Petroleum reservoir operations such as oil and gas production, hydraulic fracturing, and water injection induce considerable stress changes that at some point result in rock failure and emanation of seismic energy. Such seismic energy could be large enough to be felt in the neighborhood of the oil fields, therefore many issues are recently raised regarding its environmental impact. In this research we analyze the magnitudes of microseismicity induced by stimulation of unconventional reservoirs at various basins in the United States and Canada that monitored the microseismicity induced by hydraulic fracturing operations. In addition, the relationship between microseismic magnitude and both depth and injection parameters is examined to delineate the possible framework that controls the system. Generally, microseismicity of typical hydraulic fracturing and injection operations is relatively similar in the majority of basins under investigation and the overall associating seismic energy is not strong enough to be the important factor to jeopardize near surface groundwater resources. Furthermore, these events are less energetic compared to the moderately active tectonic zones through the world and usually do not extend over a long period at considerably deep parts. However, the huge volume of the treatment fluids and improper casing cementing operation seem to be primary sources for contaminating near surface water resources.
Due to the increasing demands on the traditional oil and gas resources, hydraulic fracturing became an important technology applied for enhancing production from hydrocarbon reservoirs, particularly the unconventional ones. Recently, the conjugated practices of horizontal wells with multi-stage hydraulic fracturing not only have increased the well productivity dramatically, but also lead to enormous increase in hydraulic fracturing. This increase approaches the size of massive fracture treatments carried out in the 1970 [
To successfully monitor a hydraulic fracture treatment, the location of the monitoring array, determination of an adequate velocity structure, and management of noise are important issues to be considered [
To capture good quality spectral responses with minimal signal interference in a microseismic record, the downhole geophones should possess sufficient sensor response, minimal tool resonances, and suitable frequency response. In addition, the acquisition system should enable sampling rate between 0.5 and 0.25 msec that corresponds to Nyquist frequency 1000 and 2000 Hz respectively. Seismic attenuation is usually encountered at high frequency signals from far events compared to the low frequency and near ones that are usually mitigated during signal processing. In most cases, microseismic monitoring utilizes downhole sensors, while near surface sensors are in some cases deployed. Being close to the source in borehole microseismic, the depth of the microseismic event is usually accurately determined, but surface arrays proved to be more efficient in determining the spatial location of hypocenter. Dynamic microseismic images are obtained as a live streaming to the fracture propagation using the time history of the microseismic activity [
It is well-documented that long-term injection of fluids into deep formations induces earthquakes [
Typical microseismic array deployed in vertical well (a) and horizontal well (b)
arguments have been raised concerning the seismicity associating hydraulic fracturing operation [
Drinking water could be contaminated during hydraulic fracking operations through hidden pathways connecting the producing horizon with shallow aquifer, casing-cement failure, and flowback water used during the treatment. Reference [
Faults are physically static if the in-situ stresses are creating enough frictional forces along fault planes. Fluid injection results in shear stresses within the rock by increasing the pore pressure and therefore weakening the rock fabrics. When the shear stress increases enough to overcome the in-situ stresses, the rock initiates a fracture followed by a slip or directly starts slip on a pre-existing fault plain, resulting in an earthquake.
Reference [
where:
µstatic is the static friction coefficient; Fnormal is the normal force.
Since microseismic records represent a graphical demonstration to stress decay, fracture geometry and growth behavior can be identified using standard earthquake seismology principles [
Reference [
where:
ρ is the density, Vs is the shear velocity, R is the distance from the receivers to the event, Ωo is the low-fre- quency amplitude of the displacement spectrum, and Fc is a radiation pattern factor. Since the hydraulic fracturing occurs within a small intervals of the producing horizon, the source-receiver distance {R} and radiation pattern factor {Fc} are most likely the influential parameters in Equation (2) to determine the Mo value.
To determine Ωo value for each event, the amplitude spectrum is plotted versus frequency of the microseism after correcting for attenuation. Then the corner frequency is graphically identified by the intersection of the power-law decay at high frequency with the line that approximately represents the low frequency amplitude (
Graphical determination of the corner frequency using amplitude spectrum and frequency after correction for attenuation
Reference [
where:
Kc is a constant, and fc is the corner frequency. The value of the constant Kc has been used to equal (~2.2) by Reference [
The seismic energy (E) released during the slippage along the fault plane is approximated by Reference [
Analogous to Richter scale of earthquakes, the moment magnitude (Mw) is a more convenient value to represent seismic moment and/or seismic energy that is obtained by:
where Mo and E in this equation are expressed in dyne-cm.
The slip displacement and area can be determined as a function of the seismic moment by:
where:
μ is the shear modulus of the rock (typically 2.2 × 106 psi for shale), d is the slip distance, and A is the slippage area.
In the present work the calculated moment magnitude from different unconventional reservoirs is subjected to statistical analysis and plotted against various parameters related to the reservoir (e.g. depth) and treatment parameter (e.g. injection rate and volume). This helps understanding the capability of hydraulic fracturing operation to disturb the surface and subsurface environment at the vicinity of stimulated wells.
