BR 
EN In binary multi-objective well placement optimization, multiple conflicting objective functions must be optimized simultaneously in reservoir simulation models containing discrete decision variables. Although multi-objective algorithms have been developed or adapted to tackle this scenario, such as the derivative-free evolutionary algorithms, these methods are known to generate a high number of duplicated strategies in discrete problems. Duplicated strategies negatively impact the optimization process since they: (i) degrade the efficiency of recombination operators in evolutionary algorithms; (ii) slow the convergence speed as they require more iterations to find a well-distributed set of strategies; and (iii) perform unnecessary re-evaluations of previously seen strategies through reservoir simulation. To perform multi-objective well placement optimization while avoiding duplicated strategies, this paper investigates the application of a newly proposed algorithm named MOEA/D-NFTS, with a modified diversity preservation mechanism that incorporates prior knowledge of the problem, on a multi-objective well placement optimization problem. The proposed methodology is evaluated on the UNISIM-II-D benchmark case, a synthetic carbonate black-oil simulation model in a well placement optimization problem using a binary strategy representation, indicating the presence or absence of a given candidate well position in the final strategy. The objective functions are the maximization of the Net Present Value, the maximization of the Cumulative Oil Production, and the minimization of Cumulative Water Production. The modified MOEA/D-NFTS performance is compared with a baseline algorithm without diversity preservation, and the evidence shows that the MOEA/D-NFTS produces statistically significant superior results, and is suitable for binary multi-objective well placement optimization.
Mês: agosto 2023
Prototyping visualizations as a support for selecting representative models of petroleum reservoirs
Petroleum engineers usually create a set of hundreds of models of a given oil reservoir under analysis to represent its uncertainties. Assisted optimization approaches may help engineers to select a subset of these models (a.k.a. representative models, or RMs for short), which is used in computational flow simulations in replacement of the original set, aiming to reduce the total simulation runtime without changing the quality of these results. Despite the power of visualization techniques to help people to understand multidimensional datasets like the ones provided in this scenario, we noted a few efforts that use these techniques to help petroleum engineers to assist the selection of RMs. In this context, our research aims to test the hypothesis that it is possible to improve how interactive visualization resources are currently used to aid decision-making regarding the selection of RMs, mainly in the presence of multiple sets of RMs or multiple variables (obtained from running model simulations). This work presents our first steps towards this goal: (a) literature review and (b) definition of visualization prototypes that aim to help the analysis of RMs regarding the values of the variables provided by simulation outputs, and the risk curves associated with these variables. As preliminary results, we present our proposed interactive visualizations and briefly point out the design rationale behind these prototypes.