A Full Field Simulation Model of the Minagish Oolite Reservoir in Kuwait

Abstract

The Minagish Oolite is a thick undersaturated carbonate oil reservoir in the Minagish field in West Kuwait (Fig. 1) containing several billion STB. It is a mature but relatively undeveloped reservoir. Since discovery in 1959, it has produced 10% of its OOIP under a combination of natural depletion, gas re-injection and aquifer drive. Initial reservoir pressure had declined by about 450 psi prior to the Gulf war in 1990. The well blowouts following the war caused a significant pressure drop of another 700 psi. Following the blowout, plans were made to redevelop the West Kuwait fields and increase the production rate starting in 2001 and to sustain the plateau for at least 5 years. This strategy called for three-fold increase in the production rate of Minagish Oolite reservoir. Since the existing well inventory and the loss of the gas re-injection facility could not sustain the desired plateau rate, additional field development was required. To achieve the production target, a multidisciplinary team was formed to evaluate options. The recommended plan required the drilling of additional producers and installing a field-wide peripheral waterflood. The reservoir, however, presented a number of significant challenges to waterflooding, such as the presence of a substantial and not well defined tarmat near the oil/water contact, and uncertainties of lateral and vertical heterogeneities. In 1997 a full-field simulation model was developed, but this model didn't capture the water movement properly because of insufficient reservoir data at that time. As new core was obtained, a refined reservoir description was developed. Building on lessons learned from the previous full-field model and sector models, a new full-field model was developed which significantly improved well-by-well history matches. Although containing twice as many grid cells, the new model ran up to four times faster than the previous model by making use of the Analytical Aquifer option within the model, improved relative permeability curves and other model refinements. This paper traces the history of the field and the systematic evolution of the development plan. The reservoir simulation efforts including modeling strategy, history matching events, prediction runs, future direction and challenges are also addressed. Introduction Numerical simulators are an important tool for reservoir management, providing management the ability to observe how alternate development plans and operating strategies will affect future oil production and recovery. As additional information is acquired and new technologies are developed, it is necessary to periodically update the reservoir simulation tools. This report identifies the reasons for building a new model, the differences between it and the previous model, and documents the data-sources, files and the methodology used to construct the new model. The previous model (FFM 97) was constructed and initialized in 1997. The model was based on a course 12-layer reservoir description and history matched reservoir performance up through the start of dumpflood water injection. In predictive mode, however, the model did not adequately predict the rapid water movement in the northeast quarter of the field or the arrival of initial water in the peripheral producers. Sector models constructed at the same time indicated that a refined reservoir description that incorporated the observed barriers and high permeability streaks should provide an improved match of the observed water movement.