An improved method of well-to-well connectivity evaluation is proposed. The method is based on combination of two approaches developed by now - capacitance model (CM) and Multiwell Productivity Index (MPI). Multiwell productivity index is evaluated independently of production data basing on well location geometry and average reservoir properties. Connectivity coefficients, derived from MPI are used as constrains when searching for CM solution. Following the physical meaning of the problem the capacitance model is improved by adding a constraint which was possibly overlooked by previous researches. These improves convergence of optimization problem which is inherent part of CM algorithm and enables to apply it to evaluate waterflood for real reservoir with more than 60 wells.
An injection and production rates as well as bottom-hole pressure data can yield a lot of valuable information about well interaction and therefore, reservoir characteristics, if they are analyzed properly. For example, such analysis can reveal a preferential flow direction throughout the field or quantify interaction of injector to surrounding producers that enables to reduce ineffective water circulation. Proposed method combines CM and MPI approaches which both got their own advantages and drawbacks. Capacitance Model provides quite detailed analysis and quantifies the well interaction. Thus, it allows us to learn more about the reservoir structure as well as to optimize the waterflood, but it turns out to be unstable when applied to real data. Also it misses an important constraint connected with a physical meaning of CM parameters. The stability of CM can be increased by using an MPI approach. To this effect, we use analytical MPI values in construction of a regularizing functional within the CM's algorithm.
The method was applied to one of Rosneft fields to establish well interaction pattern. Recommendations were given to improve waterflood efficiency.