Practical Kinetic Modeling of Petroleum Generation and Expulsion

John G. Stainforth
Shell International E and P, Houston, Texas

Twenty years ago, understanding and modeling of primary migration and expulsion was regarded as one of the outstanding problems of petroleum geoscience. But the industry has apparently more or less given up on research into this area, except for a few notable exceptions. As a result, the standard industry models for petroleum generation are seriously hampered by:
     1   Unsuitable laboratory pyrolysis methods for studying hydrocarbon generation
     2   Inappropriate mathematical models for petroleum generation
     3   Inappropriate methods to derive kinetic parameters from the laboratory data
     4   Inadequate models for primary migration and its coupling with petroleum generation
     5   Little experimental work to determine the parameters of primary migration
     6   Insufficient use of natural subsurface data
These problems are present in the standard open-system pyrolysis method for deriving hydrocarbon generation kinetics. The experimental method is inaccurate for compositional kinetics; the parallel first-order model is not a good representation of reality; the mathematical way of fitting the data is flawed, leading to false compensation effects between activation energies and frequency factors and poor extrapolations to geological conditions. Simple switch or adsorption models of expulsion are insufficient to describe the residence time of species in the source rock. Yet the residence time controls the thermal stresses to which the species are subjected for cracking to lighter species.
We have built on the Shell Genex model since 1988 in which primary migration is regarded as rate-limited by the slow diffusion of the petroleum through and out of the kerogen and recently have improved the handling of the heavy (SARA) species in the model. Two roles are crucial: (1) the free radicals in the cracking kinetics of these species, and (2) the evolving free volume of the kerogen in controlling their transport.
The result is a model of unprecedented ability to predict petroleum composition from different source rocks in different settings. By coupling it with well-calibrated thermal and trap filling models, and running these against a wide range of laboratory and geological conditions, we continually test the model. We check the predictions of the model against key measured data, such as the transformation ratio of kerogen; the evolution of the chemical composition of the kerogen; the distribution of oil and gas in basins (strongly overprinted by trap filling and leaking effects); the properties of the petroleum such as the API and GOR gravity of oils, and the compositions of gases.

AAPG Search and Discover Article #90066©2007 AAPG Hedberg Conference, The Hague, The Netherlands