The Day Job.....

With the quality of writing on offer, many of you might helpfully suggest that I don't give up the day job. Don't worry, I don't intend to. Instead I thought my blog this week would be a good chance to give you a flavour of what my day job actually entails.

My motivation for doing so: I've just had a paper accepted (well almost, a few minor corrections to do) to Geophysical Prospecting, one of the major international applied geophysics journals. Link to the paper here, in its almost finished form (it'll be copy-edited and re-formatted by GeoProsp before they publish). It's also a good opportunity for me to practice explaining my research to a more general (although still highly intelligent, and rather good-looking, I've no doubt ;-) ) readership.

In this paper, I develop a method to improve our ability to image the fractures formed during hydraulic fracturing for shale gas. This is important for both the operators, who want to know as much about the fractures as possible to maximise production, and for regulators, who want to make sure the fractures create will not provide a pathway for fluid contamination. So how does my method work?

You'll remember in this post I talked about how geophysicists deploy geophones in boreholes to listen out for the popping and crackling of the rock as it fractures, and we use the recorded data to identify where the fractures are going. This is pretty standard, the bread-and-butter for many service companies who offer variants on this technology.

The new bit, where my colleagues at Bristol and I come in, is to realise that as the seismic waves travel from the source (the fracture) to the geophone, they will be travelling through previously fractured rock. Therefore, polarisations and arrival times of the recorded waves will be controlled by the properties of not only the rock, but also of the fractures. In particular, we measure splitting of shear waves. When the S-waves move through fractured rock, they become split into faster and slower waves with a 90 degree polarisation difference. We measure the fast wave polarisation, and the delay between fast and slow waves. These measurements can tell us about the properties of the fractures.

In particular, in this latest paper we show that the S-wave splitting measurements can tell us about the ratio of normal to tangential compliance of the fractures. In layman's terms, that is the ratio of how easy it is to squeeze the fractures versus how easy it is to slide the fractures. Lab experiments have seemed to show that the presence of proppant (the sand particles injected to 'prop' the fractures open) will increase the normal to tangential ratio, so if we can see changes through time, this can tell us where the proppant has gone, allowing us to predict where the greatest flow will come from during production (the better 'propped' a fracture is, the better it will flow).      

The key results picture is below:

This shows the frack-job through time. The bottom panel shows the fluid injection rate (blue), the proppant concentration (green) and the resulting microseismicity (black) through time. The frack-job is conducted with 4 phases - 3 initial injections to create fractures, and a final phase where proppant in introduced. The upper panels show the results of our S-wave splitting measurements. Of most note is the 2nd panel, for Zn/Zt - the normal to tangential compliance ratio. You can see it is low during the initial fracture stages, but after 13:00, when proppant is injected, it increases to values almost as high as 2.

Like all good scientists, we are cautious in our conclusions. We think that this increase is most likely showing the proppant entering fractures. However, further work is needed to verify our findings, and to see if this method can work on other frack-job datasets. However, the initial findings are promising. If this method takes off, it'll allow operators to gain a better understanding of the fractures they create during fracking operations.

Anyway, I hope you've stayed with me up until this point, dear reader. Any questions - do pop them in the comments below. Talking about one's actual research to a general audience can be quite tough, but I hope I've been able to convey what my day-job entails. I promise to talk about something more exciting next time.