coated CarboLite exhibited corresponding conductivities of 2,705, 1,909, and 833 md-ft, with permeabilities of 130, 93, and 43 Darcy. The coated proppant therefore demonstrated higher conductivity at 2,000 and 4,500 psi but experienced a greater reduction as closure stress increased to 6,000 psi, indicating a stress-dependent coating response. Because the existing conductivity datasets were generated using different fracture substrates, parallel proppant-pack permeability experiments were incorporated under controlled conditions to more directly evaluate the influence of the coating on flow performance. The experimental results were integrated with propped-fracture flow modeling to evaluate how coating-induced differences in permeability, conductivity, and fracture width influenced fracture-scale fluid transport. This combined experimental and modeling approach provided a framework for evaluating both the potential benefits and stress-dependent limitations of graphene–epoxy-coated proppants for high-temperature geothermal applications.

Topic: FORGE

Proppants placed within enhanced geothermal system (EGS) fractures may experience substantially different thermal histories depending on their location relative to the injection well and the advancing thermal front. Proppants within thermally disturbed regions closer to the fracture inlet can experience pronounced cooling during injection and reheating during subsequent thermal recovery, whereas material farther from the injection region may remain closer to reservoir temperature. Repeated temperature excursions may affect proppant integrity, fluid–proppant interactions, and long-term hydraulic performance. This study comparatively evaluated the durability of CarboLite and graphene-epoxy-coated CarboLite under continuous and cyclic hydrothermal conditions. A 30-day experimental program compared both proppants under three conditions: pristine, continuous hydrothermal aging, and cyclic hydrothermal exposure. Isothermal samples were maintained at 220 °C to represent sustained high-temperature reservoir exposure. Thermally cycled samples were alternated between 220 °C and a lower temperature representative of injection-induced cooling, followed by reheating toward reservoir temperature. Extended residence periods at each temperature were used to promote thermal equilibration of the fluid–proppant system. The cyclic condition was designed to investigate repeated thermal perturbations within the thermally affected region of a propped fracture rather than reproduce a uniform temperature history throughout the entire fracture. Proppant-pack permeability was measured before and after exposure to quantify hydraulic-performance retention. SEM-EDS was used to evaluate particle-surface alteration, coating integrity, cracking, delamination, and other morphological changes, while ICP-OES analysis of recovered fluids was used to quantify elemental release associated with fluid–proppant interactions. Comparison of uncoated and graphene-epoxy-coated CarboLite under identical exposure histories enabled differentiation of changes associated with continuous hydrothermal aging from those induced by repeated thermal cycling. The study evaluated whether graphene-epoxy enhancement improved the physical, geochemical, and hydraulic durability of ceramic proppants subjected to thermal perturbations representative of the near-injection region of EGS fractures.

Topic: FORGE

Hydrothermal alteration mineralogy provides an indirect, in-situ record of the thermal history of a geothermal system, complementing measured downhole temperatures and surface gas geothermometry, although it is also influenced by lithology, fluid composition and reaction kinetics. This paper characterizes the carbonate and clay alteration assemblage preserved in drill cuttings from Korosi exploration well KW-03 (Kenyan Rift), using X-ray fluorescence (XRF), X-ray diffraction (XRD), polarised-light microscopy, scanning electron microscopy with energy dispersive spectroscopy (SEM-EDS). It also tests which factors control illite-smectite (I/S) development. The proportions of illite layers in I/S were estimated from XRD peak positions using Watanabe (1988) method as a semi-quantitative indicator of alteration grade. Alteration is low-to-moderate in intensity, beginning with chalcedony at 342 m below ground level (bgl) and transitioning to quartz from 590 m bgl. Carbonate mineralogy is lithology-controlled: near-pure calcite (≥95% CaCO3) dominates in basalt and basaltic andesite, dolomite in mugearite, and Fe-rich ankerite in trachyte and benmoreite. Clay content (1.6-23.1 wt%) and loss-on-ignition (1.1-7.5 wt%) increase progressively with depth to 1200 m bgl and are dominated by randomly interstratified (R0) illite-smectite (I/S) mixed layers. Illite proportion range between 39% to 4% and correlates positively with host-rock K2O content (1.4 – 4.9 wt%) but only weakly with measured temperature (117-147 °C). Because K2O is highest in the shallow cooler trachytes (550 – 850 m bgl) the effects of lithology and temperature cannot be fully separated, but the results strongly suggest that protolith composition is the principal control on I/S development. The absence of epidote and albitized plagioclase throughout with chalcedony, quartz and I/S rather than illite corroborates sustained subsurface temperatures below ~200 °C consistent with the present day measured temperatures of a maximum of 147 °C. Illite content therefore records host rock chemistry and should not be used as a stand alone geothermometer without lithological control

Topic: Field Studies

The Eburu Geothermal Project is operated by the Kenya Electricity Generating Company PLC (KenGen). It is located inside the Eburu Forest ecosystem and its corresponding wildlife dispersal areas. The dispersal areas include much of the vast private ranches that lie adjacent to the forest. To achieve its projected geothermal energy expansion plan from 2.4 to 60MW, the Eburu geothermal project requires to access suitable sites within Eburu Forest ecosystem for drilling and well testing. Using an inter-disciplinary approach, an environmental impact assessment report for the proposed activities is developed to provide a framework to guide managers in the conservation and sustainable management of Eburu Forest. A key component of this assessment report is the management plan, which spells out the anticipated impacts on the environment and their respective mitigation measures. Environmental management minimizes conflicts between different uses and users of the forest, by mitigating potentially destructive activities while ensuring that the forest’s values and objectives are enhanced. Consequently, joint preparation and implementation of the environmental management plan by KenGen and the Kenya Forest Service (KFS) ensure that both geothermal development and natural resources are managed sustainably.

Topic: Field Studies

+Fluid flow in fractured subsurface reservoirs depends strongly on the directional hydraulic properties of individual rock fractures. Understanding these properties is particularly important in enhanced geothermal systems (EGS), where preferential flow influences fluid circulation and heat extraction. Existing numerical approaches describe shear-induced aperture evolution through geometric reconstruction, surface contact mechanics, and particle-based simulations. Although particle-based approaches explicitly represent asperity damage, further assessment is needed to distinguish how progressive damage and contact redistribution influence the magnitude and principal orientation of permeability anisotropy. This study investigates the emergence and evolution of permeability anisotropy in a single rough fracture using a sequential discrete element method (DEM) and flow modeling framework. The Subspring Network Breakable Voronoi (SNBV) model in PFC3D represents frictional sliding, dilation, and asperity damage through grain-scale contact interactions and bond failure. Fracture aperture and contact distributions evolve through mechanical interactions under prescribed normal stresses and shear displacements. Mechanical behavior is assessed through material calibration and shear s. Spatial aperture and contact maps extracted from mechanically equilibrated configurations provide the inputs for steady, incompressible flow simulations. The two-dimensional Reynolds equation, based on the local cubic law, relates local aperture to hydraulic conductance under lubrication-flow assumptions. Independent in-plane pressure gradients and the corresponding mean flux vectors are used to estimate the effective permeability tensor, principal permeability directions, and anisotropy ratio for each fixed fracture configuration. These hydraulic measures are examined alongside dilation, contact redistribution, and damage localization. The analysis examines how shear-induced deformation and damage generate and modify preferential flow pathways, with particular attention to whether progressive damage amplifies, suppresses, or redirects permeability anisotropy under different normal stresses.

Topic: Enhanced Geothermal Systems

In high-temperature geothermal systems, heat is transferred from deep magmatic and crustal sources to the shallow reservoir through fluid circulation and conductive heat transport. Conventional geothermal reservoir models are limited to subcritical conditions. As such, numerical models for liquid-dominated systems typically approximate this heat input by prescribing bottom boundary conditions in the form of fixed hot upflow and background heat flux. These simplified boundary conditions do not explicitly represent deeper recharge processes or their response to reservoir pressure changes during production. This study assesses how well conventional subcritical boundary conditions approximate deep recharge processes. This is done by comparing subcritical simulations with a supercritical base case for a synthetic geothermal system. This supercritical model represents the heat source as a deep heat flux rather than a mass input. Simulations are performed using the Waiwera simulator and its supercritical air-water energy equation of state. The study evaluates differences in mass and energy transfer at the depth of the bottom boundary of the subcritical model, and how these inputs change in response to production from the system. It considers how differences in boundary conditions affect key parameters that are observed in the subcritical reservoir, including temperature, pressure and gas saturation, and the impact simplified boundary conditions have on forecast.

Topic: Reservoir Engineering

In an enhanced geothermal system (EGS), dominant flow pathways can lead to early thermal breakthrough and premature reservoir cooling. Therefore, identifying these pathways is essential for commercial EGS operations, and distributed fiber-optic sensing (DFOS) provides an effective diagnostic approach for this purpose. This study presents a cluster-level inflow analysis of the Utah FORGE EGS using real-time fiber-optic monitoring during a week-long circulation test conducted in August 2025. We analyzed distributed temperature sensing (DTS) and Rayleigh frequency shift distributed strain sensing (RFS-DSS) data from producer 16B(78)-32 to identify flowing fractures and quantify their flow contributions. The normalized first spatial derivative of DTS and depth comparisons with RFS-DSS enabled identification of flowing fractures, while a DTS-based mixing method quantified their inflow contributions. We then used the identified fracture count as an input to an infinite fracture model to forecast thermal breakthrough. The analysis identified flowing fractures, including at least one fracture within the open-hole section. The estimated flow contributions showed good agreement with the 2024 PLT and 2025 thermal slug at the stage level, except for Stages 2 and 5. The results demonstrate that flowing fracture pathways and their flow contributions can be characterized largely from fiber-optic production logging data. The identified number of fractures and their flow contribution provide significant information for forecasting thermal breakthrough and evaluating reservoir-scale thermal behavior.

Topic: FORGE

This study presents a two-phase water/vapor wellbore production model to investigate the potential to reduce pumping power in enhanced geothermal systems (EGSs), with applications to the Utah FORGE EGS. Although Utah FORGE currently maintains single-phase liquid production, two-phase production may reduce the hydrostatic head in the production well and thereby lower the injection pumping power at the surface. We developed a two-phase wellbore flow model using the drift-flux approach to simulate pressure, temperature, and vapor quality along the production well under various production conditions. Model validation showed good agreement with available field production data, and we subsequently used it to investigate two-phase production scenarios in the Utah FORGE EGS with different vapor qualities. The resulting wellbore pressure and temperature profiles are used to evaluate changes in bottomhole flowing pressure and the associated pumping power requirement. Overall, this study provides a quantitative framework for predicting wellbore flow behavior and evaluating the potential reduction in pumping power associated with two-phase production. The framework can support the assessment of two-phase production strategies in future EGS developments.

Topic: Enhanced Geothermal Systems

We present a data-driven framework for high-resolution mapping of subsurface temperature across the United States using Google AlphaEarth embeddings combined with neural-network regression. AlphaEarth provides multi-modal geoscientific embeddings derived from large geospatial datasets, enabling the integration of geological, geophysical, and remote-sensing information into a unified representation of subsurface conditions. Using available temperature-at-depth and geothermal datasets as training data, we learn the relationship between AlphaEarth embeddings and subsurface thermal structure, and predict temperatures from the surface down to 7 km depth at significantly higher spatial resolution than existing national models. Related analysis indicates that AEF embeddings retain geophysical information unevenly across spatial scales and target types. We therefore evaluate whether the predicted temperature fields preserve both broad regional trends and localized anomalies, and identify where sparse temperature observations remain necessary to constrain unresolved heterogeneity and geographic transfer. Among several machine-learning approaches tested, neural-network regression provides the best predictive performance and spatial consistency. The resulting maps capture regional geothermal trends and localized thermal anomalies across the continental United States, providing an efficient tool for geothermal exploration and regional thermal resource assessment. This study demonstrates the potential of foundation-model embeddings to accelerate large-scale subsurface characterization and geothermal resource exploration by leveraging heterogeneous geoscience data.

Topic: Emerging Technology

Geothermal energy has the potential to support U.S. energy goals, but project development can be affected by locally specific concerns, information needs, project histories, and trust. Current engagement approaches often lack systematic ways to connect changing public discourse, local conditions, stakeholder perspectives, and engagement planning. This project is developing an integrated, data-driven framework to support transparent and context-aware engagement for geothermal project developers and government agencies. The framework connects three research components. First, a multi-source perception-monitoring workflow applies natural language processing to news and social-media data to identify prominent themes, concerns, and regional and temporal patterns. Second, a site-context and stakeholder-mapping framework integrates socio-political, community, market, technical, and regulatory evidence. Third, a strategy library and future machine-learning-enabled platform will connect site profiles with potential engagement approaches and incorporate new data and community feedback over time. This paper presents the overall project architecture and its early implementation across the three components. This work integrates peer-reviewed research, geothermal reports, regulatory records, and evidence from analogous energy projects. A human-oversight, AI-assisted workflow preserves source passages, provenance, counter-evidence, uncertainty, and quality-control decisions while supporting consistent analysis of place compatibility, perceived risks and benefits, procedural fairness, and stakeholder perspectives. Evidence is evaluated at the project level before its relevance to geothermal settings is assessed. The resulting evidence products support perception monitoring, site characterization, and strategy development, rather than automatic judgments of community acceptance. The framework is intended to help practitioners identify information gaps, relevant stakeholders and local conditions, and engagement approaches that merit further consideration. It does not infer community consent or replace direct engagement. Ongoing work will evaluate the framework through geothermal cases and use interviews and surveys to contextualize patterns identified in public data.

Topic: General

Induced seismicity poses a critical challenge for enhanced geothermal systems (EGS) reservoir management because larger events and their ground motion can affect public safety, operations, and social acceptance. Reliable probabilistic seismic hazard analysis (PSHA) requires forecasts of seismicity rates, locations, magnitude distributions, and site-specific shaking. Physics-based models require detailed subsurface characterization and can be computationally expensive, while conventional statistical approaches often struggle to capture nonlinear relations among operations, reservoir conditions, and seismicity. We develop a machine-learning-based PSHA framework that combines QuakeCastNet, a deep-learning seismicity-forecasting model, evolutionary intelligent agents for adaptive magnitude-frequency relationships, and data-driven ground-motion prediction (CGM-FAS). QuakeCastNet integrates a modified Temporal Fusion Transformer for time-series forecasting with extension to spatiotemporal prediction. It assimilates geological information, historical seismicity catalogs, and injection and production records to estimate the evolving seismicity rate and spatial distribution. Evolutionary intelligent agents explore candidate magnitude-frequency formulations conditioned on latent reservoir states inferred by QuakeCastNet. This equation-discovery approach yields interpretable relationships for how magnitude distributions evolve with operational forcing and reservoir conditions, providing probabilistic estimates of moderate-to-large events. To translate event forecasts into shaking hazard, CGM-FAS uses a conditional variational autoencoder to predict multi-frequency Fourier amplitude spectra at instrumented and uninstrumented locations. The model represents source, path, and site effects and quantifies uncertainty, allowing site-specific ground-motion estimates for candidate events. We demonstrate the framework using seismic and operational datasets from Utah FORGE and The Geysers, including ground-motion evaluation at Utah FORGE. By integrating deep learning, evolutionary equation discovery, and data-driven ground-motion prediction, the framework connects time-dependent seismicity forecasts with expected shaking for ML-based PSHA. It provides an interpretable, adaptive basis for traffic-light thresholds and safer geothermal reservoir management.

Topic: Enhanced Geothermal Systems

Enhanced geothermal systems (EGS) evolve through coupled fluid flow, fracture activation, stress redistribution, and deformation. These processes control reservoir performance and induced-seismicity risk, yet they cannot be characterized adequately by any single seismic sensor type. Conventional monitoring often emphasizes earthquake detection and location, leaving limited resolution of the evolving fracture network, source process, and surrounding rock properties. We present a multi-sensor seismic-monitoring framework that combines backbone regional seismometers, dense surface geophone arrays, deep borehole three-component sensors, and borehole distributed acoustic sensing (DAS). Each sensor type provides a distinct constraint and different advantages. Recent dense surface deployments and deep borehole instrumentation create an opportunity to move beyond event catalogs toward monitoring subsurface evolution. Joint analysis of seismicity, wavefield attributes, distributed-fiber measurements, and deformation can constrain changes in attenuation, fracture properties, rupture processes, and fluid pathways. These observations provide complementary evidence for interpreting whether stimulation activates existing structures, creates new flow paths, or redistributes stress through the reservoir. We discuss applications to EGS monitoring, including recent multi-sensor deployments, and identify the data integration and modeling requirements for converting heterogeneous measurements into actionable reservoir characterization. The framework supports improved stimulation design, induced-seismicity risk management, and data-informed adaptive traffic-light systems.

Topic: Enhanced Geothermal Systems

Make-up wells sustain generation capacity in mature geothermal fields, and a failed make-up well is a multimillion-dollar loss. At Darajat, a vapor-dominated field in West Java, Indonesia, a well is productive where it intersects feedzones, the effective fractures that carry steam at commercial rates. Feedzone occurrence is controlled by the major structures, an intrusion body, the volcano-stratigraphic unit, and wellbore orientation relative to the maximum horizontal stress (SHmax). Prior to this study, steam rate was forecast with a statistical well deliverability tool that drew on nearby wells and the general SHmax orientation. Geological concepts informed target selection through qualitative interpretation and expert judgment. Their influence was not quantitatively incorporated into the deliverability tool, because the geological model then available did not demonstrate statistically significant relationships between geological features and feedzone occurrence. New interpretations of fault dips, stratigraphy, and intrusion models between 2019 and 2023, initiated by newly available well data, motivated a reassessment. The reassessment augments the existing deliverability tool with machine learning trained on the Darajat feedzones and geological features. The drilled footage was discretized into 10,178 five-meter sections, of which 143 carry feedzones confirmed by flowing pressure-temperature-spinner surveys. Seventeen candidate geological features were screened against feedzone occurrence by chi-square testing at p below 0.05. Five passed, namely distance to the Kendang and Gagak faults, distance to the intrusion, volcano-stratigraphic unit, and wellbore orientation relative to SHmax. Productivity index was screened separately by analysis of variance at the same threshold and retained a different set, because distance to the Kendang fault governs whether a feedzone occurs but not how productive it is. Two decision tree models were then trained on the features that passed each screen. One gives the probability that a section holds a feedzone, the other the productivity index distribution of that feedzone. Trees were chosen for interpretability, and a configuration was accepted only when every terminal leaf described a geologically sound scenario. Tree outputs replace the bearing-chart feedzone count and the offset-well productivity inside the deliverability tool, and the Monte Carlo engine, the wellbore model, and the pressure-temperature profile are unchanged. The augmented workflow was applied to the five-well 2024 Darajat make-up campaign, with forecasts issued before drilling. All five wells were productive, with four performing above P50. Compared in parallel on the same wells, the augmented workflow reduced aggregate forecast error by 38% than the legacy. Well targeting improved, but predicted and actual steam rates still differ in some wells. Current work examines the geological and operational controls behind the spread between P10, P50, and P90 outcomes.

Topic: Field Studies

Thermal breakthrough, the onset of sustained cooling in produced water, determines how long an enhanced geothermal system can sustain its output. While fundamental to project economics, it is rarely recognized at the moment it begins. Two problems stand in the way. The first is that the downhole cooling that matters, a few hundredths of a degree per day, is far smaller than the changes measurable at the wellhead. Since heat lost to the formation depends on flow rate, a change in rate can look like breakthrough. At a constant production rate the rock around the wellbore above the producing zones keeps warming for a finite period of time, and this warming can mask real cooling. Secondly, the wellhead receives the combined flow of every producing zone. Cooling that begins in a single fracture may therefore remain difficult to discern at the surface until it is substantially developed. We address both problems in the 2026 extended circulation test at Utah FORGE using two complementary workflows. The wellhead workflow removes the influence of the wellbore so that reservoir cooling can be judged at surface. Wellhead temperature is first corrected with a transient wellbore heat-transfer model. Onset of cooling is then declared only when independent change-point detectors agree with a physics-based fit. The downhole workflow looks beneath the commingled signal to find where cooling begins. Fiber inside coiled tubing in the production well is interrogated with Rayleigh frequency-shift distributed strain sensing and is a surrogate for thermal changes in the tubing–casing annulus. Calibration against a bottomhole temperature gauge and measured wellhead temperature converts this response into temperature change at each perforation cluster. Unplanned shutdowns provided in-situ checks on both workflows. We test the workflows on synthetic breakthrough cases and on field records from tests with and without thermal decline and check the downhole estimates against production logs and tracers. We show whether cooling in individual zones leads to the measured commingled wellhead signals and by how much. The wellhead workflow needs no downhole instrumentation and can be applied to any EGS circulation test. The downhole workflow shows where cooling starts and how much warning the wellhead alone would miss. Together they turn production temperature from a lagging indicator into an early and defensible basis for deciding when to stop pumping in the circulation test.

Topic: FORGE

The Great Basin contains the United States' largest identified geothermal resource, with surface features like hot springs, fumaroles, and altered ground historically guiding explorers to almost every currently producing field. However, undiscovered geothermal systems leave no such surface expressions. Uncovering these hidden resources requires predicting subsurface temperatures, though temperature data alone cannot pinpoint drill sites. Because conductive heat transport through rock varies smoothly with regional crustal structure, it contrasts sharply with localized advective heat carried by fluids along permeable fault networks—which production wells must precisely intersect. Traditional mapping methods that extrapolate surface heat flow using assumed conductivity profiles lump these two mechanisms together and fail to isolate the advective component. To untangle this, we calculated temperatures at depth across 10,587 grid nodes throughout Nevada and Utah to successfully decouple the conductive background from advective signatures and rank the region for hidden potential. By feeding 407 Curie-point depths and 3,874 conductivity measurements into a physics-informed neural network, we solved the steady conduction equation across the entire crustal column while holding back 2,847 independent heat-flow observations for unbiased validation. Each node was then scored based on predicted heat flow, estimated temperatures at a 4 km depth, and local Quaternary fault density. Our analysis revealed that fluid advection—rather than simple conduction—drives regional resource distribution, with observed heat flow surpassing the conductive baseline at 65% of our validation sites (exceeding it by a median of 19 and up to 179 mW m⁻², and clearing 20 mW m⁻² across 34% of the province). Furthermore, predicted temperatures at 4 km surpass 200 °C across 48% of the region and 250 °C across 28%. Notably, our ranking successfully flagged 78 known Nevada geothermal systems at a median percentile of 75, including 28 systems completely devoid of surface expressions at a 76th percentile, proving the model's ability to uncover blind resources. We identified 25 top-tier prospects sitting squarely within the top one per cent of the province, every single one anchored to a mapped Quaternary fault with predicted reservoir temperatures of 293–400 °C across a 2,537 km² footprint. This direct spatial alignment confirms that fault permeability acts as the primary vertical conduit delivering deep heat to drillable depths. As an independent blind test using withheld data, our model accurately predicted the temperature profile of a deep well near Utah FORGE—situated 19 km from one of our identified thermal peaks—matching a 2.3 km equilibrated log within a tight 10.7 °C root-mean-square.

Topic: Geology

Indonesia’s geothermal power development has mostly relied on high-temperature (≥230°C) volcanic systems, where self-flowing wells and flash or dry-steam plants are commonly used. This experience has shaped assessment practice. However, future development in Indonesia is expected to include medium-temperature resources (190 to less than 230°C), including outflow zones, peripheral systems, and fault-controlled reservoirs. These resources should not be assessed as smaller high-temperature flash fields because their development is constrained by different reservoir deliverability and conversion characteristics. This paper examines how stored heat Monte Carlo can be applied more appropriately to medium-temperature resource assessment. The method remains useful as an early-stage screening tool, especially where subsurface data are limited, but its results can be misleading when assumptions inherited from high-temperature flash projects are applied without adjustment. Common issues include using one large resource volume, inappropriate cut-off temperatures, optimistic recovery factors, fixed or flash-like conversion efficiency for binary plants, and high reservoir temperatures across the entire prospect without support from conceptual model. Practical adjustments are proposed, including selecting cut-off temperature based on intended plant technology and final usable reservoir temperature rather than reinjection temperature, selecting recovery factors from relevant analogue fields and expected permeability regime, using binary net conversion efficiency that accounts for parasitic loads, and differentiating reservoir thickness and temperature between upflow, proximal, and outflow zones where supported by the conceptual model. A multi-volume stored heat Monte Carlo approach is recommended, in which all volumes are sampled together within each iteration and P90, P50, and P10 are derived from the total distribution. This avoids spreading high-temperature assumptions across the whole resource area and prevents incorrect summation of percentile values from separate simulations. The Kotamobagu (North Sulawesi) and Momotombo (Nicaragua) case studies suggest that this approach can provide reasonable early-stage resource estimates before detailed numerical reservoir modelling is conducted. This is indicated by general agreement between multi-volume stored heat estimates, numerical model results, and actual stable power generation. Nevertheless, stored heat Monte Carlo estimates heat-based potential and does not directly calculate permeability heterogeneity, well productivity, pressure support, thermal breakthrough, or downhole pumping requirements. Numerical reservoir modelling remains the appropriate next step for assessing these aspects in detail.

Topic: Reservoir Engineering

Hybrid Semi-Well Loop Systems (HSLS), proposed by Ott (Ott and Dorman 2026), use hydraulically fractured vertical wells in which a sealant is injected to create a zero-permeability barrier within the propped-open fracture. This barrier redirects water injection along the propped-open fracture, increasing thermal exposure to the formation while limiting leakoff and mitigating thermal short-circuits. This study evaluates an idealized HSLS under conditions representative of Utah’s Frontier Observatory for Research in Geothermal Energy (FORGE). The hydraulic-fracture geometry is initially characterized using a Perkins-Kern-Nordgren (PKN) model representation and subsequently extended using a pseudo-three-dimensional (P3D) (Wang et al. 2023) formulation to account for variations in fracture geometry as input to reservoir simulation. It assesses whether the barrier increases fluid residence time within the fracture, improves reservoir thermal sweep, and delays thermal breakthrough, and how performance depends on the length and distribution of the barrier within the fractured zone and the well path. Preliminary simulations of continuous injection show that the barrier delays thermal breakthrough and reduces the decline in produced-fluid temperature by nearly 20% compared to the injection-production configuration without the barrier, indicating improved heat exchange. The results identify geometric configurations that reduce short-circuiting and increase heat exchange between the circulating fluid and the surrounding hot rock, providing first steps forward into an HSLS at FORGE.

