3D Reservoir Simulation of Ekofisk Compaction Drive (includes associated papers 24317 and 24400 )

Abstract

Summary. In late 1984, the seafloor at the Ekofisk field in the Norwegian North Sea was discovered to have subsided by more than 10 ft as a result of production-induced reservoir compaction. Subsidence was not expected and its detection suggested that reservoir compaction was perhaps a more important mechanism for hydrocarbon production than previously assumed. This paper describes procedures developed to account for compaction drive in the 3D reservoir simulator that is used for the Ekofisk field. The procedures involve the incorporation of results from a subsidence model into the reservoir simulator to describe the compressibility of the reservoir chalk as a function of porosity, rock type, reservoir pressure, and position in the reservoir. Introduction The unexpected discovery in Nov. 1984 of production-induced reservoir compaction and subsidence at the Ekofisk field (see Ref. I for a history of the Ekofisk field) established a need to account rigorously for the reduction in PV in the 3D reservoir simulator used for the field. Rock compressibility was the parameter in the reservoir simulator that would have taken reservoir compaction into account. input data available at that time clearly understated the compaction that had accumulated. Mechanical property data for reservoir chalk obtained through laboratory testing in early 1985, coupled with observations and assumptions about the pliability of the overburden, were initially incorporated in the 3D reservoir simulator as a pressure-dependent rock compressibility. This was an important step in establishing an overall pressure match for the field. Simultaneously, a subsidence model was being developed for the Ekofisk field, and reservoir pressure vs. time information from the reservoir simulator was a critical input to the subsidence model. The subsidence model showed an "arch effect" that redistributed the overburden load throughout the reservoir. An early conclusion was that this arch effect had to be incorporated in any rigorous treatment of rock compressibility in the reservoir simulator;hence, it was necessary to modify the approach that had been taken to account for the reduction of reservoir PV. This paper describes how the reservoir and subsidence models were coupled to describe rock compressibility in the Ekorisk field as a function of porosity, rock type, reservoir pressure, and position in the reservoir. Modeling Compaction Drive During the past several decades numerous studies have been performed primarily to address the mechanical aspects of reservoir compaction and subsidence processes. Conceptual ideas about compaction drive also have been developed during the past couple of decades. One of these studies involved simultaneous treatment of compaction drive and reservoir compaction and subsidence. The simulation in that case was directed primarily at a hypothetical elastic reservoir surrounded by an elastic geological medium, although an attempt was made to apply the developed procedures to a portion of a reservoir under Lake Maracaibo in western Venezuela. Computed reservoir pressures were found to compare favorably with observations, but the subsidence results did not, which can readily be attributed to considering only a portion of the field in the model. The work we report here deals with a numerical procedure in which computed results from a reservoir compaction and subsidence simulation are used to modify the positiondependent overburden load on the reservoir rock in a reservoir model. In this way, the compaction drive process is coupled to the reservoir compaction and subsidence processes in a rigorous way. In the initial modeling of Ekofisk field, the rock compressibilities were functions of porosity and rock type, Upper Ekofisk and Lower Ekofisk/Tor(Table 1) only. They essentially were based on early special core analyses with the premise that the overburden was relatively rigid so that effective overburden loads would result only in elastic deformation. After the discovery of subsidence of Ekofisk, the overburden could no longer be considered rig. Pore collapse was occurring and was incorporated into the reservoir model in the rock compressibility term. The initial treatment of compaction drive recognized that the overburden must have some resilience, however. Rock compressibilities that were a function of porosity, rock type, and reservoir pressure were then incorporated into the model (Fig. 1). These rock compressibility curves were based on numerous rock strength tests performed on Ekofisk chalk, but direct laboratory results were modified to account for the ability of the overburden to support at least some portion of its own weight. This treatment resulted in a good overall match of reservoir performance. JPT P. 1272^