Thousands of fracturing stimulations have been monitored using a microseismic technique, in which broad variation in the magnitudes of the recorded seismicity is documented. Based on the estimated microseismic magnitudes calculated by Brune’s method [
The plot of number of events versus magnitude for seismic data typically follows a power law distribution described by the Gutenberg-Richter relationship, (logN = a − bM). However this plot for both Barnett Shale and Cotton Valley reservoirs data set (
microseismic magnitudes calculated by Brune’s method plotted versus depth in Barnett shale gas reservoir, Data from Reference [1]
994 microseismic events recorded by two monitoring well at depth between 2756 and 2838m in Cotton valley sandstone gas field, East Texas. Data from Reference [24]
Plot of the number of events versus magnitudes recorded in depth intervals 1170 - 1975 m 1975 - 2200 m and 2200 - 2750 m showed in upper, middle, and lower respectively for the data shown in Figure 3 of Barnett Shale (Data from Reference [1] )
tion, the high number of events cluster towards the right-hand side (high Mw values,
The injection rate and volume are expected to be influential factors to microseismicity, but could be the most argumentative issues of hydraulic fracking stimulations especially, handling and treatment of the flowback water. To distinguish the relationship between seismicity and both injection rate and injection volume associating hydraulic fracturing stimulations, comprehensive data sets from Barnet shale [
The relatively large average magnitudes, reported at injection rate above 10 m3/minute, are attributed to the small number of low magnitude events and the abundant moderate magnitude values. The maximum number of events was recorded at injection rate of 6.0 m3/minute, while the higher average value came slightly later at approximately 8 m3/minute. These observations are common in highly fractured tight formations such as those existing in East Texas, where very small displacements causing small magnitudes are numerous near the treatment well vicinity and remarkably decrease as moving away [
Alternatively, the plot of injection volume versus both number of events and average magnitude values (Fig- ure 8) recorded in two shale reservoirs, Marcellus and Barnett, showed marked increase in seismicity during the first 1600 - 2300 m3 of injected fluids. Through this range, the maximum number of events is recorded after injecting ~1300 m3 of the treatment fluids. Comparing the two plots of
Plot of injection rate (m3/minute) versus Moment magnitude recorded during Barnett shale gas stimulation from different fields in East Texas (Data from Reference [1] )
Comparison between production versus time plot and number of events versus time plot for a well in Clinton County, Kentucky. Data from Reference [27]
Plot of the injected volume of the treatment fluids (m3) versus both number of events and average magnitude recorded during hydraulic fracturing stimulation in Marcellus (upper) and Barnett (lower) shale. Data from Reference [1]
It is a common practice in hydraulic fracturing operations to begin with a minifrac test and adjust the injection parameters to fit the reservoir condition. Microseismicity usually starts shortly (an hour or so) after the commence of the main treatment (
Plot of number of events and Pressure (psi) versus time (hours)since injection started in well CPU2-2, Giddings, Texas. Data from Reference [27]
A typical seismicity pattern following hydraulic fracturing (
The reservoir stimulation efficiency as well as the resulting seismicity varies dramatically among reservoirs and within the same reservoir. However, post-frack high productivity correlates generally with seismicity over wide zones within the producing horizon.
To be felt on land surface, the earthquake’s magnitude should be roughly +3 with approximate energy 2.033 × 109 N-m [
N-m. In addition, the detailed discussion of the relation of microseismicity to the depth (
The seismicity recorded during hydraulic fracking operations is strongly dependent on the reservoir geology and the parameters of hydraulic treatment. The present study showed that the maximum number of recorded events was recorded at injection rate of 6.0 m3/minute, while the higher average value came slightly later at approximately 8 m3/minute and a marked increase in seismicity is recorded during the first 1600 - 2300 m3 of injected fluids. In addition, the average microseism magnitude (Mw = −2) and energy (68 N-m) presented in this study may indicate limited role of hydraulic fracking stimulations in initiating any destructive earthquakes. This indicates that mechanical processes of hydraulic fracking operations do not directly jeopardize the surface environment or the ecology, but subsurface hazards cannot be totally excluded. The huge volume of treatment fluids that return to the surface charged with chemicals and improper well completion, particularly wellbore cement, represent the important hazardous source. The results of the present study helps understand the behavior of hydrocarbon reservoirs to hydraulic fracking and production operations that may enable a better reservoir management and improve the industry practices in drilling and hydraulic fracking.
The author acknowledges very much the authors contributed with the data used in this research and the efforts of their microseismic crews. Also the fruitful suggestions and careful revisions made by Professor Dr. Abdel-Zaher Abouzeid (Professor of Mining Engineering) and Professor Dr. Fouad Khalaf (Professor Petroleum Engineering) at Faculty of Engineering, Cairo University are greatly appreciated.