Saadat Mohammad Hossein Zadeh - National Iranian South Oil Company (NISOC), Ahwaz, Iran
Shahin Kord - National Iranian South Oil Company (NISOC), Ahwaz, Iran
Siyamak Moradi - Petroleum University of Technology, Abadan, Iran
Leila Rahimi Pour - National Iranian South Oil Company (NISOC), Ahwaz, Iran

The variation of fluid composition with depth in an oil reservoir is called compositional grading e.g. as depth increases the fluid properties changes.prediction of phase behavior and accurate characterization of reservoir fluids In these reservoirs could be achieved by models that take this phenomenon intoaccount. Prediction of compositional grading make it possible to check the reliability of new fluid samples, estimation of fluid in place, initialization of reservoir simulators andconsideration of production alternatives [1]. In this study, three giant oil reservoirs with different behaviors and differentconditions and characterizations were studied. For all of these reservoirs, complete PVT reports were selected to perform fluid characterization. All experiments weremodeled by commercial PVT software. Several isothermal and non-isothermal compositional grading models weredeveloped to explain the change in reservoir fluid properties. All the models examined for the selected reservoirs to find which of the models can best explain thefluid property change in the reservoir. The most accurate reservoir fluid valid samples with the best match between modeland experiment was used to predict the reservoir fluid composition and its corresponding PVT properties at different bottom hole locations using the mostreliable compositional grading model. Results show that for reservoirs where fluid become heavier with depth, Isothermal model can describe fluid property change in the reservoir. For reservoirs where fluidbecomes lighter as depth increases, it concluded that this is not a thermodynamicphenomenon and has different reasons. For reservoirs with little or no change in fluid property with depth, Non Isothermal models can best describe the phenomenon.

Compositional grading, saturation pressure, simulation, PVT, oil reservoir, gravity, convection, diffusion