Blowing off steam: How clean energy could stabilize faults

Geothermal energy production, considered a clean source of power, might help stabilize an otherwise active fault system.

By Meghan Zulian, University of California, Davis, Temblor science writing extern (@meghan_zulian)

Citation: Zulian, M., 2021, Blowing Off Steam: How clean energy could stabilize faults, Temblor,

As drought and wildfires outpace previous annual records, California and the White House charge ahead with expansive clean energy projects. Clean energy like solar, wind and geothermal have inevitable risks and tradeoffs. One concern about geothermal energy is the potential for inducing earthquakes during production in fault-riddled places. One such location, California’s Coso volcanic field, is home to a large geothermal plant that produces electricity for some 250,000 homes. The plant lies in a seismically active area.

Now, a new study focused on Coso Geothermal Power Plant suggests that although energy production can induce small earthquakes, it reduces the chance of big earthquakes in the long term. A complex interplay between heat loss and water removal affect the amount of stress on local faults and may actually stabilize the shaky region. The study’s evidence? When the 2019 Ridgecrest earthquakes struck nearby and seismic stations throughout eastern California lit up with aftershocks, the area around the Coso Geothermal Power Plant remained strangely quiet.

But whether this study truly indicates that geothermal energy production can mitigate earthquakes is still up for debate.

White truck parked on a paved dirt road that is offset in two places.
Surface faulting crossing the road on the Naval Air Weapons Station China Lake (NAWSCL) from the magnitude-7.1 Ridgecrest rupture. Credit: Ben Brooks, USGS


Shake and Bake

Contrary to popular belief, not all earthquakes in California originate along the San Andreas Fault. Ridgecrest and neighboring Coso Valley, Calif., sit within a lesser known but major fault system: Walker Lane. Although this zone accommodates less motion than the San Andreas, it hosts moderate to large-magnitude earthquakes.

Earlier this summer, faults in the northern Walker Lane drummed up a magnitude-6.0 earthquake in Antelope Valley. Back-to-back 2019 Ridgecrest earthquakes of magnitude-6.4 and magnitude-7.1 were California’s biggest in more than 20 years. The largest on record in the fault zone was the massive magnitude-7.4 Owens Valley earthquake of 1872. Though still somewhat debated, some seismologists think that faults activated during this historic earthquake contributed to geothermal features in eastern California, such as the hot springs of Coso Valley.

Geothermal energy production

Geothermal features like hot springs and fumaroles prompted the late 1970s Coso Valley volcanic energy assessment. Deemed a viable energy source, Coso Geothermal Power Plant was born in 1987.

The plant uses steam produced underground to generate electricity. In active geothermal regions — where circulating fluids are warmed by hotter-than-usual crust — some naturally occurring steam escapes from below ground. Most of the steam generated, however, is trapped below the surface. By drilling wells deep below ground, steam and hot water are released to the surface.

At the Coso plant, rising steam rotates nine 30-megawatt turbine-generators, which produce electricity. The water, which cools when it reaches the surface, is subsequently injected into the subsurface to be naturally heated and returned upward again.

Injecting water into the subsurface — also done as part of fracking operations to produce oil and natural gas — is known to cause earthquakes because forcing water below ground increases the pressure in the pore space within rocks, which can cause small faults scattered in the subsurface to slip. But the water injected during geothermal energy production is replacing water that was previously removed. If the amount injected is equal to the amount pumped out, pore pressure should remain relatively consistent, and the seismicity should stay the same.

Golden colored field with mountain in background. Steam clouds coming from points on ground.
Fumaroles are a type of geothermal feature, shown here near Lake Myvatn in Northern Iceland. Credit: Matt Palmer, Unsplash.


At Coso, the amount of water re-injected into the ground has been decreasing since the mid-1980s. This should have meant fewer earthquakes in the area because pore pressure would have decreased overall. However, from the late 1980s to 2019, the area was relative active, seismically speaking. Only after the Ridgecrest earthquakes did Coso become oddly still, as surrounding regions were rocked by numerous aftershocks.

