No abstract

The Ekofisk Field is a naturally fractured chalk reservoir located in the Norwegian sector of the North Sea. The natural fractures clearly control the permeability distribution, as the effective permeability can reach 50 mD whereas the matrix permeability only ranges between 0.1 mD and 10 mD. Permeability mapping in this field has been challenging due to the structural, stratigraphic and mineralogical complexity, tectonic history and non-negligible matrix permeability. A detailed fault interpretation has resulted in a complex fault pattern. A fault intensity (P21) parameter calculated from the fault pattern has proved to be the critical component for permeability mapping. Correlations were found between the fault intensity (P21) values and (1) the fracture distributions from cores and logs at individual wells, and (2) the fracture component of the well test permeability ( K frac – total permeability less matrix permeability). These relationships allowed field-wide fracture permeability maps to be developed based on the P21 results. Total permeabilities were obtained by summing the matrix permeability and the calculated fracture permeability. These permeability maps were introduced in the Ekofisk Flow Model 2002 and refined to match the rate performance of the 50 initial wells run in prediction mode (well head pressure constraint). The runs in prediction mode have proved to be very effective for calibrating the permeability distributions, based on initial well performance above the bubble point. This simulation technique was extended to cover all producers (262 wells) during the entire history of the field to refine the maps further. After calibration with the performance data, a satisfactory history match was obtained by making minor changes to permeability and other dynamic parameters. Additionally, the running of the model of the mature Ekofisk Field in prediction mode for its full field life has provided a robust tool for calibrating field performance.

Over 7.8 meters of seafloor subsidence has occurred at the Ekofisk Field in the Norwegian sector of the North Sea since the start of production in 1971. Full field water injection was initiated at Ekofisk on a limited scale in 1987. The surface subsidence is a result of reservoir compaction, which is considered primarily to be due to pressure depletion until the early 1990's and water weakening thereafter. Rock compressibility was input as a function of initial porosity and increasing net effective stress (i.e. declining reservoir pressure) in earlier Ekofisk studies. In 1994, under a voidage balancing reservoir management program, water injection was increased sufficiently to stabilize reservoir pressure. However, no reduction in surface subsidence rate was seen. This, in combination with other field and lab observations, led to the conclusion that water was weakening the reservoir chalk and necessitated revising the rock compressibility functions at Ekofisk to include the effect of additional compaction due to the water weakening. The development and implementation of the water induced compaction functions at Ekofisk is presented in this paper. Rock compressibility is now input into the model as a function of initial porosity, net effective stress, and water saturation. As water saturation increases in a model cell due to water injection or water influx, the model cell transitions to a weaker stress-strain curve. The effect of increasing water saturation, and the resulting water weakening of the chalk, is that compaction and subsidence may continue in spite of stable or increasing reservoir pressure. Both laboratory and field data are presented which support the use of the water weakening functions. The development and calibration of these curves is presented, which includes the effects of fracturing, creep, water dispersion effects, hysteresis logic, and strain hardening. A comparison of the calculated and measured compaction and subsidence bowls is also presented.

Résumé -Couplage itératif de la géomécanique et des écoulements pour modéliser la compaction des roches en simulation de gisements -Les méthodes conventionnelles de simulation des réservoirs expriment l'effet de la compaction des roches sur la réduction du volume poreux grâce à la notion de compressibilité de la roche soumise à une condition de charge donnée (déformation hydrostatique ou uniaxiale). Cette approche est d'ordinaire satisfaisante dans le cas de gisements formés de roches compétentes. Cependant, pour des formations plus déformables ainsi que pour des mécanismes de compaction de roche plus complexes, un couplage des modèles géomécanique et d'écoulement des fluides est nécessaire afin d'obtenir une simulation numérique du réservoir plus rigoureuse et plus précise. En général, l'efficacité du calcul numérique et la convergence de la solution sont deux facteurs clés du succès de la méthode d'un point de vue économique et numérique pour que celle-ci puisse être appliquée à l'échelle d'un réservoir. Le présent article propose une procédure itérative pour le couplage du modèle géomécanique et des écoulements polyphasiques pour la simulation numérique des gisements, applicable à des champs 3D de grande taille. La méthode proposée est générale et efficace pour la modélisation des roches ayant des lois de compaction et de changement de perméabilité complexes, ainsi que pour la simulation de divers scénarios de production. Les descriptions des modèles, des lois de comportements, des méthodes de résolution ainsi que des stratégies de réduction des temps de calcul sont présentées. Afin de faire la démonstration des capacités de la méthode de couplage itératif proposée, plusieurs problèmes sont étudiés, parmi lesquels un exemple de réservoir à l'échelle du champ. Abstract -Iterative Coupled Analysis of Geomechanics and Fluid Flow for Rock Compaction in Reservoir Simulation

