Mechanism of Underground Fluid Injection Induced Seismicity and Its Implications for CCS Projects

CO₂ capture and storage projects must consider the potential possibility ofinjection induced seismicity. Moderate earthquakes and strong earthquakes may endanger human life and property, and even felt earthquakes and microquakes also pose a threat to seal integrity of CO₂ reservoir and increase the risk of leakage. Underground fluid injection induced seismicity usually happens in some geoengineering projects such as waste fluid disposal, EOR and EGS, and it occurs when fluid is injected along the fault. Therefore, it can be studied through stress analysis. The density of supercritical CO₂ is smaller than water, which may develop density flow in the deep strata or water-rock interactions in pre-existing structures, and cause the variation in permeability and pressure to induce a seismic activity. In this paper, we reviewed the mechanism of underground fluid injection induced seismicity with the focus of CCS, combined with fluid injection projects and seismic monitoring analysis in both commercial scale and experimental scale, to investigate its impact on the integrity of the cap rock of the reservoir. Finally, we summarized the appropriate site selection, injection methods and monitoring programs to prevent the occurrence of induced seismicity.     

美国北加州吉塞尔斯诱发地震与热汽田特征的关系

吉塞尔斯地震活动可能是因蒸汽开发引起的.水在一个承受很高构造剪应力和应变的大范围破裂体中汽化为蒸汽。 汽田地震震源机制解与区域构造应变场几乎一致,并且在该区域范围内汽田地震与别的构造地震不易区分。观测中注意到地震活动与注液历史无关,这表明孔隙水压力增高与注液不可能是吉塞尔斯诱发地震的成因。 相反,所有证据都表明诱发地震与孔隙水压力及温度降低有关。形成机制有两种最大可能:其一,是裂隙排水(汽)导致局部剪应力增加所致,其二,是由稳定滑动转化为不稳定滑动(粘滑)。没有其它记载的诱发地震机制与吉塞尔斯汽田条件相符。     

Seismicity and faulting attributable to fluid extraction

The association between fluid injection and seismicity has been well documented and widely publicized. Less well known, but probably equally widespread are faulting and shallow seismicity attributable solely to fluid extraction, particularly in association with petroleum production.

Two unequivocable examples of seismicity and faulting associated with fluid extraction in the United States are: The Goose Creek, Texas oil field event of 1925 (involving surface rupture); and the Wilmington, California oil field events (involving subsurface rupture) of 1947, 1949, 1951 (2), 1955, and 1961. Six additional cases of intensity I–VII earthquakes (M < 4.6) without reported faulting may be attributable to shallow production from other large oil and gas fields. In addition to these examples are thirteen cases of apparently aseismic surface rupture associated with production from California and Texas oil fields. Small earthquakes in the Eloy—Picacho area of Arizona may be attributable to withdrawal of groundwater, but their relation to widespread fissuring is enigmatic. The clearest example of extraction-induced seismicity outside of North America is the 1951 series of earthquakes associated with gas production from the Po River delta near Caviga, Italy.

Faulting and seismicity associated with fluid extraction are attributed to differential compaction at depth caused by reduction of reservoir fluid pressure and attendant increase in effective stress. Surface and subsurface measurements and theoretical and model studies show that differential compaction leads not only to differential subsidence and centripetally-directed horizontal displacements, but to changes in both vertical- and horizontal-strain regimes. Study of well-documented examples indicates that the occurrence and nature of faulting and seismicity associated with compaction are functions chiefly of: (1) the pre-exploitation strain regime, and (2) the magnitude of contractional horizontal strain centered over the compacting materials relative to that of the surrounding annulus of extensional horizontal strain.

The examples cited include natural systems strained only by extraction of fluids, as well as some subsequently subjected to injection. Faulting and seismicity have accompanied both decrease and subsequent increase of fluid pressures; reversal of fluid-pressure decline by injection may enhance the likelihood of subsurface faulting and seismicity due chiefly to earlier fluid pressure reduction. A consistent common denominator appears to be continuing compaction at depth; the relative effects of fluid extraction followed by injection are not easily separated.     

