Development of a high speed optimization tool for well placement in Geological Carbon dioxide Sequestration

摘要

Carbon dioxide Capture and Storage(CCS)is a promising technique for mitigating greenhouse gas emissions by capturing carbon dioxide in flue-gases from fossil-fuel combustion and storing that.Geological Carbon dioxide Sequestration(GCS)injecting the captured CO2 into deep geological reservoirs through injection wells is a part of CCS.In the case of commercial scale GCS,a large amount of CO2 over 100 million tons per year will be injected.Moreover,installing injection well is expensive in cost since the targeted reservoir is often deeper than 1km.Considering the project cost,a smaller number of wells are needed to allocate effectively to satisfy constraints of the project such as the required injection volume of CO2.For searching the optimum placement automatically,we have developed an optimization tool by combining an optimization method(CMA-ES)with a numerical simulation code(TOUGH2).However,this tool needs to be improved because searching the optimum placement with thousands of simulations is computationally demanding.In this study,the tool was improved by leveraging a parallel computing technique and supercomputer.The performance of the tool was demonstrated through a case study of well placement optimization on a heterogeneous reservoir model.As a result,the tool could find a reasonable solution within the realistic time(9 days),which is supposed to be several hundred times faster than the conventional approach.