Induced Seismicity and Energy Technologies
INTRODUCTION TO INDUCED SEISMICITY AND STUDY BACKGROUND
An earthquake is a shaking of the ground caused by a sudden release of energy within the Earth. Most earthquakes occur because of a natural and rapid shift (or slip) of rocks along geologic faults that release energy built up by relatively slow movements of parts of the Earth’s crust. The numerous, sometimes large earthquakes felt historically in California and the earthquake that was felt along much of the East Coast in August 2011 are examples of naturally occurring earthquakes related to Earth’s movements along regional faults (see also the section Earthquakes and Their Measurement, this chapter). An average of ~14,450 earthquakes with magnitudes above 4.0 (M > 4.0)1 are measured globally every year. This number increases dramatically—to more than 1.4 million earthquakes annually—when small earthquakes (those with greater than M 2.0) are included.2
Although the vast majority of earthquakes have natural causes, some earthquakes may also be related to human activities and are called induced seismic events.3 Induced seismic events are usually small in both magnitude and intensity of shaking (see the section on Earthquakes and Their Measurement later in this chapter). For example, underground nuclear tests, controlled explosions in connection with mining or construction, and the impoundment of large reservoirs behind dams can each result in induced seismicity (Box 1.1). Energy technologies that involve injection or withdrawal of fluids from the subsurface also have the potential to induce seismic events that can be measured and felt (see Kerr, 2012).
The earliest and probably most familiar documented example of an induced seismic event related to fluid injection is the activity that occurred in the Denver, Colorado, area in the 1960s in connection with liquid waste disposal at the Rocky Mountain Arsenal. An injection well at the Arsenal pumping into relatively impermeable crystalline basement
1M represents magnitude on the moment-magnitude scale, which is described in the section Earthquakes and Their Measurement, this chapter.
2 See earthquake.usgs.gov/learn/faq/?faqID=69.
3 Some researchers (e.g., McGarr et al., 2002) draw a distinction between “induced” seismicity and “triggered” seismicity. Under this distinction, induced seismicity results from human-caused stress changes in the Earth’s crust that are on the same order as the ambient stress on a fault that causes slip. Triggered seismicity results from stress changes that are a small fraction of the ambient stress on a fault that causes slip. Anthropogenic processes cannot “induce” large and potentially damaging earthquakes, but anthropogenic processes could potentially “trigger” such events. In this report we do not distinguish between the two and use the term “induced seismicity” to cover both categories.
Observations of Induced Seismicity
Seismicity induced by human activity has been observed and documented since at least the 1920s (Pratt and Johnson, 1926). The number of sites where seismic events of M > 0 have occurred that are caused by or likely related to energy development are listed below by technology. (References for these sites with location and magnitude information are in Appendix C; note that in several cases the causal relationship between the technology and the event was suspected but never confirmed.) The numbers of sites globally are listed first in the column; the world map (Figure 1) shows these sites by technology and magnitude. The numbers in parentheses are the numbers of sites, as a subset of the global totals, in which seismic events in the United States have been caused by or likely related to energy development. In addition to energy technologies that are the topic of this report, the list also shows induced seismicity due to surface water reservoirs (dams) and other activities related to mining.a Event locations are plotted on global and U.S. maps in Figures 1 and 2.
Global (United States only)
Oil and gas extraction (withdrawal)
Secondary recovery (water flooding)
Hydraulic fracturing (shale gas)
Surface water reservoirs
Other (e.g., coal and solution mining)
Note that the figures include locations where a spatial association between seismicity and human activity has suggested a causal relationship, but where a causal relationship has not been positively established. Indeed, establishing such a causal relationship often requires a significant amount of scientific effort and fieldwork in the form of temporary seismometer arrays, particularly for the remote locations at which underground activities are conducted.
aMining operations can cause seismic events, in addition to the explosions that are used to fracture rock for excavation. These seismic events may occur at shallow depths as a result of changes in crustal stress, both by removal of mining ore and by redistribution of crustal stress from fracturing sound rock. Such events are not considered further in this report.
Figure 1 Worldwide locations of seismicity reported in the technical literature caused by or likely related to human activities, with the maximum magnitude reported to be induced at each site.
