Addressing Induced Seismicity: Findings, Conclusions, Research, and Proposed Actions
Induced seismic activity attributed to a range of human activities has been documented since at least the 1920s. However, recent induced seismic events related to energy technology development projects that involve fluid injection or withdrawal in the United States have drawn heightened public attention. Although none of these events resulted in loss of life or significant damage, their effects were felt by local residents. These induced seismic events, though usually small in scale, can be disturbing for the public and raise concern about additional seismic activity and its consequences in areas where energy development is ongoing or planned. The findings, gaps, proposed actions, and research recommendations outlined in this chapter, based upon material presented earlier in the report, address
• the types and causes of induced seismicity;
• issues specific to each energy technology addressed in the study (geothermal energy, conventional and unconventional oil and gas production, injection wells for disposal of wastewater associated with energy development, and carbon capture and storage [CCS]);
• oversight, monitoring, and coordination of underground injection activities to help avoid felt induced seismicity;
• hazards and risk assessment; and
• best practices.
Although credible and viable research into possible induced seismic events has been conducted to date by industry, the academic community, and the federal government, further research is required because of the potential controversies surrounding such events. The Department of Energy, the U.S. Geological Survey, and the National Science Foundation are important organizations both for conducting and for supporting this kind of research and research partnerships with industry and academia. In addition to proposed actions to address induced seismicity, research recommendations are specifically highlighted in Box 7.1; some of these recommendations are specific to individual energy technologies, but most can be conducted with a purpose to understand induced seismicity more broadly.
BOX 7.1
Research Recommendations
Data Collection—Field and Laboratory
1. Collect, categorize, and evaluate data on potential induced seismic events in the field. High-quality seismic data are central to this effort. Research should identify the key types of data to be collected and data collection protocol.
2. Conduct research to establish the means of making in situ stress measurements nondestructively.
3. Conduct additional field research on microseisms in natural fracture systems including field-scale observations of the very small events and their native fractures.
4. Conduct focused research on the effect of temperature variations on stressed jointed rock systems. Although of immediate relevance to geothermal energy projects, the results would benefit understanding of induced seismicity in other energy technologies.
5. Conduct research that might clarify the in situ links among injection rate, pressure, and event size.
Instrumentation
1. Conduct research to address the gaps in current knowledge and availability of instrumentation: Such research would allow the geothermal industry, for example, to develop this domestic renewable source more effectively for electricity generation.
Hazard and Risk Assessment
1. Direct research to develop steps for hazard and risk assessment for single energy development projects (as described in Chapter 5, Table 5.2).
TYPES AND CAUSES OF INDUCED SEISMICITY
Findings
1. The basic mechanisms that can induce seismicity related to energy-related injection and extraction activities are not mysterious and are presently well understood.
2. Only a very small fraction of injection and extraction activities among the hundreds of thousands of energy development wells in the United States have induced seismicity at levels that are noticeable to the public.
3. Current models employed to understand the predictability of the size and location of earthquakes through time in response to net fluid injection or withdrawal
Modeling
1. Identify ways in which simulation models can be scaled appropriately to make the required predictions of the field observations reported.
2. Conduct focused research to advance development of linked geomechanical and earthquake simulation models that could be utilized to better understand potential induced seismicity and relate this to number and size of seismic events.
3. Use currently available and new geomechanical and earthquake simulation models to identify the most critical geological characteristics, fluid injection or withdrawal parameters, and rock and fault properties controlling induced seismicity.
4. Develop simulation capabilities that integrate existing reservoir modeling capabilities with earthquake simulation modeling for hazard and risk assessment. These models can be refined on a probabilistic basis as more data and observations are gathered and analyzed.
5. Continue to develop capabilities with coupled reservoir fluid flow and geomechanical simulation codes to understand the processes underlying the occurrence of seismicity after geothermal wells have been shut in; the results may also contribute to understanding post-shut-in seismicity in relation to other energy technologies.