Analytical investigation of hydraulic fracture-induced seismicity and fault activation

More recent public discourse has taken place regarding the potential correlation between seismic activity and hydraulic fracturing in shale gas reservoirs. Public fears about the risk of seismicity stem mainly from past earthquakes induced by conventional deep injections because the two types of projects share similar mechanisms of rock failure and fault activation. Although previous earthquake risks associated with fluid injection were not serious, the situation would be far more problematic if hydraulic fracturing in a shale gas reservoir triggered a similar-sized earthquake due to potential environmental issues. In fact, almost all documented injection-induced earthquakes have been associated with long-duration and high-volume injection rather than short-term (hours) pressurization (e.g., hydraulic fracturing). In general, hydraulic fracturing operations mostly induce microseismic events through rock failure and activation of small fractures. Although shale reservoirs in tectonically active zones pose a high risk of inducing large-magnitude seismic activities, the internal geological conditions and external stimulation conditions are impossible to be satisfied simultaneously to trigger activation of an entire fault and to result in a destructive earthquake during hydraulic fracturing operations.     

Seismicity and fault aseismic deformation caused by fluid injection in decametric in-situ experiments

Seismicity induced by fluid perturbations became an important societal concern since felt earthquakes (M_w up to 6) occurred after anthropogenic activities. In order to mitigate the risks associated with undesired seismicity, as well as to be able to use the micro-seismicity as a probe for in-depth investigation of fluid-driven processes, it is of crucial importance to understand the links between seismicity, fluid pressure and flow. We have developed a series of in-situ, decameter-scale experiments of fault zone reactivation by controlled fluid injection, in order to improve the near-source geophysical and hydromechanical observations. The deployed geophysical monitoring close to the injection allows one to cover the full frequency range of the fault responses from the static deformation to the very high-frequency seismic emissions (up to 4 kHz). Here, we focus on the microseismicity (M_w ∼ –4 to –3) recorded during two fluid injection experiments in low-permeable shale and highly-fractured limestone formations. In both experiments, the spatio-temporal distribution of the seismic events, the energy balance, and the seismic velocity changes of the fractured medium show that most of the deformation does not actually emit seismic signals. The induced deformation is mainly aseismic. Based on these high-resolution multiparametric observations in the near-field, we therefore proposed a new model for injection-induced seismicity: the seismicity is not directly induced by the increasing fluid pressure, but it is rather triggered by the stress perturbations transferred from the aseismic motion caused by the injection.     

Main features of the fluid regime during the development of earthquake foci

The features of a fluid regime in the focus of an earthquake are considered. During the development of the focus, the fluid regime in its zone depends on the internal movement of fluids and their entry from outside. The main inflow of fluids and the increase in their pressure occur due to ascending filtration through subvertical conductive faults. Fluids may also flow from surface segments, which causes the majority of induced earthquakes. The close relationships between the mechanism of natural earthquakes and the fluids coming from the deep high-pressure zones point to the genetic similarity of degassing and seismicity of the tectonosphere.     

Correlations in aftershock and seismicity patterns

Correlations in space and time play a fundamental role in earthquake processes. One direct manifestation of the effects of correlations is the occurrence of aftershocks due to the stress transfer in the vicinity of a main shock. Less obvious and more speculative changes in correlations may occur in the background seismicity before large earthquakes. Using statistical physics it is possible to introduce a measure of spatial correlations through a correlation length. This quantity characterizes how local fluctuations can influence the occurrence of earthquakes over distances comparable with the correlation length. In this work, the physical basis of spatial correlations of earthquakes is discussed in the context of critical phenomena and the percolation problem. The method of two-point correlation function is applied to the seismicity of California. Well defined variations in time of the correlation length are found for aftershock sequences and background seismicity. The scaling properties of our obtained distributions are analyzed with respect to changes in several scaling parameters such as lower magnitude cutoff of earthquakes, the maximum time interval between earthquakes, and the spatial size of the area considered. This scaling behavior can be described in a unified manner by utilizing the multifractal fit. Utilizing the percolation approach the time evolution of clusters of earthquakes is studied with the correlation length defined in terms of the radius of gyration of clusters. This method is applied to the seismicity of California.     

