Methodology for data assimilation in reservoir and production system to improve short- and medium-term forecast

Abstract

Numerical reservoir simulation is an important tool that is widely used in the petroleum industry for production forecast and decision-making processes. In a context of integrated production system involving the reservoir, wells, and production facilities, the quality of production forecast depends not only on the quality of reservoir models and well productivity/injectivity but also on the boundary condition for the future extrapolation which, in turn, depends on the production system characteristics. Both systems, reservoir and production facilities, are characterized by uncertain properties and behavior, such as porosity and permeability, among others, for reservoir; pipeline roughness, multiphase flow behavior in pipes, among others, for the production system. This paper proposes a methodology for dynamic data assimilation to reduce uncertainty in both reservoir and production system models to improve short- and medium-term production forecast. The methodology was applied to a benchmark case built using real data from an offshore field in The Campos Basin (Brazil). The results from this study show that it is important to perform data assimilation process in both reservoir and production systems to obtain more reliable production forecasts. We also show that, by focusing on the short- and medium-term production forecast, the production system and reservoir uncertainties cause variability in the production forecast in the same order or magnitude. Based on that, it is recommended to consider multiple solutions of the production system to obtain more realistic production forecasts.