A Practical Capacitance–Resistance Model Approach to History Matching and Full Field Chemical EOR Forecasting

Abstract

Abstract Commercial application of chemical flooding requires understanding the performance of patterns already subjected to water- or chemical flooding and the capability to extrapolate this performance to new patterns. The conventional approach that relies on complex finite-difference models to obtain oil profiles is time-consuming and often unable to provide a good match on a well-by-well level. The workflow developed here instead starts from the well-by-well level, using a Capacitance Resistance Model (CRM) and oil response type curves to allocate gross rates and generate oil profiles between injector-producer pairs. These are then matched to historical performance and used for forecasting on a pattern and field scale. In the workflow, multiple patterns are analyzed simultaneously. The analysis comprises the following aspects: (1) Gross Rate Analysis: an iterative scheme is developed to calculate allocation factors between injector-producer pairs that match the measured injection and production history; (2) Swept Pore Volume Analysis: injector-producer connections are assigned pore volumes and properties such as heterogeneity and initial recovery factor at the start of the forecast period; (3) Oil Response Analysis: type curves are used to calculate oil production from each injector-producer connection to match historical oil production and predict future oil response under different development scenarios. The workflow was first verified against synthetic finite-difference models and a good match of gross rates and oil profiles was obtained. The allocation factors and pore volumes assigned to injector-producer pairs were comparable to the values obtained from streamline calculations and simulated tracer data. The workflow was subsequently applied to an actual brownfield development with 24 waterflood patterns, 46 polymer patterns and 2 alkali-surfactant-polymer (ASP) patterns. The historical performance of individual producers and patterns were well captured. Finally, the incremental recovery obtained from EOR was evaluated by comparing the performance forecast of existing EOR activity (NFA) with the waterflood baseline. The workflow produces comparable if not more accurate results than alternative methods in a fraction of the time. In addition, it provides good physical insights into the origin of the produced oil, making it especially suited for development optimization studies. The established workflow also sets the foundation for a generalized approach that can be applied for development and reservoir management of various chemical EOR applications.