Overview

Increasing Rate of Earthquakes Beginning in 2009

Annual number of earthquakes with a magnitude of 3.0 or larger in the central and eastern United States, 1970–2016. The long-term rate of approximately 25 earthquakes per year increased sharply starting around 2009. (Public domain.)

The number of earthquakes in the central U.S. has increased dramatically over the past decade. Between the years 1973–2008, there was an average of 25 earthquakes of magnitude three and larger in the central and eastern United States. Since 2009, the average number of M3 earthquakes has jumped to 362 per year. The rate peaked in 2015 with 1010 M3+ earthquakes. Since 2015 the earthquake rate has declined, with 690 and 364 M3+ earthquakes in 2016 and 2017, respectively. Nonetheless, this rate is far higher than the average of 25 earthquakes per year. Most of these earthquakes are in the magnitude 3–4 range—large enough to have been felt by many people—yet small enough to rarely cause damage. Damage has been caused by some of the larger events, including the M5.8 Pawnee and M5.0 Cushing Oklahoma earthquakes that occurred in 2016.

This increase in earthquakes prompts two important questions:

• Are they natural, or man-made?
• What should be done in the future as we address the causes and consequences of these events to reduce associated risks?

Myths and Misconceptions

Cartoon illustration showing oil production, wastewater disposal, and enhanced oil production.(Public domain.)

Fact 1: Fracking is not directly causing most of the induced earthquakes. Disposal of waste fluids that are a byproduct of oil production is the primary cause of the recent increase in earthquakes in the central United States.

Wastewater disposal wells typically operate for longer durations and inject much more fluid than hydraulic fracturing, making them more likely to induce earthquakes. In Oklahoma, which has the most induced earthquakes in US, only 1-2% of the earthquakes can be linked to hydraulic fracturing operations. The remaining earthquakes are induced by wastewater disposal.

Fact 2: Not all wastewater injection wells induce earthquakes.

Most injection wells are not associated with felt earthquakes. A combination of many factors is necessary for injection to induce felt earthquakes. These include: the injection rate and total volume injected; the presence of faults that are large enough to produce felt earthquakes; stresses that are large enough to produce earthquakes; and the presence of pathways for the fluid pressure to travel from the injection point to faults.

Fact 3: Wastewater is produced at all oil wells, not just hydraulic fracturing sites.

Most wastewater currently disposed of across the nation is generated and produced in the process of oil and gas extraction. As discussed above, saltwater is produced as a byproduct during the extraction process. This wastewater is found at nearly every oil and gas extraction well.

The other main constituent of wastewater is leftover hydraulic fracturing fluid. Once hydraulic fracturing is completed, drilling engineers extract the fluids that are remaining in the well. Some of this recovered hydraulic fracturing fluid is used in subsequent fracking operations, while some of it is disposed of in deep wells.

Fact 4: The content of the wastewater injected in disposal wells is highly variable.

In many locations, wastewater has little or nothing to do with hydraulic fracturing. In Oklahoma, less than 10% of the water injected into wastewater disposal wells is used hydraulic fracturing fluid. Most of the wastewater in Oklahoma is saltwater that comes up along with oil during the extraction process.

In contrast, the fluid disposed of near earthquake sequences that occurred in Youngstown, Ohio, and Guy, Arkansas, consisted largely of spent hydraulic fracturing fluid.

Fact 5: Induced seismicity can occur at significant distances from injection wells and at different depths.

Seismicity can be induced at distances of 10 miles or more away from the injection point and at significantly greater depths than the injection point.

Fact 6: Wells not requiring surface pressure to inject wastewater can still induce earthquakes.

Wells where you can pour fluid down the well without added pressure at the wellhead still increase the fluid pressure within the formation and thus can induce earthquakes.

See also:

• FAQs: Induced Earthquakes
• Myths and Facts on Wastewater Injection, Hydraulic Fracturing, Enhanced Oil Recovery, and Induced Seismicity - by Rubinstein and Mahani

Observational Studies

Bryant Platt digs a hole to install seismometers at a home in southern Kansas. Seismometers are in the foreground. (Public domain.)

In response to sudden changes in seismicity that are potentially induced by human activity, the USGS may deploy temporary seismic stations to better understand the earthquakes. These deployments typically consist of 2-15 seismometers placed in the immediate vicinity of the seismicity.

Temporary seismic stations allow us to accurately pinpoint the location of the seismicity including small earthquakes that are not recorded by permanent US seismic networks. With the high-quality data from temporary seismic stations we are able to precisely locate the seismicity, identify activated fault structures, determine the faulting style of the earthquakes, and monitor for migration of the seismicity with time. All of these characteristics help us understand if the earthquakes are induced and the physical processes causing them.

The catalog of earthquakes generated by the temporary seismic stations provides key data for basic research to understand the mechanics of induced earthquakes. Scientists commonly use computer models to simulate the physical processes believed to be responsible for inducing earthquakes to better understand why they occur and how they might be prevented. The USGS also uses the observations from these seismic networks to inform forecasts of earthquake hazard due to induced earthquakes.

