A novel method for DCS reservoir modeling with consideration of oil-water relationships

Abstract

Distributary Channel System (DCS) reservoirs play a crucial role in storing oil and gas, and several modeling methods are used to study them. Among these methods, magnetic random walk is a method with unique advantages and can achieve better results in some scenarios. However, current magnetic random walk models cannot effectively consider oil-water relationship information in wells, resulting in the potential “water above oil” contradiction. To address this problem, this article proposes a novel algorithm that considers oil-water relationships. This algorithm inherited the idea of the magnetic random walk. It not only considers the distinction between channel and non-channel wells, but also considers the distinction between oil and water properties within channel wells. To avoid potential oil-water contradictions, an oil-water contradiction avoidance mechanism is introduced. In a case study, the new algorithm indeed showed better performance, effectively overcoming potential oil-water contradiction problems. Compared to existing algorithms, the new algorithm incorporates more geological constraints (structural height information and oil-water properties information) and is an important step in improving the accuracy of well-interval predictions, which can provide important support for oil and gas development and has important application prospects.