Evaluating Interwell Connectivity in Waterflooding Reservoirs with Graph-Based Cooperation-Mission Neural Networks

Abstract

Summary Interwell connectivity plays a key role in waterflooding for guiding water injection. The existing works focus on the response relationship between one injection well and one production well. No research has explored the structural information of waterflooding on a well pattern. To address this challenge, this paper proposes cooperation-mission neural networks for interwell connectivity with graph information. Specifically, we propose some assumptions based on the petroleum domain to represent the well pattern with an adjacent matrix of the graph. Then we propose two targets from the view of injection well groups and production well groups. Accordingly, we propose cooperation-mission neural networks from these two aspects to evaluate the interwell connectivity in the well pattern. We test our model from two perspectives: the accuracy of estimation with tracer and the graduality of interwell connectivity. The results demonstrate that our model makes a good performance and achieves the connectivity analysis accuracy rate of 91.4%. Moreover, this study demonstrates that it is practical to evaluate the interwell connectivity with graph.

Basic understanding of water flooding as a secondary recovery concept

“…According to Hamdy et al (2019), water flooding evaluation in a certain field is essential to be utilized as a valuation basis to measure water flooding success. The intended evaluation is a connectivity analysis of injection and production wells (Zeng et al, 2022). This connectivity analysis is also enhanced by a number of analyses such as Hall Plot analysis to acknowledge problems occurring in the injection well, Chan's WOR analysis to identify problems in the production well, and Voidage Replacement Ratio (VR) analysis to compare injection and production volumes.…”

A New Method for Calculating the Influx Index in Gas-Drive Reservoirs: A Case Study of the Kela-2 Gas Field

“…Its value has a significant impact on water invasion and rock porosity expansion. Previous studies have suggested that under high-pressure conditions in reservoirs, the reservoir water compressibility factor remains relatively constant with pressure changes, but the pore compressibility factor exhibits noticeable variations [21,22]. Currently, there are three main methods for determining the pore compressibility factor: the empirical formula method, experimental measurement method, and theoretical calculation method.…”