Understanding First, Simulation Later: Using Basic and Modern Reservoir-Engineering Techniques to Understand Reservoir Dynamics: A Proven Case

Abstract

Abstract In this paper is described a mature field in which water production is seen very early on in production life and rapidly increased to the current level of 80%. At well level a similar behavior has been a noticed, high initial oil rates with subsequent strong decline once water breakthrough occurs. This type of behavior has been observed in most of the wells completed among the different reservoirs that make up the field. Long term production data and modern decline analysis strongly indicated the presence of two systems in the reservoir. Initially it was considered normal water production behavior. Petrophysical interpretation indicated the presence of five different facies. The production data was imported directly into the simulation model and with some "adjustments" a reasonable match was achieved. This resulted in highly optimistic forecasts, never reproduced by the new wells. This paper presents an analysis and explanation of the production behavior observed in the field. Which proved to be in line with the subsequent acquisition of field data. Indicating the reservoir shows a behavior similar to a dual porosity dual permeability one, with two systems dominating the production behavior in the wells. This paper shows how important it is to have a strong understanding of reservoir dynamics before using 3D reservoir simulation models, and specially in this case, how the use of simple and modern analytical techniques, like for instance Chan plots, transient tests, conventional and modern decline analysis among others, can provide the reservoir engineer with a enough background to confidently take the next step. This paper also describes how pseudo-steady state calculations provided information regarding the characteristics of the second system. Which helped to optimize the original petrophysical interpretation in the static model. This represents a simple approach at the time to consider how detailed a geological model should be to reproduce production behavior. This work had a direct impact on completion strategies and future IOR projects to be evaluated through the use of reservoir simulation, and it clearly represents a new "way of thinking" by integrating and making sense of different analytical techniques as input for building new models (static and dynamic). Similar production behavior has also been observed in other fields in the same area, which can be explained following the methodology here presented.