Tracer-constrained stochastic reservoir characterization of limited heat exchange volume in the Gonghe geothermal reservoir, China

A conservative tracer test was conducted at the Gonghe Enhanced Geothermal System (EGS) demonstration site to evaluate inter-well connectivity in a hydraulically stimulated granitic reservoir. Breakthrough curve analysis reveals two preferential transport pathways between the injection well GH02 and the production well GH01, with strongly contrasting parameter identifiability caused by pronounced flow channelization and incomplete tracer recovery. To address the high-dimensional and non-unique nature of tracer-based reservoir characterization under field conditions, a two-stage stochastic inversion framework combining large-scale Monte Carlo screening and block-updated Metropolis–Hastings refinement was applied. The inversion results consistently indicate the presence of a localized preferential flow corridor connecting the two wells, while most of the surrounding reservoir contributes little to the tracer transport observed during the test. Heat transfer simulations constrained by the inferred porosity structure indicate that fluid circulation and heat extraction are dominated by localized preferential flow zones, while large portions of the surrounding reservoir regions with lower effective porosity remain weakly swept by the circulating fluid. As a result, only a limited fraction of the available thermal energy can be effectively extracted, leading to inefficient resource utilization and accelerated thermal decline of the actively connected flow pathways.