The enormous load exerted by dam reservoirs resulted in the strongest ever induced earthquake. An interview with Prof. Dr. Marco Bohnhoff.
The Earth's crust is nearly everywhere in a state of critical equilibrium with respect to "stored" tension. Any slight tipping of the balance between stability and stress leads to the rupture we experience as an earthquake. For centuries, people have at least been aware of the phenomenon of human-induced earthquakes that occur during mining and the extraction of natural resources such as potash, coal, crude oil or natural gas. Whether man injects pressurised fluids into deep wells or extracts hydrocarbons, the manipulation of the subsurface can disturb the critical equilibrium. But earthquakes can also be induced by activities taking place on the Earth's surface, such as the creation of artificial lakes behind dams and the enormous load they exert on the ground. Of course it can happen that nothing happens and the Earth remains calm. Predicting or let alone controlling induced earthquakes is a huge challenge. Gaining a better understanding of both the risks and the significant potentials involved would require conducting a large number of deep geothermal testing projects.
Prof. Dr. Marco Bohnhoff is a geophysicist who has spent more than 20 years of work and research in the field of natural and induced seismicity. He heads the section for 'Geomechanics and Rheology' (rock mechanics and material deformation and flow) at the German Research Centre for Geosciences and teaches experimental and borehole seismology at the Free University of Berlin. Prof. Dr. Bohnhoff is also conducting intensive research of the seismic cycle at the North Anatolian Fault in Turkey.
1. How can you as a scientist recognise whether an area is prone to earthquakes triggered by human activity?
With natural earthquakes we know that about 90% occur at tectonic plate boundaries and the remaining 10% in the interior of the plates. With induced earthquakes there is a clear spatial and temporal connection between their occurrence and human activities such as mining, the filling of dam reservoirs, or fluid injection for reservoir stimulation/fracking or underground storage.
But our experience in recent decades, starting with the insights gained from the 9 kilometre German Continental Deep Drilling Programme to the latest large round of induced earthquakes linked to shale gas development in the United States, has shown that the Earth's crust everywhere is basically in a state of critical equilibrium. So depending on how we intervene in the geological substratum, it is theoretically possible that induced earthquakes can occur anywhere, but it is not inevitable.
2. In which rock formations are earthquakes most likely to occur in Germany? Which regions are particularly threatened?
Natural earthquake activity is on the whole very low in Germany since we do not lie on an active plate boundary. The natural earthquakes that do occur in Germany are primarily seen in the Eifel region, in the southwest along the Rhine Plain, in the Alpine foothills in the south and in the Vogtland region in the southeast. The seismicity along the Rhine plain is associated with a 40 million old tectonic plate boundary that is now more or less inactive. The earthquake activity in the Alpine foothills is the result of the continuing mountain range formation process driven by Africa as the continent moves northward. Finally, the seismicity in the Vogtland region is manifested as so-called 'earthquake swarms' that are associated with formerly active volcanoes in the area.
When and where induced earthquakes occur does not depend on geography. More important is the type of rock formation that is disturbed by underground activity. For example, hydraulic stimulation or fracking operations performed in sedimentary layers are not very likely to induce any noticeable seismicity. The reason is that the rock strength in such formations is generally lower than in crystalline basement rock, and the presence of critically pre-stressed faults of sufficient size is less likely. Most induced earthquakes with magnitudes greater than 5 occur in the crystalline Earth's crust and not in sedimentary basins.
3. What exactly do you look at when you perform model calculations?
Wherever possible, we base our findings both on model calculations and on empirical measurement data from the lab and from the field. Most underground systems are highly complex, and we need to take a holistic and multi-scale approach if we are to get closer to the goal of understanding these processes in the nearest possible future well enough to be able to anticipate the occurrence of induced earthquakes, and of course natural ones as well. The ultimate aim of our research is and will remain to fully understand these processes and even control them. The point where an earthquake originates will always be hidden from us, and individual (and very expensive) boreholes drilled in the focal region unfortunately represent nothing more than stabs in the dark. This is what makes predicting earthquakes, be they natural or induced, so difficult, unlike weather forecasting, where scientists can directly observe the system they are studying.
4. At what magnitude do people even notice an earthquake? Many earthquakes occur that are not even felt by the people who live in the region.
