Computational Science Issues in Modeling Oil and Gas Production
ECMOR VIII - 8th European Conference on the Mathematics of Oil Recovery
We address major computational science issues arising in modeling oil and gas production . We describe a new methodology based on mortax spaces which allow for multiple models and for dynamic upscaling not requiring pseudo functions . Allo, we introduce a new Discontinuous Galerkin discretization which is higher-order, locally conservative and allows for non-conforming grids . 1 Introductio n In oil and gas production, the major objective is to maximize return on investment . The challenges
... lve the ability to treat large detailed flow models, geologie uncertainty, and operational flexibility since infinitely mant' production strategies are possible . These challenges clearly point to the requirement of accurate and efficient parallel simulators which can be coupled to geostatistical and economie models within a flexible and friendly computational infrastructure . Geologie uncertainty and production strategies Weed to be evaluated simultaneously . This involves multiple realizations of multiple geostatistical models and the number and localization of Wells as well as the coupling to economie models which are functions of production data, tost/ price parameters, rate of return on investment, etc . mom a computational point of view, one must be able to treat these uncertainties in a seamless fashion and to test variations on production strategies, evaluate sweep efficiency, and bypassed oil .