Research Specific to CCS with Potential to Understand Induced Seismicity Broadly
1. Use some of the many active fields where CO2 flooding for enhanced oil recovery (EOR) is conducted to understand more about the apparent lack of felt induced seismic events in these fields; because CO2 is compressible in the gaseous phase are other factors beyond pore pressure important to understand in terms of CCS?
2. Develop models to estimate the potential earthquake magnitude that could be induced by large-scale CCS.
3. Develop detailed physicochemical and fluid mechanical models for injection of supercritical CO2 into potential storage aquifers.
require calibration from data from field observations. The success of these models is compromised in large part due to the lack of basic data at most locations on the interactions among rock, faults, and fluid as a complex system.
4. Increase of pore pressure above ambient value due to injection of fluids or decrease in pore pressure below ambient value due to extraction of fluids has the potential to produce seismic events. For such activities to cause these events, a certain combination of conditions has to exist simultaneously:
a. Significant change in net pore pressure in a reservoir
b. A preexisting, near-critical state of stress along a fracture or fault that is determined by crustal stresses and the fracture or fault orientation
c. Fault-rock properties supportive of brittle failure
5. Independent capability exists for geomechanical modeling of pore pressure, temperature, and rock stress changes induced by injection and extraction and for modeling of earthquake sequences given knowledge of stress changes, pore pressure changes, and fault characteristics.
6. The range of scales over which significant responses arise in the Earth with respect to induced seismic events is very wide and challenges the ability of models to simulate and eventually predict observations from the field.
Gaps
1. The basic data on fault locations and properties, in situ stresses, pore pressures, and rock properties are insufficient to implement existing models with accuracy on a site-specific basis.
2. Current predictive models cannot properly quantify or estimate the seismic efficiency and mode of failure; geomechanical deformation can be modeled, but a challenge exists to relate this to number and size of seismic events.
Proposed Actions
The actions proposed to advance understanding of the types and causes of induced seismicity involve research recommendations outlined in Box 7.1. These recommendations also have relevance for specific energy technologies and address gaps in understanding induced seismicity.
ENERGY TECHNOLOGIES: HOW THEY WORK
Overarching Findings for All Technologies
1. Injection pressures and net fluid volumes in energy technologies, such as geothermal energy and oil and gas production, are generally controlled to avoid increasing pore pressure in the reservoir above the initial reservoir pore pressure. These technologies thus appear less problematic in terms of inducing felt seismic events than technologies that result in a significant net increase or decrease in fluid volume.
2. The basic data needed to fully evaluate the potential for induced seismicity—including fault locations and properties, in situ stresses, fluid pressures, and rock properties—are very difficult and expensive to obtain.
3. Existing regional seismic arrays may not be capable of precisely locating small induced seismic events to determine causality and better establish the characteristics of induced seismicity.
4. Temporary local seismic arrays can be installed to find faults, determine source mechanisms, decrease error in location of seismic events, and increase resolution of future events.
GAP
Simple geometric considerations to help visualize subsurface problems and identify cases that deserve further attention are in most cases absent. Developing these kinds of simple analyses could, for example, be applied to understand the length scale affected by a single well or by multiple wells relative to depth or proximity to major faults and to the surface.
PROPOSED ACTION
In locales where a causal relationship may exist between subsurface energy activities and seismicity (even for small earthquakes of M between 3 and 4), a local seismic array should be installed for seismic monitoring. An appropriate body to determine whether such an array is necessary may be the permitting agency for the well(s) thought to be involved in the seismicity. Installation of such an array may require significant resources (including instrumentation and analysis). Existing groups, such as the U.S. Geological Survey, national laboratories, state geological surveys, universities, and private companies have the expertise necessary to install arrays and conduct the necessary analyses. Full disclosure of the data and results of such monitoring is required.
