Uncertainty and Sensitivity Analysis for Geothermal Reservoir Performance and Techno-Economic Assessments: A Software Package for GEOPHIRES

Abstract

Stochastic simulation of geothermal reservoir models and economic assessments allows for probabilistic interpretations of reservoir favorability that may be used to inform geothermal project decisions. In 2018, the GEOPHIRES software was published as an open-source Python library for geothermal reservoir modeling and techno-economic calculations. This paper introduces an uncertainty and sensitivity analysis package for GEOPHIRES. Uncertainty in GEOPHIRES inputs may be specified for geological, thermal, utilization, and/or economic parameters. A Monte Carlo analysis is used to propagate those uncertainties through a specified geothermal reservoir model and economic calculations. Graphics of the reservoir performance over time are provided for the Monte Carlo replicates. Estimated reservoir performance and economic values are provided as summary statistics. Translation of the output variable uncertainties into decision-relevant information is provided by the estimated probability of achieving a specified temperature or heat production over time. We demonstrate the use of this uncertainty analysis package with a case study that evaluates the feasibility of geothermal deep direct-use technology to meet the heating demands of the Cornell University main campus in New York State. A hypothetical target reservoir is known as the Trenton-Black River play, at approximately 2.3 km depth below Cornell. We evaluate several utilization scenarios consisting of different reinjection temperatures and flow rates. We use simple analytical reservoir models available in GEOPHIRES to illustrate the functionality of the package, although use of the package is not limited to the built-in reservoir models.