Seismicity is undoubtedly an integral part of Geodynamics, since seismic data, from large-scale geophysical monitoring, can provide many valuable insights regarding the state of the Earth’s crust; seismicity, however, is not always natural, it can also be induced. In this week’s blog, we explored the subject of fluid injection-induced seismicity mainly through the lens of hydraulic fracturing (HF; hydrofracking or simply fracking), a process used in the petroleum industry to extract oil and gas from tight rock formations (e.g., shales – schists); brief discussions were facilitated pertaining to the two primary modern conundrums concerning the induced seismicity from hydrofracking operations, namely, the dominant crack source mechanisms, as well as the differentiation of the so-called ‘’wet’’ and ‘’dry’’ seismic events, along with their implications in the accurate estimation of the stimulated reservoir volume (SRV).
Dimitrios just wrapped up his bachelor in mining engineering from NTUA; in the next few months he will begin his PhD journey in the Department of Geology & Geological Engineering of the Colorado School of Mines.
Induced vs natural seismicity; what’s the difference and should we even care?
Induced seismicity, often referred to as induced microseisms, is a phenomenon where microearthquakes (i.e., seismic events of low magnitude) are triggered due to man-related activities that affect the natural stress – strain fields of the Earth, in comparison, natural earthquakes can be caused by geological processes, such as tectonic plate movements. Originally, the scientific community was interested in induced seismicity due to mining activities (e.g., rock blasting etc.); however, in the past few decades, this interest was rekindled through the scope of fluid injection-induced seismicity, given the abrupt rise of (a) enhanced geothermal systems (EGS), (b) injection of CO2 for permanent carbon capture and storage, (c) HF for oil and gas recovery, (d) injection of either water or CO2 into depleted reservoirs for enhanced oil recovery, and (e) disposal of waste water into deep formations, due to the global sustainability goals for renewable geothermal energy, environmental protection, economic optimization, amongst others.
Zang et al. (2014) classified injection operations into two separate categories, based on their time-scale and total injected volumes, i.e., long-term injection operations (b) and (e) (where the injected volumes > 100,000 m3), and short-term injection operations (a), (c), and (d) (where the injected volumes < 100,000 m3); today we will mainly discuss the latter class of injection operations, and particularly, the process of HF. In contrast to popular belief, hydrofracking, or the ‘’hydrafrac job/process’’ as it was referred in the original publication of Clark (1949), is an old school method for increasing the productivity of oil and gas wells in ultra tight rock formations, such as schists, that dates back to the late 40s. The HF method requires the drilling of a well in an oil-bearing formation of low permeability, and the subsequent pressurization of a sealed-off section of the borehole until the rock formation fails and ruptures abruptly.
During the pressurization process, which causes the progressive failure of the surrounding rock, microcracks (which eventually lead to macrocracks) form that generate elastic acoustic waves, that can have a varying degree of energy, these elastic waves, that propagate through the rock medium, are often denoted as acoustic emission (AE) events; they essentially represent induced microseismic events, that can be effectively captured via the usage of specialized high frequency AE sensors with typical frequency bandwidths ranging from 100 kHz to 1 MHz.
Overall, the induced seismicity from hydrofracking operations lies in the micro scale, i.e., it has very low recorded moment magnitudes (as low as -8 – -6 up to -3 – -1); it has been shown, however, that the scale of the induced seismicity can drastically increase across multiple orders of magnitude provided that active, or even passive, faults exist in the near vicinity of the HF. For instance, Bao and Eaton (2016) observed that a fault was reactivated during a shale gas stimulation due to the conduction of HFs. Generally, apart from the implications of microseismicity regarding the potential to create large-scale hazardous earthquakes, microseismic data can provide a great variety of insightful information relating to the fracturing process of the pressurized rock, as well as the SRV.
Crack source mechanisms; how does the pressurized rock fail?
Crack source mechanism analysis (or focal mechanism analysis) of microseismic data can provide a clear picture of the dominant fracture type of the rock, meaning tensile, shear, compressive, and/or mix-mode. This information can have great implications towards mitigating microseismicity, since tensile microcracks (type I) radiate elastic waves with substantially less energy relative to shearing microcracks (type II); very often researchers aim to generate an abundance of type I microcracks and as few type II microcracks as possible. For the most part, it appears that the dominant cracking mode is heavily dependent on the method used for the determination of the crack source mechanisms (e.g., moment tensor analysis, polarity, tensile angle etc.), the examined rock type (e.g., shale – metamorphic, granite – crystalline igneous, sandstone – sedimentary), and the density of pre-existing micro- or macrocracks in the rock volume, amongst other factors.