Geothermal Energy
FINDINGS
1. The induced seismic responses to injection differ in cause and magnitude with each of the three different forms of geothermal resources. At the vapor-dominated Geysers field hundreds of earthquakes of M 2 or greater are produced annually with one or two of M 4, all apparently caused principally by cooling and contraction of the reservoir rocks. The liquid-dominated field developments generally cause little if any induced seismicity because the water injection typically replaces similar quantities of fluid extracted at similar pressures and temperatures. The high-pressure hydraulic fracturing into generally impermeable rock associated with the stimulation operations at enhanced geothermal systems (EGS) projects can cause hundreds of small microseismic events and an occasional earthquake of up to M 3 due mainly to the imposed increased fluid pressures.
2. The mitigation of the effects of induced seismicity is in some instances clearly necessary to maintain or to restore public acceptance of the geothermal power generation activities. The early use of a “best practices” protocol and a “traffic light” control system indicates that such measures can provide an effective means to control operations so that the intensity of the induced seismicity is within acceptable levels. Further information on implementation of a protocol and control system is outlined under the final section in this chapter, Best Practices.
GAPS
1. Suitable coupled reservoir fluid flow and geomechanical simulation codes are not currently available to understand the processes underlying the occurrence of seismicity after geothermal wells have been shut in (ceased operation).
2. Field operators currently do not have ready access to downhole temperature and pressure recording instruments capable of making accurate measurements where reservoir conditions reach 750°F.
PROPOSED ACTIONS
1. Adopt and use a matrix-style “best practices” protocol by developers as outlined in Chapter 6: Such a protocol is appropriate to use in those cases where there is a known probability of inducing seismicity at levels that could pose a concern to the public. In those cases where induced seismicity occurs but was previously unanticipated, the developer should consider adopting the protocol procedures needed to complete the project in a manner more satisfactory to the public.
2. Fully disclose and discuss a “traffic light” system in a public forum prior to the start of operations when such a system is to be adopted or imposed. Such disclosure and discussion will ensure that these safeguards are clearly known and understood by all concerned.
Conventional Oil and Gas Development Including Oil and Gas Withdrawal, Secondary Recovery, and Enhanced Oil Recovery
FINDINGS
1. Generally, withdrawal associated with conventional oil and gas recovery has not caused significant seismic events; however, several major earthquakes have been associated with conventional oil and gas withdrawal.
2. Relative to the large number of waterflood projects for secondary recovery, the small number of documented instances of felt induced seismicity suggests such projects pose relatively small risk for events that would be of concern to the public.
3. The committee has not identified any documented, felt induced seismic events associated with EOR (tertiary recovery). The potential for induced seismicity is low in EOR operations as pore pressure is not significantly increased beyond the original levels in the reservoir because injected fluid volumes tend to be balanced by fluid withdrawals.
Unconventional Oil and Gas: Hydraulic Fracturing for Shale Gas Development
FINDINGS
1. The process of hydraulic fracturing a well as presently implemented for shale gas recovery does not pose a high risk for inducing felt seismic events. Thirty-five thousand wells have been hydraulically fractured for shale gas development to date in the United States. To date, hydraulic fracturing for shale gas production was cited as the possible cause of one case of felt seismic events in Oklahoma in 2011, the largest of which was M 2.8. The quality of the event locations was not adequate to fully establish a direct causal link to the hydraulic fracture treatment. Hydraulic fracturing for shale gas development has been confirmed as the cause of induced seismic events in one case worldwide—in Blackpool, England (maximum M 2.3).
2. One case of induced seismicity (maximum M 1.9) was documented in Oklahoma in the late 1970s as being caused by hydraulic fracturing for oil and gas development for conventional oil and gas extraction.
PROPOSED ACTION
When a seismic event occurs that appears to be associated with hydraulic fracturing and is considered to be a concern to the health, safety, and welfare of the public, an assessment is needed to understand the causes of the seismicity (see protocol that follows).
Injection Wells for the Disposal of Water Associated with Energy Extraction
FINDINGS
1. The United States currently has approximately 30,000 Class II wastewater disposal wells; very few felt induced seismic events have been reported as either caused by
or likely related to these wells. Rare cases of wastewater injection have produced seismic events, typically less than M 5.0.
