| Title | Characterizing Fracture Flow in EGS Collab Experiment Based on Stochastic Modeling of Tracer Recovery |
|---|---|
| Authors | Hui WU, Pengcheng FU, Joseph P. MORRIS, Earl D. MATTSON, Adam J. HAWKINS, Yuran ZHANG, Randolph R. SETTGAST, Frederick J. RYERSON, EGS Collab Team |
| Year | 2019 |
| Conference | Stanford Geothermal Workshop |
| Keywords | enhanced geothermal systems, EGS Collab, fracture flow, tracer test, stochastic modeling |
| Abstract | The characterization of flow behavior in fracture network is important for the optimal design and long-term management of enhanced geothermal systems (EGS). In this study, we investigate the feasibility of characterizing fracture flow patterns in the EGS Collab experiment through stochastic modeling of tracer tests. A fracture model involving connected hydraulic and natural fractures is first developed based on applicable field data, including core logs, seismic events, DTS and flow measurements. An injection well, a production well and multiple monitoring wells are also included in the model according to experiment design. We then use a brute-force, Monte Carlo approach to perform massive realizations to simulate tracer transport processes in the hydraulic and natural fractures, and tracer breakthrough curves at multiple monitoring wells are obtained for each realization. A fitting criterion is then employed to select viable solutions that yield optimal fits of the measured tracer breakthrough curves. Through these selected solutions, we estimate the aperture distributions in the hydraulic and natural fractures and try to identify flow patterns in the fracture model. The modeling study indicates that the aperture of the hydraulic fracture is about 100 microns, and the aperture of the natural fracture is about one magnitude larger. The flow field in the natural fracture might be more channelized than that in the hydraulic fracture. We also find that the fluid/tracer may leak from the hydraulic fracture to matrix through a sink to the left of the hydraulic fracture. |