It's always difficult to state a specific magnitude value. This has to do with the fact that the magnitude is a measure of the energy released underground at the hypocentre. An earthquake of magnitude 3 that occurs at a depth of 3 to 4 kilometres will certainly be felt on the Earth's surface, but the same quake occurring 10 kilometres down would go unnoticed. This is why we have now switched to defining limits for acceptable tremors based on the ground motion measured at the surface.
5. When talking about induced earthquakes, what important events have occurred in Germany? Which events were important from a scientific standpoint?
A large portion of induced earthquakes in Germany occur in connection with salt and coal mining, a phenomenon that has been known and lived with for centuries. Magnitudes of up to 5.6 have been seen in isolated cases in connection with salt extraction, while coal mining seldom produces magnitudes greater than 4. Beyond that, there have in recent decades been a few cases of induced seismicity caused by gas production and deep geothermal projects.
The world's largest ever event clearly recognised as an induced earthquake was the one at the Koyna Dam in western India with a magnitude of 6.5. Some 200 people lost their lives as a result of this earthquake. An international research project is currently working at the site in an attempt to drill a 6 to 7 kilometre borehole into the earthquake's fracture zone in order to better understand what geomechanical parameters are decisive in causing such man-made earthquakes.
6. Are there activities that are especially risky in terms of inducing earthquakes?
As a general rule we can say that the probability of strong induced earthquakes correlates with the volume of fluid injected. So statistically speaking, the longer that water, for example, is injected into the ground the more likely it is that noticeable seismicity will be induced. This correlation is currently being seen in the United States, where most induced earthquakes are caused not by fracking per se but rather during the high-pressure injection of large quantities of water or during the secondary or tertiary recovery of conventional hydrocarbons. The decisive question is always whether pre-existing faults are present in the subsurface that may have accumulated stress caused by tectonic motion and which therefore could produce perceivable earthquakes. In the great majority of cases, the energy released through induced seismicity was already stored prior to the inducing activity. The additional pressure produced by humans is then simply the final trigger that releases the energy.
7. Smaller induced earthquakes frequently occur during the extraction of raw materials. Why does natural gas extraction top the list here?
We generally understand why induced seismicity occurs: There is a direct connection to changes in underground stresses caused by human actions. Seismicity can be triggered by the withdrawal of a volume of subsurface material or through the injection or storage of fluids underground. We cannot, therefore, say that earthquakes occur primarily in connection with natural gas production. This is currently a hot topic in the Groningen gas field in the Netherlands because several noticeable earthquakes of around magnitude 3 occurred during the many years of gas production at the site. But the larger induced earthquakes are more closely linked to the injection of large quantities of pressurised water – similar to the water injection done in the United States - or to large overlying loads on the surface, as seen for example at the Koyna dam in India. In such cases, magnitudes of 5 or even higher are reached, which can often result in serious seismic hazards.
8. Hydraulic stimulation involves the high-pressure injection of water into a rock bed in order to fracture the rock or enlarge existing fractures. What pressures can be produced during this procedure? Can you provide a comparison to illustrate how high such pressures are?
During hydraulic stimulation, or fracking as it is called, the pressures used are immense and may reach several hundred bar. Such pressures are required to increase the permeability in the deposits found several kilometres down with the aim of bringing hydrocarbons or hot water to the surface. To illustrate it helps to know that water pressure increases by about 0.1 bar per metre of depth. You can feel this pressure in your ears even in shallow water. So anyone who has ever gone diving can imagine what several hundred bar would mean.
9. Deep geothermal energy (400 metres and deeper): Is an earthquake more likely to occur during drilling or during the hydraulic or chemical treatment of rock formations? What is the most critical moment in the process?
We can't talk about 'the one moment' or 'the one earthquake'. Thousands of small earthquakes take place during hydraulic stimulation or fracking, reflecting the fact that the idea here is to create flow paths in the reservoir. These microquakes can neither be felt nor measured on the surface, but they are extremely important because they produce the tiny shear movements that create the underground flow paths in the first place. In purely statistical terms, the number of earthquake events increases about ten fold at each higher step of the magnitude scale. So for every one thousand microquakes of magnitude 0 about one magnitude 3 quake occurs, the effects of which could very well be felt on the surface assuming an average depth of some 4 kilometres. If we set up a measurement grid to record as many possible small quakes as possible, this can help us to estimate the probability of an earthquake occurring that would be noticed on the surface. We can then feed this information into what is known as a traffic light system so that engineers can respond by reducing pumping rates during stimulation in order to prevent possible larger earthquakes in future.
