From Bureau of Economic Geology, The
University of Texas at Austin (www.beg.utexas.edu).
For more information, please contact the author.
Bureau Seminar, October 24, 2008
Euclid Visits an Outcrop
Steven Bryant
Department of Petroleum and Geosystems Engineering The University of Texas at Austin
It is axiomatic that the natural processes that build sedimentary rocks also determine the engineering transport coefficients of those rocks  permeability, resistivity and so on. But relating those coefficients quantitatively to geologic processes has challenged scientists for a century. On the one hand, the physics is well known (e.g. the NavierStokes equation for fluid flow) and often simple (e.g. Ohm's law for DC electricity). On the other, the boundary conditions for the equations are intractably complicated. A million pores can occupy a cubic centimeter of rock, and every pore will have rough walls, a different, irregular shape and different number of connections to neighboring pores. Thus quantitative, a priori prediction of rock properties was the province of dreamers for decades. Even if one could measure or describe those walls, how could one use that information in a computation? Correlation and empiricism became the mainstay.
Certain geometric idealizations, of which Euclid would be proud, have enabled remarkable progress. The essential idea is to represent the result of any particular process (e.g. sedimentation) in terms of a grainbased, geometrically determinate model. Different processes will affect the porelevel geometry in different ways. With some mathematical idealizations that capture the geometry and topology of the pore space in the model rock, it is possible to solve the relevant physics equations. In this fashion the quantitative effect of, say, epitaxial cement, on trends of properties (e.g. permeability vs porosity) can be predicted accurately for conventional clastic rocks.
It should not be surprising that correlations and empiricisms for formation evaluation or reservoir quality prediction in conventional reservoirs do not work so well for unconventional rocks. Since unconventional reservoirs are now the foundation of the domestic natural gas industry, the question arises: can we develop grainbased, geometrically determinate models of things like tight gas sandstones or mudrocks? Such models could provide much needed insight into the petrophysical behavior of these reservoirs.
In this talk I summarize what Euclid saw at the outcrop and how it can be used to predict trends of rock properties from a geologic description. I then discuss several extensions of this idea which provide insight to geologic controls on producibility in tight gas sandstones.
