Abstract
Identifying promising areas for the placement of new wells in a mature oilfield presents a challenging task, requiring a complex integrated analysis of geological and production data. Full field reservoir simulation model reflects all relevant information about subsurface properties and historical production, making it a great tool to quickly evaluate relative performance of potential infill locations. However, evaluation of all possible scenarios and well patterns for infill drilling using dynamic simulator can be a time-consuming and computationally prohibitive process.
Historically, this problem was commonly addressed by leveraging optimization algorithms such as "Genetic Algorithm", "Particle Swarm Optimization" or "Ensemble based optimizations". These algorithms help to minimize the required number of simulation runs for assessing optimal well placement to a few hundred cases. However, a significant drawback of these approaches is their lack of transparency, when the final recommended scenario with the best well performance might lack a coherent underlying rationale. For example, it could be unclear why all infill wells should be drilled in a certain region of the oilfield or positioned very close to other offset producers.
This paper introduces an innovative approach to streamline the selection process for drilling infill wells. By conducting a focused set of simulation runs, our method employs a unique design that enables enhanced post-processing of the results. This leads to the generation of clear visualizations highlighting the most promising areas within the field. The effectiveness of this methodology has been demonstrated through its application in determining optimal infill producer locations within the Korolev oilfield in Western Kazakhstan.