Seismicity changes associated with reservoir loading

Changes in seismic activity have been related to the filling of large reservoirs in over thirty cases. These changes range from variations in the level of micro-earthquake activity detectable only with instruments of high sensitivity to destructive earthquakes with magnitudes greater than 6. On the other hand, the filling of many other large reservoirs has not been accompanied by increased seismicity.

A number of factors may contribute to the generation or absence of post-impounding seismicity. Increased vertical stress due to the load of the reservoir and decreased effective stress due to increased pore pressure can modify the stress regime in the reservoir region. Whether or not these stress changes are sufficient to generate earthquake activity will depend on a complex interaction of the induced stress with the state of pre-existing stress near the reservoir, and on the geologic and hydrologic conditions at the site. The combined effect of increased vertical load and increased pore pressure will have the greatest tendency to increase activity in regions where the maximum compressive stress is vertical (normal faulting). In regions where the minimum compressive stress is vertical (thrust faulting) increased stress due to a vertical load should have a minimum effect. For all of the larger reservoir-induced earthquakes the stress system determined from fault plane solutions is in agreement with the pre-existing stress field in the region of the reservoir. These earthquakes are all of strike-slip or normal type, there being no reported cases of large induced earthquakes with thrusting mechanisms.

The potential for major changes in seismicity may be highest in regions of moderate strain accumulation (low to moderate natural seismicity). In areas of high strain accumulation and high levels of natural seismicity, the stress changes induced by the reservoir will be small compared to natural variations. In aseismic areas, with low strain accumulation, the reservoir-induced stresses may be insufficient to raise the stress level to a state of failure.     

Integral equation solution of heat extraction‐induced thermal stress in enhanced geothermal reservoirs

During fluid injection in enhanced geothermal systems, thermo‐mechanical processes can play an important role. In fact, the phenomena of reservoir seismicity and the variation of injectivity with respect to injection water temperature can be attributed to the induced thermal stresses. In this paper, a three‐dimensional integral equation formulation is presented for calculating thermally induced stresses associated with the cooling of a fracture in a geothermal reservoir. By utilizing Green's function in the integral equation, the three‐dimensional heat flow and stresses in the reservoir are modelled without discretizing the reservoir. The formulation is implemented in a computer program for the solution of injection into an infinite fracture as well as for the injection/extraction in an arbitrarily shaped fracture. Copyright © 2005 John Wiley & Sons, Ltd.     

从断层中心论向块体中心论转变--论活动块体在地震活动中的作用

讨论活动块体在地震活动中的作用。板缘地震在空间上呈线性分布,而中国大陆地震在空间上呈片状分布。大陆强震往往涉及两个以上方向断层的活动,且强震往往沿块体边界迁移或在其两侧断层上对迁。一些强震前后由中小地震震源机制解反映的Ｐ轴方向往往发生近９０°的转向,这种现象难以用区域应力方向变化来解释,但可从块体活动角度出发,用块体两个边界断层的先后错动来解释。地震前异常的远程效应、震后烈度异常分布图像以及地震序列特点等也显示了活动块体的作用。不同地区由地震活动性推测的块体活动方式有所差别,其原因可能和区域主压应力轴与块体两个边界断层走向的夹角不同有关。据此认为,在分析中国地震活动时要把视角从以活动断层为中心转变为以活动块体为中心     

Hydraulic Fracturing-induced Seismicity at the Hot Dry Rock Site of the Gonghe Basin in China

Hot dry rock is becoming an important clean energy source. Enhanced geothermal systems (EGS) hold great promise for the potential to make a contribution to the energy inventory. However, one controversial issue associated with EGS is the impact of induced seismicity. In August 2019, a hydraulic stimulation experiment took place at the hot dry rock site of the Gonghe Basin in Qinghai, China. Earthquakes of different magnitudes of 2 or less occurred during the hydraulic stimulation. Correlations between hydraulic stimulation and seismic risk are still under discussion. Here, we analyze the hydraulic stimulation test and microseismic activity. We quantify the evolution of several parameters to explore the correlations between hydraulic stimulation and induced seismicity, including hydraulic parameters, microseismic events, b-value and statistical forecasting of event magnitudes. The results show that large-magnitude microseismic events have an upward trend with an increase of the total fluid volume. The variation of the b-value with time indicates that the stimulation experiment induces small amounts of seismicity. Forecasted magnitudes of events can guide operational decisions with respect to induced seismicity during hydraulic fracturing operations, thus providing the basis for risk assessment of hot dry rock exploitation.     

