The Dolphin Field is located in the East Coast Marine Area, approximately 60km off Trinidad. It comprises a 5,000 feet succession of stacked, unconsolidated, Pleistocene shoreface-deltaic sandstones and mudstones lying within a three-way dip, fault closure. The field is complicated by poor seismic data due to gas attenuation, fault shadow effects, large velocity variations and noise in the overburden. As a result the use of seismic amplitudes for fluid determination is unreliable and fault interpretation problematic. Despite over 15 years of production, 5 exploration wells and 13 development wells there remains a high level of subsurface uncertainty. Specific areas of uncertainties include: zones of poor correlation as a result of large-scale sediment deformation; the configuration and sealing properties of both structural and syn-sedimentary faults; the effects of thin-bedded deposits (forming over 50% of the Dolphin reservoir intervals) on reservoir pay and connectivity; and the absence of all but one proven Gas Water Contact (GWC). Defining the range of these uncertainties has a key impact on the range of GIIP outcomes, the resultant development plan and subsequent reservoir management. This paper will discuss the challenges and impact of these uncertainties on the Dolphin Field. The uncertainties identified on Dolphin form the basis for uncertainty planning in the development of subsequent analogous greenfields also within the Greater Dolphin Area.

The Enterprise gas field is a discovery in Shell's operated acreage in the Columbus Basin off the east coast of Trinidad. It is comprised of two major fault blocks, with an exploration well in the smaller one and an appraisal well in the larger one. The larger fault block is further broken up by minor faults. The potential compartmentalization of this fault block is the major uncertainty in the development of the Enterprise field. The development plan called for one or two wells to be drilled in the Enterprise field. Detailed mapping of each of the minor faults and analysis of log and pressure data from the wells was used to determine how well connected the various segments are. Using relationships derived from a global database and Vshale logs from offset wells, a range of fault transmissibility multipliers was derived for each fault based on calculated shale gouge ratios and mapped fault throws. Results of the dynamic fault seal analysis were integrated with dynamic simulation and showed that using base case fault transmissibility multipliers, for all segments where there is reservoir-reservoir juxtaposition across the minor faults, there will be connectivity and the larger fault block can be drained by a single development well. Various combinations of well placements were tested against low, base and high case geological realizations and these were used to determine the optimal development scenario for the field.