Figure 2 Locations of seismic events caused by or likely related to human activities within the coterminous United States and portions of Canada as documented in the technical literature.
rock caused induced earthquakes (three M 5.0 to M 5.5 earthquakes4), the largest of which caused an estimated $500,000 in damages in 1967 (Nicholson and Wesson, 1990) (Box 1.2).
More recent public attention to the potential correlation between seismic events and energy technology development began with several felt seismic events: in Basel, Switzerland, in 2006; at The Geysers, California, in 2008; and near the Dallas-Fort Worth airport in 2008. During the course of this study, several additional seismic events with potential correlation to energy development have occurred in different parts of the United States and in several other nations (see later in this chapter and in Chapters 2 and 3 for details of some of these events). The potential for induced seismic events has also been highlighted in the context of ongoing public discussion of shale gas development through hydraulic fracturing operations. Although none of these recent events resulted in loss of life or significant structural damage, their effects were felt by local residents, some of whom also experienced minor property damage. Particularly in areas where tectonic (natural) seismic activity is uncommon or historically nonexistent and energy development is ongoing, these seismic events, though small in scale, can be disturbing for the public and can raise concern about further seismic activity and its consequences.
This report addresses induced seismicity that may be related specifically to certain kinds of energy development that involve fluid injection or withdrawal. The study arose through a request made in 2010 by Senator Bingaman of New Mexico, chair of the Senate Energy and Natural Resources Committee, to Department of Energy Secretary Stephen Chu (Appendix D). The senator asked the secretary to engage the National Research Council to examine the scale, scope, and consequences of seismicity induced by energy technologies and specifically associated with four energy technologies: geothermal energy, shale gas,5 enhanced oil recovery (EOR), and carbon capture and storage (CCS). The study’s statement of task is presented in Box 1.3.
The aim of this report is to provide an understanding of the nature and scale of induced seismicity related to energy technologies and to suggest guidance as to how best to proceed with safe development of these technologies in terms of any potential induced seismicity risks. The report begins with an examination of the types and potential causes or mechanisms for induced seismicity (Chapter 2), reviews the four energy technologies that are the subject of the study and the ways they may induce seismic activity (Chapter 3), and discusses government roles and responsibilities related to underground injection and induced seismicity (Chapter 4). Chapter 5 considers the hazard and risk for induced seismicity and identifies some paths for understanding and managing induced seismicity, with steps toward
4 The initial reports of the magnitudes of the events at the Rocky Mountain Arsenal did not have details about the magnitude scale being used. Subsequent detailed analysis of seismograms (Herrmann et al., 1981) indicated that the magnitudes of the largest earthquakes were actually M 4.5 to M 4.8, slightly smaller than the initially reported magnitudes. See Box 1.2 for details.
5 When the committee uses the term “shale gas,” it is referring to dry gas, gas, and some liquids.
best practices for mitigating induced seismicity risk in Chapter 6. Chapter 7 contains the report’s findings, conclusions, proposed actions, and research recommendations, including identification of information and knowledge gaps and research and monitoring needs. The remainder of this chapter briefly reviews earthquakes and their measurement, introduces the four energy technologies that are the subject of this report, and presents several historical examples of induced seismic activity related to energy development.
The significance of understanding and mitigating the effects of induced seismicity related to energy technologies has been recognized by other groups as well, both internationally and domestically. The International Partnership for Geothermal Technology Working Group on Induced Seismicity6 under the auspices of the International Energy Agency, for example, has been addressing the issue as it relates specifically to geothermal energy development. International professional societies such as the Society of Petroleum Engineers and the Society of Exploration Geophysicists are coordinating a public technical workshop on the topic.7 Within the United States, government agencies such as the Department of Energy and U.S. Geological Survey have also been engaged in explicit efforts to understand and address induced seismicity in technology development. The Environmental Protection Agency has been facilitating a National Technical Working Group on Injection Induced Seismicity8 since mid-2011 and anticipates releasing a report that will contain technical recommendations directed toward minimizing or managing injection-induced seismicity.