Seismicity around Brazilian dam reservoirs

More than 30 cases of seismicity associated with dam reservoir sites are known throughout the world. Despite the lack of data in some areas, where seismicity occurred after reservoir impounding, there have been distinct seismic patterns observed in seismic areas after dam projects implantation. This has demonstrated that reservoir loading can trigger earthquakes. A mechanism of earthquake generation by reservoir impounding is proposed here with particular application to the Brazilian cases and to areas subject to low confining stress conditions in stable regions. Six artificial lakes are described and the associated earthquake sources are discussed in terms of natural or induced seismicity. Earthquake monitoring in Brazil up to 1967, when Brasilia's seismological station started operation, was mainly based in personal communications to the media. Therefore, there is a general lack of seismic records in relatively uninhabited areas, making it difficult to establish a seismic risk classification for the territory and to distinguish natural from induced seismicity. Despite this, cases reported here have shown an alteration of the original seismic stability in dam sites, after reservoir loading, as observed by the inhabitants or records from Brasilia's seismological station. All cases appear to be related to an increase in pore pressure in permeable rocks or fracture zones which are confined between impermeable rock slabs of more competent rock. It is apparent that some cases show some participation of high residual stress conditions in the area.     

A finite difference method for earthquake sequences in poroelastic solids

Induced seismicity (earthquakes caused by injection or extraction of fluids in Earth’s subsurface) is a major, new hazard in the USA, the Netherlands, and other countries, with vast economic consequences if not properly managed. Addressing this problem requires development of predictive simulations of how fluid-saturated solids containing frictional faults respond to fluid injection/extraction. Here, we present a finite difference method for 2D linear poroelasticity with rate-and-state friction faults, accounting for spatially variable properties. Semi-discrete stability and accuracy are proven using the summation-by-parts, simultaneous-approximation-term (SBP-SAT) framework for discretization and boundary condition enforcement. Convergence rates are verified using the method of manufactured solutions and comparison to the analytical solution to Mandel’s problem. The method is then applied to study fault slip triggered by fluid injection and diffusion through high-permeability fault damage zones. We demonstrate that in response to the same, gradual forcing, fault slip can occur in either an unstable manner, as short-duration earthquakes that radiate seismic waves, or as stable, aseismic, slow slip that accumulates over much longer time scales. Finally, we use these simulation results to discuss the role of frictional and elastic properties in determining the stability and nature of slip.     

Plate subduction,and generation of earthquakes and magmas in Japan as inferred from seismic observations: An overview

A dense nationwide seismic network recently constructed in Japan has been yielding large volumes of high-quality data that have made it possible to investigate the seismic structure in the Japanese subduction zone with unprecedented resolution. In this article, recent studies on the subduction of the Philippine Sea and Pacific plates beneath the Japanese Islands and the mechanism of earthquake and magma generation associated with plate subduction are reviewed. Seismic tomographic studies have shown that the Philippine Sea plate subducting beneath southwest Japan is continuous throughout the entire region, from Kanto to Kyushu, without disruption or splitting even beneath the Izu Peninsula as suggested in the past. The contact of the Philippine Sea plate with the Pacific plate subducting below has been found to cause anomalously deep interplate and intraslab earthquake activity in Kanto. Detailed waveform inversion studies have revealed that the asperity model is applicable to interplate earthquakes. Analyses of dense seismic and GPS network data have confirmed the existence of episodic slow slip accompanied in many instances by low-frequency tremors/earthquakes on the plate interface, which are inferred to play an important role in stress loading at asperities. High-resolution studies of the spatial variation of intraslab seismicity and the seismic velocity structure of the slab crust strongly support the dehydration embrittlement hypothesis for the generation of intraslab earthquakes. Seismic tomography studies have shown that water released by dehydration of the slab and secondary convection in the mantle wedge, mechanically induced by slab subduction, are responsible for magma generation in the Japanese islands. Water of slab origin is also inferred to be responsible for large anelastic local deformation of the arc crust leading to inland crustal earthquakes that return the arc crust to a state of spatially uniform deformation.     

