What can we learn about deep-water reservoirs from fields currently in production? In frontier and immature deep-water provinces, where well data are scarce, seismic coverage is patchy and risks are high, explorationists are increasingly turning to analogue provinces and fields for guidance in predicting trap configuration, reservoir character and production performance. Development strategies can also benefit from the experiences - and mistakes - of other operators in deep-water fields. Such an empirical approach has been adopted by C&C Reservoirs Ltd in an ongoing global review of the exploration and development characteristics of nearly 100 deep-water fields. Here, 'deep-water' denotes the environment of reservoir deposition rather than the bathymetry of the present-day field location.

The studied fields have a wide geographic distribution, with 40 from North America, 34 from the UK and Europe, 10 each from Latin America and Africa and three from the Far East. A total of 30 different basins are covered that are grouped into three main types: passive-margin basins (e.g., Gulf of Mexico, Campos and Lower Congo basins), transform-margin basins (e.g., Los Angeles, Bredasdorp and Sabah basins), and intracontinental basins (e.g., Central North Sea, Gulf of Suez and Carnarvon basins). By comparing the key characteristics of deep-water fields in these three basin types, important exploration and development lessons can be learned that can be applied to prospects and new discoveries.

Focusing initially on deep-water fields in passive-margin basins, recurring patterns of trapping and reservoir architectural styles are recognized. Overwhelmingly, passive-margin fields are contained in combination structural-stratigraphic traps, with 80% showing some element of stratigraphic closure, principally depositional pinch-out, which generally tends to diminish in importance upslope with increasing structuration and sand occurrence. Basinwide mobile substrate (salt or shale) thickness controls regional deformation style, the likelihood of structural trapping and reservoir distribution. Among the studied fields, six principal deep-water reservoir types are recognized, whose abundance varies across the continental slope: (1) channel-dominated reservoirs, which are commonest on the middle slope; (2) sheet-dominated reservoirs - middle-lower slope; (3) leveed-channel reservoirs - upper slope; (4) canyon-fill reservoirs - upper slope-shelf; (5) debrite reservoirs - upper slope; and (6) contourite reservoirs - lower slope-abyssal plain.

The studied fields indicate that development strategies and recovery efficiencies in deep-water reservoirs are controlled primarily by hydrocarbon type, sand-body connectivity, reservoir permeability and natural drive mechanism. The presence of strong aquifer drive in conventional-oil fields leads to higher recoveries and may obviate the need for water injection facilities that are extremely costly in deep-water locations. Deep-water reservoirs with solution-gas drives and those with primary gas caps yield lower ultimate recoveries. Aquifer drive tends to be weak where individual sand volumes are small, faulting is intense and/or reservoirs are, or have been, deeply buried. The high porosities (commonly >25%) of most deep-water reservoirs, particularly in passive-margin basins, yield high well potentials, but control of production of these poorly consolidated sands is often key to optimizing well recoveries, especially where oils are viscous. Although few extended-reach horizontal wells have so far been drilled, there is no technological barrier to their drilling in ultra-deep waters and they are likely to prove essential for maximizing recovery from variably interconnected deep-water reservoirs.