Caltech geologist Jean-Phillipe Avouac and his colleagues wanted to figure out what caused this peculiar behavior near Coso.

Ground subsidence due to energy production

Previous InSAR satellite studies — which allow scientists to compare before and after imagery to calculate how Earth’s surface changes — showed a drop in ground level of about 5 inches (13 centimeters) around the Coso Geothermal Power Plant between 1993 and 1998. Subsidence continues today. Based on this past work, Avouac and his team knew that there was both seismicity and subsidence before the Ridgecrest earthquakes.

Significant subsidence following the onset of geothermal plant operations, the authors claim, indicates the pumping and injecting of water changed the orientation and amount of stress on the faults below ground. This is not uncommon in other geothermal fields, they noted in the study.

They reckoned the surprising lack of seismicity after Ridgecrest reflected less total stress available in the fault system. Less stress would mean fewer aftershocks. Yet, exactly how this could happen in the Coso geothermal area remained unclear.

Cool clues to changing stress

At Coso, removing hot water and replacing it with colder water ultimately decreased the volume of water below ground and the surface therefore was lowered. But, the team calculated that this change in volume related to water temperature could only explain a small amount of the observed subsidence.

Plus, the fluid pressure decrease due to the net extraction of fluids should have stabilized the faults years before Ridgecrest, yet earthquakes occurred up until those events. This was an important clue that something else was driving deformation and subsidence, according to Avouac.

Injecting cold water has another important effect: cooling the rocks. “Thermal contraction [or cooling] releases normal stress, which unclamps the faults and drives seismicity,” says Avouac. To confirm this could be the case at Coso, the team included thermal contraction in their calculations.

Their results show that the Coso area experienced no major aftershocks following the 2019 quakes because geothermal electricity production had relieved much of the stress in the system.

Their results show that contraction of the rocks due to injection of cool water caused some of the faults to slip in small magnitude earthquakes (or even without producing a quake), which explains an initial increase in seismicity observed near Coso when the plant opened. These smaller events released stress on these faults. In 2019, the nearby Ridgecrest earthquakes increased stress on the faults in Coso, but not enough to overcome the decrease in stress they experienced due to cooling.

Their calculations show that this cooling effect could explain both the observed subsidence and seismicity before Ridgecrest.

Warehouse buildings and large pipes in a hilly desert landscape.
Coso Geothermal Plant. Credit: U.S. Department of Energy


Geo-engineering earthquake-less regions

For most Californians, the prospect of using clean energy to reduce earthquakes could be appealing, and according to Avouac, it’s feasible at relatively small scales. “In any area where we are producing geothermal energy, the same process would happen,” he said.

However, Emily Brodsky, a seismologist at the University of California Santa Cruz, points out that the processes stabilizing Coso’s faults still caused seismicity, particularly when the plant was just starting production, so geo-engineering could be a slippery slope. “Cooling is destressing the faults, but it may also be making faults more available to rupture,” she says. Such effects as those observed at Coso may not be ubiquitous, she says, noting that “the best test of a hypothesis in observational science is when we see it somewhere else.”


Collett, L. W. and C. (2021, May 10). The 2019 Ridgecrest, California, Earthquake Sequence. ArcGIS StoryMaps.

Guglielmi, Y., Cappa, F., Avouac, J.-P., Henry, P., & Elsworth, D. (2015). Seismicity triggered by fluid injection–induced aseismic slip. Science, 348(6240), 1224–1226.

Im, K., Avouac, J.-P., Heimisson, E. R., & Elsworth, D. (2021). Ridgecrest aftershocks at Coso suppressed by thermal destressing. Nature, 595(7865), 70–74.

Pierce, Ian. (2021, July 9). Tectonic mystery swirls as earthquake rocks California-Nevada border. Temblor.Net.

Temblor. (2016, May 19). Eastern California’s intense seismic and geothermal activity: Chicken or egg? Temblor.Net.