EARTHQUAKES AND THEIR MEASUREMENT
The process of earthquake generation is analogous to a rubber band stretched to the breaking point that suddenly snaps and releases the energy stored in the elastic band. Earthquakes result from slip along faults that release tectonic stresses that have grown high enough to exceed a fault’s breaking strength. Strain energy is released by the Earth’s crust during an earthquake in the form of seismic waves, friction on the causative fault, and, for some earthquakes, crustal elevation changes. Seismic waves can travel great distances; for large earthquakes they can travel around the globe. Ground motions observed at any location are a manifestation of these seismic waves. Seismic waves can be measured in different ways: earthquake magnitude is a measure of the size of an earthquake or the amount of energy released at the earthquake source, while earthquake intensity is a measure of the level of ground shaking at a specific location. The distinction between earthquake magnitude and intensity is important because intensity of ground shaking determines what
6 See http://internationalgeothermal.org/; http://www.iea-gia.org/documents/Switzerland_Inducedseismicity_IPGT_IEA_201105031.pdf
7 See http://www.spe.org/events/12aden/documents/12ADEN_Brochure.pdf
8 See http://www.gwpc.org/meetings/uic/2012/proceedings/09McKenzie_Susie.pdf; P. Dellinger, presentation to the committee, September 2011.
The Rocky Mountain Arsenal Earthquakes
During the spring of 1962 seismological stations in Colorado began recording a number of small earthquakes near Denver. Although Denver had previously been considered to be in an area of low seismicity, between April 1962 and August 1967 over 1,500 earthquakes were recorded at the seismograph station at Bergen Park, Colorado. Some of the earthquakes were noticeable to local residents and exceeded M 3 and M 4. The earthquakes were eventually attributed to the underground injection of fluid using a deep well drilled on land known as the Rocky Mountain Arsenal approximately 6 miles northeast of downtown Denver.
The Rocky Mountain Arsenal was used by the U.S. Army from 1942 through 1985 for both the manufacture and the disposal of chemical weapons. In 1961 the army drilled a well on the arsenal grounds for the disposal of chemical fluid wastes by underground injection. The well was drilled to a depth of 12,045 feet into Precambrian crystalline rocks (rocks greater than about 700 million years old) beneath the sedimentary rocks of the Denver basin. Fluid injection began in March 1962, and from that time through September 1963, fluid was injected at an average rate of 181,000 gallons per day (gal/day). Injection was stopped in October 1963, but commenced again from August 1964 through April 1965. During this second injection cycle the fluid was not injected under pressure but was fed to the well under gravity flow at a rate of 65,800 gal/day. In April 1965 pressure injection resumed at a rate of 148,000 gal/day. The maximum injection pressure at any time was 72 bars (1,044 pounds per square inch [psi]).a
In April and May 1962, two seismological observatories in the Denver area began recording a series of small earthquakes.
In June of 1962 several earthquakes occurred which were large enough to be felt by residents and caused considerable concern. By November of 1965 over 700 shocks had been recorded and, although 75 of these had been felt, no damage was reported….” (McClain, 1970)
Research conducted in the mid-1960s on the deep injection well located on the Arsenal grounds detailed the correlation between the amount of fluid injected into the Arsenal well and the number of Denver earthquakes (Evans, 1966). This research indicated a strong relationship between injection volumes and earthquake frequency (see Figure). More detailed investigation by several local universities and the U.S. Geological Survey (USGS) gave further support to this conclusion. The research showed the majority of the earthquakes had epicenters within 5 miles of the Arsenal’s injection well. The depths of the earthquakes varied from 12,140 to 23,000 feet (3,700 to 7,000 meters) below the surface, which is the depth of Precambrian rocks in the area. Research also showed that the epicenters for the earthquakes aligned in a generally northwest-to-southeast direction, similar to the orientation of a system of natural vertical fractures found in the Precambrian rocks in the area.
Although injection into the Arsenal well ceased in February 1966, earthquake activity continued for several more years. The strongest earthquakes actually occurred after injection into the well was discontinued. A detailed analysis of seismograms (Herrmann et al., 1981) indicated seismic moments of the largest earthquakes that can be converted to M 4.5 (April 1967), M 4.8 (August 1967), and M 4.5 (November 1967). These magnitudes are more accurately determined and somewhat smaller than the magnitudes reported in earlier papers on the
Figure Histograms showing relation between volume of waste injected into the Rocky Mountain Arsenal well and earthquake frequency. SOURCES: Adapted from Evans (1966); Healy et al. (1968); McClain (1970); Hsieh and Bredehoeft (1981).