For instance, Butt et al. (2024) performed true-triaxial HF tests on cubical granite rock specimens, they generally observed tensile dominated events using both a low and a high viscosity fracturing fluid. Moreover, Naoi et al. (2020) conducted simple uniaxial HF tests on Eagle Ford shale specimens, they noticed an extreme domination of tensile events. These conclusions, come into contrast with the more commonly encountered shear dominated events observed in actual production fields (e.g., Maxwell and Cipolla 2011) and large-scale in situ experiments (e.g., Ishida et al. 2019). Overall, by uncovering the influence of each parameter on the derived focal mechanisms a deeper understanding can be gained towards the necessary steps to decrease the larger magnitude shear events.
Wet and dry microseismic events; towards an accurate estimation of the SRV
Microseismicity induced by HF stimulations can be mainly attributed to either pressure or mechanical changes/perturbations; given this simple distinction, the seismic events can be divided into: ‘’wet’’ microseismic events, which are caused due to fluid-flow related pressure changes (they are directly connected with the main HF), and remote ‘’dry’’ microseismic events, which are a product of stress changes at considerable distances away from the borehole (they are usually not connected with the main HF). Provided that for many years, it is a common practice to estimate the SRV by inferring to the AE cloud, the inclusion of isolated and distant ‘’dry’’ events, which are not actually connected with the main HF and hence do not really contribute to the SRV, into the AE seismic cloud can result in significant overestimations of the SRV.
Although there exists no unified way to directly differentiate between ‘’wet’’ and ‘’dry’’ events, one of the following two paths is usually adopted; namely, Maxwell et al. (2015a,b) observed that ‘’dry’’ and ‘’wet’’ events have noticeably different b-values, with former having a b-value of around 1, whereas the latter has a b-value of around 2. Finally, a less effective way is to create a distance – time from injection plot, in an attempt to locate early-stage distant events. A method capable of precisely distinguishing between ‘’dry’’ and ‘’wet’’ microseismic events can be truly beneficial to the progression of the field, since it will allow for the accurate determination of the SRV using the derived microseismic data.
References Bao X, Eaton DW (2016) Fault activation by hydraulic fracturing in western Canada. Science 354(6318): 1406 – 1409. Butt A, Hedayat A, Moradian O (2024) Microseismic Monitoring of Laboratory Hydraulic Fracturing Experiments in Granitic Rocks for Different Fracture Propagation Regimes. Rock Mech Rock Eng 57: 2035 – 2059. Clark JB (1949) A Hydraulic Process for Increasing the Productivity of Wells. J Pet Technol 1(1): 1 – 8. Ishida T, Fujito W, Yamashita H, Naoi M, Fuji H, Suzuki K, Matsui H (2019) Crack expansion and fracturing mode of hydraulic refracturing from acoustic emission monitoring in a small-scale field experiment. Rock Mech Rock Eng 52: 543 – 553. Maxwell SC, Cipolla C (2011) What does microseismicity tell us about hydraulic fracturing? In: SPE Annual Technical Conference and Exhibition, Denver, Colorado, SPE146932. Maxwell SC, Mack M, Zhang F, Chorney D, Goodfellow SD, Grob M (2015a) Differentiating Wet and Dry Microseismic Events Induced During Hydraulic Fracturing. In: SPE/AAPG/SEG Unconventional Resources Technology Conference, San Antonio, Texas, USA. Maxwell SC, Chorney D, Goodfellow SD (2015b) Microseismic geomechanics of hydraulic-fracture networks: Insights into mechanisms of microseismic sources. The Leading Edge 34(8): 904 – 910. Naoi M, Chen Y, Yamamoto K, Morishige Y, Imakita K, Tsutumi N, Kawakata H, Ishida T, Tanaka H, Arima Y, Kitamura S, Hyodo D (2020) Tensile-dominant fractures observed in hydraulic fracturing laboratory experiment using eagle ford shale. Geophys J Inter 222: 769 – 780. Zang A, Oye V, Jousset P, Deichmann N, Gritto R, McGarr A, Majer E, Bruhn D (2014) Analysis of induced seismicity in geothermal reservoirs – an overview. Anal Induc Seism Geotherm Oper 52: 6 – 21.