As part of these observational studies, the USGS works closely with scientists at local universities, state geological surveys, state regulatory agencies, and other scientific institutions. Working with these partners is critical for our work. They bring local knowledge and understanding of these areas, and complement the technical capabilities of the USGS. Local partners frequently assist or lead our deployments of temporary seismic stations, as they are often able to deploy the instrumentation more quickly than scientists from the USGS offices in California, Colorado, and New Mexico. Local homeowners are also critical for our studies, as they often host our seismic monitoring equipment.

Map of current USGS temporary seismic deployments to monitor potentially induced earthquakes. (Public domain.)

Hazard Estimation

Cumulative number of M ≥ 2.7 earthquakes in the central and eastern United States (CEUS) and five select zones of induced seismicity since 1980. (Public domain.)

The USGS National Seismic Hazard Model (NSHM) forecasts earthquake hazard, or the potential strength and frequency of ground shaking from future earthquakes, in the continental United States. The USGS releases a forecast about every six years, the last one was released in 2014. The forecasts are used in the design of buildings, bridges, highways, and other structures. They also provide critical information about areas of higher earthquake hazard for use by governmental disaster management agencies, industry, and the public for use in developing earthquake risk reduction plans and actions.

In previous NSHM forecasts, earthquakes that were attributed to human activity, including induced earthquakes from underground fluid injection or extraction, were not included. However, because of the recent increase in induced earthquakes in some areas of the central and eastern United States (CEUS) (e.g., Oklahoma) (Ellsworth, 2013) (Figure 1) and because the largest induced earthquakes have caused damage to buildings and other structures, induced earthquakes now need to be considered. Most induced earthquakes in the CEUS are thought to be caused by deep wastewater-disposal related to industrial activity (Andrews and Holland, 2015).

One-Year Forecasts

Chance of damage forecasted from an earthquake during 2016. (Public domain.)

Chance of damage forecasted from an earthquake during 2017. (Public domain.)

In March 2016, the USGS released its first induced earthquake hazard model, the 2016 One-Year Seismic Hazard Forecast for the CEUS from Induced and Natural Earthquakes. This model forecasted the strength and frequency of potential ground shaking from future induced and natural earthquakes for a one-year period, based primarily on earthquake data from the previous year (Figure 2). Areas of high induced earthquake hazard were identified in Oklahoma-southern Kansas, the Raton Basin (CO/NM border), north-central Texas, and north-central Arkansas. Near some areas of these active induced earthquakes, hazard is higher than in the 2014 NSHM by more than a factor of 3 and is similar to the chance of damage caused by natural earthquakes at sites in parts of California.

In March 2017, the USGS released an updated induced earthquake forecast for the CEUS for 2017. The 2017 one-year forecast uses the same earthquake hazard model and methodology as the 2016 one-year forecast, but incorporates an updated earthquake catalog that includes earthquakes from 2016 (Figure 3). The 2017 forecasted earthquake rates are lower in regions of induced earthquakes due to lower rates of earthquakes in 2016 compared to 2015, which may be related to decreased wastewater injection, caused by regulatory actions or by a decrease in unconventional oil and gas production. Nevertheless, the 2017 forecasted earthquake hazard is still significantly elevated in areas of the CEUS compared to the earthquake hazard calculated from only natural earthquakes in the 2014 NSHM.

Continuing Research

Current research activities include: development of new ground motion models (GMMs) specifically for induced earthquakes, analysis of catalog statistics, assessment of differences in induced and natural earthquake sources, estimation of potential induced earthquake magnitudes, consideration of alternative seismic rate models that are constrained by physics, and tests of models to identify parameters are all areas where research could improve the methodology. The USGS welcomes feedback from industry, academia, and users of the forecasts as this work continues.

Non-USGS Publications

• Andrews, R. D. and A. Holland (2015). Statement on Oklahoma seismicity, April 21, 2015, Oklahoma Geological Survey, 2 p. (last accessed February 2017). (PDF)

Numerical Modeling

The USGS uses computer simulations to evaluate the physical relationships between fluid injection (or extraction) and earthquakes. We can only indirectly study these relationships using observations, so computer simulations help us gain a better physical understanding for the processes that are likely responsible for occurrence of these earthquakes. Scientists can also use these simulations to understand complex patterns in seismicity collected by either temporary or permanent seismic stations.

One Earth process the USGS simulates is how fluids flow from an injection well. Simulating fluid flow allows us to see how the fluid pressure underground changes in response to injection. Fluid pressure increases within faults are believed to be the main cause of induced earthquakes. A simulation of a scenario similar to injection in Oklahoma shows a detectable rise in fluid pressure out to 5 miles away from the well. After ten years that same pore pressure change can be seen at nearly 15 miles. Based on simulations like this, the USGS can evaluate whether pressure changes at a given location such as a known fault would be large enough to cause an earthquake.

The USGS also models many other physical processes to better understand induced earthquakes. These include: temperature, pressure, and stress changes associated with geothermal energy production; observable surface deformation associated with reservoir compaction around geothermal plants, and injection of liquid carbon dioxide at carbon sequestration plants.

Simulation of fluid pressure increases from injection into a reservoir over time. (Public domain.)