The spatial correlation between earthquakes and landslides in Hokkaido (Japan), a GIS-based analysis of the past and the future

Although earthquakes are thought to be one of the factors responsible for the occurrence of landslides in Hokkaido, there exist no enough records which can allow correlating many of the old slope failures in the island with earthquakes. In the absence of these records, an attempt was done in this study to use the abundance, frequency, magnitude, depth, and distribution of historical earthquakes to deduce that many of the slope failures in the region were triggered by strong and continuous seismicity. The determination of the zones of influences of selected earthquakes using an existing empirical function has also supported this conclusion. Moreover, the use of a 10% probability of exceedance of earthquake intensity in 50 years, and the geological and slope maps has allowed preparing a landslide hazard map which explains the role of future earthquakes in the formation of slope failures. The result indicates a high probability of occurrences of landslides in the hilly regions of the southeastern part of Hokkaido due to expected strong seismicity and earthquake intensities in these areas. On the other hand, the low level of intensity in the north has given rise to low probability of landslide hazard. There are also places in the center of the island and high intensity regions in the east where the probability of landslide hazard was influenced by the contribution of the geological and slope maps.     

Use of tsunami waveforms for earthquake source study

Tsunami waveforms recorded on tide gauges, like seismic waves recorded on seismograms, can be used to study earthquake source processes. The tsunami propagation can be accurately evaluated, since bathymetry is much better known than seismic velocity structure in the Earth. Using waveform inversion techniques, we can estimate the spatial distribution of coseismic slip on the fault plane from tsunami waveforms. This method has been applied to several earthquakes around Japan. Two recent earthquakes, the 1968 Tokachi-oki and 1983 Japan Sea earthquakes, are examined for calibration purposes. Both events show nonuniform slip distributions very similar to those obtained from seismic wave analyses. The use of tsunami waveforms is more useful for the study of unusual or old earthquakes. The 1984 Torishima earthquake caused unusually large tsunamis for its earthquake size. Waveform modeling of this event shows that part of the abnormal size of this tsunami is due to the propagation effect along the shallow ridge system. For old earthquakes, many tide gauge records exist with quality comparable to modern records, while there are only a few good quality seismic records. The 1944 Tonankai and 1946 Nankaido earthquakes are examined as examples of old events, and slip distributions are obtained. Such estimates are possible only using tsunami records. Since tide-gauge records are available as far back as the 1850s, use of them will provide unique and important information on long-term global seismicity.     

Fracture Network Volume Fracturing Technology in High-temperature Hard Formation of Hot Dry Rock

It is more difficult for a hot dry rock to form a fracture network system than shale due to its special lithology, physical and mechanical properties under high temperature. The essential characteristics, rock mechanics and in-situ stress characteristics of a hot rock mass have been systematically studied by means of laboratory tests and true tri-axial physical simulation. The fracture initiation and propagation characteristics under different geological and engineering conditions are physically simulated, and the main controlling factors for the formation of a complex fracture network are revealed. The technology of low displacement for enhancing thermal cracking, gel fluid for expanding fracture and variable displacement cyclic injection for increasing a fracture network has been applied in the field, and good results have been achieved. Micro-seismic monitoring results demonstrate that complex fractures were formed in the field test, and the stimulation volume for heat exchanging reaches more than 3 million cubic meters. The research results play an important role in the stimulation technology of an enhanced geothermal system (EGS) and realize a breakthrough for power generation.     

