The representation and characterization of the heavy traction of reservoir oils are crucial for accurate predictions of multiple-contact gas injection processes using cubic equations of state. The mechanisms in most rich-gas processes were recently identified as combined condensing/vaporizing phenomena, which are more complex than the traditional condensing phenomena. Similar interpretations apply to CO2 processes. To capture these phenomena, current techniques involve the use of a sizeable number (4 or more) of pseudocomponents to represent the heavy fraction (e.g. CO2) making field-scale simulation prohibitively expensive.
A technique is proposed to represent the heavy fractions by a small number of components (one or two), while maintaining accurate predictions of condensing/vaporizing phenomena. The highlights of the technique are:
the allowance for the critical temperature of the heaviest pseudocomponent to vary with composition and
the reproduction of the critical point on pressure-composition diagram.
Test runs show that the proposed method gives predictions of CO2 and rich-gas multiple-contact processes with 4 to 6 components with the same level of accuracy as with 14 components.
Rich-gas and CO2 injection arc import.1nt enhanced-oil recovery processes, where multiple contacts between the injected gas and the reservoir oil result in efficient displacement of oil. The traditional interpretation of the mechanisms occurring in rich-gas drive corresponds to a condensing process where the intermediate components condense from the injection gas and enrich the reservoir oil towards the point of miscibility (Stalkup. 1984). However, recent studies indicate that the mechanisms in most rich-gas drives correspond lO a combined condensing/vaporizing phenomena or a liquid-extraction phenomena (Zick, 1986; Mansoori and Gupta, 1988; Novosad and Costain, 1988, 1989). Thus, the phenomena which occur in rich-gas drives are in many ways similar to those in CO2 processes (Novosad and Costain, 1989).
The above phenomena arc usually modelled with a cubic equation of state (EOS). With this approach, the number of components representing the hydrocarbon system has a large impact on accuracy and computational cost. The computational costs arc an overriding factor if EOS calculations are used in a compositional simulator to predict reservoir performance. Thus, it is essential to reduce the number of components in field-scale simulation, while still preserving the accuracy of the phase behavior predictions.
The most important factor in reducing the number of components is the representation of the heavy fraction (e.g. C6.) of the reservoir fluid. A number of methods have been proposed to split the heavy fraction into several pseudocomponents and to characterize these pseudocomponents. This characterization involves the estimation of component critical properties and acentric factors. Some of these techniques are described in Whitson (1983), Pedersen el al (1984), Li el al (1985), Coats and Smart (1986), Wu and Batycky (1988), and Newley and Merrill (1989). These techniques usually require a sizeable number of pseudocomponents to represent the heavy fraction (four or more) in order to obtain accurate predictions of the phase behavior or multiple-contact processes.