Topic: Enhanced Geothermal Systems

The Paradox Basin of southeastern Utah, southwestern Colorado, southeastern Arizona, and northwestern New Mexico hosts deep sedimentary brines with lithium concentrations that may support brine-based extraction. This DOE-funded project integrates existing data compilation, new field sampling, geochemical modeling, and techno-economic analysis to evaluate the basin's lithium resource potential, with emphasis on the Mississippian Leadville Limestone and the Pennsylvanian Paradox Formation. Compilation of publicly available data identifies the Leadville Limestone as the most promising target, combining favorable permeability with lithium concentrations spanning a median of 13 mg/L to a 95th percentile of 78 mg/L. A measured value of 75 mg/L from the Lisbon B 815 well exceeds commonly cited economic thresholds and occurs in a structurally active area where salt-cored anticlines and listric faults connect the Leadville to overlying Paradox Formation source rocks. The organic-rich black shales of the Paradox Formation are interpreted as the primary lithium source, releasing lithium through smectite-to-illite conversion and kerogen cracking during burial, with subsequent downward migration and fault-controlled delivery into the Leadville reservoir. Isotopic and groundwater-age data indicate brines are predominantly meteoric with a preserved ancient paleo-seawater component, while radiogenic strontium ratios suggest additional brine contributions from Precambrian basement. New field sampling confirms strong spatial variability in lithium content. A Leadville brine recovered by wireline swabbing from the shut-in Evelyn Chambers #1 well yielded a lithium concentration of approximately 54 mg/L, approaching economic thresholds, whereas six produced-water samples from the Greater Aneth area returned low concentrations ( less than 10 mg/L), and brines from the Sagebrush 1 helium well ranged 7.5–17.7 mg/L. In parallel, a modular techno-economic framework based on WaterTAP is being developed to evaluate direct lithium extraction pathways under dedicated-well and co-production scenarios. Future work will apply equilibrium geochemical modeling to quantify rock-to-brine lithium transfer and refine resource and economic assessments.

Topic: Geochemistry

Effective management of a mature geothermal field requires a good understanding of how injected fluids move through the reservoir and interact with producing wells. At Olkaria, reinjection is an important component of resource management, but the highly heterogeneous and fractured nature of the reservoir makes it difficult to predict subsurface fluid movement using well data alone. Tracer testing therefore provides an important means of investigating reservoir connectivity and assessing the possible implications of reinjection on production. This study presents the application of naphthalene disulfonate (NDS) tracers in the Olkaria geothermal field, Kenya, with emphasis on their use in understanding subsurface fluid movement and supporting reinjection management. Different NDS tracer species were injected into selected reinjection wells and monitored at production wells over time. The resulting concentration–time responses were evaluated to identify tracer breakthrough, estimate relative travel times and assess the degree of hydraulic connectivity between injection and production wells. The tracer responses demonstrate the heterogeneous nature of fluid movement within the Olkaria reservoir, with different production wells showing varying breakthrough characteristics. The observed responses provide useful information on preferential flow paths, reservoir connectivity and the potential influence of reinjected fluids on production areas. The study further demonstrates that tracer information becomes more valuable when interpreted together with well production history, injection rates, temperature and pressure data, and the geological understanding of the reservoir. The findings highlight the practical value of NDS tracers as a reservoir management tool rather than merely as a means of establishing well-to-well connectivity. Proper interpretation of tracer responses can support decisions on the location and distribution of reinjection, help identify areas where early returns may present a risk of thermal breakthrough, and provide information for improving conceptual and numerical reservoir models. Continued integration of tracer testing with routine reservoir monitoring is therefore recommended to support adaptive reinjection strategies and sustainable long-term utilization of the Olkaria geothermal resource. Keywords: Olkaria, geothermal reservoir, Naphthalene Disulfonate (NDS), tracer testing, reinjection, reservoir connectivity, resource management, sustainability.

Topic: Tracers

Hyperspectral image spectroscopy (HIS) has revolutionized the spectral interpretation of minerals by providing non-destructive, high-resolution spectral data for precise mineral identification and mapping. This study explores the application of (HIS) in determining the wavelength of minimum reflectance, a critical spectral parameter for distinguishing minerals based on their unique absorption features. Wavelength mapping using HIS enhances the accuracy of mineral classification and spatial distribution analysis, making it a valuable tool in geological exploration and remote sensing applications. Compared to traditional analytical techniques such as x-ray diffraction (xrd) and x-ray fluorescence (xrf), (HIS) offers a rapid, large-scale, and non-destructive means of mineral characterization. While xrd is highly precise in identifying crystalline structures and phase compositions, it requires physical samples and extensive laboratory preparation. Xrf, on the other hand, though non-destructive is effective in determining elemental compositions but lacks the capability to directly identify mineral phases. In contrast, HIS enables remote, high-throughput mineral identification based on spectral signatures, making it particularly advantageous for field-based exploration, planetary studies, and resource assessment. The integration of (HIS) with xrd and xrf provides a comprehensive approach to mineral characterization, leveraging the spectral mapping capabilities of HIS alongside the structural and elemental insights offered by xrd and xrf. This synergy enhances the accuracy of geological interpretations and supports advanced mineral exploration and environmental studies. This study was focused on the analysis of a lithium pegmatite rock sample and the findings underscore the growing importance of hyperspectral data in complementing conventional mineralogical techniques for more efficient and precise geoscientific analyses. The minerals that were identified from the sample included; lepidolite, muscovite, topaz, quartz and albite

Topic: Geology

Blind (or hidden) geothermal systems present a major exploration challenge, because they lack obvious active surface manifestations commonly used to identify geothermal resources. Granite Springs Valley (GSV), located in northwestern Nevada within the Great Basin, contains one such hidden prospect. Previous investigations have documented elevated subsurface temperatures and delineated both paleo-geothermal deposits (sinter) and structural features associated with the geothermal system. Geological, geophysical, thermal, and modeling datasets have subsequently been integrated to improve characterization of the geothermal system. However, hyperspectral imaging has not been included in previous investigations at GSV. This study evaluates hyperspectral remote sensing imagery to provide additional surface mineralogical indicators for exploration of the blind GSV geothermal resources. A preliminary assessment has been conducted of the NASA AVIRIS (classic archive identified candidate flight) line footprints intersecting the study area. The publicly available Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) imagery is being used to characterize spatial variations in surface mineralogy across the prospect. The surface reflectance spectra are evaluated for diagnostic absorption features in the visible, near-infrared, and shortwave-infrared wavelengths, with particular attention to selected clay and other hydroxyl-bearing minerals, carbonates, sulfates, silica-bearing materials, and iron-bearing minerals. The hyperspectral-derived mineralogical patterns are being evaluated in the context of independent geological, structural, thermal, paleo-geothermal, and geophysical information. Additionally, an emphasis is placed on whether spatially coherent mineralogical assemblages correspond with independently identified features of the geothermal system rather than interpreting individual minerals as diagnostic of geothermal activity. This distinction is important, because similar surface materials may result from lithology, weathering, evaporative processes, or other non-geothermal processes. This study tests the value of hyperspectral imagery to provide key mineralogical context and augments established exploration datasets at a well-characterized hidden geothermal prospect. Results will help assess whether surface mineralogical information derived from imaging spectroscopy can provide an additional exploration constraint for hidden geothermal systems in the Great Basin.

Topic: Geology

In geothermal wells in the Olkaria, Kenya Rift Valley, smectite clays such as montmorillonite and nontronite remain the dominant cause of wellbore instability. When exposed to water-based drilling fluids, these clays rapidly hydrate, swell, and induce wellbore constriction, bit balling, differential sticking, sloughing, and pack-off. The resulting stuck-pipe events generate significant non-productive time and inflate the already elevated costs of geothermal drilling. Conventional mitigation relies on mechanical freeing techniques and spotting fluids, augmented by calcium chloride to promote clay shrinkage via cation exchange. Although inexpensive, calcium-based exchange yields only modest volume reduction in smectite interlayers, limiting its effectiveness. Potassium chloride delivers markedly superior shrinkage but is routinely excluded on economic grounds. Laboratory and field data demonstrate that ammonium chloride offers a superior, cost-effective alternative. When ammonium ions occupy the interlayer sites of smectite, they induce substantially greater lattice contraction than calcium ions, producing a more stable, collapsed clay structure. At the same time, ammonium chloride is considerably less expensive than potassium chloride, enabling routine deployment without compromising project economics. This paper presents ammonium chloride as a practical mitigation measure that directly addresses the root cause of clay swelling while maintaining operational affordability. Implementation in Olkaria-type rift settings is expected to reduce stuck-pipe incidents, shorten drilling times, and lower overall well-construction costs, advancing the economic viability of East African geothermal development.

Topic: Geochemistry

Coastal aquifers hydraulically connected to cold deep seawater offer a potential cooling resource for data centers through geologic seawater air conditioning (GeoSWAC). Sustainable cooling performance depends on subsurface connectivity, thermal replenishment, hydraulic response, and the temperature requirements of the cooling system. The study combines regional geological modeling, coupled groundwater flow and transport simulations, and analytical well-design estimates to assess hydrogeologic controls and operational implications for GeoSWAC in the Miami region. A conceptual geological model informs a two-dimensional numerical model of density-dependent groundwater flow, salinity transport, and heat transport. The analysis examines simulated hydraulic head, salinity, temperature, and pressure distributions, with qualitative comparison against published regional temperature patterns. Sensitivity tests assess the influence of hydraulic conductivity, porosity, thermal conductivity, and recharge, while analytical doublet calculations provide preliminary estimates of well spacing and production drawdown. Under the current model assumptions, preliminary steady-state simulations indicate landward penetration of cool saline water through deep permeable formations, with a modeled Boulder Zone temperature of approximately 8.5°C at the coastline. Sensitivity results identify hydraulic and thermal conductivity as influential controls on temperature at a selected offshore location, while salinity remains near seawater concentration. Idealized analytical calculations yield doublet-spacing estimates of 0.61–1.06 km for a circulation rate of approximately 50 kg/s over a 30-year design horizon, with small estimated aquifer drawdown under the assumed high-transmissivity conditions. The results highlight the importance of matching site-specific source temperatures and hydraulic conditions to data center cooling requirements and support the assessment of alternative subsurface-access configurations, supplementary cooling, and seasonal GeoSWAC utilization, providing a preliminary basis for ongoing transient modeling and subsequent techno-economic analysis.

Topic: Emerging Technology

Geothermal energy is a cornerstone of Kenya’s renewable energy portfolio, contributing to over 40% of the country’s electricity generation and supporting national efforts toward a low-carbon energy transition. As geothermal development expands across the Kenyan Rift, understanding its environmental performance is essential for ensuring sustainable resource utilization. This study applies a cradle-to-grave Life Cycle Assessment (LCA) framework to evaluate the environmental impacts associated with geothermal energy development in Kenya, with a focus on Menengai Field. The assessment was performed using the Geothermal Energy Impact Estimator (GEIE) and the Environmental Footprint 2.0 methodology. Sixteen environmental impact categories were evaluated, encompassing well drilling and construction, surface infrastructure, operation and end-of-life activities. The study aimed to identify the main contributors to environmental impacts and compare the results with published geothermal LCA studies from other geothermal producing regions worldwide. These findings provide a baseline for assessing the sustainability of geothermal development in Kenya and highlight opportunities for impact reduction through improved drilling practices, material optimization and long-term asset management. Lessons drawn from global geothermal case studies can support more sustainable geothermal expansion in Kenya and other emerging geothermal regions.

Topic: General

Power system is a major component of a geothermal drilling rig, its analytical data is a key element in drilling rig operations planning, well drilling costing and is a remarkable cause of non-productive time in drilling ahead operations. Paka geothermal field-Kenya utilises four (4) CAT 3512B diesel powered generators to provide the required energy for the high-power consuming geothermal drilling equipment and erratic loading encountered during drilling ahead operations. Power system monitoring records obtained during drilling operations indicate that the generator power system is operating under very low generated power factor, low active power (KW) and high apparent power (KVA). These power system inefficiencies coupled with erratic generator loading, generator derating factors, power conversion harmonics always requires the drilling crew to engage two generators at early stages of drilling operations resulting in increased fuel consumption by the generators and subsequent increase in overall operational and maintenance costs. This study was therefore carried out to help reduce fuel costs in a geothermal drilling rig by introducing reactive power compensator in the power system to improve the generated power factor. This improved power factor will thus enable running only one generator instead of two generators. The system was then monitored and evaluated on generator output parameters, reliability of equipment during drilling operations, effect on generator loading changes, operational cost, maintenance cost and generators availability in comparison to the older configuration of running two generators during the drilling ahead operations. Application of result of the study will assist in maximum utilization of capacity of geothermal drilling power generators in remote areas, reduction in overall operating cost by reducing fuel costs of drilling operations and reduction in overall maintenance cost

Topic: Drilling

The UK’s first operational geothermal power plant, located at United Downs, Cornwall, started generating in February 2026 with the finalisation of connection and the ability to output to the UK’s National Grid. The project has been developed from conception by Geothermal Engineering Ltd (GEL). The installed plant is a 3 MW gross capacity binary cycle ORC facility which utilises a geothermal doublet drilled into a fractured granite reservoir. Natural, hot ( greater than 180 °C) saline waters are pumped to surface via a 5,275 m (measured depth) production well. Spent brine is reinjected into the system via a 2,393m (measured depth) injection well. Both wells were directionally drilled into the subvertical, NNW-SSE aligned Porthtowan Fault Zone, one of a number of similar ‘cross-course’ fault structures which span the region. The geothermal plant at United Downs provides a proof of concept for further projects across Cornwall and other geologically equivalent systems. The development proves the existence of inherent permeability within the fractured granite at depth, the presence of high temperature fluids within the granite and that low-pressure hydraulic stimulation can enhance reservoir properties. Furthermore, the system design offers a basis for the planned enhancement of future developments, through the drilling of sidetracks and better well cleaning/stimulation activities. Downstream of the power plant, GEL have developed a 100 tonnes-per-annum lithium carbonate extraction and production facility. Testing of the geothermal doublet in 2021 first confirmed high lithium concentrations within the saline waters. Regular analysis of the geothermal brine has proved lithium concentrations exceeding 350 ppm. The lithium extraction facility uses a portion of ‘spent’ geothermal fluid from the power plant at c. 50 °C, prior to its reinjection. The facility utilises an adsorption based DLE technology to extract lithium after pre-treatment and impurity removal. All removed material is reincorporated into the spent brine prior to reinjection. The lithium carbonate produced onsite is of ‘technical grade’, and possesses greater than 99.3 % purity, meaning that the product does not require subsequent refinement before being sold to market. GEL are currently in the process of scaling up lithium extraction on site (to c. 1000 tpa), supported by the UK Government’s DRIVE35 grant scheme. This is scheduled to come online in late 2029.

Topic: General

Application of Integrated Geophysical Methods for the Analysis and Interpretation of the Complexity of the Olkaria Geothermal Field Reservoir System Philip Omollo1, Jun Nishijima2, Colvin Murila1 1Kenya Electricity Generating Company (KenGen) PLC, Naivasha 20117, Kenya 2Faculty of Engineering, Department of Earth Resource Engineering, Kyushu University, Fukuoka 819-0395, Japan pomollo@kengen.co.ke Keywords: Resistivity, Gravity, Microgravity, Olkaria geothermal, reservoir system Abstract The Olkaria geothermal field, located within the Great Rift Valley in Kenya, is characterized by diverse geothermal manifestations, including geothermal grass, silica sinters, altered ground, and prominently steaming grounds and fumaroles. These fumaroles exhibit high temperatures ranging from 70°C to 90°C, indicating a substantial subsurface potential for geothermal energy harnessing and utilization. Most drilled wells in the area show high temperatures at reservoir depths greater than 200 ℃. Integrated geophysical methods have been employed to identify subsurface geological structures that govern the geothermal reservoir. These methods relate resistive and dense signatures, aiding in the retrieval of extensive subsurface geological information. The geothermal gradient in the Olkaria field is estimated at about 45°C per kilometer. The formations exhibit a resistive propylitic reservoir zone with high-density bodies, characterized by resistivity values greater than 70 Ωm and densities greater than 2600 kg⁄m^3 . Although most production wells in the Olkaria geothermal field have been drilled to depths of about 3 km, 3-D resistivity inversion models suggest the presence of a heat source at depths of approximately 5 – 8 km below the surface and a reservoir depth of about 2.5 – 4 km. This indicates potential for deepening production wells in the future. A structurally conspicuous anomaly trending NNE-SSW has been observed, connected to some high-producing wells in the Olkaria Domes field. Microgravity observations reveal mass changes across the field, with the primary natural recharge appearing to come from the northwest and east of the Domes field. The potential recharge zone is indicated by a high resistive signature, which may be associated with cold water infiltrating into the reservoir system.

Topic: Geophysics

The Olkaria I Additional Unit (OIAU) power plant, situated in Kenya’s Olkaria Geothermal Field, consists of three units with a combined installed capacity of 220 MW. Unit 4 was commissioned in October 2014 and unit 5 in January 2015 while unit 6 followed in July 2022. In recent years, Units 4 and 5 have experienced persistent operational difficulties, including turbine blade damage, frequent trips on high scrubber levels due to brine carryover, loss of condenser vacuum, and scaling and deposition primarily on the first and second turbine stages. Performance analysis has proven to be an effective condition monitoring technique, enabling early detection of degradation and supporting optimized operation, overhaul planning, and maintenance strategies. This approach shortens critical maintenance paths and informs timely decision making. Notably, vibration trend analysis prior to Unit 4’s unplanned 2024 outage for blade replacement clearly indicated progressive turbine degradation. Despite reservoir well flow tests confirming adequate steam supply to meet installed capacity, the units consistently fail to reach full load. Specific steam consumption has risen, requiring steam flows exceeding design values to achieve current output levels. Performance analysis is instrumental in determining whether this shortfall results from turbine inefficiencies caused by age related degradation, fouling, and damage, or from other limitations in steam delivery. This paper presents a detailed performance evaluation of the Olkaria IAU units based on key condition monitoring parameters that include vibration trends, available steam, Specific steam consumption capacity, steam quality, turbine chamber pressures, condenser vacuum among others. It also assesses the effectiveness of implemented modifications and corrective measures in mitigating identified issues and improving overall plant efficiency.

Topic: Production Engineering

Carbonate-hosted geothermal reservoirs can generate and release CO2 through coupled brine-CO2-rock reactions, particularly where calcite dissolves under high-pressure and high-temperature conditions. These reactions simultaneously govern scaling tendency, corrosion risk, permeability evolution, pore structure, and mechanical response, making CO2-equivalent generation an integrated reservoir problem rather than a purely geochemical one. This study develops new chemical treatments for reducing CO2 emissions from carbonate geothermal systems. Multiple chemical treatments were explored, including the injection of silica additives (silica fume, silica flour) and chelating-agent pre-flushes (EDTA, GLDA), using Red Sea water (RSW) and CO2-rich RSW as base main fluids. This work combined static chemical screening, HPHT disk aging at 70°C, and staged core flooding in dry carbonate cores (10 in length and 1.5 in diameter) at 100°C, 2000 psi confining pressure, 1500 psi backpressure, and 0.5 cm³/min injection flow rate. Phase 1 injected RSW baseline, a chemical slug, and an RSW post-flush; Phase 2 repeated this sequence using CO2-rich RSW as the baseline and post-flush fluid. Effluent was analyzed by ion chromatography and total inorganic carbon (TIC), with CO2-equivalent generation cross-checked against Ca2+/Mg2+ stoichiometry. Pre- and post-flood CT imaging, porosity, pore volume, dry mass, and dynamic elastic properties linked fluid chemistry to rock alteration, while outlet-temperature records were interpreted using local thermal equilibrium (LTE) and non-equilibrium (LTNE) models. Silica fume and silica flour emerged as the most effective mitigation additives. In HPHT screening, 0.1wt% silica fluids did not increase Ca2+/Mg2+ release relative to the base case and maintained near-neutral pH, whereas EDTA produced the strongest carbonate attack. In Phase 1 core flooding, 0.1wt% silica fume lowered cumulative TIC-based CO2-equivalent from 105.20 mg (RSW baseline) to 50.10 mg during the additive stage and 28.13 mg during post-flush; silica flour likewise sustained low CO2-equivalent output. In Phase 2, under CO2-rich RSW, 0.1wt% silica fume reduced the cumulative CO2-equivalent from 357.51 mg to 129.69 mg, and 0.1wt% silica flour reduced it from 248.42 mg to 109.65 mg. By contrast, 5wt% EDTA increased CO2-equivalent from 221.60 mg to 783.33 mg, confirming its role as an active carbonate-dissolving pre-flush rather than a suppressant, while 5wt% GLDA reduced CO2-equivalent from 399.57 mg to 279.30 mg but showed complexation-affected free-ion behavior. Post-flood CT and petrophysical data indicated only limited alteration and no severe sustained plugging across the silica runs, and LTE and LTNE models produced nearly identical outlet-temperature predictions, supporting LTE as an adequate outlet-scale descrption. The obtained results show that CO2-equivalent generation in calcite-dominated geothermal core floods is controlled by the combined effects of CO2-rich brine, carbonate dissolution, sulfate-related ion partitioning, chemical speciation, and thermal stabilization. Silica fume and silica flour reduce net inorganic-carbon release without causing severe flow impairment, while EDTA should be used only cautiously as a short pre-flush requiring metal complexation, and GLDA occupies an intermediate position requiring joint interpretation of TIC, pressure response, and rock-property change. The findings support a reservoir-centered chemical-selection framework for carbonate geothermal systems that weighs scale control against CO2-equivalent generation, injectivity, and rock integrity.

Topic: Reservoir Engineering

Rock integrity is a critical factor in the long-term performance of geothermal reservoirs. Chemical alteration of the host rock during brine circulation can progressively weaken the mechanical strength, reduce injectivity, and compromise wellbore stability over the life of a field. This study assesses the impact of four candidate treatment fluids (0.1 wt.% silica fume, 0.1 wt.% silica flour, 5 wt.% GLDA, and 5 wt.% EDTA) on the structural and mechanical integrity of carbonate rocks flooded at 100°C under both CO2-free and 5v/v% CO2-rich Red Sea water (RSW). The rock samples were characterized before and after the flooding experiments by conducting medical CT imaging, compressional and shear wave velocities (Vp/Vs), porosity, and permeability measurements. Medical CT imaging showed no wormholes or dissolution channels in core samples treated with silica fluids. GLDA-flooded cores showed a significant reduction in the CT number, indicating severe face dissolution. Flooding the core samples with EDTA fluid led to localized rock dissolution and a slight reduction in the CT number. However, the rock porosity changed by less than 0.3 percentage points in all tested cores, indicating that porosity alone cannot capture the loss of rock integrity. The dissolution mainly occurred within the intergranular cement rather than throughout the rock matrix. The elastic properties (Vp/Vs) proved more sensitive to the chemical injection. Young's modulus was reduced by an average of 7.5% post-flood across the treated cores. Young’s modulus decreased by an average of 7.5% after flooding across all treated cores. Poisson’s ratio increased from 0.27 to 0.32 for the EDTA-treated samples and from 0.18 to 0.21 for the GLDA-treated samples. The obtained results are consistent with grain-contact cement dissolution. Moreover, permeability showed a similar trend: it remained nearly unchanged after silica-fume treatment, decreased by 35% after silica-flour treatment due to fluid retention, and increased by 44% after EDTA treatment due to mineral dissolution. Overall, this study introduces an integrated approach to evaluate how fluid chemistry controls rock integrity during geothermal treatment. CT imaging and elastic properties were the most reliable indicators for evaluating rock integrity. Silica-based fluids preserved carbonate rock integrity better than chelating agents. These findings can guide fluid selection to maintain wellbore stability and support safe, long-term geothermal energy production

Topic: Reservoir Engineering

Influence of pH and Temperature on Silica Polymerization and Scale Formation in Olkaria Geothermal Fluids, Kenya Benedict Ouma Otieno Kenya Electricity Generating Company PLC (KenGen), Olkaria Geothermal Project, P.O. Box 785, Naivasha, Kenya Email: botieno@kengen.co.ke Abstract Silica scaling is a major geochemical challenge in geothermal power production, where changes in temperature, pH, dissolved silica concentration, and fluid composition can promote silica polymerization and mineral deposition. This study investigates the influence of pH and temperature on silica polymerization and scale formation in geothermal fluids from the Olkaria Geothermal Field, Kenya. Four production wells representing Olkaria II, IAU, IV, and V will be investigated. Geothermal-fluid samples will be characterized for temperature, pH, dissolved silica, and other relevant physicochemical parameters. Controlled laboratory experiments will evaluate silica polymerization and precipitation under selected pH and temperature conditions. Changes in dissolved silica concentration with time will be used to assess polymerization kinetics and scaling tendency. Where analytical facilities permit, reactive/monomeric silica and total silica will be determined to estimate the polymerized silica fraction. Aluminum and iron will also be investigated because of their potential influence on silica precipitation and scale formation. Previous studies have confirmed silica scaling in Olkaria and identified amorphous opal-A, quartz, and aluminum-enriched deposits in production systems including OW-34 and OW-35. Experimental studies have also examined silicic-acid and aluminum precipitation using Olkaria geothermal water, while other investigations have focused on thermodynamic scaling prediction. However, limited research has systematically evaluated the effects of controlled pH and temperature on silica polymerization using fluids from multiple Olkaria production sectors. This study addresses this gap by integrating field characterization, controlled laboratory experiments, silica-polymerization kinetics, and comparative geochemical analysis. The results are expected to identify conditions that promote silica polymerization and scale formation and provide a basis for improved silica-scale prediction, monitoring, and control in geothermal power production. Keywords: Geothermal fluids; silica scaling; silica polymerization; pH; temperature; dissolved silica; mineral precipitation; aluminum; geothermal geochemistry; Olkaria.

