ABSTRACT
Phase envelopes are routinely employed by reservoir engineers for fluid characterisation. These envelopes are controlled by reservoir fluid composition, pressure and temperature. As a result of increasing source-rock maturation, fluids with decreasing molecular weights and densities and increasing gas-to-oil ratios (and hence different phase envelopes) are generated, which are thus linked to fluid history. In addition to their importance for exploration, charge models can play a key role in constraining reservoir models and optimising field development, particularly when pressure-volume-temperature (PVT) data are properly integrated with fluid geochemistry. Two contrasting scenarios of fluid phase evolution from two different fields are presented, and their relations to charge analysis and reservoir models are discussed. The first example discusses the identification, based on hydrocarbon geochemistry complemented by overlapping modeled phase envelopes, of compartmentalised filling cycles in what was initially considered a single oil-rimmed gas accumulation. The second example presents an opposite scenario where two wet gas accumulations 20-km apart laterally and 400-feet average depth difference appear to represent a single more-expansive accumulation spread over areas of variable PVT conditions and reservoir qualities. The wet gas across both accumulations is characterised by a continuous phase evolution pattern that shrinks systematically (cricondentherm shifts to lower temperature and cricondenbar to lower pressure), suggestive of phase fractionation of a charge of single maturity. The proposed gas distribution model represents a discovery of a hybrid conventional and unconventional (tight sand) system, with potential for basin-centered gas. These findings provided better understanding of observed and projected fluids, impacting the development and completion plans by locating new gas producers. A recent well drilled midway between the two accumulations indeed tested wet gas, confirming fluid connectivity. Future work will attempt to link the gas distribution model with seismic attributes.
