TSK 11 Göttingen 2006 Ehrlich et al. Quantification and sensitiv- ity of fault seal parame- ters demonstrated in an in- tegrated reservoir modelling work flow. A case study on the Njord Field, Halten Ter- race, Norway Vortrag Ralf Ehrlich1,3 Einar Sverdrup1,3 Vibeke Øye2 Jorunn Sjøholm2 Kari Smørdal Lien2 Roald Færseth2 The primary objective of this paper is to present a fault seal case study from the Njord Field, offshore Norway. The study utilised analogue field studies as well as core descriptions and petrophys- ical well data in order to evaluate the sealing potential of large to medium scale faults that segment the reservoir. Dynamic data and 4D seismic informa- tion was used to calibrate the results Figure 1: Reservoir simulator grid. Fault zones will be represented as grid cell bound- aries with no volume. through multiple fault seal scenarios. 1 Roxar AS, PO Box 165 Skøyen, N-0212 Oslo, Norway 2 Norsk Hydro ASA, PO Box 7190 Bergen, Norway 3 present address: Ener Petroleum ASA, Lysaker Torg 5, 1325 Lysaker (Oslo), Norway The study was performed using the FaultSeal module of Irap RMS. Intensive faulting has in nearly all cases a strong influence on fluid flow in reser- voirs. As methods to investigate fault seal have evolved and the tools are be- coming integrated in the reservoir mod- elling work flows, the effect of faults now have an increased influence in nu- merical models of reservoir communica- tion. Also, several recent articles have increased our understanding of faults and their effect on fluid flow in differ- ent types of lithology and reservoirs (e.g. Yielding et al. 1997, Manzocchi et al. 1999, Sperrevik et al. 2002). In reser- voirs where faults form complex fault patterns and comprise effective seals, a good representation of the faults re- quires the three-dimensional geometry, displacement pattern, fault zone thick- ness variation and deformation charac- teristics. In current reservoir simulator grids, such properties must be captured in a regularised coarse grid on the grid cell boundaries (Fig. 1). The Njord Field is one of the most chal- lenging oil fields offshore Norway, par- ticularly in terms of prediction of struc- tural deformation and flow communica- tion (Dart et al. 2004, Rivenæs et al. 2005). The field is located on the Hal- ten Terrace, ca. 130 km NW of Kristian- sund, Norway. The reservoir sequence of the Lower Jurassic Tilje Formation is approximately 120m thick, and com- prises alternating sandstones and shales, which were deposited in a tidal to estu- arine paleoenvironment (Dalland et al. 1988). The structural development of the field is defined by several rifting phases throughout the Triassic, Juras- sic and Early Cretaceous (Blystad et al. 1995, Ehrlich & Gabrielsen 2003). The field is structurally deformed by a com- 1 Ehrlich et al. TSK 11 Göttingen 2006 Figure 2: Cross section through a part of the Njord field showing intensive faulting and reservoir segmentation plex pattern of segmented and linked ex- tensional faults, which have led to a high degree of compartmentalization in the reservoir zone (Fig. 2). Since production start-up in 1997, dy- namic well data, 4D seismic surveys, and tracer data have shown that faults have a significant effect on fluid flow and communication patterns within the Njord field. Faults with displacement in excess of 25meters are commonly re- garded as sealing in relation to pro- duction timescales (Dart et al. 2004). The sealing mechanism on these faults has been assumed to be caused by shale/clay smearing. Previous unpub- lished studies have shown that the most important factors in fault seal assess- ments of the Njord field are the stratig- raphy involved in faulting as well as the fault throw. Analogue field studies have therefore focussed on the mechanisms of incorporating incompetent lithology from the host rock into fault zones, and to quantify the sealing capacity of such faults (Fig. 3). Dynamic well