Thursday, October 9, 2014

Measuring the "maturity" of a hydrocarbon fluid

Geochemical analysis of a gas, condensate or oil can provide a lot of useful information about its origin. In the best cases this can include the type of organic matter in the source rock, whether it is clastic or carbonate, the depositional environment (terrigenous, lacustrine, marine, evaporitic) and whether it is has been altered by secondary processes (e.g. phase separation, biodegradation). In some cases, the presence of absence of certain marker compounds (e.g. oleananes, 24-isopropyl cholestanes) places broad limits on the age of the source rock.

Many geochemical reports also speak about the "thermal maturity" of a fluid. It is not always clear what is meant by this term but generally it seems to mean the maturity of the source rock from which the fluid was expelled. There are several kinds of geochemical maturity indicators, but most are either based on systematic relationships among gas isotopes (carbon or hydrogen), on ratios formed from isomeric compound pairs having differing thermal stability (e.g. sterane or methyl phenanthrene isomers) or on the concentration of thermally resistant compounds such as the diamondoids. Fluids with more of the thermally stable species or isomers are considered to have been expelled from a source rock at a more advanced stage of kerogen conversion.

There are several fundamental problems with the concept as described above - but that is a topic for another day. For now, I would like to describe just one method of saying something about the "maturity" of a fluid sample. This is a method based on the relative abundance of two tetramethyl benzenes and two tetramethyl naphthalenes. These are mono- and fused diaromatic compounds containing 10 and 14 carbon atoms respectively. These compounds have boiling ranges straddling the range where we most commonly see a change from a gas-condensate fluid to a volatile oil. For convenience we call them the "semi-volatile aromatics" or "SVA". The method has been in use in a few companies for several years but has never been published, except in the form of a poster presented by Ben Van Aarssen at the International Meeting on Organic Geochemistry (IMOG) in 2007. A description of the method can be found in the book of abstracts for that conference. At that time, the SVA "maturity index" had not been calibrated to an equivalent vitrinite reflectance for the expelling source rock. There is also an SVA source index - derived from the same compounds and analysis - which gives an indication of degree of "algal" vs. "terrigenous" organic matter contribution to the source rock.

Some aspects of the SVA method are described below. I hope to get the time to write it up properly over the next year (in conjunction with Ben Van Aarssen). The bottom line is that it gives (we believe) a more robust indication of relative fluid maturity than most other methods. This arises from (a) a clear separation of the source and maturity effects (b) a basis in compounds which are present in reasonably high concentration in both gas condensates and oils and (c) a response range that extends from the earliest onset of fluid expulsion to well beyond the point at which gas becomes the dominant product. Calibration to absolute source rock maturity is more problematic - and may not even have a simple physical meaning,  as described in the last few few figures. This is an issue of particular relevance to shale oil/gas plays where - as per the previous blog entry - we often want to know how much of the fluid we produce was generated locally as opposed to being migrated in from more mature section elsewhere.

Happy Modeling :)

Saturday, October 4, 2014

How fast can oil and gas migrate in the Eagle Ford shale ?

In view of increasing evidence for migration in some of the shale plays (since it has become common practice, should we still call it unconventional?), I wanted to try to  estimate the migration rates in the shales, using he Eagle Ford as an example.

Darcy's law states that at 1 Darcy (D) permeability, a fluid with 1 centipoise (cP) viscosity will flow 1 centimeter per second under a pressure gradient of 1 atmosphere (14.7 psi) per centimeter. That translates to about a rate of 0.000073 ft/sec at a gradient of 1 psi/ft.

At typical reservoir conditions, gas viscosity is around 0.01 cP, and the permeability of the shales are known to range between 1 nano darcy (nD) to 10 micro darcies (μD) or more. At hydrostatic conditions, the buoyancy gradient for typical gas (static water gradient is 0.43 psi/ft  and 0.13 psi/ft for gas) is 0.3 psi/ft. There are about 3.16 x10 13 seconds in a million years. The Eagle Ford dips at about 30 meters per kilometer, meaning the lateral buoyancy gradient is about 0.009 psi/ft. Given all these, gas migration rate up dip along a 1 μD permeability zone would be

Migration rates (ft/my) = (0.000073*0.009) (psi/ft) x 10-6 D x 3.16x1013 (sec/my) /  0.01 cP = 2082 ft/my (635 m/my) **.

The gas window part of the Eagle ford is quite over pressured, and the pressure gradient may be around 5000 psi over 50 km (=0.1 psi/ft, or about 10 times higher than the buoyancy gradient) which can result in 10 times the above migration rate. Higher rates would be possible in zones if the permeability is 10 μD or higher. The Eagle ford has been generating gas for more than 10 million years so this could add up to fairly long distance migration. Imagine what would happen in the Permian basin where the source rock was mature since 200 million years ago.

Since the incline (dip) is (30 m/km) and the permeability anisotropy (kx/kv) is is at least 100 to 1000 times, the process still favors lateral migration rates by 3 to 30 times. Perhaps lateral migration in the shales is more common than we have thought?

This could also partly explain some of the higher than expected GORs we see at relatively low maturity. Some of the gas condensates may have migrated up dip to mix with the oil. Since oil viscosity is 100 times that of gas, and a one third in buoyancy gradient - so it would migrate 300 times slower. The vast difference in migration rates in the oil window and the gas window creates a super sized natural dynamic trap. The difference in over pressure gradients may also provide additional help in this regard.

** note an earlier version of this post had a mistake in conversion that resulted in much higher rates. Thanks to Andrew for finding the mistake!