Compositional modelling of saturated two-phase hydrocarbon reservoirs is a difficult task. When both oil and gas phases are present, the characterization of one phase may not be the same as the other phase. Proper characterization will, therefore, become an issue. In this paper, we examine the compositional data of four two-phase hydrocarbon reservoirs. An attempt is made to characterize the reservoir fluids in both the gas cap and the oil column. In addition to the characterization of the plus fractions, we also review reservoir fluid sampling and validation.
Equations of state are commonly used in compositional modelling of both gas condensate and oil reservoirs. A key to the successful application of an equation of state (EOS) is the proper characterization of the heavier hydrocarbons.
Traditionally, the characterization has focussed on one single hydrocarbon phase, either oil or gas. For oil reservoirs, the effect of injected gas on the mass transfer with the in situ oil for the characterization is accounted for by including laboratory swelling data in the match between calculated values from the EOS and measurements. Similarly, for gas condensate reservoirs, the constant composition expansion, constant volume depletion, and the mixing of injected lean gas and the in situ gas condensate PVT experiments are believed to be adequate for the characterization.
When a reservoir contains both the oil column and the gas cap, the characterization of the reservoir fluid for the EOS should describe the phase behaviour of both the oil and the gas phases. Generally, there is a significant difference in the molecular weight (and other properties) of the plus fractions of the oil and the gas phases that are in equilibrium. Therefore, different characterizations may become necessary for each phase. In other words, one may need to characterize the heptane-plus fraction of the oil column to be different from that of the gas cap.
In this paper, we will study the fluid characterization of four different two-phase hydrocarbon reservoirs of significant size. The main objective is to develop a simple practical approach for the study of two-phase reservoirs. A large number of PVT samples for the gas cap and the oil column of these four reservoirs will be used in this study. Since many PVT samples may not be representative, we will present a discussion on sampling and its validity.
Reservoir fluids, whether sampled at the bottomhole or at the wellhead, should be representative of the fluid in the reservoir. The procedures to obtain representative samples have been addressed in the literature. In 1954 and later, Reudelhuber(1-3) emphasized the need for proper well conditioning to ensure representativeness of samples from oil reservoirs. He recommended minimizing the differential pressure for bottomhole samples by restricting the flow rate. This procedure applies to undersaturated reservoirs with high permeability. In saturated reservoirs and in reservoirs with high drawdown, the flowing pressure may be less than the bubblepoint pressure. Therefore, the in situ reservoir fluid has to be reconstituted from the wellhead samples; the well producing gas-oil ratio (GOR) should be stabilized.