Topic: Geochemistry

Structural frameworks are central to the development and sustainability of active geothermal systems because they control permeability, fluid circulation, heat transfer, and reservoir compartmentalisation. This study investigates how structural features and their intersections enhance or restrict fluid flow within the Olkaria Geothermal Field, Kenya. The analysis integrates geological and geochemical data, temperature and pressure profiles, feed-zone characteristics, and well-performance data. It evaluates the relationship between structural orientation, fracture intensity, hydrothermal alteration, and the occurrence of productive permeability, with particular attention to the regional NW-SE trending structures and younger N-S and NE-SW structural corridors, as well as other structurally controlled sectors of the field. The results indicate that structures at Olkaria exhibit variable hydraulic behaviour and cannot be treated uniformly as either fluid conduits or barriers. Favourably oriented, fractured, and interconnected structures provide pathways for deep up-flow and lateral outflow, whereas unfavourable stress conditions restrict flow and compartmentalise the reservoir. The findings demonstrate the importance of integrating structural interpretation with well and reservoir data to distinguish productive structures from non-productive or sealing features. This approach improves geothermal conceptual models, reduces drilling uncertainty, optimises production and reinjection well targeting, and strengthens long-term reservoir-management strategies for sustainable geothermal development at Olkaria.

Topic: Geology

The geothermal wellbore is where reservoir heat becomes deliverable energy or fails to, flashing depth, heat loss, and flow stability all negotiated over a few kilometres of pipe. We review geothermal wellbore flow and heat transfer: the two-phase modeling record against measured profiles, flashing-depth prediction and its sensitivity chain, flow-stability boundaries in cycling service, and the completion-design levers, insulation, diameter staging, that the physics rewards. The wellbore is the component where geothermal projects convert reservoir promise into plant feed, and its modeling record deserves the validation grading it receives here. Wellbore-simulation approaches were assembled with validation against measured pressure-temperature profiles, the scarce but decisive data, flashing-depth sensitivity examined across the input chain, productivity, enthalpy, wellhead pressure, stability evidence reviewed from cycling and throttled operation, and machine-learning wellbore-model surrogates surveyed. Validation against measured pressure-temperature profiles is the organizing standard, scarce data given the decisive role it deserves. Model skill was genuine for stable production and degraded exactly where operations get interesting, transients, low rates, near stability boundaries, where correlation-based two-phase closures showed their non-geothermal ancestry. Flashing depth proved most sensitive to inputs operators know worst, feed-zone enthalpy distribution above all. Design levers mattered asymmetrically: diameter staging moved deliverability more than insulation in the compiled cases, except in cold-climate low-enthalpy service where the ranking flipped. The sensitivity finding directs instrumentation: feed-zone enthalpy distribution, the input operators know worst, moves flashing depth most. The first validation-graded synthesis of geothermal wellbore modeling with the sensitivity chain exposed and design levers ranked by demonstrated effect. Well designers gain closure-selection and instrumentation guidance for the component where reservoir performance becomes, or fails to become, plant feed. Well designers gain closure-selection and instrumentation guidance for the kilometres of pipe where deliverability is won, lost, and usually under-modeled.

Topic: Production Engineering

Binary geothermal plants are designed for a resource condition that begins drifting the day production starts, and the reservoir and plant optimization literatures have mostly drifted apart the same way. We review coupled reservoir-plant optimization for binary systems: ORC design-point matching against resource-decline trajectories, off-design performance evidence, reinjection-temperature coupling where plant choices feed back into resource life, and the value of optimizing the system rather than its halves. Plants are designed for a resource condition that starts expiring at first production, and the optimization literatures on either side of the wellhead have barely met. Coupled optimization studies were assembled and compared against sequential design practice, off-design ORC performance evidence reviewed from operating plants, resource drift moves plants off design early, the reinjection-temperature feedback examined, plant thermal choices set injection temperatures that reshape reservoir evolution, and machine-learning coupled surrogates surveyed. Coupled studies are ed against sequential design practice directly, so the value of closing the loop is measured rather than asserted. Coupled optimization shifted design points away from day-one resource conditions toward trajectory-weighted conditions, sacrificing initial output for lifetime energy in every compiled study, and the value grew with resource-decline uncertainty. Off-design evidence showed real plants losing more to drift than design-point comparisons capture. The reinjection feedback was material and almost universally ignored: plants optimized in isolation chose injection temperatures that cost reservoir life their own economics never saw. The feedback finding is the buried one: plant-chosen injection temperatures reshape reservoir evolution in ways isolated plant optimization never sees. NThe first synthesis to close the reservoir-plant loop for binary geothermal with the reinjection feedback quantified and trajectory-weighted design shown to beat design-point practice. Developers gain a coupled-design case stated in lifetime energy, the number their financing actually discounts. Developers gain the coupled-design case in lifetime energy terms, the number financing discounts, and designers the trajectory-weighted method that beats day-one optimality.

Topic: Reservoir Engineering

The Darajat Geothermal Field is one of the few vapor-dominated systems in the world, located in West Java, Indonesia. This high enthalpy resource is currently generating up to 279 MWe through power generation units. To sustain production levels and meet long-term generation targets, Make-Up Well (MUW) drilling campaigns are routinely conducted across the field. During MUW drilling, monitoring of offset wells conducted by the geochemistry team was originally aimed at ensuring that NCG concentrations remain within power plant operating limits, as drilling fluid injection could increase NCG levels in connected offset wells. Further evaluation also suggests that drilling induced NCG responses can be used to characterize inter-well connectivity. By integrating NCG response analysis with key geological feature controls on feedzone occurrence, the interpretation can be extended beyond general well-to-well connectivity to define specific feedzone-to-feedzone connectivity and relationships within the reservoir. A representative example was observed during the drilling of DRJ-E0. As DRJ-E0 penetrated its second feed zone, NCG concentrations in adjacent production wells DRJ-CB, DRJ-CC, and DRJ-BE increased to as much as five times their baseline levels. The second feedzone of DRJ-E0 and the affected feedzones in these production wells are interpreted to be structurally controlled by the Kendang Fault. This structural correspondence, supported by the aligned NCG responses observed during drilling, indicates possible feedzone-to-feedzone connectivity through fault-controlled permeability pathways, further supporting the interpretation that the Kendang fault acts as a major pathway for reservoir fluid and a key target for make-up well drilling. This combined approach also provides independent support for the interpreted 3D model of subsurface permeable fault zones associated with the Kendang Fault as key geological features. Consequently, it enhances field development strategies and decision making, including well placement for future drilling campaigns and injection management strategies.

Topic: Geochemistry

The Utah Frontier Observatory for Research in Geothermal Energy (FORGE) is a field-scale laboratory to develop the technology to make enhanced geothermal systems (EGS) commercially viable. Induced seismicity is associated with EGS operations, and seismic monitoring and traffic light protocols are key elements of developing the technology related to EGS risk mitigation and reservoir development, including critical information regarding the development of fracture networks at depth. At Utah FORGE, we have explored many different network configurations for seismic monitoring and provide real-time seismic monitoring during operational phases. The most recent operational phase is a multi-month circulation test designed to look for thermal break through in the reservoir. As we have done for previous operations, we are developing three independent earthquake catalogs. The first utilizes the local near-surface seismic network. This monitoring focuses on the traffic light protocols, and events are submitted into the publicly-available U. S. Geological Survey Comcat database. The second catalog utilizes instrumentation, both geophones and DAS, in deep boreholes. This catalog is focused on the microseismicity within the reservoir to understand fracture permeability and flow paths. At the completion of operations, this catalog will be contributed to the geothermal data repository. We additionally generate an automatic catalog using machine learning picks and a grid-based stacking and migration algorithm. The purpose of this catalog is for situational awareness. Since events are not reviewed, this catalog is not shared. In this paper, we will compare these catalogs to each other and analyze the seismic response to Utah FORGE operations. Additionally, we will share the details of a multi-month nodal geophone campaign. Following the completion of the circulation, the nodal data will be integrated with the shallow and deep seismic networks to refine details of the seismic response. We will also share details on any events that exceed the traffic light alarms and share how an adaptive traffic light protocol worked during circulation operations.

Topic: FORGE

Enhanced Geothermal Systems (EGS) comprise several critical components that require continuous monitoring and real-time analysis along their full length to ensure the safe and efficient extraction of geothermal energy. These components include high- and extreme-temperature geothermal wells; pipelines that convey superheated fluids from the wells to the energy conversion system; the energy conversion system itself; and the reinjection system, which returns the remaining fluids to the geothermal reservoir to maintain an adequate fluid supply. We will demonstrate how a High Temperature, 320°C tested, Fiber Optic Seismic Vector Sensor (FOSVS), Fiber Optic Pressure Sensors (FOPS), Enhanced Distributed Acoustic System (EDAS), Distributed Temperature (DTS) and Distributed Strain Sensor (DSS) system deployed in boreholes and on pipelines can provide a continuous, real-time, and full system length assessment of the geothermal energy production and fluid injection. This will allow for the optimization of energy extraction by providing the critical data needed to optimize and safeguard geothermal energy production.

Topic: Production Engineering

Cornwall, a county located in the southwest of the United Kingdom, has long been recognised as holding the highest potential for geothermal power development. From early research into EGS systems as part of the 1980’s Hot Dry Rocks Project at Rosemanowes Quarry, to the more recent development of the United Downs Project, which utilises a fractured granite reservoir and natural saline waters, the enhanced geothermal gradient in the region (c. 36 °C/km), generated by the radiogenic Cornubian Granite Batholith, provide temperatures suitable for power production at accessible depths (c. 4 - 5 km). The United Downs Project demonstrated that NNW-SSE aligned faulting, known locally as ‘crosscourses’, can be used to extract hot, mineral-rich waters for both geothermal power production and critical mineral extraction. The co-production model financially derisks future developments and avoids the requirement for extensive hydraulic stimulation associated with EGS developments. Aligned proximal to the maximum tectonic stress in the region, cross-course faults have long been associated with permeability and subsurface fluid flow, often being reported as the source of fluid inflow into 19th and 20th century subsurface mine workings in the county. Recorded in detail by the miners at the time, and in contemporary surveys by the British Geological Survey, many of the crosscourses are known to extend kilometres to tens of kilometres in length, offering the conditions needed for an extensive fractured reservoir, capable of sustaining long-term power production and critical minerals extraction. While the lateral extents of crosscourses are well understood, the challenge is predicting which crosscourse structures extend deeper into the subsurface, beyond the deepest mine records, where subsurface data becomes sparse. At United Downs, the width and dip of the Porthtowan Fault Zone (PFZ) is relatively well constrained near-surface (up to 500 m depth) due to a high density of subsurface mine workings in the area, however, this still required projection to c. 5 km depth. Traditional geophysical exploration was limited, either due to lack of data or lack of application in identifying faulting within homogenous granite. Despite this, the project demonstrated that in the case of the PFZ, the fault zone was confirmed at depth in its inferred location. This paper discusses the potential effectiveness and value in the use of both traditional and novel exploration methods for fractured geothermal systems in the granite in Cornwall. As well as developing the mapping of fault location, lateral and vertical extent, and dip, work is being conducted to develop an understanding of age, connectivity and total volume of the geothermal systems. Isotopic analysis of deep geothermal brines in conjunction with analysis of trace hydrocarbon contents provides initial data addressing the source and nature of long-term migration of these geothermal brines.

Topic: Geology

Through Congressional mandates the U.S. Geological Survey is required to develop geothermal assessments of Alaska and Hawaii. However, both states are sparsely covered with geophysical data. A large geophysical data collection effort is developing as a collaboration between state geological surveys, universities, and federal and state agencies. Described in this abstract are types of geophysical data being collected, locations, and how the data will be used to provide information to geothermal assessments. Because Alaska covers a vast area most of the data collection will be focused on airborne platforms including magnetics, radiometrics, and electromagnetics. Large regional gravity and magnetotelluric surveys are planned for the interior of Alaska to provide regional understanding of the subsurface. A focused study on Augustine Island in the Cook Inlet of Alaska is planned to characterize the volcanic hydrothermal system that will include magnetotellurics, gravity, drone-based magnetics and thermal-infrared, passive seismic, and airborne electromagnetics. Hawaii spans a much smaller area, therefore much of the geophysical data collection will be land-based and drone-based in targeted areas. The geophysical data will be modeled in 3D jointly interpreted. From the geophysical models a first-order approximation of physical properties such as porosity, permeability, and temperature will be estimated to provide valuable information for geothermal assessment.

Topic: Geophysics

Next-generation geothermal energy is a major economic and technical opportunity for Canada. Canada’s extensive petroleum and mining expertise provides a strong foundation for advancing next-generation geothermal energy. A nationally consistent assessment of Canada’s subsurface thermal conditions is needed to support geothermal resource evaluation, inform policy, and guide investment. To this end, we present the first stage of our results for the Canadian Thermal Model (CTM), revealing Canada’s 3D temperature, thermal conductivity, and heat flux regimes for western and central Canada. The model extends from the surface to the base of the crust at 25 km² spatial resolution and 10 m to greater than 5 km depth resolution. The CTM is implemented using a physics-constrained neural network framework based on the Stanford Thermal Model, and is informed by more than twenty geophysical, geological and thermal variables. Thermal properties and conditions of the upper crust are constrained by borehole temperatures, thermal conductivity and radiogenic heat-production measurements, together with spatial predictors such as proximity to volcanic centres. Deeper crustal conditions influencing thermal regimes are captured by geophysical features such as seismic velocity models, Moho heat flux, and gravity potential fields. The model is bounded by conductive and convective heat transfer laws, and validated against real borehole measurements.    The CTM captures the spatial and depth-wise variability of thermal properties which govern geothermal potential, advancing existing thermal models of Canada. However, data sparsity is a pervasive limit to model certainty, that can only be improved by targeted data acquisition. To account for these limitations, we assess model uncertainty and include spatial indicators where thermal property estimates are probabilistic. We are also compiling a hierarchical list of data collection needs for western Canada to encourage further investigation in Canada’s geothermal assets. Overall, this study demonstrates the potential for data-driven, statistically grounded thermal modelling techniques in estimating temperature regimes at depth and in data-sparse regions.

Topic: Modeling

The DeepHeat research programme is investigating the scientific and engineering requirements for developing superhot geothermal resources beneath New Zealand’s Central Taupō Volcanic Zone. Led by Earth Sciences New Zealand in partnership with the University of Auckland, Victoria University of Wellington, international research organisations, industry, and Māori partners, the programme seeks to improve understanding of fluid movement near and within the brittle ductile transition. Its wider activities include geophysical characterisation, geodynamic and fracture network modelling, laboratory investigation of rock and fluid behaviour, reservoir and wellbore simulation, stimulation and well integration technologies, environmental monitoring, techno-economic assessment, and the co-design of Māori led development pathways. This paper presents the planned contribution of the reservoir modelling team to DeepHeat. The work will develop an integrated numerical framework for representing the natural state and subsequent development of superhot geothermal systems. An initial three-dimensional digital conceptual model will be constructed for a selected study area using available geological, geophysical, geochemical, and thermodynamic information. This interpretation will be transferred systematically into a Waiwera reservoir model, providing traceability between the geological conceptualisation and the numerical representation. A central component of the work is the extension and verification of Waiwera for simulations spanning conventional, near-critical, and supercritical conditions. The resulting model will represent conductive and convective heat transfer, meteoric recharge, deep heat and fluid inputs, pressure and temperature evolution, and fluid flow through heterogeneous permeability structures. Sensitivity and uncertainty analyses will be used to identify the parameters and observations that exert the greatest control on model predictions. This will help prioritise data acquisition and guide the progressive refinement of the conceptual and numerical models. As results become available from the wider programme, geophysical interpretations, geodynamic simulations, fracture characterisation, geomechanical experiments, and geochemical constraints will be incorporated into the reservoir modelling workflow. Longer-term developments will investigate AI-assisted calibration, permeability forecasting, and the integration of reservoir and wellbore models to evaluate drilling targets, well performance, stimulation strategies, and sustainable production scenarios. The intended outcome is a progressively refined and uncertainty-aware modelling framework that connects multidisciplinary observations with predictions of fluid movement and permeability evolution in superhot geothermal systems. This framework will support DeepHeat’s objective of reducing exploration and development risk and establishing New Zealand as an international test bed for next generation geothermal technology.

Topic: Modeling

Production wells X-1, X-2, X-3, and X-4, located in the Tuchang Geothermal Prospect Area in northeastern Taiwan, were logged on behalf of CPC Corporation Taiwan using a suite of borehole logging tools. These included the Borehole High-Resolution Resistivity Imaging Tool, the Advanced Neutron-Gamma Logging Tool, the High-Resolution Resistivity Array Tool, and the Pressure-Temperature (PT) Logging Tool. Logging operations focused primarily on the 8-1/2-inch diameter production intervals. These data were carefully analyzed and interpreted to improve the classification of subsurface structures—specifically fractures and faults—by identifying their orientations and determining whether they are open (hydraulically conductive) or healed (non-conductive). This analysis is essential for understanding the structural complexity and hydraulic behaviour of these structures. Subsurface structures and in-situ stress orientations within the prospect generally conform to the broader regional tectonic framework, yet some local variations are evident. These variations reflect the complex architecture of large-scale fracture and fault systems present in the area, which may be intersected by the production wells. Data obtained from the High-Resolution Resistivity Imaging Tool offer valuable insights into fluid flow pathways, allowing for the identification of structurally controlled permeability zones. Conductive fractures and faults are critical to reservoir performance, as they provide the primary conduits for geothermal fluid circulation. All four production wells exhibit evidence of such conductive features, underscoring their relevance to the system’s productivity. Large conductive fractures and faults with strike orientations of NE–SW, NW–SE, and N–S have been correlated with key permeability indicators, such as thermal plumes and abrupt temperature increases. These correlations confirm that such features are important contributors to the overall permeability network. The integration of borehole imaging, geophysical logging, and expert interpretation offers a robust framework for characterizing subsurface complexity. This enhanced understanding of fracture systems and stress regimes supports improved reservoir characterization and informs more effective well targeting and development strategies within the Tuchang Geothermal Prospect.

Topic: Geology

Mineral dissolution and precipitation can alter injectivity and heat recovery in enhanced geothermal systems by changing fracture conductivity, matrix permeability, and reservoir stresses. These effects depend on both mineral-volume changes and the mechanical response of the surrounding rock. This interaction is particularly relevant in reservoirs stimulated by hydraulic fracturing, where fractures must remain conductive under compressive in-situ stresses. This study uses a recently developed reactive transport framework integrated into a three-dimensional hydraulic fracture and reservoir simulator. The formulation couples aqueous speciation and kinetic mineral reactions with fluid flow, heat transport, and rock deformation to simulate the evolution of pore space and fracture aperture. Mineral dissolution at fracture walls changes the space available for flow and the mechanical response to closure, modifying fracture conductivity and stress interactions among fractures. In the matrix, the model partitions reaction-induced changes in solid volume between pore volume change and deformation of the rock skeleton, coupling permeability evolution with chemoelastic stresses. We assessed coupled mineral and hydraulic evolution against the published Xie et al. reactive transport B3 . We then investigated silica dissolution and precipitation in an idealized EGS circulation case comprising an injection-production well pair connected by multiple hydraulic fractures. Controlled comparisons under specified thermal and chemical conditions quantified how direct chemical alteration, fracture stress redistribution, and matrix chemoelasticity affect injection pressure, fracture conductivity, flow partitioning among fractures, and heat recovery. The study examines when chemical alteration can be represented through hydraulic-property changes alone and when its mechanical consequences materially affect predictions of geothermal reservoir performance.

Topic: Modeling

This study updates a 2026 Stanford Geothermal Workshop techno-economic analysis of superhot rock (SHR) enhanced geothermal systems with several methodological advances. A coupled inflow and wellbore model (after Scott, 2026) was implemented in the open-source GEOPHIRES simulator. Each scenario now resolves phase change, flashing and the need for pumping between the reservoir and the wellhead, and the power cycle is chosen to suit the resulting wellhead conditions. A well integrity lifetime parameter was added to GEOPHIRES to trigger redrilling alongside thermal drawdown, and it is parameterized from a review of field evidence on high-temperature and superhot well failures. The resource basis was rebuilt from the Stanford Temperature Model, with gradient bands weighted by land area and high-gradient western hotspots added. Project economics reflect a western US investor case with federal tax incentives, state taxes, federal royalties, power purchase agreement pricing, and grid interconnection and transmission costs. Updated well designs and drilling cost cases span first-of-a-kind to mature-industry conditions. Levelized cost of electricity, after-tax returns and well requirements are evaluated for utility-scale projects at bottom-hole temperatures from 200 to 450 °C. The analysis examines where and under what cost conditions superhot resources become competitive with lower-temperature EGS and with firm power prices, and quantifies how SHR economics depend on assumptions about well life, well maintenance and drilling cost.

Topic: Enhanced Geothermal Systems

Utah FORGE is a field-scale enhanced geothermal system testbed developed to characterize how a discrete fracture network (DFN) controls fluid and heat transport between an injector and one or more producers. This paper reports a ten-year, full-physics forecast study, built on the FALCON/MOOSE PorousFlow framework, that varies the number and placement of producing wells around a fixed injector (16A) to quantify their effect on cumulative recovery efficiency (produced/injected water mass). Twenty-two configurations – one to eight producers, placed either evenly or weighted by a DFN-connectivity screen – were run to completion at ten years. Recovery efficiency rises monotonically with producer count, from 67–72% for a single well to 85% for eight, with steeply diminishing returns past three to four wells; well placement matters most exactly where well count is lowest.

Topic: Enhanced Geothermal Systems

Water access continues to be a critical question in the development and expansion of Enhanced Geothermal Systems (EGS) in the Western United States where subsurface conditions can be optimal, however, water availability can be scarce. Building on previous work using Large Language Models (LLMs) to collect data regarding groundwater availability in Texas, we expanded the scope of our research to additional states where there is potential for EGS (Nevada, Utah, New Mexico, California, and Colorado). The expansion presented challenge, as each state employs a distinct groundwater management structure (or lack thereof) with varying levels of oversight between state and basin/local level management agencies. We utilize COMPASS, a software package developed by the National Laboratory of the Rockies (NLR), to scrape the web for relevant documents and prompt the LLM to return a set of target fields using a defined JSON schema (consistent across all states), an update from previous extraction methods. We present the results of these findings across the target states, validation results, and potential future research opportunities. 

Topic: Enhanced Geothermal Systems

Estimating enhanced geothermal system (EGS) well construction costs requires accounting for well geometry, drilling performance, completion design, and the sequence of construction activities. This study presents an interval-based techno-economic calculator that integrates these factors to estimate construction time and cost for EGS wells. The framework allows users to define vertical, curved, and lateral intervals with distinct dimensions, operational parameters, and completion configurations. The model organizes inputs into four categories: well geometry and completion design, operational performance, activity scheduling, and unit costs. Inputs include interval depths, borehole and casing dimensions, rate of penetration, drill-bit life, logging requirements, circulation frequency, equipment operating rates, and material prices. The calculations distinguish measured depth from true vertical depth and account for drilling, tripping, logging, casing, cementing, and wellhead activities. Outputs include cumulative construction time and cost versus measured depth, total well cost, cost per drilled foot, and activity-level time and cost contributions. Application to a representative EGS well illustrates the calculation workflow and construction cost breakdown. Sensitivity analysis examines how selected design, performance, and economic inputs influence total construction cost, extending the assessment beyond rate of penetration alone. The framework provides a transparent basis for comparing well construction scenarios, identifying influential cost drivers, and evaluating potential operational improvements. It supports technology assessment and integration of well construction economics into broader EGS lifecycle techno-economic models.

Topic: Drilling

Enhanced geothermal system (EGS) performance depends on operating conditions and the extent to which stimulated fractures participate in reservoir circulation. Increasing circulation rate can increase heat extraction while also increasing pumping requirements, makeup-water demand, and thermal depletion. This study evaluates how circulation rate, effective fracture participation, water loss, and operating availability jointly influence the performance and economics of an EGS doublet. The analysis uses an integrated techno-economic framework that couples analytical reservoir heat transfer, hydraulic pressure losses, pumping demand, surface electricity generation, and lifecycle cost accounting. An effective fracture participation factor represents incomplete utilization of the stimulated reservoir by adjusting active fracture count and equivalent spacing while retaining the installed stimulation expenditure. A specified water-loss fraction determines makeup-water requirements and associated costs, while operating availability defines annual circulation and generation hours.The study examines these factors individually and through paired sensitivity analyses, with particular emphasis on the interaction between circulation rate and effective fracture participation. Outputs include production temperature, gross and net electrical power, lifetime electricity generation, cumulative makeup-water demand, water consumption per unit of net electricity, and levelized cost of electricity. The analysis holds well geometry and completion design constant to distinguish operational effects from changes in infrastructure. The study aims to identify operating conditions that balance electricity generation, pumping requirements, water use, and thermal longevity. By comparing initial power output with lifetime energy and economic performance, the framework supports evaluation of operating strategies and quantifies the economic implications of incomplete fracture participation and resource losses in EGS production.

Topic: Enhanced Geothermal Systems

Indonesia hosts numerous active and Quaternary volcanic systems that are closely associated with geothermal resources. Young volcanic centers are commonly considered favorable indicators of geothermal potential because recent magmatic activity may provide a heat source. However, volcanic age alone may not adequately represent the thermal conditions of a magmatic system. This study presents a literature-based review of Indonesian volcanoes to examine the relationship between volcanic age and rock composition and to discuss their implications for geothermal exploration. The study compiles published information on volcanic age, volcano type, and rock composition from Indonesian volcanic systems. Generalized magma emplacement temperature ranges associated with different rock types are also compiled from existing literature and used as supporting information to evaluate first-order thermal implications. The analysis focuses on comparing the distribution of volcanic rock composition between different volcanic age groups and relating these compositions to the literature-derived temperature ranges. The results are used to assess whether younger volcanic systems consistently exhibit characteristics indicative of relatively hotter magmatic conditions. The review provides a preliminary regional perspective on volcanic age and rock composition as geological indicators for geothermal exploration in Indonesia. These parameters can help identify areas that warrant further investigation, but should be integrated with evidence of heat, fluid, and permeability to better assess geothermal potential. Overall, the review highlights how regional understanding of Indonesian volcanism can support the interpretation of geothermal exploration targets.

