Probing dominant flow paths in enhanced geothermal systems with a genetic algorithm inversion model

Chunwei Zhou,Gang Liu,Shengming Liao
DOI: https://doi.org/10.1016/j.apenergy.2024.122841
IF: 11.2
2024-02-16
Applied Energy
Abstract:Dominant flow path has been proven to be the main reason causing the short-circuit effect which rapidly declines production temperature in the short term and worsens electricity production. Therefore, a key solution to avoid failure is to clearly explore fractured reservoir information. Here, we proposed a genetic algorithm inversion model combined with a path planning algorithm (cubic Bezier curves) to probe dominant flow paths in fractured reservoirs. The Bezier curve, generate complex curves by a few control points, is embedded into the thermal-hydrologic (TH) model as dominant flow paths. The TH model is embedded into the genetic algorithm as a fitness function to obtain the temperature differentials under different dominant flow paths. The genetic algorithm optimizes and iterates the control points by fitness value. It was discussed the feasibility of genetic algorithm inversion model in various dominant flow fractures (shape, uneven distribution, width, discontinuity distribution, and multiple branching distribution). Results show that the genetic algorithm inversion model has high inversion precision ( β≥82% ) when the single dominant flow path is smooth distribution. The width, unevenness, and multiple branching of dominant flow fractures do not affect the inversion results in dominant flow path. In addition, the genetic algorithm inversion model still obtains a part of dominant flow path ( β≥38% ) when the dominant flow fracture is a discontinuity distribution.
energy & fuels,engineering, chemical
What problem does this paper attempt to address?