Lessons From Eyehill

Abstract

Abstract In-situ combustion has been promoted as an enhanced oil recovery method since the turn of the century. Laboratory testing and reservoir modelling typically predict high recovery factors for this process. Field implementation in complex reservoirs is required to evaluate the predictive ability of these models and to establish successful equipment design and operating techniques. This study compares the performance of the Eyehill Thermal Project with model predictions and presents valuable lessons learned from the operation of this pilot. Introduction The Eyehill lower Cretaceous pool is located roughly 70 miles south of Lloydminster near the western border of Saskatchewan, as shown in Figure 1. The reservoir was discovered and developed by Murphy and Canadian Reserve Oil and Gas Ltd. Canada-Cities Service. Ltd., now Canadian Occidental Petroleum Ltd., joined the group after it was determined that primary recovery was limited by high water production and an enhanced oil scheme would be required to increase the probability of significant recovery. The Eyehill reservoir is typical of many heavy oil and oil sands occurrences complicated by bottom or edge water. High oil saturation and relatively thick net pay values contained in a contoured structure of significant volume were thought to be sufficient incentive to evaluate recovery technique and tackle operating problems in view of the possibility that positive results could be applied to similar tracts. History The Cretaceous Cummings sand pool was discovered with completion of 52 feet of highly saturated oil sand pay overlying 28 feet of water in the well location 1-16-40-28-W3M in June 1971. The oil reservoir was then developed with 12 primary oil completions and defined further with five unproductive holes. The channel configuration is truncated on the updip flank unconformably with Devonian topography and on the downdip side by bottom water. With oil pays ranging to 65 feet, early mapping indicated heavy oil-in-place of approximately 160 million barrels with low connate water saturation averaging 16 per cent and with the reservoir for the most part over a free water aquifer. This configuration is not unlike many other occurrences in the area and, in fact, scattered throughout the Lloydminster region, the south reaches of Cold Lake, Peace River and Wabasca. The primary completions, all perforated near the top of the oil column, produced promising initial rates averaging 50 barrels of clean oil per day. Anticipating water production problems, coning models were run at a very early date to project primary recovery, potential. These evaluations projected cone breakthrough after roughly 100 days of continuous production, followed by a rapid and serious increase of water cut. Field experience closely conformed to the coning studies and water cuts soon exceeded 95 per cent. A typical primary recovery performance is illustrated in Figure 2. High water cuts tended to limit commercial recovery even at early experimental high-volume rates using submersible pumps. Primary recovery projections indicated that less than one per cent of initial oil-in-place would be recovered without an enhancement scheme.