Title:

A Thermal-Hydrological-Mechanical Model of Patua Geothermal Field, Nevada

Authors:

Torquil SMITH, Eric SONNENTHAL, Nori NAKATA, Trenton CLADOUHOS, Michael SWYER

Key Words:

Patua geothermal field, hydrological simulation, thermal-hydrological-mechanical, native state model

Conference:

Stanford Geothermal Workshop

Year:

2023

Session:

Modeling

Language:

English

Paper Number:

Smith

File Size:

959 KB

View File:

Abstract:

A Thermal-Hydrological-Mechanical (THM) model is being developed for the Patua geothermal field in Nevada. The purpose of the model is to help integrate observed surface deformation (LIDAR, GPS), MEQs, subsurface stress, and regional strain rate data to evaluate deformation of the field for potential stimulation. A new permeability model was constructed to approximate native-state temperatures, first by reevaluating existing geologic and fault data, and then by calibrating TH properties. A set of 71 fault picks from well logs (Pollack, 2021) combined with a set of drilling lost circulation zone (LCZ) observations (Cladouhos et al., 2017) produced 76 prospective fault-well intersection points. This was analyzed using a gridded search for the orientation of planes with the maximum number of fault picks and LCZ depths in a plane (within a 25 m tolerance). After fitting five of the six points closest to Patua Hot Springs with a plane (within 25 m tolerance), a plane striking N 108.8 E dipping 68.9 was used to divide the remaining picked points into a northeastern set and the remainder for separate analysis. In subsequent stages, at each stage the plane with maximum weighted number of picks in a 50 m thick zone was selected, and the remaining data on either side was analyzed separately. In the maximizations, the number of points fit was weighted by 1/[tan(dip) +4.2/tan(dip)] + 1/ 4.2npts) to lessen sampling bias due to roughly common well orientation (near vertical) and to limited well depth compared to lateral sampling extent. This yielded a set of 12 planes fitting 3 to 10 points each, fitting 69 of the 76 points total. The faults were assumed to be centered on the mean of the fitted intersection points in strike and dip directions, and fault strike length and dip breadth was taken as sqrt(12) times the square root of the intersection point variances in strike and dip directions respectively, somewhat arbitrarily for faults sampled by well intersection, but rigorous for rectangular fault patches sampled randomly (uniformly). Initially, unfaulted reservoir rock was given 10^-15 m^2 (E-W, vert) and 10^-16 m^2 (N-S) permeabilities, with 10^-15 m^2 horizontal permeability at elevations above 100 m depth in the central part of the study area (i.e., above 1150 m elev.), with fault permeabilities from 20 x 10^-15 to 125 x10^-15 m^2, in a simple 5 fault model based on an extension (M. Swyer, 2017, Cyrq Energy internal document) of the Cladouhos et al. (2017) conceptual model. Faults were modelled as connected sequences of regular grid blocks (100 x 100 x 100 m at grid center). Initial permeabilities were selected based on the Garg et al. (2017) and Murphy et al. (2017) reservoir models for Patua. A 251.5°C temperature at -3550 m elevation (~4800 m depth) was assumed based on extrapolation of one of the deeper wells. Following earlier studies, an inhomogenous zone of increased temperature was presumed at the base, here with 10°C amplitude maximum increase above background, smoothly decreasing to no increase 3 km laterally from the inhomogeneity center. Permeabilities were modified to a 2 x 10^-15 to 10^-14 m^2 range for the faults and 1.2 x10^-16 to 1.9 x10^-16 (E-W, vert.), 3.8x10^-16 (N-S) background, with 10^-15 m^2 horizontal permeability above 1150 m elevation, through many intermediate values, in an attempt to fit interpolated native-state (e.g., soon after drilling) well temperature profiles at -1200 m elevation (~2450 m depth at the center of the study area), and +1150 m elevation, as well as the general trends of the set of well temperature profiles. Indifferent agreement led to reconsidering the underlying fault model, re-analysis of the fault pick and LCZ data, and subsequent adoption of the 12 fault plane model above, initially with the same background permeabilities as attained in the simpler 5 fault model, and fault permeabilities of 0.5-2 x 10^-15 m^2. Calibrated fault permeabilities range from 2-60 x 10^-15 m^2, with surficial layers having a horizontal permeability reduced to 7.5 x 10^-16 m^2. To facilitate convective upwelling with a smaller artificial bottom boundary temperature perturbation (10°C maximum increase over background), seven of the modelled faults were extended downwards to -2925 m elevation: those not intersecting others on extension. Simulation temperatures at reservoir depth (1005 m below sea level) match observations within about 10°C, at shallow depth (1107 m elev.) to about 20°C.


ec2-18-234-55-154.compute-1.amazonaws.com, you have accessed 0 records today.

Press the Back button in your browser, or search again.

Copyright 2023, Stanford Geothermal Program: Readers who download papers from this site should honor the copyright of the original authors and may not copy or distribute the work further without the permission of the original publisher.


Attend the nwxt Stanford Geothermal Workshop, click here for details.

Accessed by: ec2-18-234-55-154.compute-1.amazonaws.com (18.234.55.154)
Accessed: Thursday 28th of March 2024 02:05:54 PM