Summary. This paper describes the design, implementation, and initial results of the full-field water-injection program in the Ekofisk field of the North Sea. Two pilot waterfloods, injection-well-pattern design, verticalconfinement considerations, optimization of production-well sidetracks, corefracture analysis and orientation, production-well sidetracks, core fractureanalysis and orientation, regional per-meability variations, reservoir geologyand faulting, and overall anisotropy are discussed. Results of a comprehensive waterflood surv-eillance program are presented as well as 3D-model predictionsfor ultimate recoveries. Introduction The Ekofisk field is located in the Norwegian sector of the North Sea and iscomposed of two naturally fractured chalk formations, the Ekofisk and the Tor. This essentially volumetric, solution-gas-drive reservoir was initially undersaturated, with an initial pressure of 7,120 psi and a bubblepoint pressure of 5,545 psi at 268 degrees F. Initial psi and a bubblepoint pressureof 5,545 psi at 268 degrees F. Initial solution GOR at producing separatorconditions was 1,530 scf/STB, and initial oil gravity was 33 degrees API. Production started from four subsea producers in 1971 and switched to three permanent production platforms in 1975. Natural-gas injection (basically swinggas) was platforms in 1975. Natural-gas injection (basically swing gas) wasimplemented in 1975 and is expected to continue through 2011. This natural-gasinjection along with oil expansion, solution gas, gravity drainage, andcompaction drive would have yielded primary recoveries of about 24% in terms ofoil equivalent, primary recoveries of about 24% in terms of oil equivalent, assuming 6 Mcf = 1 bbl oil. The Ekofisk field waterflood was designed to enhance recovery from anaturally fractured, low matrix permeability, solution-gas-drive, chalkreservoir, which contained more than 6.8 × 109 bbl original oil in place(OOIP). Enhanced recovery potential from waterflooding was investigated through extensive laboratory experiments and pilot waterfloods to examine recoveries byspontaneous imbibition. A Tor formation pilots began in April 1981 andcontinued through June 1984. The results basically confirmed the laboratory results and were used to justify the field waterflood in the Tor formation. Approval to proceed with the Tor waterflood was granted in Oct. 1983, affectingthe northern two-thirds of the field. An additional platform (Platform 2/4 K)was constructed for the injection facilities and included well slots for 30wells. The total cost for the waterflood development was $1.5 billion. Fieldwater-flood injection began in Nov. 1987. A pilot in the lower Ekofisk formation began in June 1986 and continues todate. This pilot has performed beyond expectations and has supported the higherrange of laboratory measurements of spontaneous imbibition. The success of thispilot coupled with the early results of the Tor waterflood led to approval ofthe Ekofisk waterflood expansion project in June 1988. This project includedexpansion of water injection into the lower Ekofisk formation and the remaining Tor formation in the southern one-third of the field. It also included acomprehensive infill-drilling program. The initial Tor waterflood was expected to increase reserves from Ekofiskfield by 160 X 106 bbl oil equivalent. The waterflood expansion project isexpected to increase reserves by an additional project is expected to increasereserves by an additional I go X 106 bbl oil equivalent. In addition toincreased recovery, the waterflood became important in providing pressuresupport to help mitigate the reservoir subsidence identified in 1984. This paper describes the design, implementation, and monitoring program ofthe Ekofisk field waterflood. Pilot performance is compared to program of the Ekofisk field waterflood. Pilot performance is compared to an extensivedata-acquisition program, structured to solve critical unknowns in thewaterflood regions. Response curves are presented, and the experience gained inthe waterflood project is summarized. Reservoir Description Geology Overview. The two oil-producing formations in Ekofisk field, Ekofiskand Tor, are composed of chalk sediments made up mostly of skeletal carbonatematerial. The Ekofisk formation can be split into three layers: the upper andlower sections and the "tight zone" . The upper section containsalternating sequences of autochthonous deposition and reworked Danian-agematerial and has an average thickness of 400 ft. This section containsporosities of 25 to 48 % and moderate natural fracturing. The lower Ekofiskformation is primarily reworked Maastrichtian-age sediments of 120 ft averagethickness. primarily reworked Maastrichtian-age sediments of 120 ft averagethickness. The lower section contains uniform porosity in excess of 30% andintense natural fracturing, The tight zone is composed of an average of 70 ftof autochthonous chalk. This layer is characterized by low porosity andpermeability and restricts communication between the Ekofisk and Torpermeability and restricts communication between the Ekofisk and Tor formationsin the majority of the field. The Tor formation contains Maastrichtian-agereworked chalk sediments. Porosities between 25 and 40 % are typical, withoil-bearing sections up to 500 ft in the crestal region . Natural Fracture Description and Trends. Four major fracture types exist in Ekofisk: tectonic, stylolite-associated, irregular, and healed. Tectonicfractures predominate in the Ekofisk formation, while most of the fractures inthe Tor formation are stylolite-associated. The tectonic fractures in the Ekofisk form well-developed parallel andconjugate sets. The highly fractured zones typically have spacings as small as 2 to 6 in. Zones of lower fracture intensity have spacings of 6 to 40 in., with40-in. spacings rarely encountered. The dip of the tectonic fractures variesfrom 65 to 80 degrees. Stylolites in the Tor formation are parallel to beddingplanes and are usually only a few feet apart. Stylolite-associated fracturesdevelop perpendicular to the stylolite seams and are essentially vertical. Fracture lengths vary from 4 to 8 in. These fractures form permeable zonesparallel to bedding planes that extend laterally for large distances. The mosthighly fractured zones correspond to areas with the greatest rate of change instructured dip. Two major fracture trends exist in Ekofisk field. The dominant trend in themajority of the field is a basement-faulting-dominated, north-northeast/south-southwest trend. This trend is especially pronounced inthe north and northwest portions of the field and is largely responsible forthe prolific nature of these regions. A secondary radial trend resulting fromstructural uplift is present throughout the field and is related to the rate ofchange in structural dip. This fracture trend becomes most important in regionsor the field where the basement-dominated trend becomes less pronounced. Theseareas demonstrate overall lower fracture intensities, and pronounced. Theseareas demonstrate overall lower fracture intensities, and thus lowerproductivity, and are commonly encountered in the eastern and southernflanks. Effective permeabilities have been calculated from well tests up to a factor of 50 times the 1 - to 2-md matrix values. These permeabilities have beenobserved in the most intensely fractured regions permeabilities have beenobserved in the most intensely fractured regions of both formations and aredirectly related to the fracture intensity. Fig. 1 depicts the importance ofnatural fracturing to well productivity in a typical log section of an Ekofiskwell. SPEFE P. 284