6 Febbraio, 2013 14:00 in punto

Mechanisms of Petroleum Trap-Filling & Compositional Implications

J. Stainforth, Geological/Geochemical consultant

Aula B5.2 Dipartimento di Matematica Politecnico di Milano

The problem of compositional grading in petroleum fields has received renewed interest in the last ten years, especially in the context of reservoir compartmentalization, and there are now two rival hypotheses to explain the phenomenon. The old one is gravitational segregation acting on petroleum that was well-mixed by the trap-filling process. The newer one is that the grading merely reflects the filling history of the trap with little mixing in the trap (Stainforth, 2004).
These two hypotheses have become to be regarded as opposite end-members.
In this presentation, I outline the theories behind the hypotheses and then test them against
classic field cases showing strong compositional grading. The petroleum distribution in all these
field cases can be explained entirely by the trap-filling hypothesis coupled with the petroleum system logic of source rock kitchen, reservoir volume and shape. I also show how the same cases can be fitted with a model of molecular clumping (mainly of asphaltenes), as proposed by Mullins and co-workers at Schlumberger, but only by making some rather arbitrary assumptions.
An added difficulty with the latter hypothesis is that the cases presented straddle a complete
spectrum of petroleum from heavy tarry oils to light gas-condensate in which the latter contain
virtually no asphaltene.
Neither of these hypotheses, as it stands, sits easily with the diffusivities of petroleum
compounds. These are generally too low for petroleum fields to have achieved complete
equilibrium between gravity and diffusion in the time since trap-filling began. But they are large
enough that there should generally have been some noticeable mixing in that time, especially in the older fields: curiously, though, there is no indication that older petroleum accumulations are better mixed than younger ones.
Finally, I present a way out to this quandary, which reconciles these two rival hypotheses that appear diametrically opposed to one another. This requires the addition of a new conjecture,
which is very simple, yet seems to eliminate all the difficulties posed by the observations.