Reservoir Characterization of Ekofisk Field: A Giant, Fractured Chalk Reservoir in the Norwegian North Sea-History Match

Authors:

Abstract
Summary  The Ekofisk reservoir is a high porosity, low matrix permeability naturallyfractured chalk. Fluid flow is largely governed by the distribution,orientation, and interconnectivity of the natural fracture system associatedwith complex structure and reservoir distribution. Because of the impactheterogeneity has on preferential fluid-flow direction, significant attentionwas given to capturing as much of the intrinsic heterogeneity as possible, bothlaterally and vertically, in the new three-dimensionally geological model. Anon-uniform simulation mesh was defined for the fluid-flow model with a gridsize of 450 ft×450 ft in the crestal area of the field and increasing towardsthe flanks. A flow-based upscaling technique was then applied to preserve theheterogeneity from the geological to the fluid-flow model. Because of thecomplexity of the Ekofisk field, with its numerous faults and fracturenetworks, anisotropy was one of the primary attributes calibrated to achieve anindividual well and field history match. However, faults and fault sealingfactors, vertical permeability, pseudorelative permeability curves, bubblepointpressure correlations, local permeability, and rock compressibility were alsokey parameters in the history match, and are presented in this paper. A briefdiscussion on the preliminary implementation of water-induced compaction isincluded. Introduction  The Ekofisk field is a prolific field discovered in 1969 that is located inthe Norwegian sector of the North Sea. The reservoir consists of twofine-grained limestone producing formations, the Ekofisk formation (Danian Age)and the Tor formation (Maastrichtian Age), separated by a thin, impermeableTight Zone. The reservoir was initially overpressured and contained anundersaturated oil at 7,120 psi and 268°F at a datum elevation of 10,400 ftsubsea. The bubblepoint pressure was approximately 5,545 psig. Productionstarted from the chalks in 1971. Current estimates from the reservoircharacterization project indicate about 7 billion barrels of oil originally inplace. Production as of the end of 1998 from 76 deviated and horizontal wellswas 310,000 BOPD and 510 MMcf/D of gas. Reinjection of natural gas in excess ofsales has been ongoing in Ekofisk since 1975 with 1.3 Tcf of gas injected todate. Eight wells were initially completed for gas injection service, with fiveof the original gas injection wells being recompleted as production wellsthrough time. A pilot water injection project was initiated in 1981 in thehighly fractured Tor formation1 and in the Lower Ekofisk in1986.2 Fieldwide water injection began in 1987.3 Currentwater injection rates are 800,000 BWPD into 37 active water injection wells. Anumber of additional improved oil recovery techniques are being evaluated, anda water-alternating-gas (WAG) pilot in the southern area of the field wasattempted. Fig. 1 shows a structure map of the Ekofisk field drawn onthe top Ekofisk formation. Reservoir characterization of Ekofisk was directed at gaining a detailedunderstanding of reservoir hydrocarbon volumes, the architecture of thereservoir, and at fully describing the heterogeneity and anisotropy ofreservoir parameters.4,5 Because water breakthrough not consistentwith expectations has been observed in areas of the field, it is important thatthe highest degree of heterogeneity be represented in the flow model. This isespecially significant given that the Ekofisk field is currently undergoing amajor field redevelopment in which 45 new wells will be drilled before the endof 2003. To date a total of 25 new wells have already been drilled and arecurrently on production. The history match of the Ekofisk reservoir characterization fluid-flow modelwas completed in September 1997 after a period of approximately 12 months ofintense work. The complexity of the Ekofisk field, with its numerous faults andfracture networks, provided quite a challenge in matching the 25 years worth ofproduction and performance data. The heterogeneity that was captured in thethree-dimensional (3-D) geological model, and preserved in the upscalingprocess to the fluid-flow model, proved to be the key to being able to matchindividual well performance. In general, a very good history match was achievedon both a field and platform basis, and on an individual well basis. Fluid-Flow Model