data from the Njord field, however, show that the pressure differences across the faults are higher than those predicted from anal- yses of clay/shale smear. Based on core observations and measurements, it Figure 3: Fault core (ca. 1m thick) consist- ing of sand lenses, fault gouge, and smeared coal and shale. The fault throw is ap- proximately 15m. Locality: Hartley Steps, Northumberland, England is concluded that cementation processes have been active along the faults and have increased the fault seal capacity for the faults in the Njord Field (Fig. 4). In cases where cementation processes act along faults, fault seal analysis evaluat- ing smear effects only become ambigu- ous. This study focussed on integrating the fault seal calculations in a common reservoir modelling workflow in order to quickly investigate the effects of chang- ing fault seal parameters and algorithms with respect to reservoir performance (Figure 5). The fault seal module of 2 TSK 11 Göttingen 2006 Ehrlich et al. Figure 4: Micro-fault cemented by calcite RMS has the ability to create and orga- nize a large number of fault rock prop- erty predictions, and to export final results for reservoir simulation. The algorithms include Shale Gouge Ra- tio (SGR), Shale Smear Factor (SSF), Clay Smear Potential (CSP), and user- defined SGR curves, as well as the pub- lished algorithms of Manzocchi et al. (1999) and Sperrevik et al. (2002). The module supports stair-step implementa- tion of faults, and include numerous and advanced cell face visualization that al- low answers to be viewed and checked. The results from this study demonstrate the importance of evaluating the de- formation mechanisms of fault prior to fault seal analyses, and the effect of a ge- ological fault seal tuning process. The study further shows the value of inte- grating fault seal as one work task in a reservoir modelling work flow, enabling multiple sensitivity runs and rapid as- sessment of the fault seal results. The integrated workflow will certainly en- courage an even tighter communication between geologists and reservoir engi- neers. References Blystad P, Brekke H, Færseth RB, Larsen BT, Skogseid J & Tørudbakken B (1995) Struc- tural elements of the Norwegian continental shelf: Part II. The Norwegian Sea Region. Norwegian Petroleum Directorate Bulletin, 8 Dalland A, Worsley D & Ofstad K (1988) A lithostratigraphic scheme for the Mesozoic and Cenozoic succession offshore mid- and northern Norway. Nowegian Petroleum Di- rectorate Bulletin, 4 Dart C, Cloke I, Herdlevær Å, Gillard D, Rivenæs JC, Otterlei C, Johnsen E & Ekern A (2004) Use of 3D visualization techniques to unreveal complex fault patterns for pro- duction planning: Njord field, Halten Ter- race, Norway. In: Daviers RJ, Cartwright JA, Stewart SA, Lappin M & Underhill JR (eds.) 2004. 3D Seismic Technology: Ap- plication to the Exploration of Sedimentary Basins. Geological Society, London, Mem- oirs, 29, 249–261 Ehrlich R & Gabrielsen, RH (2004) The com- plexity of a ramp-flat-ramp fault and its ef- fect on hanging-wall structuring: an example from the Njord oil field, offshore mid-Norway. Petroleum Geoscience, 10, 305–317 Manzocchi T, Walsh JJ, Nell P & Yielding G (1999) Fault transmissibility multipliers for flow simulation models. Petroleum Geo- science 5, 53–63. Rivenæs JC, Otterlei C, Zacheriassen E, Dart C & Sjøholm J (2005) A 3D stochastic model integrating depth, fault and property uncer- tainty for planning robust wells, Njord Field, offshore Norway. Sperrevik S, Gillespie PA, Fisher QJ, Halver- son T & Knipe RJ (2002) Empirical estima- tion of fault rock properties. In: Koestler AG & Hunsdale R (eds) Hydrocarbon Seal Quantification, NPF Special Publication 11, 109–125 Elsevier, Amsterdam Yielding G, Freeman B & Needham T (1997) Quantitative fault seal prediction: AAPG Bulletin, 81, 897–917 3 Ehrlich et al. TSK 11 Göttingen 2006 Figure 5: 3D grid of the east flank, Njord Field 4