COMBINING MIXED-COMPOSITION PETROLEUM STREAMS USING PROXY MODELS OF EQUATION OF STATE FOR COMPOSITIONAL INTEGRATED PRODUCTION MANAGEMENT

Abstract

When two distinct offshore reservoirs produce in the same region, the same gathering system for both reservoirs may be used. It is recommended that the gas/oil composition is modeled because its effects can impact reservoir forecasting. An incorrectly-modeled Integrated Production Management (IPM), i.e. one with insufficient data, can impact system management negatively due to the lack of a rigorous compositional modeling impact assessment. To fully study this problem, there need to be a method to model mixed-composition petroleum streams. Most proposed methods to carry this task are Pseudo-Compositional (based on K-values) or are based on Black-Oil models. The objective of this work is to develop a method for mixed-composition petroleum streams to be used in future compositional IPM optimization studies. This paper implements, validates, and discusses the limitations and reproducibility of a mixed-composition petroleum stream. It models a methodology based on an Equation of State (EoS) and PVT data, and combines concepts proposed by Carpio (2012) to automate the process. This work compares the proposed method with two mixing methods from the literature. The three mixing methods studied are based on concepts of tuned EoS and the well-known methods of EoS from Peng-Robinson (1978), and volume translation from Jhaveri-Youngren (1988). The analysis and comparison of the methods are based on conventional data and simulated tuned experiments, when applicable. The results of this study show a small but significant variability of mixed-stream properties from the three proxy methods with the potential to impact optimal reservoir control conditions. Therefore, the deviation potential of mixed-stream models is worth further investigation, thus, justifying a subsequent sensitivity study, with focus on closed-loop IPM.