Prof. Manfred Joswig, Universität Stuttgart (to be held in the working group, 7.3.)
Earthquakes in Germany are tied to pre-fractured tectonic regions in the Upper Rhine valley, the Lower Rhine area, the Swabian Alb, and the Vogtland region. The strongest earthquakes of last century reached intensity VII to VIII, and caused medium to significant damage at not-reinforced masonry. In the ‘aseismic’ northern German sedimentary basin, Earth crust in under stress, too, and experiences strain by post-glacial uplift and salt tectonics. This caused weak earthquakes like the 2000 Wittenburg (east of Hamburg) and 2005 Bremerhaven event. These intra-plate earthquakes are hard to predict due to their long recurrence periods, but forced a design compliance for seismic hazard of intensity VI for critical technical infrastructure in northern Germany.
Earthquakes are also observed for gas production from conventional reservoirs. They happen close to reservoir, with clear normal or reserve faulting mechanism, and related to a significant drop of reservoir pressure by some years of depletion. Gas fields in the northern Netherlands provide a good example for this weak, shallow, induced seismicity which exclusively originates in the immediate proximity of gas-prone sand-stone formations at some 2.000 m depth. The first event happened some fifteen years after start of production, and no event exceeded ML 3.5 nor 4 km depth until now. However, due to shallow source depth the ML 3.5 caused exceptional intensity VI (people got scared, light damage by masonry cracks, and falling tiles) at few places.
The situation is more complex for gas production in northern Germany. Reservoirs and earthquakes are significantly deeper, and a poor instrumental coverage results in further uncertainties for hypocenter depth estimates. The largest events in Soltau 1977 and Rotenburg 2004 are up to a magnitude stronger than the Dutch upper magnitude limit, lie close to mapped, pre-existing faults, and in case of Soltau just happened at the start of gas depletion. This makes a distinction from natural, intra-plate earthquakes challenging, and a subject of scientific dispute until now.
Earthquakes are fracture processes in Earth crust, and these fractures also happen by fracking conventional and unconventional gas reservoirs. However, these fractures are created intentionally to enhance production rates, and the resulting single-fracture lengths of meters correspond to events of ML 0 and smaller, which is an energy factor of 1.000 below felt earthquakes. Fracking is applied for decades, and due to the shale gas boom hundreds of thousands of times. Only in rare, notable exceptions tectonic stress was released by fracking, and felt earthquakes were triggered. In the actual 2011 Blackpool, GB episode, fracking of a shale gas reservoir hit a pre-existing fault, and high permeability enabled large injection volumes which caused extended pore-pressure perturbations. In an environment of critical pre-frack stress conditions, the fracking caused Mohr-Coulomb failure of the pre-stressed fault at a length sufficient for a felt earthquake. Suited microseismic monitoring can unveil such situation already during build-up, and invoke counter-measure of modified fracking parameters.
Some further cause of seismic hazard comes by the injection of waste water. Injecting large volumes of fluid over many years to decades into subsurface can be much more efficient to cause large-scale diffusion of pore pressure perturbations than short-term, small-scale fracking does. Injecting waste water into depleted reservoirs is standard in oil and gas production, but the related triggering of earthquakes depends on critical stress regimes on pre-existing faults, again. Thus earthquakes from injection, like the 2011 Youngstown, Ohio sequence are rare exceptions for a widely used technique. For the northern German sedimentary basin, ExxonMobil supplied data from five selected, representative injection wells related to gas production, with operation times from ten to thirty years, and cumulative injection volumes from 0.3 to 3 Mm3. In an extended radius of ten km each, no earthquake was reported neither before, during, or after operation.
To access the secondary risk of a leaking borehole caused by sheared tubing or damaged cementum, the ML 4.5 Rotenburg event (intensity V to VI) shall act as reference. Its fracture length was 4.6 km, and average slip was 1.6 cm. For maximum damage the slip should be inclined against the borehole but directly touching it.
