Methodology for Reservoir Complexity Determination from Downhole Fluid Analysis and EOS Predictions

Abstract

Abstract Reservoir hydrocarbons often present complicated properties and phase behaviors in single columns. It is of great importance to determine fluid heterogeneities and compartmentalization at the early exploration and appraisal stage. However, gathering information on fluid heterogeneities and phase behaviors is a difficult process that conventionally requires numerous fluid samples and associated laboratory analysis. An effective and reliable way is highly demanded. Pressure gradients are traditionally used to evaluate in-situ fluid density, contacts, or connectivity. This process may be misleading as fluid compositional changes and compartmentalization give distortions in the pressure gradients. The Downhole Fluid Analysis (DFA) measurements provide basic compositions, gas-oil ratio, density, viscosity, and saturation pressures in real-time to reveal fluid heterogeneities. Therefore, it is needed to incorporate a rigorous mathematical approach in respect of all data available, so as to allow an objective assessment of reserves and reservoir architectures. This paper demonstrates a methodology for evaluating fluid variations with depth through an equation-of-state (EOS) approach by interpreting DFA measurements. Using a proposed method, the topmost DFA data in the column was delumped and characterized, and then an EOS model is established for the reservoir fluids. The EOS parameters can be tuned to match DFA measurements at different depths. The tuned EOS is applied to predict compositional and property gradients with depth. If the difference between the predictions and DFA data is significant, then something happens between the top pay zone and current depth must be investigated; if the predictions are close to DFA data, the process is repeated for the next depth. Case studies are presented to illustrate how this methodology is applied. The case studies indicate that the methodology of integrating DFA data with other logs provides a powerful tool of characterizing reservoir architecture, which is invaluable for optimal reservoir management and development. Introduction This paper relates to a methodology for interpreting downhole fluid analysis, for estimating fluid compositional and property variations with depth, and for determining compartmentalization of reservoirs through an equation-of-state (EOS) approach. In this method, the DFA data in the top reservoir column was delumped and characterized by the method proposed by Zuo et al. [1], and then an EOS model is established for the reservoir fluids. To obtain better EOS for the fluids, the EOS parameters can be tuned to match DFA measured data such as live fluid densities, compositions and formation pressures at different depths. The establish EOS model is applied to predict compositional and property gradient with depth. If the difference between the EOS predictions and DFA data is significant, then something happens between the top pay zone and current depth (maybe flow barrier). More DFA stations should be required to figure out what is going on there. On the other hand, if the EOS predictions are close to DFA data, the process is repeated for the next depth. Case studies are presented to illustrate how this method is applied to reservoirs to identify flow barriers, and predict fluid properties. The case studies indicate that the methodology of integrating downhole fluid analysis with other logs provides a powerful tool of characterizing reservoir architecture, which is invaluable for optimal reservoir management and development plan.