Session 3 - 17 July Geostatistics and Geomechanics

Kieran Blacker - For Dan - can you predict a-b values using downhole dynamic properties (from sonic logs)?

It would be an extremely useful thing to be able to do and save time-consuming lab measurements. Given that a-b appears to be related to clay content, then some combination of sonic logs and gamma ray logs might be used to determine the clay content and give an idea of a-b. The sonic logs measure the dynamic elastic properties, and often the presence of clay reduces the compliance and this can show up in the logs. Gamma ray logs will be sensitive to ‘radioactive’ clays such as illite and smectite.

Tiago Alves - Dan. Let's think pore-fluid pressure. Would smaller fluid pressures (smaller than the original pressures pre-drilling) correlate with enhanced frictional stability? Or is it clay content the sole control?

The effect of pore-fluid pressure in a-b is not certain. There are surprisingly few experiments on common fault gouges at a wide range of conditions (P, T, Pf, shear strain rate, etc). Scuderi and Collettini (2016, Scientific Reports) did look at the frictional stability with varying pore-fluid pressure and found changes, but these experiments were only on carbonate gouges. More investigation on this topic, and on a wider range of common gouges, is needed.

Helen Lewis - The carbonate basin shape appearing in your analytics is a really nice illustration of the "whole system" behaviour. It is also unsurprising once pointed out. Do you know if anything like it was observed by the geological workers?

The carbonate high was observed by the geologists, but it’s relation to the modelled property was not introduced into the model initially as it was not apparent. Furthermore, the initial model was a single rock type while the update model had multiple rocktypes distributed globally though. Partial correlation analysis helped to identify the relation and a neural network was used to map the region attributed the relation.