MODELING AND ANALYSIS OF THERMO-POROMECHANICAL EFFECTS ON RESERVOIR STIMULATION, DISTRIBUTED STRAIN SENSING, AND REACTIVE TRANSPORT IN EGS

Loading...
Thumbnail Image

Date

Authors

Ratnayake, R Mudiyanselage Dhyan Ruwantha Bandara

Journal Title

Journal ISSN

Volume Title

Publisher

University of Oklahoma – Graduate College

Abstract

Enhanced Geothermal Systems (EGS) offer a sustainable pathway to commercial scale geothermal energy extraction by creating reservoirs in hot, low-permeability rocks. However, accurately predicting long term performance requires understanding the strong coupling between thermal, hydraulic, mechanical, and chemical processes within fractured rock systems. This work develops a comprehensive numerical model to examine thermo-poromechanical behavior and reactive transport effects on fractures during long term fluid circulation. A hybrid Displacement Discontinuity Method–Finite Element Method (DDM-FEM) framework is formulated to capture fracture deformation, fluid and heat transport, solute migration, and fracture propagation. The model also includes forward simulation of distributed fiber-optic strain responses to support diagnostic interpretation of EGS stimulation and circulation behavior.The numerical model is first applied to analyze the response of a stationary penny shaped fracture under stress, hydraulic, and thermal loading. Stress loading increases aperture, pore pressure diffusion induces poroelastic back stress resulting in fracture closure, and fracture cooling causes delayed aperture growth. Stress and hydraulic effects dominate early time scales (hours/days), while thermal stresses control long term behavior (weeks/months). Fiber-optic Distributed Strain Sensing (DSS) diagnostic simulations were performed to interpret field-recorded strain and strain rate signatures when shear deformation is present. The signatures differ from the classical “heart-shaped’’ response associated with Mode I fracture opening. By studying shear deformation of natural fractures or faults using five fracture–fiber orientation scenarios, a strain catalog was developed. In addition, the fiber signature of a stimulated fracture network was elucidated. Furthermore, it has been shown that the influence of pore pressure and poroelastic stresses significantly alter DSS signatures in low permeability reservoirs with impact on common practice of fracture height growth estimates using fiber signature. The research also considered the temperature dependent fluid properties in circulation operations, in contrast to fracture simulations which assume constant values. The effects of temperature dependent density, viscosity and gravity on fluid temperature distribution and fracture propagation was studied. The results show that incorporating gravity significantly affects the temperature distribution, causing enhanced cooling in the downward direction. Additionally combined effects of thermal stresses and gravity contributed the fracture to grow primarily in downward direction. Effect of water viscosity was more significant due to its higher sensitivity to temperature. Reduced viscosity at high temperatures increased leak off into the surrounding rock matrix, modified the pressure distribution, and ultimately limited fracture growth. The results indicate that in EGS, the reduction in viscosity can reduce the fracture size while possibly enhancing secondary fracture reactivation. Reactive transport has been shown to significantly modify fracture apertures and consequently, overall permeability in EGS reservoirs. In this study, reactive transport simulations were carried out using two minerals, calcite and silica. The results show that undersaturated cold water injection causes silica dissolution and aperture increase, while supersaturation leads to precipitation and aperture decrease. Calcite dissolved only near the injection well before cooling slowed reactions, while supersaturated cold water injection caused rapid precipitation ahead of the thermal front. Overall, because both reaction rates and equilibrium concentrations depend strongly on temperature, the occurrence of dissolution or precipitation depends on whether the solute or thermal front arrives first at a given location. After the 2022 stimulation of Stages 1-3 at Utah FORGE, six additional stimulation stages were planned for 2024. The idea was to stimulate injection well 16A(78)-32, monitor the stimulation using DSS fiber installed in well 16B(78)-32 and based on DSS results perforate well 16B. In this study forward modeling was conducted for Stages 4,5 and 7 to predict the resulting hydraulic fracture geometries and the distributed strain and strain rate patterns they would generate. The simulations showed that under the planned pumping schedule, Stage 4 fracture does not intersect with 16B whereas Stage 5 and two of four clusters in Stage 7 are expected to intersect. Corresponding DSS strain/strain rate signatures given by these interactions were also successfully simulated. Additionally, the results highlight how low fracture/cluster spacing creates strong stress shadow effects significantly influencing the fracture propagation paths. Field fiber strain rate results obtained during stimulation were also interpreted, and it was observed that the field results differed from the simulated ones. Several fracture driven interactions were identified including a response in the same region of the fiber during the stimulation of Stages 3R–5, indicating that the same fracture system was accepting fluid across all three stages. An unusual strain rate pattern was observed in Stage 8, and an attempt was made to reproduce this signature our fracture simulator. While signature pattern was successfully created, height of the simulated signature was considerably shorter than the field results. Another Utah FORGE application involved simulating the 30-day fluid circulation test carried out at the site. A fracture model was built using a filtered DFN derived from field data from the Utah FORGE team. The simulated results closely matched field observations, specifically, a 90% recovery rate and stable production temperatures around 440K (330 degrees F). Minor differences in injection pressure and temperature profiles were observed and attributed to model simplifications. An extended 100-day circulation simulation was performed using the calibrated model and the results showed a gradual temperature decline of about 4-5K, with deeper zones contributing to higher temperature flow throughout the period. Overall, the study has advanced prediction and diagnostic interpretation of coupled thermo-chemo-poromechanical behavior in EGS, supporting improved reservoir design and long-term performance assessment.

Description

Citation

Related file

Notes

Endorsement

Review

Supplemented By

Referenced By

DOI

Collection Detail

# of Isolates from RBM

# of Isolates from TV8