Topic: Geology

Geophysical interpretation plays an important role in developing geothermal conceptual models and supporting exploration decisions. However, the relationship between a geophysical response and its geological cause is often non-unique. For example, a conductive anomaly may reflect clay-rich alteration, conductive fluids, lithological variations, or a combination of these effects. Consequently, reducing uncertainty in the geophysical model does not necessarily result in a corresponding reduction in uncertainty in the geological interpretation. This paper proposes a practical framework for managing interpretational uncertainty in geothermal geophysical workflows. Rather than treating uncertainty as a limitation to be reported after interpretation, the framework considers how uncertainty can be identified, evaluated, tested, and acted upon throughout the interpretation process. The approach distinguishes uncertainty associated with data quality and coverage, geophysical modelling and resolution, and the geological interpretation of geophysical responses. Particular attention is given to situations where multiple geological interpretations remain consistent with the available geophysical evidence. The framework introduces an interpretation based decision pathway in which uncertainty is first characterized, alternative interpretations are explicitly considered, and independent geological, geochemical, and geophysical constraints are used to evaluate competing hypotheses. The remaining uncertainty is then assessed against the objective of the exploration stage. This recognizes that an uncertainty level acceptable for regional prospect screening may not be acceptable for prospect evaluation or well targeting. Where uncertainty remains decision-critical, the framework provides a basis for identifying whether additional data acquisition, alternative modelling, or retention of multiple conceptual models is warranted. The proposed approach provides a practical means of linking geophysical interpretation uncertainty to exploration action. By making the reasoning between geophysical observations, geological interpretations, confidence, and subsequent decisions explicit, the framework aims to support more transparent and defensible geothermal conceptual models and reduce the risk of over-interpreting non-unique geophysical responses

Topic: Geophysics

Idle geothermal wells may retain substantial subsurface heat after becoming uneconomic for conventional production, creating opportunities for closed-loop heat recovery without additional reservoir stimulation. However, retrofit feasibility cannot be assessed from temperature alone because residual wellbore flow, feed-zone connectivity, pressure-dependent phase behavior, and completion geometry can strongly control both heat availability and hydraulic performance. This study evaluates a thermally active idle geothermal well in West Java, Indonesia using a field-constrained thermal-hydraulic framework that integrates long-term temperature-pressure recovery surveys, flowing production logs, multi-speed spinner data, injection and falloff testing, surface well-test measurements, thermodynamic reconstruction, and the as-built well completion within a common depth reference. The reconstructed well state shows that the well remains both thermally and hydraulically active, with temperatures of approximately 221°C at 1500 m measured depth after 241 days of recovery and approximately. 223°C at 7.4 bara under fully flowing conditions. Integrated interpretation identifies an upper permeable exchange interval at approximately 1030–1220 m and a deeper thermally active interval at approximately 1450–1580 m, while multi-rate injection data yield an apparent whole-well injectivity of approximately 5.54 L s^-1,bar^-1 with R²=0.983. Thermodynamic reconstruction further shows a strong operating-state dependence, ranging from a dense liquid-equivalent recovery state to superheated vapor during full production and near-saturation conditions during throttled flow. Initial closed-loop simulations indicate megawatt-scale heat-recovery potential, with approximately 4.56MWth extracted at a representative 3 kg/s water-circulation rate and an equivalent power potential approaching 0.80MWe under the assumed conversion conditions. When combined with the actual completion geometry, the field-constrained design envelope indicates that a closed-loop depth near 1500 m provides the most favorable compromise between access to the deep thermal source and increasing hydraulic and mechanical restrictions below the 7-in liner. These results demonstrate that operationally idle geothermal wells can remain dynamically active thermal assets and that their closed-loop reuse is best evaluated through coupled field, thermodynamic, hydraulic, and completion constraints rather than temperature or conductive heat transfer alone.

Topic: Modeling

Idle geothermal wells may retain substantial subsurface heat after becoming uneconomic for conventional production, creating opportunities for closed-loop heat recovery without additional reservoir stimulation. However, retrofit feasibility cannot be assessed from temperature alone because residual wellbore flow, feed-zone connectivity, pressure-dependent phase behavior, and completion geometry can strongly control both heat availability and hydraulic performance. This study evaluates a thermally active idle geothermal well in West Java, Indonesia using a field-constrained thermal-hydraulic framework that integrates long-term temperature-pressure recovery surveys, flowing production logs, multi-speed spinner data, injection and falloff testing, surface well-test measurements, thermodynamic reconstruction, and the as-built well completion within a common depth reference. The reconstructed well state shows that the well remains both thermally and hydraulically active, with temperatures of approximately \(221^{\circ}C\) at 1500 m measured depth after 241 days of recovery and approximately \(223^{\circ}C\) at 7.4 bara under fully flowing conditions. Integrated interpretation identifies an upper permeable exchange interval at approximately 1030–1220 m and a deeper thermally active interval at approximately 1450–1580 m, while multi-rate injection data yield an apparent whole-well injectivity of approximately 5.54Ls^-bar^-1 with R²=0.983. Thermodynamic reconstruction further shows a strong operating-state dependence, ranging from a dense liquid-equivalent recovery state to superheated vapor during full production and near-saturation conditions during throttled flow. Initial closed-loop simulations indicate megawatt-scale heat-recovery potential, with approximately 8.56MWth extracted at a representative 12 kg/s water-circulation rate and an equivalent power potential approaching 0.80MWe under the assumed conversion conditions. When combined with the actual completion geometry, the field-constrained design envelope indicates that a closed-loop depth near 1500 m provides the most favorable compromise between access to the deep thermal source and increasing hydraulic and mechanical restrictions below the 7-in liner. These results demonstrate that operationally idle geothermal wells can remain dynamically active thermal assets and that their closed-loop reuse is best evaluated through coupled field, thermodynamic, hydraulic, and completion constraints rather than temperature or conductive heat transfer alone.

Topic: Modeling

Closed-loop geothermal systems (CLGS) have emerged as an alternative approach to extracting subsurface heat without directly producing native geothermal fluids, offering potential opportunities for geothermal development in formations with limited permeability or unfavorable fluid conditions. Despite increasing global interest and recent progress toward commercial deployment, uncertainties remain regarding sustained thermal performance, operational reliability, and the extent to which reported project capacities reflect demonstrated field performance. This study evaluates 14 publicly documented CLGS projects and field cases across North America, Europe, and Asia, covering field demonstrations, technical validation, heat-focused applications, and commercial developments. A comparative review of published technical reports, field-test results, and project disclosures was conducted to assess technology configurations, operational experience, thermal performance, and development maturity, while distinguishing measured results from estimated and announced capacities. The findings indicate that CLGS technologies have progressed beyond early conceptual demonstrations, although the availability of long-term operational evidence remains limited. At The Geysers, a closed-loop demonstration sustained approximately 7.4 MWth at 330 GPM during a 30-day test, with peak thermal output reaching approximately 9 MWth. Meanwhile, the Geretsried project in Germany began supplying electricity to the grid in December 2025, with approximately 8.5 MWth subsequently reported from its first, partially completed loop, highlighting both commercial progress and remaining well-construction challenges. Field observations from medium-deep coaxial systems in China further demonstrate the importance of thermal decline and heat-recharge limitations during continuous operation. These findings suggest that future CLGS deployment should prioritize sustained thermal performance, wellbore integrity, operational optimization, and transparent reporting of net energy output. Heat-focused applications, repurposing of existing wells, and site-specific technology selection represent promising pathways for expanding CLGS deployment while addressing persistent technical and economic uncertainties.

Topic: Enhanced Geothermal Systems

**Toward Zero Surface Discharge: A Field-Based Assessment of Water Reuse and Operational Constraints in Geothermal Power Plants** Water management in geothermal power plants involves balancing process-water requirements, condensate handling, reinjection obligations, and environmental discharge limitations. Although condensate recovery and water reuse are established practices, opportunities for reducing surface discharge remain highly dependent on actual plant water balances, water quality compatibility, and operational constraints. This study develops a field-based assessment of integrated water management using operational information from an anonymized geothermal power plant in Indonesia. The assessment combines process-level water balance reconstruction, water quality characterization, source-to-use compatibility screening, and scenario-based evaluation of water allocation alternatives. Condensate, stormwater, and supplemental water sources are assessed in relation to existing plant demands, storage availability, treatment requirements, and reinjection constraints. Particular attention is given to distinguishing genuinely available surplus water from flows already allocated to cooling, reinjection, or other essential plant functions. Alternative management scenarios are formulated to examine discharge reduction opportunities under existing infrastructure, improved water allocation, and potential treatment or storage modifications. Variations in water supply, operational demand, and reinjection availability are considered to identify conditions that limit further reuse. The analysis provides a structured approach for comparing achievable discharge reduction against implementation requirements and operational reliability. By integrating field-specific water balance information with quality and operational constraints, this study aims to establish a practical methodology for identifying technically feasible water reuse pathways in geothermal power plant operations. The proposed approach supports operational decision-making toward reduced surface discharge while maintaining compatibility with plant and reservoir management requirements.

Topic: Injection

Geothermal hot springs represent one of the most established forms of direct geothermal utilization, offering opportunities for local economic development, tourism, and broader applications of geothermal heat. However, many hot spring facilities remain primarily dependent on conventional bathing activities, with limited integration of complementary services and productive heat utilization. Expanding these facilities into more diversified geothermal destinations requires an understanding of site-specific resource conditions, operational limitations, and market readiness rather than adopting a uniform development model. This study examines geothermal direct-use development opportunities based on field observations at more than 15 hot spring facilities across West and Central Java, Indonesia, supported by stakeholder discussions, literature review, and international ing. A comparative qualitative assessment was conducted to identify recurring operational challenges, existing utilization patterns, and opportunities for integrating geothermal resources with additional commercial and community-based applications. Field observations indicate that many facilities remain highly dependent on entrance-ticket revenue, with limited supporting services, relatively short visitor stays, and uneven weekday utilization. These findings highlight opportunities for development through diversified visitor services, geothermal education, local enterprise participation, and selected cascaded heat-use applications where resource conditions permit. Based on the assessment, a site-adaptive development framework is established, integrating geothermal resource suitability, existing infrastructure, market accessibility, implementation complexity, and stakeholder readiness. The framework identifies three progressive development pathways, ranging from low-investment operational improvements to integrated destination services and more advanced geothermal wellness applications. The results suggest that geothermal direct-use expansion should be guided by site-specific technical and commercial conditions rather than standardized infrastructure development. This approach provides a practical basis for prioritizing development opportunities, improving resource utilization, and supporting scalable geothermal direct-use initiatives in Indonesia and other geothermal regions.

Topic: Direct Use

Geothermal well testing has undergone significant evolution since the early production tests that established geothermal energy as a viable source of power generation. Today, well testing remains a critical tool for evaluating drilling success, reservoir performance, well productivity, injectivity, and stimulation effectiveness across conventional geothermal developments, Enhanced Geothermal Systems (EGS), and emerging next-generation geothermal technologies. This paper presents an overview of recent advancements in geothermal well test facility design and data acquisition systems, with a focus on technologies and methodologies that have improved data quality, reliability, flexibility, and operational efficiency of well testing facilities. Topics include developments in atmospheric separation equipment, multiphase flow measurement, digital instrumentation, automated control systems, and real-time data transmission. The application of these innovations across a broad range of geothermal resource types is discussed, along with the resulting benefits for reservoir characterization, operational decision-making, and project risk reduction. Key challenges, lessons learned, and emerging industry trends are also examined, highlighting how modern well testing facilities are enabling more comprehensive data collection and supporting the continued advancement of global geothermal development.

Topic: Reservoir Engineering

Paleosol intervals in volcanic geothermal sequences can be drilled uneventfully in one well yet progress to severe wellbore instability, pack-off, and stuck pipe in another. Although paleosol-related drilling problems are commonly associated with reactive clay, lost circulation, and hole-cleaning difficulties, field criteria for distinguishing a manageable encounter from progressive wellbore deterioration remain limited. This study reconstructs paleosol-related drilling events from multiple high-temperature geothermal wells in Indonesia, including original-hole, sidetrack, managed-deterioration, and successful-comparator cases. Daily drilling reports, end-of-well reports, mudlogs, cuttings descrptions, methylene-blue index data, drilling-fluid properties, directional surveys, and intervention histories were integrated using an event-based chronology from first paleosol evidence through deterioration, intervention, pack-off, recovery, or isolation. The cases show that paleosol presence alone did not determine failure. Lost circulation, high inclination, elevated methylene-blue index, KCl concentration, and individual filtration values were neither necessary nor sufficient to explain severe outcomes. A more consistent indicator of worsening wellbore condition was the persistent or repeated generation of formation-derived solids after circulation or cleanout, particularly during prolonged open-hole exposure and re-entry. Severe cases progressed from abnormal cavings or recurrent fill to increasing drag, torque and/or standpipe-pressure abnormalities before pack-off, whereas the successful comparator crossed and isolated the paleosol while mechanical conditions remained stable. Clean wiper trips and temporary improvement after stabilization were found to demonstrate short-term runnability rather than durable formation stability. Based on the cross-case comparison, an operational decision framework is proposed to distinguish manageable exposure, progressive deterioration, recoverable wellbore restriction, and catastrophic failure. These conditions are linked to monitoring, hole conditioning, stabilization, and casing or liner isolation decisions. The framework provides a practical basis for determining when continued open-hole management remains effective and when further exposure should be limited through physical isolation.

Topic: Drilling

Indonesia has one of the world’s largest geothermal resource bases, yet many projects face delays before exploration, drilling, construction, or expansion due to unresolved social risks. National planning documents can identify priority fields and commercial operation targets, but they do not necessarily indicate whether a project is socially ready to move. This paper argues that geothermal project readiness should be assessed not only by subsurface potential, permitting, financing, and engineering progress, but also by measurable social-readiness indicators. This study develops an upgraded Social License to Operate framework for geothermal project delivery using evidence from Indonesian field cases. The paper expands the previous case base by combining 17 maturity-coded geothermal cases with additional publicly documented social-risk cases from Indonesia. The analysis identifies recurring risk drivers, including land and livelihood disruption, water and environmental concerns, cultural and customary-space issues, safety and H₂S anxiety, local labor expectations, misinformation, and weak grievance response. The proposed framework introduces a Social Readiness Index, supported by dashboard indicators, FMEA-based risk prioritization, and phase-gate decisions. The framework integrates four mechanisms: benefit-sharing compact, direct-use demonstration, local talent and gatekeeper participation, and grievance service-level agreement. These mechanisms are assessed across geothermal project phases, from pre-survey and land access to drilling mobilization, construction, operation, and expansion. The expanded review shows that many Indonesian geothermal projects already contain useful engagement elements, including CSR programs, livelihood support, local recruitment, supplier participation, stakeholder engagement, and grievance mechanisms. However, these practices remain uneven and rarely integrated into a measurable project-readiness system. Direct-use is also found to be under-integrated: several Indonesian examples exist, but they are not commonly used as formal social-license instruments. The paper concludes that Indonesia does not only need a geothermal pipeline; it needs a socially ready geothermal pipeline.

Topic: General

Superhot Enhanced Geothermal Systems (SH-EGS), which operate above the critical point of water ( greater than 374°C, greater than 22 MPa), represent a promising clean energy solution capable of delivering significantly more energy than conventional geothermal technologies. However, simulating these systems remains a challenge due to the extreme coupling of thermal, hydraulic, mechanical, and chemical (THMC) processes under high-pressure, high-temperature conditions. Key uncertainties include fracture permeability evolution in ductile rock and the stability of numerical solvers in supercritical regimes. We present ongoing work to extend LANL’s Amanzi and dfnWorks codes to enable long-duration, fully coupled THMC simulations under supercritical conditions. Our current efforts are focused on implementing a high-resolution EOS framework for supercritical fluids and stabilizing solver performance in steep P–T gradients. Preliminary results demonstrate improved convergence and numerical stability using spline-based EOS interpolation in synthetic test cases. The modeling framework is being designed to support realistic, multi-physics simulations of SH-EGS scenarios, including fluid–fracture–chemistry interactions and long-term flow behavior. We outline a validation approach using component-level tests, synthetic s, and analog comparisons to existing geothermal field data. The resulting capability will enable more predictive design and risk assessment for SH-EGS development and contribute to a broader understanding of subsurface systems operating in extreme conditions.

Topic: Enhanced Geothermal Systems

The Dieng geothermal field in Indonesia is two-phase liquid-dominated geothermal field that was developed by PT Geo Dipa Energi (Persero) in 2002 with an installed plant capacity of 60 MWe. Several numerical models have been developed over the last 2 decades. The latest numerical model of the Dieng geothermal field was constructed in 2026 with an aim of quantifying lithium resources within the Dieng reservoir as well as estimating how lithium concentration within the reservoir and in the production fluids would change when a lithium extraction facility is put into operation. This paper describes the process of numerical simulation via model calibration considering the production and injection data of wells in Dieng Unit 1 and includes additional wells drilled for Dieng Unit 2 wells. A robust model that includes calibration to natural-state and operational history for 22 years of production from the field is described. The calibrated numerical model is used to forecast the field’s performance under various operating conditions with the aim of identifying long-term challenges and opportunities to maintain stable production of lithium from this field.

Topic: Modeling

High-temperature geothermal well cements experience thermal and mechanical stresses that may result in loss of integrity and well damage. To fulfill their functions cementitious materials must combine sufficient chemical resistance, mechanical properties and resistance to damage propagation. Reinforcement provides one approach to modifying mechanical properties and resistance to damage propagation, but its effectiveness and stability under high-temperature hydrothermal conditions depend on its physical properties (size distribution, pre-processing, etc.), and reinforcement and binder chemistry. This study compares performance of two chemically different cementitious binder systems reinforced with carbon, glass, basalt fibers, and several grades of acicular wollastonite, and halloysite. Samples were autoclaved at 250 °C, and compressive strength, Young's modulus, porosity, and energy absorption (compressive toughness) from stress-strain measurements were determined. The results show that different reinforcements affect strength, stiffness, and toughness differently and that these effects are strongly dependent on the cement matrix. Carbon fibers provided the most consistent improvement in toughness in both binder systems. Glass fibers predominantly increased strength and stiffness, while the response to wollastonite depended strongly on wollastonite type, concentration, and binder chemistry. The tested grade of basalt fibers and halloysite provided limited benefit under the investigated conditions. Selected carbon-, glass-, and wollastonite-reinforced formulations were further evaluated after extended exposure at 250 °C to determine whether the initial mechanical effects were retained as the cement matrices and reinforcement–matrix interfaces evolved. SEM/EDX, XRD analyses were used to examine reinforcement stability, interfacial reactions, and possible degradation mechanisms. These results were complemented by previous laboratory testing of carbon-fiber-reinforced cementitious composites at 300 °C and geothermal well exposures at approximately 350 °C for periods up to nine months. The results demonstrate that reinforcement selection for high-temperature geothermal cements cannot be based on reinforcement type or compressive strength alone. Binder–reinforcement interactions control the balance among strength, stiffness, and resistance to damage propagation and may evolve during prolonged hydrothermal exposure.

Topic: Reservoir Engineering

In fractured geothermal reservoirs, injected fluid can short-circuit through dominant fractures and cause premature thermal breakthrough. Tracer tests probe these pathways, but a good fit to the outlet response does not establish that the inferred fracture geometry is identifiable, especially in branching systems where several paths feed one outlet. This work conditions a thermal tracer inversion on the full solute posterior, recovers aperture as a posterior output, and scores every surface-area and aperture estimate by the relative width of its posterior. The score uses the posterior alone, so it needs no independent knowledge of the true geometry. Tested on 3D-printed single-path and branching s of known geometry, the screen separated usable from unusable estimates where the tracer fits were equally good, and surface area and aperture on the same path could receive opposite verdicts. Conditioning on the solute posterior made estimates usable that independent thermal inversion left unusable. Carrying unscreened estimates into an illustrative doublet forecast biased it 77% high inside a deceptively narrow interval. Tracer-derived fracture geometry should therefore be screened estimate by estimate before use.

Topic: Tracers

Free thermal convection in geological formations is attracting renewed interest for geothermal heat production and the recovery of economically valuable solutes. In heterogeneous porous media, linear stability analyses have been reformulated for a wide range of permeability distributions, yielding relations between the instability threshold and the heterogeneity parameters. These relations are generally complex and require a case-by-case numerical solution, while few studies have sought a more systematic interpretation based on averaged estimates of the heterogeneous parameters. Rayleigh-Bénard convection is therefore extremely difficult to predict in Darcy-scale reservoir simulation due to the very high resolution required, large time scale, as well as the lack of validation based on the expected threshold. Upscaling workarounds would be useful to overcome this bottleneck. In fact, in the simplest configuration of two superposed horizontal porous layers, different effective permeabilities have been proposed to estimate the instability threshold, leading to contrasting conclusions in favor of either the harmonic or the arithmetic mean. In the present study, we combine a finite-element generalized eigenvalue formulation, implemented in FEniCS/SLEPc, with transient finite-volume simulations in OpenFOAM to determine the threshold. We show that neither estimator accurately predicts the instability threshold and propose the geometric mean as an alternative. The resulting prediction is highly accurate when both layers contribute to convection, and the transition to convection localized in the more permeable layer is clearly identified. The results are further extended to systems with multiple layers, for which the geometric-mean estimator remains accurate. In addition, the developed tools allow us to characterize the resulting convective regimes.

Topic: Modeling

Exploration methods are hard to test. Whether a well produces is rarely published alongside the data that justified drilling it. Public datasets usually end at exploration, and operating records stay proprietary. Zanskar has released both for the Lightning Dock geothermal field in the Animas Valley, Hidalgo County, New Mexico. The package covers 41 wells with wireline and mud logs, drilling and completion records, flow and injection tests, and PTS surveys. It also includes four tracer campaigns run between 2012 and 2019, magnetotelluric data from the 2010 and 2014 surveys (135 stations, two inversion models), aeromagnetics, gravity, and LiDAR. Groundwater monitoring contributes 46,187 depth-to-water readings from 25 wells since 2013. Geochemistry contributes 14,108 analyses from 722 samples. The operations record runs from January 2014 to August 2026. It contains 6,850 rows and 70 columns of daily plant- and well-level measurements, including power output and brine handling. Well identifiers are consistent across the package, so each well's production history links directly to its logs and to the geophysics over it. This lets an exploration method be checked against twelve years of observed reservoir behavior. The same record supports reservoir engineering work on decline, injection response, and tracer-derived connectivity between wells. The release is part of GEOGYM (DOE EE0011649), a multi-site for de-risking geothermal exploration.

Topic: Field Studies

Characterizing proppant conductivity and permeability over extended periods and a range of closure stresses is essential for hydraulic-fracture design and post-treatment performance evaluation in both geothermal and unconventional reservoirs, where proppants must sustain fracture aperture and stable fluid flow under in-situ stress and temperature. This study evaluates the long-term fracture conductivity of three proppant candidates: olivine, St. Peter sandstone (Northern White sand), and steel slag. Olivine and St. Peter sand are naturally occurring materials that require minimal processing, whereas steel slag is an industrial by-product recovered from steel recycling operations. Because all three avoid the energy-intensive manufacturing of synthetic ceramic proppants, they are potentially low-cost, and the use of steel slag additionally converts an industrial waste stream into a value-added product. Experiments were conducted in an API-recommended conductivity cell in which a proppant pack is confined between two low-permeability granite slab, allowing high confining stress, temperatures up to 230 °C, and continuous fluid flow for several days. Tests were performed with both smooth and rough granite surfaces, the latter representing the topography of actual reservoir fractures, to assess the influence of fracture-surface roughness on conductivity. The results from the two surface types will be compared to determine how roughness affects the long-term conductivity of each proppant. Conductivity decline varied among the tested materials, but most retained their conductivity under severe thermal and mechanical loading. Our earlier testing showed that resin-coated, petroleum coke, and low-density ceramic proppants sintered and degraded severely under elevated temperature and closure stress. In contrast, the natural proppants examined here underwent grain crushing but showed no sintering or similar degradation under comparable conditions. These findings indicate that low-cost natural and by-product materials can maintain fracture conductivity under realistic reservoir conditions, making them viable candidates for geothermal and unconventional applications.

Topic: Enhanced Geothermal Systems

Reliable reservoir-temperature estimation is essential for geothermal exploration, yet solute geothermometry is often limited by the absence of suitable boiling chloride springs in many Indonesian geothermal systems. Sulfate-water oxygen isotope geothermometry has been proposed as an alternative approach; however, its reliability under Indonesian geothermal settings remains poorly constrained. This study evaluates the performance of sulfate-water oxygen isotope geothermometry using 13 samples from seven geothermal fields in Sumatra, including both surface manifestations and geothermal wells. Reservoir temperatures were calculated using the formulations of Mizutani and Rafter (1969) and Lloyd (1968), and subsequently compared with reported reservoir temperatures and the geochemical characteristics of each field. The calculated temperatures range from 54 to 586°C and exhibit highly variable agreement with reported reservoir temperatures. The most reliable results were obtained from four geothermal well samples at Sibayak, yielding temperatures of 249-285°C, which closely correspond to reported reservoir temperatures of 275-300°C. In contrast, most surface manifestations produced significant underestimation or overestimation. These deviations are interpreted to reflect isotopic modification during fluid ascent and discharge, including H2S oxidation, steam-heating processes, evaporite mineral interaction, and meteoric-water mixing. Although sulfate-water oxygen isotope geothermometry was evaluated using steam-heated sulfate and fumarolic manifestations that are unsuitable for solute geothermometry, reliable reservoir-temperature estimation still requires boiling chloride springs. The results did not support the hypothesis that the method could overcome the limitations of solute geothermometry through its application to non-chloride manifestations. Therefore, sulfate-water oxygen isotope geothermometry faces similar challenges to solute geothermometry.

Topic: Geochemistry

Hydrothermal resource assessments in the western United States have mostly produced favorability maps, which rank likeliness of resource presence but doesn’t resolve for site-level capacity estimates. Technical-potential assessment frameworks need site-level capacity in megawatts as input, which favorability scores cannot supply, and the most widely used capacity , the USGS assessment of Williams et al. (2008), reports undiscovered potential only as aggregate regional totals. We address this gap with a two-stage hurdle model that predicts both the presence and the expected capacity of hydrothermal resources on a 5 × 5 km grid across the western United States (122,914 sites). The model draws on 30 geological and geophysical predictors used in Stanford Thermal Earth Model and is trained on 65 operating hydrothermal power plants, labeled with nameplate capacity, and 779 negative sites identified from residual heat-flow data as zero megawatt sites. In the first stage, a classifier estimates the probability of resource occurrence in each cell. These probabilities are calibrated with isotonic regression, and thresholds are optimized for the F2 score. Two operating threshold are defined, a low threshold (calibrated probability = 0.25) and a high threshold (= 0.65), and the classifier reaches an F2 score of 0.81. In the second stage, a regressor trained only on existing power plants predicts log1p-transformed capacity under squared-error loss, with a Duan smearing correction to remove retransformation bias. On a held-out 20% of producing sites, the mean absolute error (MAE) is 19.79 MW. Across all labeled sites, the MAE is 9.77 MW at the low threshold and 9.95 MW at the high threshold, and aggregate predicted capacity falls within about 11% of observed (predicted/actual ~1.11) in both cases. Applied region-wide, the model identifies 13,390 sites (10.89%) under the low threshold and 2,051 sites (1.67%) under the high threshold, corresponding to total hydrothermal resource potential of 132.63 GW and 41.65 GW, respectively. The result is a spatially explicit, capacity-resolved hydrothermal resource dataset that can feed technical-potential and capacity-expansion modeling directly, where favorability maps cannot.

