Premium
Estimating the permeability distribution and its uncertainty at the EGS demonstration reservoir Soultz‐sous‐Forêts using the ensemble Kalman filter
Author(s) -
Vogt C.,
Marquart G.,
Kosack C.,
Wolf A.,
Clauser C.
Publication year - 2012
Publication title -
water resources research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.863
H-Index - 217
eISSN - 1944-7973
pISSN - 0043-1397
DOI - 10.1029/2011wr011673
Subject(s) - ensemble kalman filter , geothermal gradient , data assimilation , borehole , permeability (electromagnetism) , tracer , kalman filter , soil science , geology , monte carlo method , environmental science , petroleum engineering , geotechnical engineering , extended kalman filter , geophysics , statistics , mathematics , meteorology , geography , membrane , biology , genetics , physics , nuclear physics
We present an estimation of the permeability fields of the reservoir at the Enhanced Geothermal System (EGS) at Soultz‐sous‐Forêts, France, based on the data assimilation technique Ensemble Kalman Filter (EnKF). To this end, we assimilate data from a tracer circulation experiment performed in 2005. Using a 3‐D numerical simulation of fluid transport and chemical tracer dispersion, we advance the tracer in time and control the concentration. With the EnKF we obtain reliable fits for concentration data recorded in both existing production boreholes, GPK2 and GPK4. As an alternative to discrete fracture networks, our heterogeneous equivalent porous medium approach thus can also characterize the hydraulically fractured zone of the engineered geothermal system. We present best estimates for permeabilities (10 −14 m 2 –10 −12 m 2 for the fracture zone) and the corresponding uncertainty which is about one order of magnitude. After comparing our results to results from a massive Monte Carlo and from a gradient‐based Bayesian approach, it becomes clear that only the EnKF of this three approaches is able to fit concentrations at GPK2 and GPK4 simultaneously. Based on the EnKF estimates obtained, a long‐term performance prediction including an uncertainty analysis for the reservoir (as it was in 2005) yields no thermal breakthrough in the system within at least 50 years of operation. Our study demonstrates the efficiency of the EnKF when estimating the permeability distribution in an EGS reservoir even with sparse data available.