Earthquakes in stable regions seem paradoxical. In areas such as Groningen, the Upper Rhine Valley, and Utah, the shallow crust is considered “velocity-strengthening,” meaning it resists sudden slip. According to traditional geology, faults in these zones should creep slowly, not rupture violently.
Yet seismic networks show otherwise. These regions have produced small to moderate earthquakes for decades, shaking ground that should be motionless. Such tremors are often linked to gas extraction, fluid injection, or geothermal drilling. Each event seems to defy textbook logic, prompting scientists to ask: what reawakens these sleeping faults?
“Faults can be found almost everywhere. Faults in the shallow subsurface are usually stable, so we do not expect shock movements to occur along them,” said Dr. Ylona van Dinther, who supervised the Utrecht study. Her team focused on faults buried just a few kilometres deep, far from any tectonic boundary. They found that, over time, these faults don’t weaken; they actually become stronger.
This quiet strengthening happens over millions of years. As mineral grains along the fault line weld together, friction increases, forming an invisible store of potential energy. When modern human activities change the stress balance, for instance, when gas or fluids are extracted, that locked-up strength can finally give way.
The hidden process of fault healing
Although inactive faults appear motionless, their surfaces are far from inert. Through a process known as fault healing, tiny mineral grains gradually bond and recrystallize, increasing the fault’s resistance to slip. The longer a fault remains dormant, the more cohesive and strong it becomes.
Dr. Van Dinther explained: “Although these faults do not move, we still observe a very slow growth of the surface that connects them. This sort of fault healing gives rise to additional strength. It is this extra fault strength that can cause an acceleration once a fault has been set in motion.”
Using numerical models that spanned millions of years, Utrecht scientists simulated how friction evolves in long-resting faults. Their results showed that after roughly 30 million years of inactivity, a fault could gain up to 0.25 in static friction, equal to about 4 megapascals (MPa) of extra strength. When disturbed by pressure changes caused by human operations, this stored energy is released in one short burst, an earthquake.
The study demonstrated that even faults considered “velocity-strengthening” can become unstable once healed. The first slip releases the built-up energy, after which the fault returns to stable motion. In other words, the quake is a one-time event: the fault cannot store that same energy again in human timescales.
Why shallow earthquakes are so damaging
The earthquakes that occur in stable regions differ from typical tectonic ones. They are much shallower, often only 1 to 4 km (0.6 to 2.5 miles) below the surface. That proximity means shaking at the surface is stronger, even if the quake’s magnitude is moderate.
“As such areas do not have a history of earthquakes, the people living there are more at risk as infrastructure has not been built to withstand earthquakes,” said Van Dinther. She noted that this makes induced quakes potentially more damaging than deeper natural ones of similar size.
Induced earthquakes also occur in different locations from natural ones. Utrecht’s study included a global illustration showing natural quakes clustered along tectonic boundaries, while induced quakes appeared as red dots scattered across supposedly stable interiors of continents. These include geothermal fields in France, gas reservoirs in the Netherlands, and waste injection sites in the United States.
Because these earthquakes arise in regions not designed for seismic safety, even small events can cause significant structural damage. The shaking is short but intense, affecting buildings, pipelines, and industrial facilities. In Groningen, repeated tremors led to the costly reinforcement of thousands of homes and ultimately to the government’s decision to shut down gas production.
Understanding the shallow depth of such events is key to mitigating risk. Seismic hazard models must now account not only for plate boundaries but also for areas where long-dormant faults might be reactivated by human activity.
The Groningen paradox resolved
For decades, Groningen has puzzled scientists. Its subsurface sandstone behaves as a velocity-strengthening material, which should resist sudden fault slip. Yet since the 1980s, earthquakes up to magnitude 3.6 have struck the area, damaging homes and infrastructure.
The Utrecht study offers an explanation. Groningen’s faults had been inactive for between 30 and 65 million years, long enough for their surfaces to heal significantly. When gas extraction began in the early 1960s, the pressure drop within the reservoir, several MPa, destabilized these healed faults.
Model simulations by Utrecht researchers showed that this process would lead to the first major earthquake roughly 35 years after production began. In reality, the first damaging event occurred in 2012, about 49 years later, remarkably close to the prediction.
“This potential acceleration, in the form of an earthquake, occurs only once,” Van Dinther said. “As soon as that extra fault strength, which has been built up over millions of years, finds a way out, the situation becomes stable again. As a result, there is no more earthquake activity at that spot.”
After the initial rupture, the same fault transitions into a stable sliding mode, where it slowly releases stress without generating new quakes. This finding means that while induced earthquakes are serious, their recurrence potential at any single healed fault is very limited.
Lessons for geothermal and energy storage safety
These findings carry major implications for the future of energy and subsurface technology. Projects involving geothermal heat extraction, hydrogen storage, or CO2 sequestration all involve injecting or removing fluids underground. Such operations can alter pressures and stresses around ancient faults.
If these faults have been inactive for millions of years and have healed significantly, they can rupture once when reactivated. This underscores the need for detailed geological screening before any project begins. Determining a fault’s last movement, mineral composition, and healing rate can help identify whether it poses a one-time risk or is likely to remain stable.
Van Dinther emphasized that understanding the healing process is essential for safe energy use. “We need to gain a deeper understanding of the behaviour of faults — will they accelerate or slow down? — the role of fault healing, and how this translates into movements on the fault,” she said. “Then we will also be able to better estimate and anticipate one-off risks, and improve how we communicate this information.”
Utrecht University has already begun developing new calculation models that combine geological dating, laboratory friction data, and computer simulations. These models aim to predict when and where induced earthquakes might occur and to quantify the resulting ground motion. The ultimate goal is to make geothermal and energy storage projects both sustainable and safe.
A quiet crust, not a dead one
The broader message of the research is that Earth’s crust is more dynamic than it appears. Even regions that seem geologically silent can harbor faults capable of sudden movement if given the right trigger. But these are not endlessly repeating dangers.
Once a healed fault slips, it becomes part of the system’s self-regulating behavior. The surface may continue to creep slowly, releasing stress in a harmless way. The same fault is unlikely to host another large earthquake for millions of years.
This understanding shifts how scientists view seismic hazard. Instead of categorizing regions as simply active or stable, they now recognize that stability depends on history. A fault’s behavior reflects both its geological past and its present human influence.
The Earth, in this sense, remembers. Its faults carry records of time, silence, and stress. And when humans disturb that balance, even the quietest ground can briefly roar back to life.
References:
1 Why earthquakes sometimes still occur in tectonically silent regions – Utrecht University – October 27, 2025
2 Analysis of Shear Wave Splitting Patterns in Alaska: Evidence for Strong Intra-Slab Anisotropy – Meng Li et al. – Nature Communications – October 15, 2025 – https://doi.org/10.1038/s41467-025-63482-3 – OPEN ACCESS