Topic: Modeling

This work presents the continuation of the J-DEI project, advancing from conceptual development toward numerical validation. Following the initial definition of the system architecture and its underlying geothermal and hydrostatic principles, the current stage focuses on testing the physical and engineering assumptions that support the proposed technology. Numerical modelling is used as a necessary intermediate step to evaluate whether the system can operate within feasible thermal, hydraulic, structural, and material limits before progressing toward experimental and physical validation. The current development of J-DEI therefore focuses on numerical simulations designed to evaluate heat-transfer performance, pressure conditions, structural integrity, and material behaviour under representative operating scenarios. Particular attention is given to the use of molten-salt buffers as thermal and pressure-management elements, as well as to high-temperature and corrosion-resistant materials intended to withstand the combined effects of pressure, thermal gradients, and repeated operating cycles. These simulations are also intended to identify potential failure mechanisms, refine geometry and material assumptions, and define the operating ranges that should later be tested experimentally. Unlike conventional geothermal systems, which generally depend on natural reservoir permeability and fluid circulation through geological formations, J-DEI is conceived as a more controlled engineered heat-extraction system in which pressure management, heat transfer, and fluid containment are incorporated into the designed architecture. This work therefore represents a further step in the scientific validation of the J-DEI concept, using numerical analysis as a bridge between theoretical design and future experimental demonstration. The broader objective is to determine whether this approach can improve geothermal efficiency, broaden the range of usable resources, and contribute to the development of more scalable and potentially lower-cost geothermal technologies.

Topic: Enhanced Geothermal Systems

In this work the impact of fracture propagation on heat production in enhanced geothermal systems (EGS) is investigated using a numerical framework that accounts for thermo-mechanical effects during long-term cooling. As fluid circulation cools the fracture surface, thermal contraction increase the fracture aperture and generate tensile stresses near the fracture tip. Over extended time periods, these thermal stresses can become sufficiently large to induce fracture propagation including wing crack growth. This process may influence heat production in different ways, including improved permeability, enhanced hydraulic connectivity, and access to hotter regions of the reservoir. At the same time, fracture growth may also increase the risk of fluid loss if propagated fractures intersect pre-existing natural fractures. To evaluate these effects, multiple simulation cases with different fracture orientations and cluster numbers are considered under the same reservoir and formation conditions. This analysis is intended to clarify how fracture growth induced by thermal stresses affects production temperature and overall thermal performance in EGS reservoirs.

Topic: Enhanced Geothermal Systems

Geothermal energy systems generates significant volumes of geothermal brines enriched with dissolved minerals and metals as a by product of power production. These high temperature, saline fluids represent both environmental management challenge and an untapped resource. Properly managed Geothermal brines can support sustainable energy generation while enabling resource recovery, reducing waste and improving overall technoecomic viability of geothermal operations.Thus large volumes of geothermal brines are enriched with dissolved metals and nutrients, creating extreme physiochemical environments characterised by high temperature, salinity and pH. These conditions support extremophylic microorganisms capable of metal tolerance, transformation and bioaccumulation. Harnessing extremophile-driven processes in combination with phytomining approaches, offers a sustainable pathway for recovering valuable and critical metals such as li thium (Li), cesium (Cs), rubidium (Rb), strontium (Sr), manganese (Mn), zinc (Zn), copper (Cu), and i ron (Fe), along with rare earth elements (REEs) present in trace concentrations from Geothermal brines while mitigating environmental impacts.These metals represent a strategic resource for energy storage, electronics, and clean technologies, making geothermal brines an attractive target for bio-based recovery.Overall, extremophile-driven phytomining aligns strongly with circular economy principles and represents a low-carbon pathway for critical metal recovery in future geothermal resource management Keywords: Geothermal energy , Geothermal brines, Phytomining, Extremophile,Sustainable energy resource, Sustainable energy.

Topic: Geochemistry

Ormat’s Raft River geothermal facility in Cassia County, Idaho produces 10.2 MW. Five production wells feed the plant with about 3000 kph of a circumneutral pH sodium-chloride brine with temperatures ranging from 130 to 150 ºC. During replacement of a failed pump in production well RRG-1 in early 2025, caliper logs identified casing damage at a depth of 3340 ft. While the damage suggested the potential ingress of an external fluid, its influence on production chemistry and scaling risk remained uncertain. In May 2026, production from RRG-1 declined by approximately 20%, and inspection of the pump revealed severe calcite scale deposition. Two competing mechanisms were proposed to explain the scaling event. First, production pressure prior to the damage fluctuated near the gas breakout pressure (GBOP). Temporary pressure drops below GBOP could have resulted in CO2 escape and calcite supersaturation. Alternatively, influx of a chemically incompatible fluid through the damaged casing may have altered fluid composition and increased calcite saturation. Distinguishing between these mechanisms was important because each would require a different mitigation strategy, with different operating costs. To evaluate these hypotheses, a downhole sampler was deployed to collect fluid samples at 3700 ft, 3340 ft (the depth of casing damage), and 1960 ft (pump intake depth). Geochemical analyses identified localized changes in fluid composition near the damaged casing. Fluid compositions from the three sampling depths were incorporated into a mixing model to evaluate the impact of fluid ingress on calcite saturation. Results indicate that mixing with the external fluid increased calcite saturation indices but did not generate sufficient saturation to explain the observed scale accumulation. Consequently, pressure induced CO₂ degassing is interpreted as the most probable cause of the scaling event. This case study demonstrates how targeted depth sampling can be integrated with well integrity and performance data, together with geochemical modeling, to distinguish between competing scaling mechanisms in geothermal wells. The results highlight the value of depth-resolved fluid chemistry for diagnosing production challenges, reducing uncertainty in mitigation strategies, and supporting cost-effective operational decision-making.

Topic: Geochemistry

Shear deformation of rough fractures can reorganize aperture fields into channelized flow structures and generate hydraulic anisotropy, potentially exerting a strong control on production in Enhanced Geothermal Systems (EGS). However, the scale dependence of shear-induced hydraulic anisotropy and its implications for Sweet Spot Orientation (SSO) of production wells have not been systematically quantified. Here, we develop a multiscale framework linking shear-modified fracture hydraulics to reservoir-scale production-well optimization. Unsheared and sheared fracture aperture fields are reconstructed from experimental observations to determine equivalent permeability tensors, and their scale dependence is evaluated by expanding the fields from 0.2 to 1 m using statistical upscaling strategies. The resulting hydraulic properties are incorporated into analytical and numerical doublet models for porous and fractured reservoirs. By systematically varying the angular position of the production well around a fixed injection well, we investigate the effects of permeability anisotropy, principal permeability direction, interwell distance, and well penetration length on steady-state production. Isotropic reservoirs show little dependence on production-well angle, whereas in anisotropic reservoirs, SSO is governed primarily by the principal permeability direction. Stronger anisotropy increases the benefit of well-placement optimization, with the relative flow-rate difference reaching approximately 42.7% at a permeability anisotropy ratio of 100:1. In heterogeneous fractured reservoirs, flow channeling can shift SSO away from the bulk principal permeability direction. The proposed framework provides a quantitative basis for determining SSO from flow responses while accounting for permeability anisotropy and flow channeling.

Topic: Enhanced Geothermal Systems

The Salton Sea Geothermal Reservoir contains lithium-rich hypersaline brines that are currently produced for geothermal power generation and could support future lithium extraction. Assessing this opportunity requires dynamic reservoir forecasts that account for production, reinjection, chemical breakthrough and uncertainty in the subsurface system. We present an updated three-dimensional numerical model of the Salton Sea geothermal reservoir developed using the Waiwera simulator. The model incorporates revised geological interpretations, historical production and reinjection data, downhole temperatures, produced-fluid enthalpy and chemistry, and a dual-porosity representation of fracture-dominated flow. Uncertainty was evaluated by generating 1,000 realizations of uncertain reservoir properties and boundary conditions. These were filtered against natural-state and production-history observations to obtain an ensemble of 100 models for forecasting. The ensemble was used to simulate 30 years of continued operation under three reinjection strategies: continuation of the current well configuration, relocation of injection to peripheral areas of the reservoir, and constrained relocation of selected injectors within existing concession areas. All forecasts assumed 95% recovery of lithium from produced brine and continuation of operator-reported production and reinjection rates. Under the current configuration, the median forecast cumulative lithium recovery is approximately 360 kt by 2055, substantially below the previously reported static estimate of approximately 760 kt Li. Relocating injection to peripheral areas increases median cumulative recovery to approximately 407 kt by delaying the return of lithium-depleted brine to production wells. Concession-constrained relocation provides a smaller field-wide increase, to approximately 368 kt, and redistributes recovery among neighboring production areas. These results show that effective production–reinjection connectivity and chemical breakthrough are important controls on lithium recovery. They also demonstrate that an injection strategy that benefits an individual concession may reduce recovery elsewhere in the hydraulically connected reservoir. The updated model provides a probabilistic framework for evaluating geothermal and lithium co-production. The results support coordinated field-scale management and indicate that lithium-resource assessments should account explicitly for reinjection, reservoir connectivity and operational constraints rather than relying solely on static brine-in-place estimates.

Topic: Modeling

Enhanced geothermal systems (EGS) rely on complex fracture networks to provide the permeability needed for commercial-scale heat extraction. Reservoir simulation of such systems can represent fractures either explicitly, as a discrete fracture network (DFN), or implicitly, through upscaling into an equivalent continuum representation. Explicit DFN modeling is generally considered more physically rigorous, but carries substantially greater computational and workflow complexity, and the practical trade-off between the two approaches has not been systematically quantified against real EGS field performance data. This study develops two reservoir simulation models of the Utah FORGE site, built from a common geological dataset but differing only in how the site's fracture network connecting injection well 16A(78)-32 and production well 16B(78)-32 is represented. The first model represents fractures explicitly, with fracture geometry and conductivity defined from site characterization, while the second represents the same fracture network through an upscaled continuum formulation. All other model inputs are held identical, isolating the effect of fracture representation method alone. Both models are validated against observed pressure, flow rate, and temperature data from Utah FORGE's extended circulation test, providing a against real EGS field performance rather than a synthetic test case. This study highlights how fracture representation methods affect simulated geothermal reservoir performance, including the sensitivity of results to grid resolution. The findings guide EGS modelers on when explicit DFN simulation's added complexity is justified by improved accuracy, and when continuum approaches remain an adequate, more tractable alternative.

Topic: Enhanced Geothermal Systems

Amid extended timelines for transmission interconnection, temporarily or permanently islanded microgrids are increasingly viewed as pathways to faster deployment of large electric loads. This study investigates the feasibility, economics, and optimal design of island grids built around next-generation geothermal power plants. It models and optimizes the design and operations of a binary-cycle enhanced geothermal facility alongside supplementary lithium-ion batteries and solar photovoltaics to support a co-located large electric load while meeting strict standards for availability and reliability. Results indicate that dry-cooled geothermal plants are poor candidates for island grids when operated inflexibly. While variability in dry-cooled geothermal output with changing ambient air temperature does not have a substantial impact on the value of grid-connected geothermal power plants, the need for excess capacity to maintain output on the hottest days increases the cost of dry-cooled island systems by nearly 50% compared to a wet-cooled alternative. Pairing with solar power reduces system cost moderately due to complementary patterns in solar and geothermal production profiles. Flexible operation emerges as a pathway to economically viable dry-cooled geothermal island grids, enabling a reduction in cost to near-parity with wet-cooled systems. Flexible plants produce geofluid at higher rates during summer months to compensate for reduced plant efficiency while reducing production during the winter to maintain reservoir longevity. They also prioritize injection pumping during cooler nighttime periods to minimize parasitic load on hot afternoons when efficiency is lowest. Both flexible and inflexible dry-cooled geothermal plants are optimally paired with moderate amounts of lithium-ion battery storage to help bridge low-efficiency hours.

Topic: Modeling

Firm cyclic (huff-n-puff) operation of Enhanced Geothermal Systems (EGS) relies on at least a pair of wells that are not connected in the subsurface, each attached to its own hydraulically stimulated reservoir. Fractures are held open by fluid pressure, allowing injected fluid to be produced back through the same well. Because pressurized fractures (from both reservoirs) also contribute to production, such systems may require fewer and smaller pressurized fractures than continuously circulated systems, and fluid that enters the surrounding rock during injection can partially return during production. Each well–fracture system is hydraulically isolated from the others, and fracture pressure oscillates within a bounded window rather than being continuously sustained. Water availability is increasingly recognized as a constraint on EGS deployment, particularly in arid regions. Field tests of continuously circulated doublets, from early hot dry rock experiments to recent multistage horizontal wells, have reported that a notable fraction of the injected water is not recovered, requiring make-up water over the life of a project. Historical records suggest that much of this loss originates in pressurized fracture regions that contribute little to production. This paper examines water loss in EGS under both modes of operation, in light of the long-recognized trade-off in which higher operating pressures reduce flow impedance but increase water loss. The main loss mechanisms associated with hydraulic fractures and surrounding natural fractures are considered, and their behavior under sustained and cyclic pressurization is compared using analytical and reduced-order models informed by historical field observations. The results indicate that, under cyclic pressurization, water loss is governed more by the time spent near peak pressure than by the peak value itself. They also indicate that the volume of water not returned is controlled more by the extent of the pressurized area and the pressure level relative to the in-situ stress than by the operating mode itself. These findings suggest how cyclic operation of wells disconnected in the subsurface may reduce make-up water requirements by limiting the pressurized fracture area and the time spent near peak pressure.

Topic: Enhanced Geothermal Systems

The Salton Sea Geothermal Field (SSGF) is characterized by high-temperature, high-salinity brines, a stratified reservoir system with a low-salinity cap above a hypersaline zone, and complex thermodynamic and geochemical behavior. This study presents the development and calibration of a numerical reservoir model focused on the Hudson Ranch area, with the goal of validating prior conceptual and geochemical interpretations (including double-diffusive convection models) through numerical simulation. The model aims to capture the natural-state behavior of the system and provide a framework to evaluate long-term changes associated with both geothermal energy production and critical mineral extraction using new commercial numerical modeling software (Volsung). In recent years, growing interest in lithium and other critical mineral recovery from geothermal brines has emphasized the need to understand how post-extraction reinjection alters reservoir chemistry, pressure, and temperature. This model incorporates mineral depletion and fluid re-equilibration mechanisms to assess the potential reservoir impacts of large-scale lithium extraction. The work supports future feasibility studies by quantifying the thermodynamic and chemical evolution of the system under dual-use scenarios of power generation and mineral recovery. This model provides critical insights for sustainable co-production of energy and minerals from hypersaline geothermal systems.

Topic: Modeling

Countries lacking favourable volcanic or high-enthalpy geothermal systems face significant uncertainty when evaluating the feasibility of deep geothermal power generation. Limited temperature measurements, uncertain thermal properties, and poorly constrained deep geological conditions substantially increase exploration risk and project costs. This study presents an integrated workflow that links subsurface temperature modelling, uncertainty quantification, and techno-economic assessment to identify the information most critical for reducing development uncertainty in emerging geothermal regions. The methodology uses available borehole temperature measurements, rock thermal property datasets, and geological constraints from Singapore's geothermal exploration programme to construct calibrated subsurface temperature models. Monte Carlo simulations propagate uncertainties in thermal conductivity, radiogenic heat production, geological architecture, and deep boundary conditions to generate probabilistic temperature predictions rather than deterministic outcomes. The resulting temperature distributions are integrated into geothermal techno-economic evaluations to assess how geological uncertainties affect estimates of resource quality, drilling requirements, project economics, and the likelihood of reaching temperatures associated with superhot geothermal systems (~375°C). The objective is not to predict the occurrence of a commercial resource, but to determine which datasets provide the greatest reduction in uncertainty and therefore offer the highest value for future exploration programmes. By explicitly linking geological uncertainty to economic decision-making, the study provides a framework for prioritising data acquisition in regions that are traditionally considered less favourable for high-temperature geothermal development. While demonstrated using Singapore, the methodology is intended to be transferable to other emerging geothermal jurisdictions seeking to reduce exploration risk and optimise investment decisions.

Topic: Modeling

The KenGen Green Energy Park is envisioned as an integrated industrial development anchored on the geothermal resources of the Olkaria field, Kenya. This study evaluates the technical potential for integrating geothermal electricity, direct heat, water, and associated geothermal resources into industrial applications beyond conventional power generation. The assessment focuses on cascading and sequential utilization of geothermal energy, resource efficiency, industrial symbiosis, and opportunities for value addition through geothermal-based industries. Potential applications include process heating, greenhouse agriculture, aquaculture, agro-processing, mineral processing, and other heat-intensive industries. The study further evaluates opportunities for shared infrastructure, waste-heat utilization, water reuse, and optimization of geothermal resource flows. A technical framework is proposed for matching geothermal resource characteristics—including temperature, enthalpy, flow rate, and availability with industrial energy demands. The proposed approach demonstrates how integrated utilization of geothermal resources can improve overall resource efficiency, increase the economic value of geothermal development, and reduce the environmental footprint of industrial activities. The KenGen Green Energy Park provides a potential model for geothermal-based industrial development and diversification of geothermal utilization in Kenya.

Topic: Direct Use

In two-phase geothermal fields, monitoring the enthalpy of produced fluids is important in under-standing the reservoir performance. Decreasing enthalpy can indicate breakthrough of injection water or invasion of cooler groundwater, while increasing enthalpy can indicate reservoir boiling and the formation of a steam cap. Enthalpy and mass flow rate govern the amount of steam availa-ble from each well and ultimately the energy output of the powerplant. Mass flow rate of steam and water phases and total enthalpy of the flow can be measured directly for individual geothermal wells that produce to dedicated separators. However, due to the high capital cost of production separators, most geothermal fluid gathering systems are designed with satellite separation stations in which several wells produce to a single separator. In many cases all of the two-phase fluids produced from a field are combined by the gathering system and separated in a large vessel at the power plant. Without dedicated production separators for each well, the steam and water mass flow rates and total enthalpy of individual wells cannot be measured during pro-duction. We have thus incorporated a method of mass balance to ascertain whether the production from each well is still maintained and the output tallies after separation in the field. This informs whether we experience losses or not after separation.

Topic: Field Studies

Directional drilling, a key technology for exploration and exploitation of deep geothermal resources offers a number of significant advantages: it maximises wellbore exposure through productive zones in the reservoir, enables drilling to inaccessible locations like built up areas or beneath mountainous area, sidetracking, minimises environmental damage and is economical to drill several wells from a single well pad. Directional drilling techniques enables the wellbore to be deflected from the vertical and directed towards a desired direction in order to hit a predetermined target below the surface of the earth. Sidetracking directional drilling technique is applied in a number of situations such as to bypass fish or junk in the hole, sidetracking through casing, and to bypass an unusable section of the original wellbore. This paper highlights an unconventional directional sidetrack performed during drilling of Menengai well MW15A at a depth of 1335 metres to bypass a fish in the hole. The fish was left in hole after the drill string was backed off at 1567metres as a result of stuck pipe at a total drilled depth of 1688 metres. After unsuccessful fishing, it was decided that the wellbore be filled with sand to a depth of 1400 metres, set a cement plug to 1200 metres and sidetrack. The procedure was successfully performed and the well drilled to a final depth of 2340 metres, attaining the desired design objective of achieving horizontal displacement of 600 metres . The unconventional option of filling the well with sand before setting a cement plug and sidetracking proved a viable option. The option also presented significant cost saving compared to terminating the well at that depth before drilling through all the productive zones or waiting for contractor to mobilise whipstock to side track the well.

Topic: Drilling

Growing interest in geothermal operations in highly heterogeneous or fractured reservoirs (such as EGS) shows the detrimental impact of thermal short circuits between injection and production wells, prompting investigation into remediation methods, similar to oil-production conformance-control operations. The use of bulk aqueous gels to seal these short circuits appears promising, particularly compared with preformed gel particles. Injected in the sol phase, these formulations can penetrate in-depth the reservoir before gelation, after which they plug the entire fracture, regardless its aperture size. Applying bulk-gel treatments in geothermal operations requires addressing the effects of both high temperatures and marked temperature gradients. The two main challenges concern the placement of the gel, namely the control of the temperature-dependent gelation kinetics, and the gel’s long-term stability, which in particular requires retarding syneresis effects. This study presents a new strategy to achieve optimal placement with polymer-based bulk gels suitable for temperatures above 100°C. It relies on using sulfonated polyacrylamide polymers with an inorganic crosslinker. As gelation requires interactions between crosslinkers and acrylate monomers that are not initially present in the polymer's structure, these systems remain in the sol phase during injection. This means that there is no risk of injection well plugging. When the sol self-propagates in-depth towards the high conductivity fractures, it is exposed to increasing temperatures: acrylamide hydrolysis leads to formation of acrylates, enabling gelation. The experimental results are reported in three parts. First, anaerobic thermal-aging tests demonstrate, using 1H NMR spectroscopy, the progressive hydrolysis of sulfonated polyacrylamide polymers and show that this process depends on the molar fraction of sulfonated units in the polymer backbone. Second, a series of parametric gelation-tube tests shows that the gelation kinetics are governed by both the crosslinker concentration and the molar fraction of sulfonated units. This finding indicates that, in an EGS where the temperature gradient between the injection and production wells is known, the location of gel formation can be controlled through the formulation’s chemical composition. Finally, coreflood experiments conducted on artificially fractured sandstone plugs at 130°C demonstrate that the gel can effectively seal the fracture and do remain stable over at least several weeks, with the results suggesting potential for longer-term stability beyond the experimental timeframe.

Topic: Emerging Technology

Microseismicity provides an opportunity to resolve how earthquake source processes and fracture-network behavior evolve between hydraulic stimulation and circulation in an enhanced geothermal system. We analyze seismicity from the 2024 stimulation and circulation experiments at Utah FORGE using event geometry, space-time migration, waveform cross-correlation, frequency-domain attributes, source mechanisms, and nearest-neighbor statistics. Stimulation seismicity delineated two distinct planar zones rather than a single distributed fracture network. Early-stage events concentrated on the eastern plane and followed a space-time relationship consistent with pore-pressure diffusion. Later-stage seismicity expanded onto a larger western plane, departed from the diffusive trend, and exhibited source characteristics consistent with slip on a pre-existing fault. Independent waveform observations support this transition. Cross-correlation grouped events into spatially coherent families, and frequency-domain attributes of characteristic family members separated into the same spatial and temporal regimes: early eastern families had lower peak frequencies, whereas later western families had higher peak frequencies. Available moment-tensor solutions further indicate a greater opening component in the eastern cluster and more shear-dominated behavior in the western cluster. Together, these observations identify a transition from pressure-driven fracture activation to structurally controlled fault reactivation. Circulation preferentially reactivated portions of the stimulated volume while also activating previously unobserved pathways. Compared with stimulation, circulation produced more clearly separated nearest-neighbor populations and larger coherent event families, indicating stronger event-to-event interaction along selective structures. The largest circulation family also exhibited systematic along-strike migration, suggesting progressive activation of a persistent structure. These results show that the engineered reservoir developed as two separated, mechanically distinct planar zones rather than a broadly connected fracture network. This structural organization concentrates the hydraulic response, limits reservoir-wide activation, and may leave substantial rock volume bypassed. More broadly, integrating migration patterns, waveform similarity, spectral attributes, source mechanisms, and nearest-neighbor statistics provides a real-time framework for distinguishing diffusive fracture activation from fault-controlled slip during subsurface fluid injection.

Topic: Geophysics

In structurally controlled rift valleys, steeply dipping normal faults and shear zones act as conduits for upwelling geothermal fluids, hydraulically connecting deep production intervals to the shallow subsurface. Through these pathways, deep geothermal extraction can draw from shallow hydrothermal systems and connected surface waters. We investigate these dynamics using the MODFLOW 6 Groundwater Energy (GWE) simulator that represents fully coupled groundwater flow and heat transport in a representative rift-valley setting. Predictive simulations show that deep pumping induces thermal and hydrologic drawdown, altering shallow groundwater temperatures, reducing discharge to springs, and depleting tributary groundwater. To evaluate predictive uncertainty associated with the geologic framework, we test the sensitivity of the simulated impacts to Precambrian basement depth and fault-zone architecture. Results show that a shallower basement constricts deep circulation pathways and influences the propagation of pressure and temperature changes to the shallow subsurface. In contrast, simulations suggest that the fault breccia and damage zones play a minimal role in modulating shallow impacts. This modeling framework provides a quantitative basis for evaluating geologic uncertainty and assessing potential conflicts between geothermal development and existing water rights.

Topic: Modeling

Understanding the long-term performance of Enhanced Geothermal Systems (EGS), including those modeled after the Utah FORGE project, requires accurately predicting time-dependent permeability changes within propped fracture networks. This research presents a critical sensitivity study to quantify short- and long-term permeability evolution resulting from proppant compaction and creep. We developed a coupled modeling framework, integrating instantaneous mechanical collapse (Cap-Yield model) with time-dependent hydrothermal creep (Norton-Bailey model), and validated this framework against laboratory data to simulate reservoir performance over a 10-year production horizon. Our analysis reveals that permeability loss is highly conditional on the operational environment. Specifically, we demonstrate that under baseline conditions—assuming an effective stress of 1000 psi (6.9 MPa) and an injection temperature of 75C at the injection well—creep compaction and long-term permeability reduction remain relatively small, even for fractures propped with natural sand. However, the heterogeneity inherent in real reservoir conditions suggests that localized zones of significantly higher stress and temperature can induce substantial mechanical crushing and accelerated creep. For example, severe permeability loss (potentially below 10% retention over a decade) is estimated for fracture propped with natural quartz sand under high stress of up to 5000 psi (34.4 MPa) and temperatures up to 200C. The study concludes that fracture permeability degradation is strongly dependent on the synergistic effects of stress and temperature on the proppant pack. While our modeling successfully captures this stress-thermal sensitivity, we acknowledge inherent uncertainties in extrapolating transient decay functions over multi-year timescales. Therefore, to confidently constrain these long-term effects and ensure reliable forecasting, we strongly recommend targeted, long-term experimental validation. The study concludes that fracture permeability degradation is strongly dependent on the synergistic effects of stress and temperature on the proppant pack. While our modeling successfully captures this stress-thermal sensitivity, we acknowledge inherent uncertainties in extrapolating transient decay functions over multi-year timescales. Therefore, to confidently constrain these long-term effects and ensure reliable forecasting, we strongly recommend targeted, long-term experimental validation.