Rocky Mountain Arsenal earthquakes, which did not have details about the magnitude scale being used. After November 1967 earthquake activity steadily declined and virtually ceased by the late 1980s.
Initial theories postulated that the Denver earthquakes were caused by fluids being pumped into the ground by pressure injection in the disposal well; the fluids were suggested to have acted as a lubricant, allowing large blocks of rock in the subsurface to shift more easily. However, further analysis showed earthquakes triggered by fluid injection are not caused by lubrication of a fracture system but suggested instead that the earthquakes were caused by increasing the pressure of the existing fluid in the formation through high-pressure injection, which lowered the frictional resistance between rocks along an existing fault system; lowering the frictional resistance allowed the rocks to slide relative to each other.
aNote: Throughout the report we cite the units presented in the original reference followed by a conversion in parentheses to U.S. measures, metric, or units that might be more familiar to the general reader.
Statement of Task
The study will focus on areas of interest related to CCS, enhanced geothermal systems, production from shale gas, and EOR, and will
1. summarize the current state-of-the-art knowledge on the possible scale, scope, and consequences of seismicity induced during the injection of fluids related to energy production, including lessons learned from other causes of induced seismicity;
2. identify gaps in knowledge and the research needed to advance the understanding of induced seismicity, its causes, effects, and associated risks;
3. identify gaps and deficiencies in current hazard assessment methodologies for induced seismicity and research needed to close those gaps; and
4. identify and assess options for interim steps toward best practices, pending resolution of key outstanding research questions.
we, as humans, perceive or feel and the extent of damage to structures and facilities. The intensity of an earthquake depends on factors such as distance from the earthquake source and local geologic conditions, as well as earthquake magnitude. Throughout this work we refer to earthquake magnitudes using the moment-magnitude scale (Hanks and Kanamori, 1979), which is a scale preferred by seismologists because it is theoretically related to the amount of energy released by the Earth’s crust. The common symbol used to indicate moment magnitude is M.9
The earthquake magnitude scale spans a truly immense range of energy releases. For example, an earthquake of M 8 does not represent energy release that is four times greater than an earthquake of M 2; rather, an M 8 releases 792 million times greater energy than an M 2. For tectonic (“natural”) earthquakes, magnitude is also closely tied to the earthquake rupture area, which is defined as the surface area of the fault affected by sudden slip during an earthquake. A great earthquake of M 8 typically has a fault-surface rupture area of 5,000 to 10,000 km2 (equivalent to ~1,931 to 3,861 square miles or about the size of Delaware,
9 The moment magnitude scale, designated M, is the conventional scale now in use worldwide because it is related to the energy or “work” done by the Earth’s crust in creating the earthquake. An earthquake magnitude scale was first published by Richter (1936) and was based on the amplitudes of ground motions recorded on standard seismometers in Southern California. The desire was to assign a numerical magnitude value to earthquakes that was logarithmically proportional to the amount of energy released in the Earth’s crust, although it was recognized by Richter that the available data were inadequate for developing a direct correlation with energy. The original scale for Southern California achieved widespread use, was designated “Richter” or local magnitude, and was adapted for other areas with modifications to account for regional differences in earthquake wave attenuation. The moment magnitude has the ability to represent the energy released by very large earthquakes. Moment magnitude, where available, has been used throughout the report.
which is 2,489 square miles). In contrast, M 3 earthquakes typically have rupture areas of roughly 0.060 km2 (about 0.023 square miles or about 15 acres, equivalent to about 15 football fields). “Felt earthquakes” are generally those with M between 3 and 5, and “damaging earthquakes” are those with M > 5. The maximum velocity of ground shaking is a measure of how damaging the ground motion will be near the fault causing the earthquake. The intensity of shaking at any location is usually expressed using the Modified Mercalli scale and varies from III10 (felt by few people and would cause hanging objects to sway) for M 3, to X (when severe damage would occur). A large earthquake located onshore will generate intensity X near the fault rupture, intensity III at far distances, and all intensities between at intermediate distances.
