Reservoir Architecture in the Mars Field, Deepwater Gulf of Mexico, USA: The Implications of Production, Seismic, Core and Well Log Data

Reynolds, Tony, BP-Amoco, 501 Westlake Park Blvd, Houston, TX 77079

Abstract

Mars Field was discovered in 1989 in 2,900' of water some 56 miles south southeast of the mouth of the Mississippi River. Production commenced in 1996 and since that time the field has produced via natural depletion over 100 million barrels. Today (November, 1999), the field produces from 15 wells completed in eight reservoir horizons at a rate of around 180,000 barrels of oil equivalent a day. All the reservoir sands have been deposited in deepwater in an intraslope basin. The need to understand pressure data collected during production and the requirement to predict two-phase flow and evaluation of a waterflood scheme has prompted review of the architecture of major reservoir horizons. The largest of these reservoirs, the 100m thick Yellow horizon, is discussed here.

Appraisal mapping revealed two key geometries: (i) a sheet sand, the Lower Yellow, which was described in two parts as a lower "amalgamated sheet" and an upper "layered sheet", and (ii) an overlying channel complex, the Upper Yellow. Al though initially in pressure equilibrium, it was thought that an intervening shale would compartmentalise these intervals into two distinct reservoirs during production.

The present study has refined the appraisal view to reveal eight first-order architectural elements, AE1 to AE8, defined and mapped on the basis of seismic data. These architectural elements have been divided into second order, sub-seismic elements that can be correlated on the basis of wireline log and core data, and characterised in terms of facies proportions, facies successions, reservoir properties, and likely inter-well architecture.

Key findings are that (1) small areas of contact between first order architectural elements have allowed significant pressure communication throughout the Yellow reservoir; (2) pressure data do not indicate low vertical permeability in interbedded sand and shale successions previously described as a layered sheet, and an alternative inter-well view of a low net to gross channelised sheet is suggested; (3) aquifer influx has not been observed, suggesting that the original field-wide OWC is perched and that the true OWC is deeper; and (4) the bulk of the reservoir drive energy is due to time varying compaction.