This work proposes a methodology based on a metaheuristic that combines Branch-and-Bound,Tabu Search and Scatter Search in order to optimize the production of a petroleum reservoir by determining well types(producer or injector)and locations.After analyzing reservoir geometry and drainage area,possible places were proposed where wells could be drilled.With that,40 binary variables were created to decide whether or not a producer or an injector should be drilled.Numerical simulations,guided by the metaheuristic,were performed in order to maximize the objective function.The objective function took into account not only cumulative oil production,but also well drilling costs.As posed,this problem has a solution space of almost 3 and a half trillion possible simulation cases,thus the need of an efficient optimizer.By the time objective is the best solution and as reservoir engineering functions are not so easy to solve,a simulation-based optimization is then required.The methodology was implemented in a synthetic field,resulting in a development plan that corresponded to a 12%increase on the recovery factor over a field life of 25 years.
展开▼