Optimal experimental design for finding optimally located boreholes in geothermal engineering.




An obstacle for further commercial utilization of deep geothermal resources is the high cost for deep drilling in regard to the productivity risk of a well. Minimizing the risk by optimizing the location of a well is thus a question of primary importance. For a given geological model of a reservoir, we demonstrate how this problem may be tackled by a mathematical optimization approach called optimal experimental design. We show how this problem can be mathematically derived, numerically formulated, and practically implemented using distributed computing. We discuss an example for a sedimentary geothermal reservoir with multiple lithological units. We show how optimized positions of boreholes can be determined such that uncertainty of estimating the hydraulic permeability of a target rock unit from temperature measurements is minimized.