Most earthquakes, whether natural or induced, that are recorded by seismometers are too small to be noticed by people. These small earthquakes are often referred to as microearthquakes or microseisms. This report adopts the latter term for all seismic events with magnitude M < 2.0. Microseisms as small as M -2 (see Appendix E for an explanation of negative magnitudes) are routinely recorded by local seismometer arrays during hydraulic fracturing operations used to stimulate oil and gas recovery. At M -2 the rupture areas are on the order of 1 m2 (a little less than 11 square feet).
Most naturally occurring earthquakes occur near the boundaries of the world’s tectonic plates where faults are historically active. However, low levels of seismicity also occur within the tectonic plates. This fact, together with widespread field measurements of stress and widespread instances of induced seismicity, indicate that the Earth’s crust, even in what we may consider geologically or historically stable regions, is commonly stressed near to the critical limit for fault slip (Zoback and Zoback, 1980, 1981, 1989). Because of this natural state of the Earth’s crust, no region can be assumed to be fully immune to the occurrence of earthquakes.
Induced seismicity may occur whenever conditions in the subsurface are altered in such a way that stresses acting on a preexisting fault reach the breaking point for slip. If stresses in a rock formation are near the critical stress for fault rupture, theory predicts and experience demonstrates that relatively modest changes of pore fluid pressures can induce seismicity. Generally, induced earthquakes are not damaging, but if preexisting stress conditions or the elevated pore fluid pressures are sufficiently high over a large fault area, then earthquakes with enough magnitude or intensity to cause damage can potentially occur.
Identifying whether a particular earthquake or microseism was caused by human activity or occurred naturally is commonly very difficult; often, inferences are made based on spatial and temporal proximity of the earthquake and human activity, on seismic history in the region, and on whether general models of induced seismicity would support a connection. For example, a small amount of fluid injected into the crust at shallow depths (e.g., during
10 The Mercalli scale uses Roman numerals.
a hydraulic fracturing operation) would not be considered the cause of a M 7 earthquake that was initiated at 10 km depth, even if the hydraulic fracturing and earthquake were close in space and time.
The earthquake history of a region also plays a role in inferring whether a particular earthquake was induced. If a certain earthquake appears to be related to human activity, but similar earthquakes have occurred in the past in that region, the connection with human activity is more tenuous than if the correlation between earthquake and human activity occurred in a previously aseismic region. In the latter case, an important indicator might be the rate of occurrence of multiple earthquakes, compared to the historical rate (Ellsworth et al., 2012). The important point is that there often is no definitive proof that a particular earthquake was induced; conclusions are usually based on inference.
ENERGY TECHNOLOGIES AND INDUCED SEISMICITY
Geothermal energy production captures the natural heat of the Earth to generate steam that can drive a turbine to produce electricity. Geothermal systems fall into one of three different categories: (1) vapor-dominated systems, (2) liquid-dominated systems, and (3) enhanced geothermal systems (EGS). Vapor-dominated systems are relatively rare. A major example is The Geysers geothermal field in Northern California. Liquid-dominated systems are used for geothermal energy in Alaska, California, Hawaii, Idaho, Nevada, and Utah. In both of these types of hydrothermal resource systems, either steam or hot water is extracted from naturally occurring fractures within the rock in the subsurface and cold fluid is injected into the ground to replenish the fluid supply. EGS are a potentially new source of geothermal power in which the subsurface rocks are naturally hot and fairly impermeable, and contain relatively little fluid. Wells are used to pump cold fluid into the hot rock to gather heat, which is then extracted by pumping the fluid to the surface. In some cases a potential EGS reservoir may lack sufficient connectivity via fractures to allow fluid movement through rock. In this case the reservoir may be fractured using high-pressure fluid injection in order to increase permeability. Permeability is a measure of the ease with which a fluid flows through a rock formation. (See Chapter 2 for detailed discussion of permeability and its relevance to fracture development and fluid flow.) In each of these geothermal systems, the injection or extraction of fluid has the potential to induce seismic activity. Further description of these technologies and examples of induced seismic activity are provided in Chapter 3.
Oil and Gas Production
Oil and gas production involves pumping hydrocarbon liquids (petroleum and natural gas), often together with large amounts of aqueous fluids (groundwater) that commonly contain high amounts of dissolved solids and salts (“brine”), from the subsurface. In the United States, oil and gas operators are required to manage these aqueous fluids through some combination of treatment, storage, disposal, and/or use, subject to government regulations. Commonly, these fluids, if not reused in the extraction process (see also Carbon Capture and Storage, below), are disposed of by injection into the deep subsurface in wells that may be located at some distance from the site of the oil or gas extraction (see also Chapters 3 and 4).