2. Injected fluid volume, injection rate, injection pressure, and proximity to existing faults and fractures are factors that determine the probability to create a seismic event. High injection volumes in the absence of corresponding extractions may increase pore pressure and in proximity to existing faults could lead to an induced seismic event.
3. The area of potential influence from injection wells may extend over several square miles, and induced seismicity may continue for months to years after injection ceases.
4. Reducing the injection volumes, rates, and pressures has been successful in decreasing rates of felt seismicity in cases where events have been induced.
5. Evaluating the potential for induced seismicity in the location and design of injection wells is difficult because no cost-effective way to locate unmapped faults and measure in situ stress currently exists.
GAPS
1. Effective and economical tools are not available to accurately predict induced seismic activity prior to injection.
2. No capability exists to predict exactly how reducing volumes, pressures, and rates can lead to reduction in seismicity after it has begun. The models discussed in Chapter 2 are critical to developing the capacity to make such predictions.
PROPOSED ACTIONS
The actions proposed by the committee to address the potential for induced seismicity related to injection wells for disposal of wastewater are similar to those suggested for geothermal energy technologies:
1. The adoption and use of a matrix-style “best practices” protocol as outlined in Chapter 6 in those cases where there is a known probability of inducing seismicity at levels that could pose a concern to the public. In those cases where the need becomes apparent only after disposal has begun, the developer should adopt the protocol procedures needed to complete the project in a manner that protects public safety.
2. When a “traffic light” system is to be adopted or imposed to control operations that could cause unacceptable levels of induced seismicity, full disclosure and discussion of the system at a public forum is necessary prior to the start of opera-
tions. Knowledge and understanding of these safeguards by all concerned are of great importance. Further information is outlined under the final section in this chapter, Best Practices.
Carbon Capture and Storage
FINDINGS
1. The only long-term (~14 years) commercial CO2 sequestration project in the world at the Sleipner field off the shore of Norway is of a small scale relative to commercial projects proposed in the United States. Extensive seismic monitoring at this offshore site has not indicated any significant induced seismicity.
2. Proposed injection volumes of liquid CO2 in large-scale sequestration projects (> 1 million metric tonnes per year) are much larger than those associated with the other energy technologies currently being considered. There is no experience with fluid injection at these large scales and little data on seismicity associated with CO2 pilot projects. If the reservoirs behave in a similar manner to oil and gas fields, these large volumes have the potential to increase the pore pressure over vast areas. Relative to other technologies, such large affected areas may have the potential to increase both the number and the magnitude of seismic events.
3. CO2 has the potential to react with the host/adjacent rock and cause mineral precipitation or dissolution. The effects of these reactions on potential seismic events are not understood.
GAPS
1. The short- and long-term effects of supercritical CO2 in influencing rock strength and rock slip strength are not well understood.
2. The potential earthquake magnitudes that can be induced by the injection volumes being proposed for CCS are not known.
3. The complexities of hydrochemical-mechanical effects on CO2 injection and storage are not thoroughly understood.
PROPOSED ACTIONS
Because of the lack of experience with large-scale fluid injection for CCS, continued research supported by the federal government is needed on the potential for induced seismicity in large-scale CCS projects. Some specific research recommendations are outlined in Box 7.1. As part of a continued research effort, collaboration between federal agencies
and foreign operators of CCS sites is important to understand induced seismic events and their effects on CCS operations.
Findings
1. Induced seismicity may be produced by a number of different energy technologies and may result from either injection or extraction of fluid. As such, responsibility for oversight of activities that can cause induced seismicity is dispersed among a number of federal and state agencies.
2. Recent, potentially induced seismic events in the United States have been addressed in a variety of manners involving local, state, and federal agencies, and research institutions. These agencies and research institutions may not have resources to address these unexpected events, and more events could stress this ad hoc system.
