Turbidite reservoirs of the Sele Formation, Central North Sea: geological challenges for improving production

Abstract

Turbidite reservoirs of the Sele Formation in the Central North Sea produce from fields such as Forties, Nelson, Montrose–Arbroath, Scoter, Pierce, the Gannet cluster, Guillemot A, Mirren and are under appraisal/development in fields such as Merganser, Phyllis, Starling and Blane. These reservoirs form part of the ‘Forties’ submarine fan system that was sourced from feeder channels in the northwest and west. Blanket 3D seismic coverage tied to wells shows that nearer their updip sources the turbidites are thicker, higher in net:gross, and more channelized. Downdip, the turbidite reservoirs are thinner bedded, finer grained and with depositional architectures that are characterized by stacked lobes and minor channels controlled by accommodation space and salt movement. In updip fields, such as Nelson and Forties, production is mature. High-quality reservoir channel sands are mostly drained. Recovery factors of 60–62% are achievable by application of water injection and 4D seismic technologies. The remaining challenge in these reservoirs is to (1) find by-passed oil in the channel facies and (2) to identify/model oil remaining in non-channelized facies, such as channel margin and interchannel facies. Portions of the channelized reservoirs have not been swept because of non-uniform water ascent and breakthrough or shale barriers. Non-channelized reservoirs have not been swept because their sands are thin bedded, interbedded with shales and of lower reservoir quality. Remaining oil zones are not correlatable between existing wells, emphasizing the uncertainties of lateral heterogeneities away from the wellbore and the importance of static reservoir modelling techniques to properly model, identify and plan infill targets with multiple realizations. Higher recovery factors (possibly as high as 70%) may be achievable with higher resolution seismic interpretations and reservoir models, application of enhanced oil recovery and sharpshooter drilling techniques and improved water management. In downdip fields, such as Scoter, Merganser, Pierce and Guillemot A, production is not as mature and wells are not as abundant. Consequently, development planning is more dependent on high-quality static and dynamic reservoir models to predict the interwell reservoir character, volumetrics and performance. In general, turbidite reservoir sandstones are thinner and finer grained. The thickest sands may be more correlatable as lobes/sheets. Permeability is lower than in updip reservoirs of equivalent porosity (tens of mD as opposed to hundreds or thousands of mD), probably due to reductions in mean grain size and textural maturity and a corresponding increase in detrital clay matrix. The greatest challenges lie in modelling lateral changes in net:gross, bed thickness, facies (especially shale bed architecture) and cementation. Where depositional slopes were oversteepened by the growth of salt diapirs, slumps and slides introduce lateral heterogeneities and reduce predictability in the reservoir section. Multiple faults radiate from such salt diapirs and may compartmentalize the reservoir. These effects may be mitigated by the construction of robust conceptual models of reservoir accumulation and architecture, supported by relevant field analogues and detailed biostratigraphy which may delineate multiple widespread marker horizons. For liquid hydrocarbons, recovery factors are typically 40–45%. Gas recovery factors are projected to be 60–65%.