Fluids may also be produced from a well during “flow-back operations” after a well has been hydraulically fractured. Hydraulic fracturing is a method of stimulating an oil- or gas-producing geologic formation by injecting fluid underground to initiate fractures in the rock to aid oil or gas production from the well. A portion of the fluid is later recovered from the well and may be reinjected for additional hydraulic fracture treatments or managed through storage, permanent disposal in an injection well, or treatment for disposal or beneficial use similar to aqueous fluids that are normally produced directly from an oil or gas reservoir. Injection of fluids related to hydraulic fracturing and injection of waste fluids into the subsurface for permanent disposal are two different processes described in detail in Chapter 3.
Oil and gas production (withdrawal) often includes fluid reinjection. The reinjected fluid may be natural gas, aqueous fluids, or carbon dioxide (CO2) used to help push more oil and gas out from the rocks and to the surface; such reinjection is termed secondary recovery. Enhanced oil recovery, also known as tertiary recovery, uses technologies that also aid in increasing the recovery of hydrocarbons from a reservoir by changing the properties of the oil (primarily aiming to lower the viscosity of the oil so that it flows more easily). The most common EOR techniques involve injecting CO2 or hydrocarbons, or heating the oil through steam injection or combustion. The injection of fluid to facilitate oil and gas production, similar to fluid injection for geothermal systems, has the potential to generate induced seismic activity. To date, EOR has not been associated with induced seismicity, although felt seismic events have been documented in connection with waterflooding for secondary recovery. The withdrawal of oil and gas has also been associated with induced seismic activity. All of these technologies and examples of induced seismic activity are described further in Chapter 3.
Carbon Capture and Storage
Carbon capture and geologic storage is the separation and capture of CO2 from emissions of industrial processes, including energy production, and the transport and permanent storage of the CO2 in deep underground formations. Currently five different types of
underground formations are being investigated for permanent CO2 storage: (1) oil and gas reservoirs, (2) saline formations, (3) unmineable coal seams, (4) organic-rich shales, and (5) basalt formations.11 Carbon dioxide has been injected into oil and gas reservoirs for several decades to enhance oil recovery. Current large-scale CCS projects in the United States are focused on injection of carbon dioxide into saline brines in regional aquifers. Carbon dioxide must be in the supercritical (liquid) phase to minimize the required underground storage volume; this requires a fluid pressure of greater than 6.9 MPa (about 68 atm12) and temperature greater than 31.1°C, which can be achieved at depths greater than about 2,600 feet (~800 meters) (Sminchak et al., 2001). Because no large-scale CCS projects have been completed in the United States, no data or reports on induced seismic activity are available. Chapter 3 reviews in more detail the CCS research and development projects ongoing in the United States, as well as three small, commercial CCS projects overseas.
HISTORICAL INDUCED SEISMICITY RELATED TO ENERGY ACTIVITIES
In the United States, seismicity caused by or likely related to energy development activities involving fluid injection or withdrawal has been documented in Alabama, Arkansas, California, Colorado, Illinois, Louisiana, Mississippi, Nebraska, Nevada, New Mexico, Ohio, Oklahoma, and Texas (see Chapters 2 and 3 for details). Appendix C lists documented and suspected cases globally and in the United States of induced seismicity, including, for example, seismic events caused by waste injection at the Rocky Mountain Arsenal (Healy et al., 1968; Hsieh and Bredehoeft, 1981; Box 1.1) and in the Paradox Basin of western Colorado (see Appendix K); secondary recovery of oil in Colorado (Raleigh et al., 1972), southern Nebraska (Rothe and Lui, 1983), western Texas (Davis, 1985; Davis and Pennington, 1989), and western Alberta (Milne, 1970) and southwestern Ontario, Canada (Mereu et al., 1986); and fluid stimulation to enhance geothermal energy extraction in New Mexico (Pearson, 1981), at The Geysers, California (see Box 3.1), and in Basel, Switzerland (see Box 3.3). Suckale (2010) provides a thorough overview of seismicity induced by hydrocarbon production. Investigations of some of these cases have led to better understanding of the probable physical mechanisms of inducing seismic events and have allowed for the establishment of some of the most important criteria that may induce a felt seismic event, including the state of stress in the Earth’s crust in the vicinity of the fluid injection or withdrawal; the presence, orientation, and physical properties of nearby faults; pore fluid pressure (pressure of fluids in the pores of the rocks at depth, hereafter referred to as pore pressure); the volumes, rates, and temperature of fluid being injected or withdrawn; the pressure at which the fluid is being injected; and the length of time over which the fluid is