Topic: FORGE

High-Temperature Aquifer Thermal Energy Storage (HT-ATES) holds great promise for sustainable energy by allowing the storage and retrieval of high heat content in aquifers. However, the process of injecting hot fluids into the subsurface carries inherent geomechanical risks, such as fracturing and ground uplift, that must be rigorously managed. This paper presents a coupled Thermo-Hydro-Mechanical (THM) numerical simulation of an HT-ATES system at the Geostorage Forsthaus pilot project in Switzerland to assess performance and develop risk mitigation strategies. Using the TOUGH-FLAC simulator, we analyzed the coupled effects of heat transfer, fluid flow, and rock mechanics within a shallow, layered reservoir structure. Our results highlight that while fluid pressure drives large-scale changes, temperature variations dictate localized stress and strain responses. Critically, we found that by implementing targeted well perforation and strict pressure control, the potential for formation damage and surface uplift can be effectively minimized. Furthermore, we established that subsurface strain is largely governed by thermal expansion, providing a strong link between temperature and mechanical deformation. This study concludes by proposing a crucial monitoring strategy: the deployment of continuous fiber optic temperature and strain measurements. These tools can effectively track subsurface responses and detect localized anomalies—such as the opening of weak geological planes—long before catastrophic failure occurs. This integrated THM modeling approach provides engineers with the necessary tools to safely design, operate, and monitor high-temperature energy storage systems.

Topic: Reservoir Engineering

Geothermal Energy Networks (GENs) provide a critical tool for the transition to electric heating and cooling of domestic building stock. Comprised of a single-pipe, ambient temperature hydraulic loop, connecting individual building’s water source heat pumps and centralized ground heat exchanger fields, the technology allows for utilization of otherwise wasted thermal energy, provides microgrid resilience, and lowers the electrical peak demand for each building compared to other electrification pathways. Geothermal networks also minimize levelized cost to consumers by eliminating fuel demand, and neighborhood scale transition facilitates minimization of transition costs to consumers. Multiple pilot projects have demonstrated the efficacy of the technology and have provided data which can be used to inform optimization of system designs, which in turn will maximize cost and grid benefits. HEATNETS, a first of its kind, reduced order technoeconomic model developed in partnership between HEET and the National Lab of the Rockies, was developed as a tool to explore the design and operational optimization of GENs as the technology is scaling, without the need for computational intensity of proprietary full physics models. In this study, HEATNETS was validated using operational SCADA data from the first of its kind, utility owned and operated pilot GEN in Framingham MA. The validated model was then used to perform a parametric optimization analysis of design features. Eight different infrastructure (e.g. borehole number, horizontal loop diameter) and operational (e.g. deadband window, use algorithm of auxiliary power) parameters were varied and explored for their effect on system coefficient of performance, overall electrical use, and present value of costs. Each variable was explored individually as a lever for maximizing performance in the desired output metric within an allowable design window. For the GEN installation analyzed, this study found design parameters that reduced the number of ground heat exchangers without breaching the design window, heavily reducing capital costs, before incurring significant compromises in system COP or individual heat pump performance. While this presents one optimization of the Framingham network, the methodology also allows for exploration of future networks that feature different design constraints, as well as testing of additional design or operational metrics beyond the scope of this study. Utilization of this type of optimization modeling will ensure future GENs can inform best design and operational practices before construction, maximizing societal benefit as the technology scales. In this way the application of key learnings from existing networks will be used to provide a road map for future GEN installations within the state of MA and beyond. The HEATNETS modeling tool facilitates the rapid implementation of this process in a research-based environment.

Topic: Low Temperature

Ceramic matrix composites can serve as a promising alternative to steel for geothermal casing applications, offering enhance corrosion resistance and durability. Polymer derived ceramic matrix such as siloxane derived silicon oxycarbide (SiOC) provides cost competitiveness, and stability against high temperature aggressive environment. Moreover, the low temperature pyrolysis of polymer to convert amorphous ceramics instead of high temperature sintering (above 1000 °C) allows the use of low cost fibers/fabrics for reinforcement, such as glass and basalt. These amorphous ceramic matrix composites can withstand geothermal operating temperature while offer lower production cost by eliminating sintering process and expensive carbon fibers conventionally used for crystalline ceramic matrix composites. Herein, we have investigated the effect of fiber orientation and layup strategy on the mechanical performance of the casing structures which includes tensile properties, burst pressure and collapsing pressure. Flat coupons were fabricated with different fiber angles involving braided fabric, woven/non-woven fabrics to measure tensile strength, and modulus and determine burst pressure and collapsing pressure as a function of fiber angle using composite laminate theory. The coupon level experimental results were validated by comparing with finite element analysis (FEA) analysis conducted for tubular structure. Furthermore, the validated FEA model was leveraged to optimize fiber lay-up strategy including fiber orientation, layer thickness, and stacking sequence through an iterative design optimization process.

Topic: Enhanced Geothermal Systems

Depleted hydrocarbon fields and their associated subsurface datasets offer a valuable starting point for geothermal development and critical-mineral extraction. Well logs, seismic surveys, and geological interpretations collected during decades of hydrocarbon exploration can help characterize reservoir geometry and properties, reducing uncertainty and supporting the assessment of new subsurface uses. This study demonstrates the value of legacy data from the 1960s–1980s and 2000s for developing a high-temperature heat-storage project in the Leopoldshafen oil field, Upper Rhine Graben (Germany). We reinterpret historical spontaneous-potential (SP) and resistivity logs using modern sedimentological concepts and sequence-stratigraphic frameworks to improve the identification and correlation of reservoir units. This approach and seismic attribute analysis investigate reservoir architecture and identify structures that have been recognized in earlier interpretations. We also apply AI-based methods to predict porosity from available legacy well data, extending reservoir-property estimates beyond intervals with direct measurements. We evaluate these interpretations against data from a currently drilled well, DeepStor-1, targeted for high-temperature heat storage at temperatures of up to 140 °C. This comparison provides an independent check on the predicted reservoir properties and geological interpretation. Building on the integrated legacy well data, we are developing approaches to evaluate the safety of high-temperature heat storage at these temperatures. These approaches assess how injection and production may affect reservoir pressure and temperature, subsurface stresses, fault stability, and the integrity of the surrounding formations, helping identify operational conditions that support safe and reliable storage. Together, the results illustrate how integrating historical datasets with modern analytical methods can improve subsurface characterization, guide well targeting, forecast safety aspects and support the transition of hydrocarbon fields to geothermal applications. The approach may also be relevant to assessing reservoirs for other uses, including critical-mineral extraction and subsurface energy storage.

Topic: General

Natural rocks in enhanced geothermal energy reservoirs consist of fractures, solid and fluid phases that form complex structures at multiple scales. Explicit incorporation of discrete fracture networks into large-scale coupled models is computationally prohibitive. Therefore, efficient multi-scale and multi-physics strategies are necessary for capturing the behavior of fractured rocks under subsurface thermo-hydro-chemo-mechanical conditions. In this work, we present a multiscale framework for the coupled thermo-hydro-mechanical behavior of rocks in geothermal reservoirs. The proposed approach allows for separate micro-constitutive laws, providing an efficient way of capturing both large- and small-scale processes and smeared representations of fractures. The performance of the modeling framework in the simulation of geothermal energy reservoir is demonstrated through numerical examples.

Topic: Modeling

Microbial activity in geothermal infrastructure can have important operational consequences as biofilm formation can reduce the heat-exchange efficiency of a cooling facility, facilitate scaling and corrosion, and thus shorten the operational lifespan of equipment. In an enhanced geothermal system (EGS), where cooled fluid is reinjected, biofilm may have additional impact by impairing injectivity if present in significant amounts. Here we report the first geochemical and genomic characterization of microbial biomass and its derivatives in an operating EGS. During a multi-month circulation test at the Utah FORGE site in 2026, sludge-like microbial biomass with elevated DNA concentrations was observed within the first week of circulation. It consistently re-accumulated after flushing in the piping between the cooling tower and the injection well, where brine temperatures remained below the upper temperature limit for life. The microbial community and associated organic compounds in sludge and water samples were analyzed using a suite of geochemical and biological methods, including GC-FID, GC-MS, and whole-genome sequencing (WGS). WGS analyses were conducted on samples collected throughout the circulation system to characterize microbial community composition, assess spatial variability and potential source populations, identify functional capabilities relevant to biofilm formation and system performance, and determine the conditions controlling biomass accumulation along the circulation loop. Notably, the sludge sample was overwhelmingly dominated by a thermophilic genus, Tepidiphilus, which is frequently associated with persistent biofilm development and surface colonization. The findings will provide the basis for a microbial activity detection workflow and for strategies to prevent and mitigate biofilm-related risks in future EGS operations.

Topic: FORGE

Laboratory scale geothermal simulators are increasingly employed for testing downhole tools such as casing assemblies and packers at temperatures above 250 °C. Heating is the most important condition to be reproduced by these simulators and the selection of the heating method directly influences the representativeness of the tests. Three heating methods (infrared, ring and induction coil) were compared in this study by using finite element heat transfer analysis in ANSYS. In an effort to isolate the effect of the heating method, all three configurations used the same casing geometry and identical boundary conditions. The boundary conditions were natural convection (h = 7.5 W/m²·K), radiation (emissivity = 0.8), and an applied heat flux of 6,600 W/m². The induction coil performed best, reaching a maximum temperature of 273.8 °C, with a peak heat flux of 62.23 kW/m². While, the ring heater reached 251.35 °C and 48.2 kW/m², while the infrared heater reached 202.56 °C and 51.4 kW/m². The superior performance of induction heating is attributed to the uniform distribution of the coils around the casing, which delivers higher peak heating intensity and more uniform temperatures under identical input flux. Since geothermal simulators must transfer heat effectively into the annulus and packer zone, induction coil heating is recommended over ring and infrared heating for these applications.

Topic: Modeling

The rapid expansion of hyperscale data centers in tropical regions has intensified the challenge of managing high cooling demand under elevated ambient temperature and humidity conditions. Facilities operated by Data Center Indonesia (DCI) in Bekasi including JK1, JK2, JK3, JK5, JK6 at the H1 Bekasi campus represent one of the largest concentrated data center clusters in Southeast Asia, with a combined installed capacity of 73 MW. In such environments, cooling systems typically account for 40% of total facility power consumption, with nearly all IT electrical load ultimately converted into heat that must be continuously rejected. This study proposes the integration of a centralized geothermal-assisted cooling system based on Ground Source Heat Pump (GSHP) technology to serve multiple co-located facilities within the Bekasi campus. A quantitative thermal assessment is conducted to estimate the aggregate cooling load and heat rejection requirements across all sites. The Power Usage Effectiveness (PUE) in DCI Bekasi is 1,5, therefore the total IT load across the campus is estimated at approximately 48.7 MW, corresponding to a comparable continuous heat rejection requirement. Meanwhile, the existing cooling electricity demand is further estimated to be approximately 24 MW. System performance was evaluated using representative GSHP operating scenarios with Coefficient of Performance (COP) values ranging from 3.0 to 5.0. The assessment indicates that the proposed system could reduce cooling electricity consumption by approximately 30 to more than 50% of the total cooling load, equivalent to 9.7–16.2 MW of continuous heat rejection capacity. This integration would significantly reduce reliance on evaporative cooling, thereby minimizing water consumption and eliminating the need for chemical water treatment associated with cooling tower operations. This study demonstrates that geothermal-assisted cooling offers a technically viable pathway to address the substantial heat rejection requirements of hyperscale data center clusters in tropical regions. The findings provide a quantitative foundation for integrating subsurface thermal resources into existing data center infrastructure, with a particular emphasis on large, multi-facility campuses.

Topic: Direct Use

Fluid injection near a natural fractures and fault changes both pore pressure and the stresses acting on the fault which can lead to reactivation, depending on the fault’s orientation and position relative to the injection source. Understanding this interaction requires examining not only whether a fault slips, but also how much it slips, in which direction, and how the slipping region and microseismicity evolve. This study uses three-dimensional poroelastic simulations to investigate fault response to fluid injection into the surrounding rock matrix. The formulation incorporates induced normal and shear stresses, pressure diffusion, and coupled fault deformation and fluid flow. The analysis examines slip onset, displacement magnitude, and the spread of slipping regions across the fault. In addition, a novel technique satisfying the energy balance is utilized to simulate the induced microseismicty. The effects of injection rate, source–fault distance, and fault orientation on the microseismic cloud will be investigated.

Topic: Enhanced Geothermal Systems

Accurate prediction of fracture connectivity and fluid circulation pathways remains a key uncertainty in the development of hydraulically stimulated Enhanced Geothermal Systems (EGS) for forecasting long-term injectivity, productivity, and thermal performance. This study presents a field-scale validation of CMG's fracturing and reservoir simulator using data acquired from a geothermal pilot consisting of an injector-producer well pair stimulated through a multi-stage hydraulic fracturing program at Newberry Superhot Rock (SHR) EGS pilot in Oregon. The pilot comprises injector (55-29) and producer (55A-29), approximately 3,008 m deep within a reservoir characterized by temperatures exceeding 330°C. The stimulation campaign included six propped hydraulic fracture stages, with detailed pumping schedules, treatment pressures, fluid volumes, and completion data available for model construction and calibration. Pressure transient analyses and production/injection histories were used to characterize fracture connectivity and reservoir performance. Key comparison included pressure evolution during pumping and shut-in, inferred fracture geometry and conductivity, and connectivity between the wells. Emphasis was placed on evaluating the model's ability to capture fracture geometry evolution, fracture-driven connectivity between the injector and producer, and the resulting flow pathways within the engineered geothermal reservoir. Model predictions were compared against measured injection and production behavior and other available surveillance data. The analysis assessed how effectively a physics-based fracture simulation approach can reproduce observed reservoir responses and examine uncertainties in fracture geometry, stimulated reservoir volume, and reservoir heterogeneity. Evaluating the model with treatment records and subsequent inter-well measurements could help assess which parameters could be constrained by the available data and which fracture characteristics could remain non-unique. The resulting field comparison provides a basis supporting geothermal stimulation design and identifying additional diagnostics needed to improve predictions in high-temperature reservoirs providing guidance for future design, optimization, and validation of geothermal stimulation programs. The study offers valuable insight into the reliability of coupled geomechanical-reservoir simulation for forecasting the performance of hydraulically stimulated geothermal reservoirs and for optimizing future EGS development strategies.

Topic: Modeling

Enhanced geothermal systems (EGS) are quickly becoming a critical part of a sustainable energy future. Understanding and optimization of these systems will be crucial to their success. This study presents an investigation using simulation of a Pilot demonstration of the world’s hottest EGS. The pilot well pair is located at Newberry site in Central Oregon. The pilot program is comprised of an injector–producer well pair, 55-29 and 55A-29, completed at depths exceeding 3,050 m separated by 100 m. Each well was stimulated with six propped fractures stages, followed by a 40-day circulation test conducted at rates 1 to 3 bbl/min. The demonstration was designed to evaluate the hydraulic stimulation, inter-well circulation, and heat generated between the pair. Circulation testing achieved production temperatures approaching 231°C and demonstrated hydraulic recovery efficiencies exceeding 90%, confirming strong communication between the stimulated wells. Both particle and chemical tracers were used to diagnose the path between the wells. Multiple production logging tool (PLT) surveys were conducted to characterize the evolution of the temperature profile with time. To understand the governing flow and heat-transfer mechanisms of the demonstration, a thermal-hydraulic model was developed using a coupled reservoir-wellbore simulator. The simulation model was calibrated to field pressure, temperature, stimulation and circulation data. The work focusses on history matching the temperature post stimulation and during the initial circulation period. The subsequent four-month thermal rebound period was also modeled and evaluated. History matching required the introduction of independent stimulated reservoir volume (SRV) regions surrounding the injector and producer, as well as pressure-dependent fracture conductivity behavior. The calibrated model indicated large stimulation volumes around the wells, particularly around the producer. The modeling results also suggest reduced conductivity in selected fracture stages during circulation and highlight the importance of fracture conductivity evolution and reservoir compressibility in reproducing observed field performance By running field-calibrated heat-harvesting and thermal rebound simulations, this work provides a distinctive case study of high-temperature EGS reservoir behavior in hot crystalline formations. Insights into fluid leak-off, heat extraction, and fracture-controlled flow mechanisms and their impact on heat recovery performance were determined. This simulation study presents a comprehensive framework for assessing stimulation effectiveness and reservoir connectivity, while capturing the unique physics of EGS and providing valuable insights to support future EGS reservoir development strategies.

Topic: Field Studies

The U.S. Geological Survey is updating its assessment of conventional hydrothermal resources in the Great Basin. The assessment applies conditional statistics, which consider areas of similar geologic conditions as having similar likelihoods for hosting a geothermal resource. Hence, partitioning a study area into spatially contiguous and statistically similar regions (i.e. regionalization) can aid the computation of conditional statistics. Recent examples of regionalization include a supervised tree-ensemble leaf-distance workflow and an unsupervised principal components with k-means clustering (PCAk). Herein, we present an unsupervised alternative, Agglomerative Regionalization via Information Divergence (ARID), and compare ARID to leaf-distance and PCAk. ARID starts with small sub-regions and then repeatedly merges the most similar adjacent zones, comparing each feature distribution via Kullback-Leibler divergence rather than averages alone. All algorithms regionalize the Great Basin from 52 features (e.g., heat flow, strain). Partitions are scored internally based on the similarity of feature values within zones and the separation of feature values between zones. Partitions are scored externally by continuous ranked probability score (CRPS) against subsurface temperature estimates from two independent geothermometers (silica [n=2,678], Na–K–Ca [n=2,614]) to test if zones distinguish resource temperatures. Because geologic properties are typically spatially autocorrelated, apparent skill can arise from spatial proximity alone; to control for spatial autocorrelation, comparisons are referenced against randomized contiguous zonations. Despite being unsupervised, ARID predominantly presents the same zonation pattern as the supervised leaf-distance approach, with 63% of the area of the same zones overlapping, suggesting identification of similar structure(s) in feature space. ARID is also the only method to exceed the randomized contiguous zonations on internal criteria (p=0.005). No method exceeded the 5th percentile of the randomized contiguous zones for the external temperature score, although both ARID and leaf-distance beat the median (p=0.34 and p=0.42, respectively). These results suggest that, given the features evaluated here, geographic proximity to measured sites was more informative for characterizing subsurface temperature than the feature-based partitioning methods. ARID demonstrates advantages in producing internally coherent sub-regions without the need for labels and is independently corroborated by the supervised zonation of the leaf-distance workflow. While slight, ARID beats the median external temperature characterization of the spatial null; highlighting the skill occupied by spatial proximity alone. Future work may investigate incorporating these zones into a conditional statistic workflow.

Topic: General

A long-term circulation test is currently underway at Utah FORGE, having started August 12, 2026, in which produced water is being recycled and supplemented by makeup water from the nearby groundwater well 58B-32. The circulated water is pressurized along the full length of the pathway to suppress vapor-gas saturation and two-phase flow, and the stimulated reservoir ranges from 200 to 225 degrees C. In preparation of the test, a permanent high pressure production pipeline about 175 m long was constructed and installed with a series of sampling ports that permit monitoring of fluid chemistry and corrosion coupons upstream and downstream of a cooling tower, which replicates geothermal energy production and cools the produced water to approximately 65 deg C. Earlier circulation tests and analyses of injected tracers indicate the round trip circulation of a given packet of injected water is roughly 2-5 hours. Furthermore, the analyses of produced waters in 2024 and 2025 showed that there is risk of silica scaling on the cooling leg of the power cycle and carbonate and anhydrite scaling on the heating leg of the power cycle. The earlier geochemical data also helped to characterize the residual fluid left in the reservoir, which means the produced water at any point in time represents a combination of injected and reservoir waters. The current geochemical monitoring program is thus designed to evaluate the evolution in fluid chemistry, to test the performance of a series of additives to mitigate scaling and corrosion, and to support interpretation of reservoir processes. At the time of this submission, geochemical data from the first four weeks of circulation are available. Comparison of the injected and produced waters indicate increases in ionic components including sodium, potassium, and chloride due to dissolution of soluble salts that are being rinsed out for the reservoir flow paths, whereas decreases in calcium, magnesium, and sulfate imply sequestration as carbonates or sulfate minerals in the reservoir; both effects are diminishing with time. Dissolved silica is relatively unchanged, reflecting quartz-silica equilibrium at reservoir temperatures about 205 deg C; similarly dissolved lithium and boron are unchanged, making them useful references for understanding reservoir processes such as mineral deposition, mixing and cooling. Carbon dioxide concentrations and pH values in produced waters are closely similar (about 0.1 wt% and pH of 6) to those measured during the 2024 circulation test, which helps to mitigate carbonate and silica scaling. These attributes plus the injection of additives, comprising water treated polymers, dispersants, and acrylic polymers, appear to be minimizing mineral deposition and corrosion in the pipeline. Emerging as key parameter is the growth of microbial communities in the cooling tower and pipeline that produce hydrocarbons and perhaps affect the concentrations of other aqueous constituents. Additional data are expected to be acquired through November, 2026, and the results will be incorporated and discussed as part of the full paper submission.

Topic: FORGE

The Salton Sea Geothermal Field (SSGF) in Southern California hosts an untapped potential resource of Critical Minerals (i.e., Li, Zn, Mn, Rb, Sr) in a hot, hypersaline, metalliferous brine and its associated reservoir rocks. Here, we characterize the distribution and mobility of these elements as recorded by the reservoir rocks as a function of depth, brine temperature, and mineral composition. The California State 2-14 scientific drill core contains records of distinct prograde metamorphic mineral zones with increasing depth: calcite-chlorite at depths shallower than 2480 meters, biotite from 2480 to 3000 meters, and amphibole below 3000 meters. In the calcite-chlorite zone (T less than 350°C), the primary mineral host for Li, Zn, and Mn is chlorite, for Sr is epidote, and for Rb is K-feldspar. As depth increases, the appearance of biotite marks the biotite zone (T~350°C). Here, the primary mineral host for Li and Zn remains chlorite, though their concentrations decrease relative to the concentrations of both elements in chlorite in the lower T calcite-chlorite zone, Mn is found in the highest concentrations in epidote (and at concentrations higher than in chlorite in the lower T calcite-chlorite zone), the primary mineral host for Sr remains epidote (though concentrations are higher in epidote in the biotite zone than in the lower T calcite-chlorite zone), and Rb is found in the highest concentration in biotite (and at concentrations higher than in K-feldspar in the lower T calcite-chlorite zone). Reservoir rocks from the bottom-most portions of nine commercially drilled wells located up to 8.5 km south and 6 km west of the California State 2-14 scientific drill core have also been analyzed; four are in the calcite-chlorite zone, two are in the biotite zone, and four are in the amphibole zone. Primary mineral hosts and general distributions of Li, Zn, Mn, Rb, and Sr in the calcite-chlorite and biotite metamorphic zones are the same as in California State 2-14. In wells that reach amphibole metamorphic zones (T greater than 350°C), actinolite becomes the primary mineral host for Mn (while Li, Zn, Rb, and Sr concentrations remain low in actinolite). At these temperatures, chlorite is not a stable mineral phase and biotite hosts Li and Zn in concentrations lower than chlorite from shallower metamorphic zones. Temperature is a dominant control for Critical Mineral variation within the reservoir rocks host minerals. At temperatures lower than 350°C, Li and Zn are sequestered in chlorite at concentrations higher than those in the bulk brine. As temperatures increase, Li and Zn concentrations decrease in chlorite at concentrations lower than those in the bulk brine. The inverse is true for Mn, Rb, and Sr, where concentrations increase in their host minerals as temperatures increase. Optimized direct extraction of these critical minerals from the brine will require strategies adapted for each element: a focus on low temperature brines for the highest concentrations of Mn, Rb, and Sr, and on high temperature brines for the highest concentrations of Li and Zn.

Topic: Geochemistry

Repurposing existing oil and gas infrastructure for geothermal and waste-heat power generation offers a low-capital pathway to expand domestic clean energy production, but it introduces a distinct engineering challenge: operators must remotely monitor and control heat-to-power assets that vary widely in source type, mechanical configuration, and geographic location, often without the benefit of a purpose-built, single-site power plant's centralized instrumentation. This paper presents a unified, vendor-certified SCADA and remote operations architecture developed to address this challenge, and reports field results from its deployment across three structurally different heat-to-power source types. The architecture is built on a Unified Namespace using MQTT Sparkplug B, standardizing data flow and enabling centralized data governance across more than 100,000 monitored points spanning geographically distributed sites. A centralized Unified Control Layer integrates heterogeneous power generation equipment through OPC UA and multi-vendor programmable logic controller connectivity, providing real-time telemetry, alarming, and operator interfaces from a single platform regardless of underlying hardware or physical location. This architecture underlies three deployed projects representing distinct heat-to-power source types. First, a project supported by the U.S. Department of Energy Geothermal Technologies Office, part of the Wells of Opportunity initiative, applies the platform to generate electricity from repurposed oil and gas wells in California San Joaquin Valley, demonstrating the architecture applicability to geothermal energy recovery from existing wellbore infrastructure. Second, a waste heat to power deployment at a natural gas compressor station in Dry Ridge, West Virginia, in partnership with CNX Resources, generates approximately 0.5 MW of zero emission power from compressor exhaust heat, reducing an estimated 1,325 metric tons of CO2 emissions in its initial phase, with a scaling pathway toward approximately 4,000 metric tons of annual reduction. Third, an industrial waste heat to power system at a seafood processing facility in Dutch Harbor, Alaska, operated in partnership with UniSea Inc, generated 67 MWh of electricity during a three week startup phase alone, representing approximately 5,000 gallons of displaced diesel fuel, and now displaces an estimated 3,500 gallons of diesel per week at full operation in a remote, off grid environment. Across all three deployments, the same underlying control architecture supported integration, monitoring, and remote operation without site specific rearchitecting, demonstrating that a standardized, vendor certified SCADA and namespace approach can generalize across geothermal wellbore heat, oilfield adjacent compressor waste heat, and industrial process waste heat. This paper details the technical architecture, integration methodology, and field validated outcomes, and discusses implications for scaling remote heat to power operations across additional repurposed oil and gas assets and off grid or resource constrained sites nationally.

