Impact of structural uncertainty on tracer test design in faulted geothermal reservoirs

Evaluation of underground processes requires numerical modeling based on sophisticated and reliable meshing. Our new GeoMeshPy library focuses on the discretization of probabilistic geological structures. This study pre-sents a synthetic show-case for the capacity of this library to quantify the impact of structural uncertainty. In here, 50 models were developed taking advantage of the computational efficiency of GeoMeshPy. Assuming a geothermal doublet system embedded in a faulted reservoir with unclear structure, recovery time and magnitude of a tracer breakthrough was calculated. Even small angular variations up to +/- 15 degrees in one of the faults yield differences of up to 26 and 30 percent for peak arrival time and magnitude, respectively. An additional inversion scheme of each of the 50 curves allows quantifying the impact on Pe acute accent clet number varying from 3.4 to 4.3 due to structural variability. Analytically calculated dispersion coefficients are almost one order of magnitude higher than values used for simulations. Besides this mismatch, calculated dispersion coefficients are unable to represent the structural uncertainty (ranging from 125.6 to 129.4 m).

Automated summary

Evaluation of underground processes requires sophisticated and reliable numerical modeling techniques. The new GeoMeshPy library focuses on discretizing probabilistic geological structures. This study showcases the library's ability to quantify the impact of structural uncertainty through the development of 50 models. These models calculate the recovery time and magnitude of tracer breakthrough in a faulted reservoir with unclear structure, revealing significant differences due to small angular variations in the faults.