"Oil production optimization of black-oil models by integration of Matlab and Eclipse" - Steen Hørsholt
Abstract:
We have developed an optimization software tool that combines the simulation power and reliability of an industry standard commercial reservoir simulator (Eclipse E300) and state-of-the-art
software for constrained optimization (Matlab).
The optimization software enables us to apply ensemble-based oil production optimization under uncertainty to realistic black-oil and compositional type reservoir models. We consider oil production optimization strategies which take the uncertainty of the geological reservoir properties into account in order to mitigate the associated risk.
Such strategies include mean-variance optimization, robust optimization and conditional value at risk optimization. We present the general workflow of the optimization tool and a numerical example. The example is a mean-variance optimization of a synthetic black-oil reservoir.
"Optimal control of compositional and thermal recovery processes" - Tobias Kasper Skovborg Ritschel
Abstract
We develop numerical algorithms for optimizing the long-term profitability of oil and gas recovery. The algorithms combine simulations of the subsurface reservoir flow with a numerical optimization algorithm.
The optimization algorithm iteratively improves the production strategy until an optimal strategy is reached. We use an adjoint algorithm for the computation of gradients such that an e_cient gradient-based optimization algorithm can be used.
We develop the algorithms for thermal and compositional reservoir flow models. We use a recently developed high-performance software library, ThermoLib, for nonideal thermodynamic computations. The algorithms are implemented in the high-performance programming languages C and C++, and we present a numerical example.