Topic: Emerging Technology

Low-temperature geothermal resources are widespread and accessible but often unsuitable for electricity generation because of their low enthalpy. In Iceland, where low-temperature geothermal fields are widely used for heating and non-recyclable municipal solid waste is still exported, waste-to-energy provides both a domestic waste-management pathway and a stable external heat source. This study evaluates how externally supplied waste-to-energy heat can enhance electricity generation from Icelandic low-temperature geothermal resources and how the point of heat addition influences thermodynamic and economic performance. Steady-state first- and second-law analyses are applied to a baseline single-flash steam plant and two hybrid configurations in which external heat is added either to the geothermal brine upstream of the flash separator or to the separated steam at the turbine inlet. Brine-side heating increases the mass of steam available for expansion, while turbine-inlet heating increases the work extracted per unit of steam. Results show that upstream brine heating yields substantially higher power output by increasing steam production, whereas turbine inlet heating achieves higher exergy efficiency with considerably less external heat input. When external heat is treated as a purchased input, improved thermodynamic performance does not necessarily reduce the levelized cost of electricity. At an assumed heat price of 15 EUR/MWh thermal, the cost of external heat strongly influences the comparative economic performance. Although district heating revenue is excluded from the electricity-only cost metric, a separate revenue assessment shows that the use of recoverable heat can substantially improve the economics of hybrid configurations. Overall, the analysis demonstrates that geothermal waste-to-energy hybridization can increase electricity generation from low-temperature geothermal resources, but its value depends strongly on the integration strategy and the economics of the heat source.

Topic: Low Temperature

Borehole image logs (BHI) are a fundamental dataset collected in geothermal exploration wells, providing direct observations of structures intersecting the borehole wall, as well as constraints on the in-situ stress field. While abundant case histories document BHI interpretation and its impact at the reservoir scale, the geothermal industry lacks the shared standards and processes needed to consistently maximize the value of BHI acquisition and analysis. Drawing on the joint perspectives of a geothermal operator, Ormat, and a dedicated BHI analysis specialist, Cubic Earth, we first review modern BHI technologies and their application to problem solving in geothermal environments. Building on that foundation, we discuss how BHI data inform decisions over immediate, near-term, and long-term timescales, and we offer best practices for planning and executing BHI acquisition. Finally, we highlight special considerations for interpreting BHI in geothermal environments and examine BHI interpretation from the borehole to the reservoir scale. Together, these elements provide a practical framework for geothermal operators seeking to extract greater value from image log data.

Topic: Geology

Lone Mountain is a hidden, ~165 °C, fracture-hosted geothermal system in the Walker Lane tectonic province of western Nevada and is the site of a recent greenfield discovery by Ormat. A 2025 drilling campaign identified high-permeability production and injection zones at separate right stepovers within broadly north-trending fault systems. Here we review the regional tectonic setting, present the geologic framework of the Lone Mountain system, describe the geophysical datasets that supported structural and conceptual modeling, and summarize key drilling results. We integrate the geologic, geophysical, and drilling data to characterize the permeability architecture of the system. Permeability is concentrated near fault intersections within the stepovers and extends into deep metamorphic basement rocks beneath a regional detachment, and the reservoir is confined by a lithologic caprock rather than a hydrothermal clay cap.

Topic: Geology

There is growing interest in hydrogen as an energy resource from geothermal systems, where it commonly constitutes a significant fraction of the noncondensable gas phase. In this study we aim to develop and validate standardized, field-ready protocols for gas sampling, preservation, and analysis in geothermal systems, with the goal of improving data quality, comparability, and interpretation across sites. This includes: 1) quantifying the consistency and accuracy of in situ gas sensing technologies relative to established laboratory-based analytical methods; 2) evaluating differences between downhole and surface gas samples and establishing best practices for sampling; 3) constraining hydrogen behavior, including losses due to chemical reactions, using isotopic methods; 4) assessing the effects of operational processes — including boiling and reinjection — on hydrogen and noncondensable gas concentrations; and 5) establishing community-accepted best practices for sampling, sample preservation, and analytical workflows. A preliminary sampling campaign at the Anderson Springs geothermal site in northern California was used to build confidence in the protocols. Results, which are in broad agreement with those carried out by Lowenstern and Janik (2005), show average gas compositions in mol % (n = 9) of: CO2 = 79.69; H2S = 4.59; H2 = 6.16; CH4 = 1.67; NH3 = 6.59; N2 = 1.14; Ar = 0.02; Air = 1.65.

Topic: Geochemistry

An in situ Thermal Response Test (TRT) was conducted in San Luis, Argentina, on a low-enthalpy geothermal system composed of two exchangers connected in series: a horizontal coil embedded in the concrete floor of a swimming pool and a vertical perimeter coil installed in the surrounding soil. The test lasted 72 hours, with an average injected power of 1070 W and a total effective length of 68.36 m. Using the Infinite Line Source (ILS) model and the slope method, an effective thermal conductivity of 0.96 W/(m·K) and an effective thermal resistance of 0.63 m·K/W were determined. Comparison with individual tests of the horizontal and vertical exchangers showed a decrease in effective thermal conductivity and an increase in thermal resistance. This behavior indicates that the obtained values reflect the integrated thermal response of the exchanger–surrounding medium system, conditioned by its geometry, materials, and power distribution, rather than a variation in the intrinsic soil conductivity. Consequently, the λₑff value determined by TRT should be interpreted as an effective parameter associated with the global system response, and not as a modification of the soil’s inherent thermal conductivity.

Topic: Low Temperature

This study presents a statistical validation method applied to in situ Thermal Response Tests (TRT), comparing calibration lines from vertical and horizontal geothermal heat exchangers. Despite differences in design, time, and location, the comparison was conducted through variance homogeneity analysis and slope similarity testing. An F-test revealed non-homogeneous variances between the slopes. Subsequently, a T-test was applied: at confidence levels of 95%, 97.5%, and 99%, the slopes were statistically different; however, at 99.75% confidence, the slopes were found to be similar. This result validates the use of the horizontal exchanger method to determine the effective soil heat transfer coefficient (λₑff), confirming its robustness and reproducibility. The proposed statistical approach strengthens the reliability of TRT-based geothermal design, particularly for non-conventional exchangers adapted to existing excavations.

Topic: Low Temperature

Increased demand and requirements for sustainable geothermal operations in relation to gas emissions calls for a deep understanding of the primary criteria operators face when designing solutions. All geothermal fluids contain species that would be gases under certain temperatures and pressures. Such species, called non-condensable gas (NCG) include CO2, H2S, H2, N2, NH3, CH4. Some types of geothermal power plants do not allow NCGs to evolve, thus do not vent or otherwise deal with NCG. Other types of plants have NCG that is in a gas phase. In these geothermal plants, NCG may be collected for use or other disposition (e.g. reinjection underground). This paper explains differences between types of geothermal power plants (e.g. condensing flash plants; back-pressure flash plants; single and two-phase binary plants), including their relative efficiencies, the situations under which each type is typically used, and relative costs. Then, the paper discusses how the type of plant affects disposition of the NCG. For example, collecting NCG for use (or to eliminate venting) requires different considerations for the different types of geothermal power plants. The paper also reviews potential dispositions for NCG. For example, NCG could be re-injected back underground for sustainable operations. Alternately, species (e.g. CO2, H2S, perhaps H2) could be separated prior to venting the remaining NCG or sending the remainder elsewhere. The content should be useful for plant operators as well as those considering future geothermal developments. This paper is meant to contribute know-how from the IEA Geothermal Gas Management Task to enhance global research and geothermal development.

Topic: Production Engineering

The Matsukawa Geothermal Power Plant was the first geothermal station in Japan starting commercial operations in 1966. More than 300 hypocenters of microearthquakes were routinely determined since 2002 in the Matsukawa field using the nation-wide network data. Roughly they are distributed around the boundary of a diorite body intruding into the Miocene formation, where productive fractures are found. The resolution of the hypocentral location is too low to delineate fracture structure. All the digital data from the local volcanic observation stations as well as the nation-wide networks have been available for all researchers. Two out of the local volcanic networks are working near the Matsukawa field. The nearest station is situated in two kilometers from the Matsukawa GPP. We made waveform cross correlation and relocated microearthquake hypocenters accurately enough to delineate fracture structure. We also developed kinematic source models for the seven microearthquakes whose magnitude are larger than 2.5. The source models as well as accurate hypocenter relocations are effective to image the fracture structure of the geothermal reservoir.

Topic: Geophysics

Geothermal exploration in volcanic environments involves substantial uncertainty, particularly where structural permeability, thermal upflow, and surface geothermal indicators are not spatially aligned. Conventional exploration screening often relies on subjective expert judgement, creating a need for more transparent and systematic target prioritization. This study develops a GIS based decision support model integrating the Analytical Hierarchy Process (AHP) and fuzzy logic for the Ungaran geothermal area in Central Java, Indonesia. Four spatial criteria, comprising filtered lineament density, elevation corrected land surface temperature anomalies, multispectral alteration proxies, and geothermal manifestation proximity, were standardized using fuzzy membership functions and weighted through Chang Fuzzy AHP to generate a Geothermal Potential Index (GPI). The results identify lineament density (0.433) and manifestation proximity (0.362) as the dominant criteria, reflecting the importance of structural favorability and surface geothermal evidence. The resulting prospectivity map delineates approximately 30 km² of high prospectivity and 1 km² of very high prospectivity. These areas were further evaluated against terrain and land use constraints to distinguish technical geothermal potential from practical exploration suitability. Comparison with published geological and geothermal system interpretations indicates that favorable zones around Gedongsongo are consistent with the interpreted upflow expression, while other prospective areas may represent secondary structural pathways or outflow related features. By integrating uncertain surface indicators with geological interpretation and practical exploration constraints, the proposed framework provides a reproducible approach for prioritizing further geological, geochemical, and geophysical investigations before exploration drilling.

Topic: Geology

Space cooling demand is increasing in tropical cities as rising temperatures intensify electricity consumption. Cooling equipment already accounts for approximately 20% of global electricity consumption, while Indonesia recorded a mean air temperature anomaly of 0.8°C in 2024 relative to the 1991 to 2020 average. Conventional cooling systems reject heat into the surrounding environment, potentially contributing to urban heat accumulation. Shallow geothermal systems offer an alternative heat sink with relatively stable subsurface temperatures. Field experiments in Bangkok demonstrated approximately 30% lower electricity consumption for space cooling using a ground source heat pump (GSHP) compared with conventional air conditioning, highlighting its potential for tropical applications. This study investigates the integration of GSHP into district cooling at a selected office complex in South Jakarta, Indonesia, using an analogue based desktop approach. Published district cooling assessments, regional geological and hydrogeological information, and literature derived thermal properties are integrated with reconstructed hourly cooling loads for the selected complex. Transient borehole heat exchanger modelling over 30 years evaluates subsurface temperature evolution, heat rejection capacity, and thermal interference under different borefield configurations and operating schedules. Hybrid GSHP and conventional district cooling systems are compared in terms of cooling efficiency, electricity consumption, and long term thermal performance, supported by sensitivity analyses of subsurface properties and borefield design. The study aims to identify sustainable heat rejection limits, operating conditions, and potential energy benefits, providing a technical basis for integrating shallow geothermal cooling into tropical district cooling systems.

Topic: Low Temperature

Superhot Rock (SHR) geothermal resources, generally targeting temperatures above approximately 374 C, offer the potential for high-energy-density, geographically scalable, firm power, but commercialization depends on creating and sustaining a productive subsurface heat exchanger under extreme temperature, pressure, chemistry, and thermal-cycling conditions. This paper reviews the state of the art in SHR reservoir stimulation and evaluates hydraulic and non-hydraulic approaches by their applicability, maturity, and principal limitations. Hydraulic methods remain the most mature pathway. Hydroshearing can reactivate favorably oriented natural fractures at comparatively low injection pressures, but its predictability and long-term conductivity are limited by geological dependence and fracture closure. Multistage tensile fracturing offers greater control of fracture placement and has been demonstrated in moderate-temperature enhanced geothermal systems, yet extension to SHR conditions is constrained by surface treating pressure, high-temperature cement and zonal isolation, perforating systems, proppant transport, and long-term proppant stability. Cyclic or pulse stimulation, thermal-shock stimulation, and supercritical CO2 stimulation provide complementary mechanisms that may reduce breakdown pressure or increase fracture complexity, although additional field validation is required. Chemical and focused-energy techniques may have specialized roles. A central uncertainty is reservoir behavior near and beyond the brittle-ductile transition, where laboratory studies indicate that stimulation can generate distributed three-dimensional cloud-fracture networks rather than discrete planar fractures. Such networks may provide large heat-transfer surface area at reduced breakdown pressure, but they do not by themselves ensure injector-producer connectivity, controllable flow distribution, or durable permeability. The review therefore concludes that commercial SHR development is unlikely to rely on a single universal stimulation technology. Project-specific hybrid strategies integrating hydraulic, thermal, chemical, and potentially focused-energy mechanisms should be developed together with survivable casing, cement, metallic isolation and flow-control systems; high-temperature diagnostics; and methods to preserve fracture conductivity and manage reservoir flow. Priority research should integrate representative high-pressure, high-temperature laboratory testing, coupled modeling, and field demonstrations to validate long-term heat recovery and well survivability.

Topic: Enhanced Geothermal Systems

Natural fractures are pervasive in unconventional reservoirs and geothermal reservoirs, as evidenced by core analyses, well log interpretations, and microseismic data. When a hydraulic fracture (HF) propagates through such formations, its interaction with natural fractures (NF) can lead to arrest, crossing, or diversion of the HF, depending on local shear strength, fluid pressure, and other conditions at the intersection. In this study, we model HF propagation in the presence of natural fractures using a 2D Displacement Discontinuity Method (DDM) that incorporates height growth for both hydraulic and natural fractures. The resulting fracture trajectories vary with natural fracture orientation and the in-situ stress field. Multiple HF–NF interaction scenarios are examined to evaluate the influence of fracture geometry and stress orientation on propagation behavior. Additionally, proppant transport, settling, and embedment during shut-in are simulated within the discrete fracture network to assess their impact on long-term fracture conductivity. This work provides insights into the coupled mechanics of HF–NF interactions and the role of proppant in sustaining conductivity in naturally fractured reservoirs.

Topic: Enhanced Geothermal Systems

Kenya's electricity sector is undergoing a transformative expansion driven by the country's vast geothermal resources, estimated at over 10,000 MW along the Great Rift Valley. This paper presents a comprehensive analysis of the impacts of stepwise geothermal energy integration—from 950W small-scale binary systems to 5000MW utility-scale installations—on existing tariff structures in the Kenya power system. Using a 50MW stepwise integration framework, we examine the techno-economic implications for Kenya Power and Lighting Company (KPLC), the single-buyer utility, and various consumer categories. The analysis synthesises regulatory frameworks including the Energy Act 2019, the draft National Geothermal Strategy 2026–2036, and Feed-in-Tariff policies, while modelling tariff dynamics under Kenya's unique generation mix where geothermal already contributes approximately 40% of baseload electricity. Our findings indicate that geothermal expansion follows a fundamentally different tariff-impact trajectory compared to variable renewable sources due to its baseload characteristics and near-zero fuel costs. The levelized cost of geothermal generation in Kenya ranges from USD 0.06 to 0.08 per kWh, compared with over USD 0.20 for heavy fuel oil. The stepwise approach reveals that tariff benefits accrue most significantly through displacement of expensive thermal generation, with the Menengai model demonstrating that public-private partnerships can deliver geothermal power at approximately Sh8 per kWh. Threshold effects centre on the transition from public to private financing models and the development of geothermal fields at Menengai, Silali, Paka, and Suswa, with Kenya's entry into the 1-GW geothermal club marking a significant milestone. Policy recommendations address tariff reform, public-private partnership frameworks, and mechanisms to protect consumer affordability while ensuring continued investment in geothermal development.

Topic: Emerging Technology

Lund North is the fourth detailed study site of the INnovative Geothermal Exploration through Novel Investigations Of Undiscovered Systems (INGENIOUS) project, which integrates geologic and geophysical data to develop efficient workflows for discovering commercially viable hidden geothermal systems in the Great Basin region. Lund North is a combined fault bend and step-over zone along the Quaternary Lund normal fault at the southern end of the Wah Wah Mountains in the northern Escalante Valley, Utah. Although this region hosts known geothermal systems (e.g., Thermo), Lund North lacks active thermal features such as surface hot springs and steam vents. However, some historical data and new regional play fairway analysis for the INGENIOUS project showed that the area had relatively high potential for hosting a hidden geothermal system. New gravity, aeromagnetic, paleomagnetic, and magnetotelluric (MT) data, Lidar-based fault mapping, and 2D and 3D geologic and geophysical models defined two structurally complex sub-domains within the resource lead. In the northern sub-domain, numerous small fault scarps and gravity and magnetic gradient lineations suggest that the NNE-striking range-front fault steps to the right and/or changes to a more E-W strike at depth. Shallow conductors in the 3D resistivity model connect to deeper conductors along active faults. A favorable structural setting in the southern sub-domain is inferred from gravity gradients that diverge from mapped fault traces at the surface, indicating a concealed NE-SW basinward fault that intersects the NNE-striking range-front fault. A subtle oblate magnetic low centered on the intersection may reflect hydrothermal alteration. Drilling at Lund North in the fall of 2026 tested these models. Temperature gradient boreholes, drilled to a maximum measured depth of ~250 m (~800 ft) on Bureau of Land Management and private land, were sited using structural and geophysical models together with limited legacy heat flow and new 2-meter temperature data. Shallow gradient measurements in the area indicate background gradients of about 40 to 50 °C/km. We compared borehole lithology, structure, and temperature observations with model predictions, including the location of the concealed southern fault and the nature of shallow conductors, and assessed whether MT conductor connectivity in the northern sub-domain or a magnetic alteration signature at the fault intersection in the southern sub-domain predicted elevated gradients.

Topic: Field Studies

CO2-based geothermal power uses CO2 instead of water as the working fluid. Heating in the reservoir lowers the CO2 density and drives a thermosiphon, so the fluid circulates without pumping and a wellhead pressure difference becomes available for a turbine. The only earlier reservoir test known to the authors (Cranfield, USA), initiated the thermosiphon repeatedly but could not sustain it. This work demonstrates stable, controllable and sustained thermosiphon circulation in a real reservoir. The test used two existing wells around 1,300 m apart in a mostly depleted oil reservoir at about 2,000 m depth with near-hydrostatic pressure. The gas, a remnant of CO2-based enhanced oil recovery, contained about 74 mol% CO2 and 22 mol% CH4. A surface loop comprised a liquid separator, a throttle emulating the turbine, and a heat exchanger cooled by ambient air only. Circulation was initiated by injecting water into the injection well until the wellhead pressure difference reached about 20 bar. The throttle was then varied in two campaigns (2022-2023) of 12 days, and about six weeks. A steady-state cycle model calibrated on the first campaign was used to recalculate second-campaign operating points. In both campaigns the fluid circulated without external pumping, with a stable pressure difference across the throttle of about 10 to 19 bar and mass flow rates up to about 3 kg/s. Flow was regulated by throttling, and the system was shut down in a controlled manner. In the first campaign, surface temperatures were still rising after 12 days and hydrates in the flow-meter impulse lines, were mitigated by insulation. The model reproduced measured mass flow and cooler outlet pressure within the expected measurement uncertainty. A reservoir-based CO2 thermosiphon can thus be sustained for weeks at field scale, supporting a larger scale geothermal power plant in which a turbine replaces the throttle.

Topic: Field Studies

The development of the geothermal reservoirs involves the processes of heat transfer, fluid flow, rock deformation, and chemical reactions. This review explores the thermo‑hydro‑mechanical‑chemical (THMC) modeling as a way to capture those couplings and assess their impact on injectivity, temperature, scaling, deformation, and reservoir life. Field observations at Soultz‑sous‑Forêts highlight the importance of fracture‑controlled flow and mineral deposition, while the lessons learned from Groß Schönebeck demonstrate that how machine‑learning surrogates could make complex THMC simulations practical for uncertainty analysis and operational planning. AI tools, such as reduced‑order models, neural operators, and physics‑informed network, extend the reach of high‑fidelity simulations, enabling rapid forecasts, sensitivity studies, and digital twins. These surrogates should complement, not replace, process‑based models, with conservation checks and validation that are essential for safe and long‑term use. Together, this integrated workflow makes coupled geothermal modeling more feasible to predict the reservoir performance, sustain the reservoir injectivity, manage the scaling, and safeguard the integrity under complicated subsurface uncertainties.

Topic: Modeling

Recent laboratory experiments using fiber-optic sensors show that stresses retained in wellbore cement after setting may be only 20-30% of the hydrostatic cement-slurry pressure. These initial stresses can strongly influence cement tensile failure, interfacial debonding, and the overall mechanical response of the wellbore, particularly under the severe thermal and pressure cycles encountered in enhanced geothermal systems (EGS). However, many computational wellbore-integrity models neglect the cement initial stress state or prescribe it using assumptions that are not constrained by experimental measurements. Consequently, prior studies often emphasize stress changes rather than absolute stress values, even though absolute stresses are required to evaluate failure mechanisms. This paper introduces a staged-construction method for establishing a user-defined, uniform initial stress state in the cement annulus of two-dimensional finite element models for wellbore-integrity analysis. The method overcomes artificial stress redistribution that can occur during model initialization and allows the prescribed stress state to be informed by laboratory measurements, field observations, or systematically varied in parametric studies. The approach is demonstrated for representative EGS wellbores subjected to thermal and pressure cycling. Comparisons with ad hoc initialization methods reported in the literature show that the proposed method establishes the target stress state more accurately and consistently. The framework therefore provides a practical approach for incorporating experimentally constrained absolute initial stresses into wellbore-integrity simulations and for more reliably evaluating plastic yielding, tensile failure, and interfacial debonding under geothermal operating conditions.

Topic: Modeling

[Thorbjarnarson]

The Ring of Fire in a Global Tectonic Framework: Connections with the Atlantic, Indian and Antarctic Ocean Boundaries Steingrimur THORBJARNARSON [Nytjar Ehf, Iceland]

The principal tectonic characteristic of Iceland is its location on the divergent boundary between the North American and Eurasian plates. All seven of the country’s currently operating geothermal power plants are situated within this tectonic context, while the geologically young formations associated with active rifting and volcanism account for many of Iceland’s most distinctive landscapes. Comparison of the proposed Convection Rolls Model with both divergent and convergent plate boundaries has indicated that their spatial relationships can be examined mathematically. The geometry and dynamics of individual subduction zones can be compared with the functional and spatial characteristics of the model in the same way as the divergent volcanic zones of Iceland. Both divergence and convergence contain predominately vertical component, as mantle material rises and contributes to the formation of new lithosphere at spreading centres, whereas lithospheric material descends back into the mantle at subduction zones. A more complete interpretation, however, also requires consideration of the large-scale horizontal organization of tectonic activity. This horizontal component leads to a geometrical definition of the Ring of Fire. Its basic dimensions are derived from the Convection Rolls Model, which provides the geometrical conditions for an inner span corresponding to one-third of the Earth’s circumference along the equator. The same geometry can be repeated around the globe, immediately providing the basis for three corresponding tectonic-ring systems. The resulting Tectonic Rings Model can then be compared with observed geological structures. Such comparisons reveal geometrical relationships involving the Gulf of California, which can be compared with the Red Sea, as well as Yellowstone, Iceland, New Zealand, the Galápagos Islands and numerous other tectonically and volcanically active regions. The significance of these relationships lies not merely in the occurrence of individual active sites along calculated geometrical boundaries, but in the repeated correspondence between major tectonic structures and the dimensions, axes and intersections defined by the three-ring system, along with Antarctica. The Tectonic Rings Model can therefore be regarded as a natural horizontal counterpart to the predominantly vertical, and sectional, geometry represented by the Convection Rolls Model. The two models are closely intertwined, as one describes the organization of mantle circulation and its relationship with divergence and subduction, while the other examines the large-scale horizontal distribution of the resulting tectonic activity at the Earth’s surface. Ultimately, both must be considered within the same physical system, governed by the transfer of internal heat, organized mantle motion, plate interaction and the rotational dynamics of the Earth.

Topic: Geophysics

Mitigating induced seismicity remains a challenge in Enhanced Geothermal Systems (EGS). When seismicity increases during injection, a common mitigation is to shut in the well immediately to stop injection and limit reservoir pressurization. However, field observations show that immediate shut-in does not always prevent delayed earthquakes. At the Basel EGS site in Switzerland, where injection occurred within a fractured fault-damage zone that may exhibit strong poroelastic coupling, the largest earthquake occurred approximately four hours after shut-in, suggesting that poroelastic stress and pressure diffusion may jointly influence post-injection seismicity. We combine an analytical poroelastic solution with a seismicity-rate formulation to identify when immediate shut-in is effective and when a controlled ramp-down may provide a safer strategy. Coulomb stress resolved on a fault is calculated by convolving the constant-rate Coulomb stress response kernel with mitigation injection-rate history. We show that the sign of the kernel minimum, governed by formation properties (Biot coefficient and Poisson’s ratio) and fault properties (fault orientation and slip direction), provides a criterion for distinguishing a pressure-dominated regime from a coupled-poroelastic-pressure regime. In the pressure-dominated regime where the kernel remains positive, immediate shut-in is optimal because it minimizes injected volume and limits pressure buildup. However, in the coupled regime relevant to Basel, an abrupt injection rate reduction interacts with the kernel negative phase and produce a rapid post-injection increase in fault Coulomb stress, potentially elevating seismicity rate after shut-in. A controlled ramp-down injects more fluid but reduces the rapid increase in poroelastic response, therefore slows the early Coulomb stress increase and limits late pressure buildup. As a result, intermediate ramp-down can produce a lower peak seismicity rate comparing with immediate shut-in or a slow ramp-down. These results suggest that mitigation strategy design should consider in-situ formation and fault properties.

