Multiple reservoir models and multiple history matches: tools for finding reserves in a complex mature field

Abstract

The Nimr field is a mature, heavy oil field in South Oman, producing under bottom water drive. It has been on production since 1985 and produces 13000 m3/d (90000 bbl/d) from 320 wells, 120 of which are horizontal. The field contains 100 million m3 of recoverable oil in several structural highs, consisting of aeolian sands, which are intersected and flanked by glacial valleys (figure 1). Key challenges for further field development are: Finding new oil: The unpredictable nature of both the location and the content of the glacial valleys poses a risk for development drilling: new wells sometimes encounter glacial waste-rock instead of good sands. Economic development requires an effective strategy to manage this risk. Chasing remaining oil: The aeolian highs are currently well developed with horizontal wells down to a 172 m spacing. An important development issue is whether further infill drilling can be economically justified, either on a grid basis, or by targeting remaining unswept oil pockets. Introduction This paper discusses how these challenges are dealt with in the Nimr field. It will explain how a scenario approach quantifies the development drilling risk, while a flexible drilling strategy subsequently manages this risk The paper will also describe the multi-discipline data gathering and modelling effort which justified and directed further infill drilling in the developed parts of the field. Sweep efficiency is quantified by several methods: core measurement, observation wells, and field pilots. Horizontal observation wells were drilled underneath producers. penetrating water cones built up by the bottom water drive mechanism. The measured saturation profiles imaged macroscopic sweep efficiency directly and showed the impact of baffles and sealing faults. Basic data were inferred by matching single well simulation models to the observed saturation profiles. Over a thousand sensitivities were simulated, scanning the parameter space for the match parameters. This process resulted in a probability distribution for each parameter, quantifying not only the best match, but also the remaining uncertainty in the matching parameters. Probabilistic geological sector models were built of representative parts of the field. Several cycles of simulation and history matching by updating the probabilistic model resulted in good matches which honoured the basic data determined from the observation wells. The matched models were used to assess the recovery factor as a function of column height and the column split between good aeolian sands low in the column and poor glacial sediments on top. The study justified further infill drilling and matured about 15 million m3 (nearly 100 million bbl) of reserves and demonstrated both the power and limitations of probabilistic modelling techniques. 2. Finding new oil The Nimr reservoirs are dominated by aeolian sequences (Amin Formation) with good porosity (28%) and permeability (0.5-5 Darcy). These Cambrian sands extend about 1000 m below the oil/water contact and this large aquifer provides strong bottom water support. The top of the package is defined by a major composite erosional unconformity. After a hiatus of around 220 million years, sedimentation resumed with the deposition of Permo-Carboniferous glacial sequences (Al Khlata Formation) on a structurally complex paleo-topography modified by additional glacial valley erosion. The glacial valleys contain a heterogeneous mixture. of sand, shale and diamictite (figure 2). In some locations, the valley fill has good reservoir properties. P. 89