“Reservoir faulting, fracturing and microseismicity”
– Bristol, 20-24 June 2016 –
The geothermal field in Basel, the oil and gas reservoir in Groningen and the fluid injection site in Oklahoma, are just few examples of how human-induced seismicity is nowadays a burning topical issue. The 3rd CREEP Short Course held in Bristol consisted in five days of lectures, practicals and a field trip at Kilve beach, in which Michael Kendall and James Wookey, in cooperation with André Niemeijer, James Verdon and Katrin Plenkers, provided us with the most cutting-edge tools to cope with induced seismicity. Notably, in order to be able to assess the hazard and the risks associated to the industrial activities and to find ways to mitigate their damaging effects, a multidisciplinary approach that couples the principles of fault and fracture mechanics with seismic data has been proposed. Following this reasoning, A. Niemeijer introduced us to the outstanding world of faults and fractures: first he explained the main criteria for rock failure and the physics behind these (presence of microcracks). Then, he upscaled these principles to faults, describing the different types of Andersonian faults, their architecture, the way they can be reactivated and the role of fluids and mineralogy in fault weakening. Last, to unravel the fault motion, the concept of friction was brought up and the constitutive equation known as Rate and State friction law discussed in detail.
Fig 1. a) Material exhibiting velocity weakening frictional behaviour (a-b<0): potentially unstable slip; b) modified after M.M.Scuderi et al.: analogue of the spring-slider system to a potentially unstable fault patch in nature.
During the 1st practical, we mainly focused on the frictional behaviour of velocity weakening materials (a-b<0), by using a matlab script that simulates a spring-slider system governed by rate and state friction law.
On the second day, we attended introduction courses on seismology given by Michael Kendall and James Wookey. The first lecture covered an overview of the history of seismology, as well as more recently developed fracking techniques and induced seismicity monitoring. This part of the course focused on the UK context and present some examples of seismic surveys carried out near stimulation areas. Shale gas reservoirs represents a potential transition energy source, as they release less CO2 in the atmosphere than coal. Otherwise, the low permeability of these reservoirs tends to limit the exploitation rate of the gas. This is why the rock frame is artificially fractured by injection of a slurry, a mixture of water and chemical component which open new cracks or reactivate pre-existing ones. Monitoring the induced seismicity is important in order to prevent catastrophic events. The surveys also allow the characterisation of the crack sets around the area of interest. This point is crucial as excessive fracturing could potentially lead to the contamination of aquifers located near the stimulation zone.
In the second course James Wookey gave an introduction to seismology, which also covered aspects of processing seismic data. Firstly, we learned how to deal with time series data in general. The advantages of filtering and displaying the data as a spectrogram, which is a representation of the time series in the frequency domain, were shown.
The next two lectures focused on the actual seismology and how it can be used to discover the inner Earth at all scales and depths. The concept of anisotropy, which is the velocity dependency of a seismic wave on its direction and polarisation, as well as shear wave splitting were introduced. Seismic reflectivity was presented by showing the case of a two-layered medium, having two different seismic impedances. The reflection coefficient describes the ratio between the amplitudes of the incident and reflected wave.
During the practical we had the opportunity to apply our new knowledge, where we looked at three seismograms with different epicentre distances and amplitudes (Teleseismic earthquake, regional earthquake and a microseismic earthquake). For each dataset we processed the data and picked the p-wave and s-wave arrival. By taking the fast Fourier transform of the data, the seismogram could be shown in the frequency domain. The practical gave us an idea of how to process and interpret a seismogram.
Event detection and location practical For the last practical of the course, we were given the roles of geophysicists at the geophysics service company CREEP Geophysical Ltd. We were hired by BigFrack Corp, a large unconventional gas company, who uses hydraulic fracturing in the Bakkcellus Shale, a North American shale play, to extract gas. We are provided with the position of the horizontal well and 17 monitoring stations around it, as well as 30 minutes of raw seismic data from the stations. The first thing we learn is how to filter the data in such a way that we can more easily isolate a possible event. Then, to locate events, an automated event detection algorithm is used, a STA/LTA (short-time average/long time average) algorithm. Once we have found an event which occurs on all (or most) traces, we locate it by picking P- and S-wave arrival times. With this information we can obtain the location of the event, to see if it could be possible that this earthquake was caused by BigFrack Corp. And in fact, this does seem to be the case.
Now the question remains: What is the magnitude of this event and can they continue with their operations? We were provided with a local magnitude scale for the region:
ML = log10(A) + 1.14log10(R) +0.0575R – 3.06
Where A is the maximum observed S-wave displacement at a given station (in nm), and R is the distance (in km) between the hypocentre and that station.
Fortunately, the magnitude of the located event is well below ML 1.0, so BigFrack Corp can continue their operations. They do receive a warning however, saying that they have to keep on carefully monitoring the seismic activity, to make sure that the seismicity doesn’t increase in the future.
Written by Elenora, Gianluca, Thomas and Philipp
