Grid Drilling below Data Resolution in a Waterflooded, Faulted, Fluvial Reservoir

The Mulussa F formation in many of Al Furat Petroleum Company's mature reservoirs consists of fluvial sand bodies cut up by faults with a range of throws below seismic resolution. Predicting waterflood efficiency in this 3D labyrinth is complicated. Sands in the low N/G (<35%) column of some 350 m in one field do not water out evenly or bottom-up. Instead, water breaks through in sands higher in the column, above unsupported sands, leading to loss of reserves when plugs are set. <br>This paper investigates which knowledge is required to predict sweep in the Mulussa and to forecast ultimate recovery of infill wells. Comparison to the actual resolution of seismic and well data shows that deterministic modelling of field development is not possible. Instead, geostatistical dynamic sector models were created, with explicit representation of sand bodies and subseismic faults. A small number of these models with particular parameter sets could explain historical field behaviour. The rewards of an ensemble of wells can be predicted using these particular models. <br>The formation is now being developed by a grid drilling campaign, with well spacing and orientation defined by the study. The first 7 wells of this campaign found rewards as prognosed.<br>