11 See, for example, http://www.netl.doe.gov/technologies/carbon_seq/FAQs/carbonstorage2.html.
12 One unit of atmospheric pressure or 1 atm is equivalent to the pressure exerted by the Earth’s atmosphere on a point at sea level.
injected or withdrawn (e.g., Nicholson and Wesson, 1990). Controlled experiments both at Rangely, Colorado (Raleigh et al., 1976; see also Chapter 2), and in Matsushiro, Japan (Ohtake, 1974), were undertaken to directly control the behavior of large numbers of small seismic events by manipulation of fluid injection pressure.
Fluid withdrawal has also been observed to cause seismic events. McGarr (1991) identified three earthquakes in California caused by or likely related to extraction of oil: (1) Coalinga, in May 1983, M 6.5; (2) Kettleman North Dome, in August 1985, M 6.1; and (3) Whittier Narrows, in October 1987, M 5.9. All three events occurred in a crustal anticline close to active oil fields and on or near seismically active faults. Although seismic deformation (uplift) observed during each earthquake has been suggested to have a correlation to removal of hydrocarbon mass (McGarr, 1991), well-documented and ongoing uplift and seismicity over the entire region, related to natural adjustments of the Earth’s crust, make it difficult to determine unequivocally if these were induced seismic events. In the mid-1970s and 1980s three large earthquakes (measuring M ~ 7) were recorded near the Gazli gas field in Uzbekistan in an area that had largely been aseismic. Although precise locations and magnitudes of the earthquakes were not possible to determine, a potential relation to gas extraction was suggested based on available data and modeling (Adushkin et al., 2000; Grasso, 1992; Simpson and Leith, 1985).
Some surface effects associated with energy technologies may occur (without associated shaking at the surface) that result from surface subsidence or “creep” rather than from slip along a fault. Examples include the Baldwin Hills dam failure in California (Appendix F).
Human activity, including injection and extraction of fluids from the Earth, can induce seismic events. While the vast majority of these events have intensities below that which can be felt by people living directly at the site of fluid injection or extraction, potential exists to produce significant seismic events that can be felt and cause damage and public concern. Examination of known examples of induced seismicity can aid in determining what the risks are for energy technologies. These examples also provide data on the types of research required to better constrain induced seismicity risks and to develop options for best practices to define and alleviate risks from energy-related induced seismicity. These issues are explored in the remaining chapters of this report.
Adushkin, V.V., V.N. Rodionov, S.T. Turuntnev, and A.E. Yodin. 2000. Seismicity in the oil field. Oilfield Review Summer: 2-17.
Davis, S.D. 1985. Investigations of Natural and Induced Seismicity in the Texas Panhandle. M.S. thesis. The University of Texas, Austin. 230 pp.
Davis, S.D., and W.D. Pennington. 1989. Induced seismic deformation in the Cogdell oil field of West Texas. Bulletin of the Seismological Society of America 79(5):1477-1494.
Ellsworth, W.L., S.H. Hickman, A.L. LLeons, A. McGarr, A.J. Michael, and J.L. Rubenstein. 2012. Are seismicity rate changes in the Midcontinent natural or manmade? Abstract #12-137 presented at the Seismological Society of America (SSA) 2012 Annual Meeting, San Diego, CA, April 17-19.
Evans, D.M. 1966. The Denver area earthquakes and the Rocky Mountain Arsenal disposal well. Mountain Geologist 3(1):23-26.
Grasso, J.-R., 1992. Mechanics of seismic instabilities induced by the recovery of hydrocarbons. Pure and Applied Geophysics 139(3-4):507-534.
Hanks, T.C., and H. Kanamori. 1979. A moment magnitude scale. Journal of Geophysical Research 84(B5):2348-2350.