A classification of induced seismicity

In order to adopt the best safety procedures, man-made earthquakes should be differentiated as a function of their origin. At least four different types of settings can be recognized in which anthropogenic activities may generate seismicity: (I) fluid removal from a stratigraphic reservoir in the underground can trigger the compaction of the voids and the collapse of the overlying volume, i.e., graviquakes; the deeper the reservoir, the bigger the volume and the earthquake magnitude; (II) wastewater or gas reinjection provides the reduction of friction in volumes and along fault planes, allowing creep or sudden activation of tectonic discontinuities, i.e., reinjection quakes; (III) fluid injection at supra-lithostatic pressure generates hydrofracturing and micro-seismicity, i.e., hydrofracturing quakes; (IV) fluid extraction or fluid injection, filling or unfilling of artificial lakes modifies the lithostatic load, which is the maximum principal stress in extensional tectonic settings, the minimum principal stress in contractional tectonic settings, and the intermediate principal stress in strike-slip settings, i.e., load quakes; over given pressure values, the increase of the lithostatic load may favour the activation of normal faults, whereas its decrease may favour thrust faults. For example, the filling of an artificial lake may generate normal fault-related seismicity. Therefore, each setting has its peculiarities and the knowledge of the different mechanisms may contribute to the adoption of the appropriate precautions in the various industrial activities.     

Global strain rates in western to central Himalayas and their implications in seismic hazard assessment

The Himalayas has experienced varying rates of earthquake occurrence in the past in its seismo-tectonically distinguished segments which may be attributed to different physical processes of accumulation of stress and its release, and due diligence is required for its inclusion for working out the seismic hazard. The present paper intends to revisit the various earthquake occurrence models applied to Himalayas and examines it in the light of recent damaging earthquakes in Himalayan belt. Due to discordant seismicity of Himalayas, three types of regions have been considered to estimate larger return period events. The regions selected are (1) the North-West Himalayan Fold and Thrust Belt which is seismically very active, (2) the Garhwal Himalaya which has never experienced large earthquake although sufficient stress exists and (3) the Nepal region which is very seismically active region due to unlocked rupture and frequently experienced large earthquake events. The seismicity parameters have been revisited using two earthquake recurrence models namely constant seismicity and constant moment release. For constant moment release model, the strain rates have been derived from global strain rate model and are converted into seismic moment of earthquake events considering the geometry of the finite source and the rates being consumed fully by the contemporary seismicity. Probability of earthquake occurrence with time has been estimated for each region using both models and compared assuming Poissonian distribution. The results show that seismicity for North-West region is observed to be relatively less when estimated using constant seismicity model which implies that either the occupied accumulated stress is not being unconfined in the form of earthquakes or the compiled earthquake catalogue is insufficient. Similar trend has been observed for seismic gap area but with lesser difference reported from both methods. However, for the Nepal region, the estimated seismicity by the two methods has been found to be relatively less when estimated using constant moment release model which implies that in the Nepal region, accumulated strain is releasing in the form of large earthquake occurrence event. The partial release in second event of May 2015 of similar size shows that the physical process is trying to release the energy with large earthquake event. If it would have been in other regions like that of seismic gap region, the fault may not have released the energy and may be inviting even bigger event in future. It is, therefore, necessary to look into the seismicity from strain rates also for its due interpretation in terms of predicting the seismic hazard in various segments of Himalayas.     

Aeromagnetic data over Harrat Lunayyir and surrounding areas,western Saudi Arabia

Harrat Lunayyir is one of the smaller lava fields in western Saudi Arabia that is of current interest due to a dike intrusion episode in 2009, an ongoing swarm of earthquakes and the possible hazard that pose. In addition to seismology, other geophysical data have been used to study the structure of the area, and the available aeromagnetic information is shown and discussed here. The reduced to the pole (RTP) magnetic grid and its enhancements have been used to define some of the main anomalies, and these have been correlated with the known geology, including the Red Sea coastal dike system. There are numerous linear features that are possibly related to the dikes or faulting, but within the area of the harrat, details due to the underlying structure are largely obscured by the magnetic surface lavas. Northeasterly trends in the magnetic data may indicate old zones of weakness that intersect one of the main coastal dikes at the location of the recent seismic activity and surface fissure, suggesting that this is a point of weakness resulting in the volcanism and seismicity that also appears to be largely limited to two of the NE trends. The association of the recent seismicity with a known geological and aeromagnetic feature is important in determining the seismic hazard for the region, especially if the location of future activity can be used to reduce overall uncertainty in the analysis by identifying potential fault sources. Here, the seismicity appears to lie on one of the NNW-trending coastal dikes that have been reactivated recently along a section between two NE-trending older faults.