Topic: Enhanced Geothermal Systems

[Toth]

Trustworthy Artificial Intelligence in Geothermal Energy: Public Trust, Human Oversight, and Sustainable Well-Being Aniko N. TOTH, Oladoyin KOLAWOLE [Ana-Geo Ltd., Hungary]

Artificial intelligence (AI) offers new potential for geothermal development, including resource exploration, subsurface characterization, drilling optimization, reservoir forecasting, induced-seismicity monitoring, and power-plant operation, to reduce exploration risk and support adaptive reservoir management. However, the success of AI-supported geothermal projects is determined not only by technical performance but also by whether the technology and its operators are trusted by affected communities. The importance of such trust has been illustrated by the enhanced geothermal system (EGS) projects in Basel (Switzerland) and Pohang (South Korea), which were halted, regardless of resource potential, after public confidence was eroded by induced seismicity in 2006 and 2017, respectively. As AI is adopted to inform operational decisions such as injection-rate control and traffic light system (TLS), fears may be raised that consequential decisions will become opaque, that model predictions will be unreliable beyond the training data, that human oversight will be diminished, and that local concerns will be dismissed in favor of algorithmic outputs. In this article, the influence of such distrust on public acceptance of geothermal projects is examined, even where measurable environmental and operational benefits are offered by AI. A framework is proposed in which AI-supported geothermal development is evaluated against four interdependent criteria: (i) the quality and uncertainty of AI-informed decisions, (ii) the transparency and explainability of AI use, (iii) the extent of meaningful community participation, and (iv) the equity with which benefits and risks are distributed. It is argued that fear and distrust cannot be resolved through communication alone and that accountable human oversight is required, together with decision-making processes in which questions can be raised, model assumptions can be challenged, and outcomes can be influenced by affected communities. Through the integration of these human criteria with technical metrics, the social license on which sustainable geothermal energy depends can be strengthened.

Topic: General

Natural fractures can strongly influence hydraulic stimulation in enhanced geothermal systems, yet this influence is difficult to infer from field observations alone. Numerical models that explicitly represent fracture geometry and its coupling to fluid flow and rock deformation are needed to interpret how stimulation propagates through fractured rock. We present a field-constrained, discrete-fracture-network- (DFN) based coupled hydro-mechanical model of hydraulic stimulation at Utah FORGE (Frontier Observatory for Research in Geothermal Energy), built using the XSite modeling framework. XSite is a discrete element code developed specifically for modeling hydraulic fracturing in naturally fractured reservoirs. It uses the synthetic rock mass (SRM) method and a lattice-based approach to represent rock deformation, new fracture propagation, and shear reactivation of pre-existing fractures. We back-analyzed Stages 8-10 of the 2024 stimulation campaign at FORGE and compared the simulated results with geophysical and wellbore monitoring data. Reservoir stimulation was controlled by pre-existing natural fractures and dominated by hydro-shearing, rather than simple tensile fracture propagation. Across the three stimulation stages, the model predicted intersections of created and reactivated fractures with the production well at locations far from those expected from a planar-fracture model, consistent with field observations. These results show how DFN-based hydro-mechanical modeling can translate field signals into a mechanistic interpretation of fracture reactivation and flow-path development during geothermal stimulation, providing a basis for predictive modeling and future reservoir development at FORGE.

Topic: FORGE

Extensive geothermal literature exists for many geothermal prospects. Recent work demonstrates that large language models (LLMs) can extract relevant information from reports and peer-reviewed publications to construct numerical 3D models that are broadly consistent with published cross sections and qualitative descrptions. Building on this approach, we use an LLM-enabled workflow to generate stochastic 3D temperature models and to estimate expected temperatures at potential drilling locations across plausible parameter ranges, thus representing potential analogous greenfield sites. This efficient 3D modeling capturing the variability which are needed to estimate the 3D statistics. These 3D statistics can then be used in the updated Value of Information (VOI) application which identifies the cost-minimizing drilling sequence to confirm or refute the presence of a specified anomaly (e.g., a temperature threshold).

Topic: Emerging Technology

The successful development of Enhanced Geothermal Systems (EGS) hinges on the ability to enhance reservoir permeability and productivity through well-targeted stimulation techniques. This study presents the application of geomechanical modeling and stimulation effectiveness modeling to optimize hydraulic fracturing and zonal isolation strategies for a geothermal well: DAC 68-1RD at the Don A. Campbell (DAC) geothermal field in Nevada. These wells are part of the Department of Energy (DOE)-funded Ormat-GeothermEx Wells of Opportunity Project, which aims to stimulate an existing well to unlock additional geothermal potential, with the potential to generate several megawatts of power. The geomechanical modeling workflow integrates high-resolution Pressure-Temperature-Spinner (PTS) logs, geophysical logs, and laboratory testing of core and cuttings to characterize the mechanical and hydraulic properties of the reservoir. A detailed 1D Mechanical Earth Model (MEM) was constructed to provide critical inputs for Kinetix modeling, which simulates hydraulic fracture propagation, stress field changes, and post-stimulation production. The model accounts for key physical mechanisms, including fracture deformation, fluid flow, proppant transport, natural fracture interactions, and the stress shadow effect, to predict the effectiveness of stimulation treatments. Stimulation effectiveness modeling evaluates the impact of various completion designs and pumping schedules on well productivity improvement. For DAC 68-1RD, the modeling results informed the design of a hydraulic fracturing treatment, including acid spearhead treatments to enhance permeability in low-permeability volcanic rock. Preliminary results demonstrate the importance of tailoring stimulation strategies to the unique geomechanical and geological characteristics of the well. For DAC 68-1RD, stimulation is expected to achieve a productivity of 10 gpm/psi, enabling production of over 2,500 gpm of 280°F fluid, with an estimated generation increase of 4.9 M. This study highlights the critical role of geomechanical and stimulation effectiveness modeling in designing and optimizing EGS stimulation treatments. By integrating advanced modeling and leveraging lessons learned from previous stimulations, this project aims to address key barriers to the widespread adoption of EGS technologies and provide valuable insights for the geothermal community.

Topic: Enhanced Geothermal Systems

The recent rise in geothermal power production has required the use of corrosion resistant alloys (CRAs) for construction. This is especially critical in areas where superhot rock geothermal environments (SHR) exist. In such cases superheated and supercritical geothermal fluids (SHG) are present, which contain high chloride levels in conjunction with high pressures and high temperatures. This work reviews a research program that evaluates different CRAs for use in these aggressive environments. Alloys such as stainless steels, Ni alloys, and titanium alloys were exposed to a high chloride test fluid (100,000 ppm Cl) at both 454ºC (850ºF) and 343ºC (650ºF). The test environments also included CO2, O2, and N2 with total of 2,200 psi (152 MPa) pressures to replicate some of the more aggressive environments encountered in the geothermal field today. The higher temperature (454ºC) showed that the Ti Gr 12 had the lowest mass loss corrosion rates and the C276 had the highest corrosion rate. Pitting was observed on the Alloy 718 material only. The lower temperature (343ºC) test showed much higher corrosion rates in the liquid phase than at 454ºC. It was also observed that samples exposed to the liquid phase at 343°C had higher corrosion rates than in the gas phase. The Ti Gr 12 and the Ti 6246 once again showed the lowest corrosion rates. The highest corrosion rate at 343°C was on the 2550 duplex stainless steel. Pitting and crevice attack were observed on the 2550 duplex stainless steel, the Alloy 718, and slightly on the Ti 6246 material. Along with the corrosion evaluation on CRA materials in SHR environments was a need to determine what sort of corrosion rates are observed in-situ. Corrosion monitoring probes have been employed in production environments for many years. The use of coupled multielectrode array sensors (CMAS) have been successfully used in numerous environments including geothermal environments. A review of a field experience using CMAS probe in the Salton Sea geothermal environment will be presented.

Topic: Production Engineering

The temperature profile of a geothermal well involves multifaceted interactions among geological, fluid-flow, wellbore, and measurement effects, and heat-transfer regimes are still read and interpreted manually. Change-point detection methods can segment temperature-depth profiles automatically, but assigning a heat-transfer regime to each interval still relies on expert interpretation and, to the best of our knowledge, has not been automated or ed. To this end, this study puts forth an initial for this task. The study uses digitized temperature and pressure logs from published figures for 28 Olkaria wells, including 346 curves. A written labeling guide clearly illustrating the formula and rationale is used to assign each depth interval one of six separate classes: conductive, upflow, outflow, recharge, isothermal, and boiling, by employing factors such as gradient, curvature, recovery between runs, and the offset from the boiling-point-for-depth curve. Using leave-one-well-out validation employing 2,549 depth-interval samples, a gradient-boosted model reaches macro-F1 0.67, compared with 0.53 for the rules applied to each depth interval. Because the labels are rule-generated, these scores measure agreement with the rules. Against an independent published interpretation of eight of the wells (674 depth intervals), the gradient-boosted model and the rule labels each agree on 73% of depth, and the per-depth rules on 52%. After completing the next phase, the full paper will report hand-reviewed labels for each well segment, synthetic data generated from physics-guided heat-transfer models, a physics-guided sequence model pretrained on those profiles before eventual training on the actual well data using a leave-one-well-out validation method, code, and datasets.

Topic: Reservoir Engineering

Ground-source heat pumps are a key technology for decarbonizing the heating and cooling sector. Sizing their borehole heat exchangers requires the thermal properties of the subsurface, which are typically determined by Thermal Response Tests (TRTs). Most established evaluation methods, from the classical line-source approach to the Excitation-Modulated TRT (EM-TRT), assume purely conductive heat transport. At many sites, however, groundwater flow through permeable layers additionally transfers heat by advection, making the Darcy velocity an additional unknown. In synthetic test cases with an aquifer spanning one third of the borehole depth, a conduction-only evaluation can overestimate the effective thermal conductivity by more than a factor of two, while its fit to the temperature data is visually indistinguishable from that of the correct model. Estimating the Darcy velocity jointly with the thermal conductivity largely removes this bias, but requires a borehole model that resolves groundwater flow and is fast enough for the dozens to hundreds of forward simulations involved in a single estimation. We therefore extend our reduced-order EM-TRT model, derived from a finite-element model of the borehole cross-section by snapshot Proper Orthogonal Decomposition (POD). Groundwater flowing around the borehole adds a single term to the governing equations that scales linearly with the Darcy velocity, just as the conductive term scales with the thermal conductivity. The Darcy velocity thus enters as an ordinary model parameter, and the reduction carries over unchanged. Instead of about 10,000 degrees of freedom, thirty suffice to cover the full range from stagnant to fast-flowing groundwater, about ten more than for pure conduction. The reduced cross-section model runs about 400 times faster than its finite-element counterpart, while reproducing its temperatures to within a few hundredths of a kelvin. Building this model revealed a pitfall relevant to any reduced model used for parameter estimation: being accurate is not the same as being informative. Trained in the usual way, the model captures less than half of the subtle temperature signature of weak groundwater flow, precisely the signal from which the Darcy velocity must be inferred. Selecting training snapshots according to this signature in addition to the overall temperature error recovers it almost completely, without loss of accuracy. The reduced model will be compared with an independent three-dimensional FEFLOW model provided by GeoSphere Austria, as a step towards TRT evaluations that quantify groundwater flow instead of being misled by it.

Topic: Modeling

Projecting the multidecade thermal performance of enhanced geothermal systems (EGS) is essential for reservoir management and anticipating premature breakthrough. At Utah FORGE, circulation tests between injector 16A(78)-32 and producer 16B(78)-32 span days to weeks, leaving decadal heat sweep unobserved. This study develops machine learning surrogates of a calibrated physics model to project thermal drawdown and identify a reduced parameter set sufficient for screening. We calibrate a multifracture thermohydraulic doublet model to the August to September 2024 circulation test (GDR 1683), yielding an in-sample RMSE of 1.48 °C. Latin Hypercube Sampling across ten site specific parameters generates 500 thirty-year scenarios, split by scenario into training (350), validation (75), and test (75) sets. XGBoost models emulate the simulator, predicting time to 10% drawdown, time to a 150 °C plant cutoff, and monthly production temperature. On held-out scenarios, the surrogate predicts time to 10% drawdown with an R² of 0.971 and a mean absolute error of 0.61 years, against a 14.2-year median breakthrough time, outperforming the analytical Gringarten solution (1.48 years) and polynomial ridge regression (3.12 years). Split conformal prediction provides 90% marginal coverage within the sampled domain. Sobol indices and TreeSHAP attributions agree closely (Spearman correlation 0.976) and identify effective fracture area, injection rate, interwell spacing, and initial reservoir temperature as the main controls on decline; a four-parameter model retains 97.7% of full model R². These figures reflect simulator fidelity, not thirty-year field validation; projections remain subject to unmodeled thermo-hydro-mechanical processes and calibration equifinality.

Topic: Reservoir Engineering

Carbon dioxide emissions from geothermal power plants can be considerable, especially for fields with high carbon dioxide content such as those in Western Anatolia, Turkey. Re-injecting the produced carbon dioxide back into the reservoir is one way of reducing these emissions. Carbon dioxide can be re-injected either dissolved in the re-injection water or as a free phase. An important question that arises is where the carbon dioxide should be re-injected: to the top or to the bottom of the reservoir. The answer is not straightforward. Pressure is higher at the bottom of the reservoir, so more carbon dioxide can be kept dissolved. On the other hand, carbon dioxide re-injected near the production zone may reach the production wells more quickly. When carbon dioxide is re-injected as a free phase, it tends to move upwards. When it is dissolved in cold re-injection water, the re-injected water tends to move downwards. In this study a two layer lumped parameter model is developed to answer this question. The model keeps an inventory of carbon dioxide in each layer and accounts for the communication between the layers. The flashing point in each layer is monitored to determine when and where a gas phase forms. Both dissolved and free phase re-injection are considered. For the free phase case, the carbon dioxide is allowed to dissolve in the layers it passes through, and the excess forms a gas phase at the top of the reservoir. Various synthetic cases are presented for re-injection at the top and at the bottom of the reservoir. The effects of the conditions of the reservoir, the vertical communication between the layers and the re-injection temperature on the amount of carbon dioxide kept in the reservoir, the carbon dioxide content of the produced fluid and the reservoir pressure are demonstrated.

Topic: Reservoir Engineering

Enhanced Geothermal Systems (EGS) can extend geothermal development beyond naturally permeable reservoirs, but identifying the lowest-risk settings for early commercial deployment remains a key challenge. Mature high-enthalpy provinces may offer such settings, since temperature, lithology, drilling conditions, fluid chemistry, and infrastructure are already partly constrained. We propose a comparative EGS-readiness framework applied to two contrasting provinces. In Western Türkiye, fault-controlled systems in the Büyük Menderes and Gediz grabens (e.g., Kızıldere, Germencik, Alaşehir systems) produce from fractured metamorphic and carbonate reservoirs with CO₂-rich fluids. In the western United States, the Great Basin hosts high-temperature resources with limited natural permeability; in field EGS projects adjacent to existing hydrothermal operations and the Utah FORGE crystalline-basement site provide field-scale stimulation and circulation experience. Rather than comparing heat in place estimates, the framework evaluates transition readiness using indicators grouped into six domains: thermal, geological, geomechanical, hydrogeochemicaL and regulatory–social constraints.These indicators are organized into an EGS Readiness Matrix that scores each site on two axes: favorability and knowledge maturity in existing wells. Domain scores are weighted and aggregated into readiness tiers, with sensitivity analysis on the weights to test the robustness of site rankings. A further objective is to identify critical data, experimental, and modeling gaps that currently limit EGS screening and subsurface characterization in both provinces, with particular emphasis on coupled thermo hydro mechanical chemical (THMC) processes under high-temperature, high-pressure (HTHP) reservoir conditions. The assessment will highlight where additional well, log, core, cuttings, production, injection, geochemical, laboratory, or field data could reduce key uncertainties in fracture behavior, permeability evolution, fluid–rock interactions, stimulation response, and long-term reservoir performance. It will also identify where targeted laboratory testing and calibrated THMC-HTHP modeling could provide the greatest value. These gaps will then be linked to the readiness framework to illustrate how filling them could improve confidence in EGS readiness scores and better prioritize future characterization, testing, and field-scale demonstration activities. The analysis distinguishes site specific conditions from those that are transferable across geothermal provinces and identifies prospective transition zones adjacent to, beneath, or beyond currently exploited reservoirs. The resulting screening framework aims to define lower-risk pathways from conventional to EGS development, and to show how U.S. stimulation and reservoir-engineering experience and Türkiye's operational experience with high enthalpy, gas-rich systems can inform each other.

Topic: Enhanced Geothermal Systems

Produced-fluid chemistry records fluid mixing, thermal re-equilibration, mineral reactions, and water-rock interaction during EGS stimulation and circulation, yet it is commonly interpreted retrospectively rather than used as a dynamic monitoring signal. At Utah FORGE, interpretation of produced waters collected during the 2022–2024 stimulation and circulation tests has linked temporal changes in major-ion chemistry, silica, sulfate, pH, and dissolved gases to processes including salt dissolution, mineral equilibration, carbonate deposition, and deep-gas influx. However, different reservoir processes may produce overlapping chemical signatures, making their individual contributions difficult to distinguish from geochemistry alone. This study investigates which evolving EGS reservoir processes can be systematically identified from produced-fluid chemistry, and which remain non-identifiable. A physics-guided framework will couple PHREEQC forward modeling with explainable machine learning. Thermodynamically consistent synthetic fluids will be generated across variations in temperature, fluid mixing, water-rock reaction progress, mineral dissolution and precipitation, and CO2 interaction, constrained by measured Utah FORGE injectate chemistry, granitoid mineralogy, and published regional geothermal-fluid compositions. Model inputs will be limited to chemical variables available in the public FORGE dataset and augmented with physically derived features such as concentration changes, selected geochemical ratios, and mineral saturation indices. Explainable gradient-boosted models will estimate the relative expression of individual reservoir processes, while sensitivity to analytical uncertainty will be used to identify process combinations that cannot be uniquely resolved. The framework will then be applied, without retraining, to the public 2022–2024 FORGE produced-fluid record, and inferred process-state transitions will be compared with published geochemical interpretations and independent hydraulic, tracer, and temperature observations. The objective is not to infer fracture geometry from chemistry, but to determine the information content and limits of produced-fluid chemistry as a dynamic sensing layer for EGS reservoirs. By combining thermodynamic constraints, explainable AI, and explicit identifiability analysis, the study aims to establish a low-cost complementary approach for tracking changes in the processes controlling EGS circulation.

Topic: Enhanced Geothermal Systems

The magnetotelluric (MT) method is widely recognized as one of the key exploration tools for investigating geothermal reservoirs. With the development over the past decade of three-dimensional (3D) inversion codes for MT data that employ the finite-element method (FEM) for forward modeling, increasingly accurate resistivity models can now be obtained. Particularly, FEM inversion codes can incorporate topography into 3D meshes without losing numerical accuracy significantly. However, applying 3D FEM inversion codes to field MT data is not straightforward because various parameters, including mesh settings, must be appropriately defined. The resulting resistivity models may depend on these parameter settings. In this study, we investigated the effects of different mesh settings, including tetrahedral versus hexahedral elements and different element sizes, on 3D inversion results for MT data acquired in the Ogiri geothermal area. The dataset consisted of MT measurements at 244 stations. We used all components of the impedance tensor and tipper data at 16 frequencies ranging from 0.00293 to 96 Hz. The 3D FEM inversion code FEMTIC was used for the analysis. FEMTIC supports both tetrahedral (Tetra) and distorted non-conforming hexahedral (DHexa) elements for mesh construction. We prepared two meshes with different element sizes for each element type. For the Tetra meshes, the smallest elements around the MT sites were less than 25 m and 40 m, respectively. For the DHexa meshes, the smallest horizontal element sizes around the MT sites were 31.25 m and 62.5 m, respectively, while the vertical element size near the ground surface was 25 m. All other inversion settings, including the input data, error floors, and treatment of galvanic distortion, were identical. Comparison of the final resistivity models revealed significant differences between the Tetra and DHexa meshes in features important for geothermal interpretation, particularly the geometry of the shallow low-resistivity cap rock and resistivity anomalies within the deeper high-resistivity reservoir zones. Differences in element size within both the Tetra and DHexa meshes also produced meaningful variations in these resistivity structures. Although the causes of these differences could not be conclusively identified, our results demonstrate that the choice of mesh type and element size can significantly affect the resulting resistivity models. Therefore, the sensitivity of inversion results to mesh setting should be carefully evaluated when interpreting the resistivity structure of geothermal reservoirs.

Topic: Geophysics

The clay cap that forms over most geothermal systems in volcanic terranes creates a partial seal that constrains the upper extent of many hydrothermal system and separates the geothermal system from cooler overlying and surrounding groundwater. From imaging many geothermal systems using magneto telluric (MT) resistivity mapping, we can see this separation of hot and cold systems and also where groundwater are likely perched at elevations much higher than expected. This clay cap appears to have a major impact on local groundwater hydrology and the movement of groundwaters flow around geothermal systems. Correlating these MT patterns with the location of water seeps and known groundwater sources over known geothermal systems indicates that MT imaging probably has considerable potential to predict the location of groundwaters and a different perspective to normal hydrological considerations. We see probable fresh water aquifers perched at high elevations in several volcanic islands and in the Andes of Chile and Peru, but also in basinal settings. These may be useful sources of water for communities and for geothermal drilling, and in any case are important for understanding the interaction between geothermal and shallower groundwater, and the impact that any geothermal production may have on groundwaters. This paper presents several examples with the intent of stimulating further investigation of this concept and the opportunity it offers for development and better resource management.

Topic: Geophysics

The influence of temperature, which causes various types of hydrothermal alteration of rocks and in turn influences resistivity patterns seen in geothermal systems is now well recognised and applied in the interpretation of deep earth resistivity imaging methods such as magnetotellurics (MT). However, the direct impact of kinetic temperature properties of rocks is less well understood or emphasised and limits the opportunity for better interpretations of resistivity models in terms of concept models and certainty of the presence of current geothermal activity. The ionic conduction of water solutions and typical hydrothermal clays means that conductive activity increases with temperature and hence observed resistivity can be directly indicating temperature. In the absence of other factors, high temperature zones in rock will have lower resistivity than surrounding rocks, but in real conditions these effects will combine with associated alteration caused by geothermal conditions and variation of rock types that may be controlling fluid movement and hence the presence of the thermal anomaly. This will not only be important for improved interpretation of conventional convective systems but also particularly applicable for understanding lower temperature systems and EGS fairway mapping. In this paper, I summarise the theoretical basis for these effects and present a simulation of the impact temperature can have on resistivity and relate this to actual case examples.

Topic: Geophysics

CO2 storage security is a critical requirement for carbon capture and storage (CCS) to ensure that injected CO2 remains contained and does not migrate to the surface. Low-temperature geothermal systems in sedimentary basins, such as the Paris Basin in France, represent viable options for CCS sites. The Paris Basin has supported decades of hot brine production and reinjection for district heating, offering a highly viable setting for continuous carbonated water reinjection, where hot brine is enriched with dissolved CO2 before reinjection. However, CCS in these sedimentary aquifers often faces higher uncertainty in sealing capacity because historical field data can be limited, requiring robust seal characterization and injection pressure management. CO2 is commonly injected into sedimentary aquifers as a dense or supercritical phase, where structural and stratigraphic trapping dominate the early post-injection period, necessitating strict monitoring of mobile-phase plumes. Conversely, carbonated water injection minimizes the risk of free CO2 leakage through subsurface faults. However, quantitative assessments comparing conventional supercritical and carbonated water injection remain limited. Therefore, this comprehensive study aims to evaluate and compare supercritical CO₂ injection against carbonated water injection using a representative Dogger reservoir model in the Paris Basin. A two-dimensional axisymmetric radial model was simulated using the MATLAB Reservoir Simulation Toolbox (MRST), employing a non-isothermal multiphase formulation with aqueous CO2 dissolution. Both scenarios targeted 150 ktCO₂/yr for five years, followed by monitoring for up to 20 years, focusing on retained mass, phase distribution, pressure buildup, and mass-balance consistency. Both scenarios retained most of the injected CO2. At the end of the simulation, the supercritical case retained 99.45% of injected CO2, compared with 95.25% for the carbonated water case. However, 442.8 kt, or 59.4% of the retained CO2 in the supercritical case, remained as free-phase CO2 after 20 years. In contrast, no free CO2 phase developed in the carbonated water simulation, and all retained CO2 remained in dissolved form. This safety benefit required a much larger injected fluid volume and a higher pressure response: the carbonated water case required approximately 811.2 m3/hr of injected fluid and produced a peak pressure buildup of 39.32 bar, compared with approximately 26.7 m3/hr and 2.41 bar for the supercritical CO2 case. These findings indicate a practical trade-off between injected fluid volume, pressure demand, and free-phase CO2 reduction.

Topic: Modeling

[Vasantharajan]

An Update on Mazama 2025 Activities Sriram VASANTHARAJAN and Alain BONNEVILLE [Stanford University, USA]

An Update on Mazama 2025 Activities

Topic: Introduction

The Utah Frontier Observatory for Research in Geothermal Energy (FORGE) conducted a circulation test in 2026 between wells 16A(78)-32 and 16B(78)-32. Field data collected during the test include injection pressure, mass injection and production rates, and production temperature. In this study, a thermo-hydraulic simulation of the Utah FORGE discrete fracture network (DFN) was calibrated to these data and used to estimate recovery efficiency, pressure and temperature trends, and the time to thermal breakthrough. The DFN is modeled as a set of 2D fractures embedded in a 3D granite matrix, with injection and production applied at the fracture–well intersections. Fracture permeability follows a cubic-aperture relationship in which aperture varies with pore pressure, linking fracture porosity, permeability, and pressure. Fracture permeability, the aperture–pressure relationship, and matrix permeability were calibrated to the measured pressure and flow response. To test how strongly the inferred properties and predicted thermal response depend on how injection is distributed among the 16A perforation zones, the calibration and projections were repeated for two flow distributions, one from the August 2024 Utah FORGE circulation test and one from Texas Tech (August 2025). Using the calibrated parameters, simulations were extended to 180 days under the planned injection schedule to estimate recovery efficiency, production pressure and temperature evolution, and the onset and rate of thermal drawdown.

Topic: FORGE