Healy, J.H., W.W. Rubey, D.T. Griggs, and C.B. Raleigh. 1968. The Denver earthquakes. Science 161:1301-1310.
Herrmann, R. B., S. K. Park, and C. Y. Wang. 1981. The Denver Earthquakes of 1967-1968. Bulletin of the Seismological Society of America 71:731-745.
Hsieh, P.A., and J.S. Bredehoeft. 1981. A reservoir analysis of the Denver earthquakes: A case of induced seismicity. Journal of Geophysical Research 86(B2):903-920.
Kerr, R. 2012. Learning how to NOT make your own earthquakes. Science 335:1436-1437.
McClain, W.C. 1970. On earthquakes induced by underground fluid injection. ORNL-TM-3154. Oak Ridge, TN: Oak Ridge National Laboratory.
McGarr, A. 1991. On a possible connection between three major earthquakes in California and oil production. Bulletin of the Seismological Society of America 81(3):948-970.
McGarr, A., D. Simpson, and L. Seeber. 2002. Case histories of induced and triggered seismicity. Pp. 647-661 in International Handbook of Earthquake and Engineering Seismology, Part A, edited by W.H.K. Lee et al. New York: Academic Press.
Mereu, R.F., J. Brunei, K. Morrissey, B. Price, and A. Yapp. 1986. A study of the microearthquakes of the Gobies oil field area of southwestern Ontario. Bulletin of the Seismological Society of America 76:1215-1223.
Milne, W.G. 1970. The Snipe Lake, Alberta earthquake of March 6, 1970. Canadian Journal of Earth Science 7(6):1564-1567.
Nicholson, C., and R.L. Wesson. 1990. Earthquake hazard associated with deep well injection: A report to the U.S. Environmental Protection Agency. U.S. Geological Survey (USGS) Bulletin 1951. Reston, VA: USGS. 74 pp.
Ohtake, M. 1974. Seismic activity induced by water injection at Matsushiro, Japan. Journal of Physics of the Earth 22:163-176.
Pearson, C. 1981. The relationship between microseismicity and high pore pressures during hydraulic stimulation experiments in low permeability granitic rocks. Journal of Geophysical Research 86(B9):7855-7864.
Pratt, W.E., and D.W. Johnson. 1926. Local subsidence of the Goose Creek oil field (Texas). Bulletin of the Seismological Society of America 34(7):577-590.
Raleigh, C.B., J.H. Healy, and J.D. Bredehoeft. 1972. Faulting and crustal stresses at Rangely, Colorado. American Geophysical Union Geophysics Monograph Series 16:275-284.
Raleigh, C.B., J.H. Healy, and J.D. Bredehoeft. 1976. An experiment in earthquake control at Rangely, Colorado. Science 191(4233):1230-1237.
Richter, C.F. 1936. An instrumental earthquake magnitude scale. Bulletin of the Seismological Society of America 25:1-32.
Rothe, G.H., and C.-Y. Lui. 1983. Possibility of induced seismicity in the vicinity of the Sleepy Hollow oil field, southwestern Nebraska. Bulletin of the Seismological Society of America 73(5):1357-1367.
Simpson, D.W., and W. Leith. 1985. The 1976 and 1984 Gazli, USSR, Earthquakes—Were they induced? Bulletin of the Seismological Society of America 75(5):1465-1468.
Sminchak, J., N. Gupta, C. Byrer, and P. Bergman. 2001. Issues related to seismic activity induced by the injection of CO2 in deep saline aquifers. Presented at the First National Conference on Carbon Sequestration, Washington, DC, May 15-17.
Suckale, J. 2010. Moderate-to-large seismicity induced by hydrocarbon production. The Leading Edge 29(3):310-319.
Zoback, M.D., and M.L. Zoback. 1981. State of stress and intraplate earthquakes in the United States. Science 213(4503):96-104.
Zoback, M.L., and M.D. Zoback. 1980. State of stress in the coterminous United States. Journal of Geophysical Research 85(B11):6113-6156.
Zoback, M.L., and M.D. Zoback. 1989. Tectonic stress field of the continental United States. Geophysical Framework of the Continental United States, edited by L. Pakiser and W. Mooney. Geological Society of America Memoir 172:523-539.