3. Currently the Environmental Protection Agency (EPA) has primary regulatory responsibility for fluid injection under the Safe Drinking Water Act; however, this act does not explicitly address induced seismicity. EPA appears to be addressing the issue of induced seismicity through a current study in consultation with other federal and state agencies.
4. The U.S. Geological Survey (USGS) has the capability and expertise to address monitoring and research associated with induced seismic events. However, the scope of its mission within the seismic hazard assessment program is focused on large-impact, natural earthquakes. Significant new resources would be required if the USGS mission is expanded to include comprehensive monitoring and research on induced seismicity.
Gap
Mechanisms are lacking for efficient coordination of governmental agency response to seismic events that may have been induced.
Proposed Actions
1. In order to move beyond the current ad hoc approach for responding to induced seismicity, relevant agencies including EPA, USGS, land management agencies, and possibly the Department of Energy, as well as state agencies with authority and relevant expertise (e.g., oil and gas commissions, state geological surveys, state
environmental agencies, etc.) should consider developing coordination mechanisms to address induced seismic events that correlate to established best practices (see recommendation below).
2. Appropriating authorities and agencies with potential responsibility for induced seismicity should consider resource allocations for responding to induced seismic events in the future.
Gap
Currently, methods do not exist to implement assessments of hazards upon which risk assessments depend. The types of information and data required to provide a robust hazard assessment would include
• net pore pressures, in situ stresses, and information on faults;
• background seismicity; and
• gross statistics of induced seismicity and fluid injection for the proposed site activity.
Proposed Actions
1. A detailed methodology should be developed for quantitative, probabilistic hazard assessments of induced seismicity risk. The goals in developing the methodology would be to
• make assessments before operations begin in areas with a known history of felt seismicity and
• update assessments in response to observed induced seismicity.
2. Data related to fluid injection (well location coordinates, injection depths, injection volumes and pressures, time frames) should be collected by state and federal regulatory authorities in a common format and made accessible to the public (through a coordinating body such as the USGS).
3. In areas of high density of structures and population, regulatory agencies should consider requiring that data to facilitate fault identification for hazard and risk analysis be collected and analyzed before energy operations are initiated.
Findings
1. The DOE Protocol for EGS, which lists seven sequential steps, provides a reasonable initial model for dealing with induced seismicity that can serve as a template for other energy technologies.
2. Based on this initial model, the committee has proposed two matrix-style protocols as examples to illustrate the manner in which these seven activities can ideally be undertaken concurrently (rather than only sequentially), while also illustrating how these activities should be adjusted as a project progresses from early planning through operations to completion.
Gap
No best practices protocol for addressing induced seismicity is generally in place for each of these technologies, with the exception of the protocol recently developed for EGS. The committee suggests that best practices protocols be adapted and tailored to each technology to allow continued energy technology development. Actions toward developing these protocols are outlined below.
Proposed Actions
1. A matrix-style “best practices” protocol should be developed in coordination with the permitting agency or agencies by experts in the field of each energy technology, including EOR, shale gas production, and CCS.
2. The adoption and use of such protocols by developers are recommended in each case where there is a known or substantial probability of inducing seismicity at levels that could pose a concern to the public. In cases where induced seismicity becomes an issue at some stage in the project, the developer can adopt the protocol procedures needed to continue the project in a manner more satisfactory to the public.
3. Even with the adoption and use of a best practices protocol, induced seismicity of serious concern to public health and safety may occur. The regulatory body affiliated with the permitting of well(s) should include, as part of each project’s operation permit, a mechanism (such as a “traffic light” mechanism) for the well operator to be able to control, reduce, or eliminate the potential for felt seismic events.
4. When a traffic light system is to be adopted or imposed to control operations that may cause unacceptable levels of induced seismicity, full disclosure and discussion
of the adopted system at a public forum prior to the start of operations is advised so that these safeguards are clearly known and understood by all concerned. Simultaneous development of public awareness programs by federal or state agencies in cooperation with industry and the research community could aid the public and local officials in understanding and addressing the risks associated with small-magnitude induced seismic events.