Nonetheless, most of the strongest induced earthquakes occur after the pumps have been shut down. The details of why this happens are not yet fully understood and are the subject of ongoing research. Concerning the site of the largest earthquake we can say that it occurs within the stimulated rock, often at the margin, when a critically pre-stressed fault is practically 'tapped' underground. The problem is that in most cases we do not know in advance where these critical zones are.
10. With hydrofracturing – better known as fracking – the important thing is to seismically control the injection of the fluid – in other words to quickly reduce the pressure and flow rate of the fluids when an increase in earthquake activity is detected. If an induced earthquake does occur, weighted drilling fluid can be used to stabilize a borehole. What technical options are available to minimise the risks?
We can currently use so-called traffic-light systems to reduce the production or pump rates when the strength of induced earthquakes increases. In most cases, the seismicity drops relatively quickly. Nonetheless, noticeable seismicity can still occur (see previous question). So strictly speaking, a traffic-light system is not a control system but rather a response system. We are at the moment still not able to actually control induced seismicity. Various approaches for doing so are currently the subject of research. But full control would ultimately imply not only having to identify every fault in the subsurface prior to drilling. It would also require that we know which of these faults have been critically pre-stressed, and to what degree, by tectonic deformation millions of years ago. Then we need to know how much additional pressure it would take to actually activate these pre-stressed faults. This is something we are not yet capable of, but it does remain our primary objective as we pursue our research into these processes. It should also be noted that this aspect of our research is directly connected to research being done on large-scale natural earthquakes.
11. There is a lot of discussion surrounding the risks associated with geothermal energy. Are these risks being assessed realistically? From the scientific standpoint, which concerns are justified and which can we say may be exaggerated given the current state of research.
One reason that the hazards of geothermal energy are being so thoroughly discussed is because they are so relevant to gaining public acceptance of this technology. But another important reason is that these hazards are also of great interest to operators and potential investors, given that the profit margin of deep geothermal at its current stage of development is rather small when compared to other forms of energy. To improve this situation it would be good to have a large number of deep geothermal test projects running in order to be able to better quantify the risks and, above all, the potentials of this technology.
12. What contribution can deep geothermal make to securing Germany's future energy supplies?
The potential for deep geothermal energy in Germany is significant, no one disputes that. But if this technology is to play an important role in the energy mix of tomorrow, we need to lay the groundwork today in order to achieve this goal in the foreseeable future.
13. Final question: How important is it for researchers to be at the forefront of this technology? How relevant is your research to what is actually happening in the real world?
In the field of deep geothermal, researchers are specifically working to unite basic research with applied aspects and in so doing are opening up great opportunities for knowledge and technology transfer. So we scientists are not only looking to gain a greater understanding of the processes behind earthquakes but are at the same time helping to boost the efficiency and cost effectiveness of renewable forms of energy.
Kwiatek G, Martinez-Garzon P, Dresen G, Bohnhoff M, Sone H, Hartline C (2015): Effects of long-term fluid injection on induced seismicity parameters and maximum magnitude at northwestern The Geysers geothermal field. J. Geophysical Research, 120, 7085-7101. DOI:10.1002/2015JB012362
Martinez-Garzon P, Bohnhoff M, Kwiatek G, Dresen G (2013): Stress tensor changes related to fluid injection at the The Geysers geothermal field, California. Geophys. Res. Lett., 40, 2596-2601. DOI:10.1002/grl.50438.
Bohnhoff M, Dresen G, Ellsworth W L, Ito H (2010): Passive Seismic Monitoring of Natural and Induced Earthquakes: Case Studies, Future Directions ans Socio-Economic Relevance. In: S. Cloetingh, J. Negendank (eds.), New Frontiers in Integrated Solid Earth Sciences, International Year of Planet Earth. Springer. DOI 10.1007/978-90-481-2737-5_7
Acatech - Deutsche Akademie der Technikwissenschaften (Hrsg.) (2015): Hydraulic fracturing. Eine Technologie in der Diskussion (acatech POSITION). München.