How to Reduce Fluid-Injection-Induced Seismicity

The recent growth in energy technologies and the management of subsurface reservoirs has led to increased human interaction with the Earth’s crust. One consequence of this is the overall increase of anthropogenic earthquakes. To manage fluid-injection-induced seismicity, in this study, we propose to use an advanced fluid-injection scheme. First, long-term fluid-injection experiments are separated from short-term fluid-injection experiments. Of the short-term experiments, enhanced geothermal systems stimulations have shown a higher propensity to produce larger seismic events compared to hydraulic fracturing in oil and gas. Among the factors discussed for influencing the likelihood of an induced seismic event to occur are injection rate, cumulative injected volume, wellhead pressure, injection depth, stress state, rock type, and proximity to faults. We present and discuss the concept of fatigue hydraulic fracturing at different scales in geothermal applications. In contrast to the conventional hydraulic fracturing with monotonic injection of high-pressure fluids, in fatigue hydraulic fracturing, the fluid is injected in pressure cycles with increasing target pressure, separated by depressurization phases for relaxing the crack tip stresses. During pressurization phases, the target pressure level is modified by pulse hydraulic fracturing generated with a second pump system. This combination of two pumps with multiple-flow rates may allow a more complex fracture pattern to be designed, with arresting and branching fractures, forming a broader fracture process zone. Small-scale laboratory fluid-injection tests on granite cores and intermediate-scale fluid-injection experiments in a hard rock underground test site are described. At laboratory scale, cyclic fluid-injection tests with acoustic emission analysis are reported with subsequent X-ray CT fracture pattern analysis. At intermediate scale, in a controlled underground experiment at constant depth with well-known stress state in granitic rock, we test advanced fluid-injection schemes. The goal is to optimize the fracture network and mitigate larger seismic events. General findings in granitic rock, independent of scale, are summarized. First, the fracture breakdown pressure in fatigue hydraulic testing is lower than that in the conventional hydraulic fracturing. Second, compared to continuous injection, the magnitude of the largest induced seismic event seems to be systematically reduced by cyclic injection. Third, the fracture pattern in fatigue testing is different from that in the conventional injection tests at high pressures. Cyclic fracture patterns seem to result from chiefly generated low energy grain boundary cracks forming a wider process zone. Fourth, cyclic injection increases the permeability of the system. A combination of cyclic progressive and pulse pressurization leads to the best hydraulic performance of all schemes tested. One advantage of fatigue testing is the fact that this soft stimulation method can be applied in circumstances where the conventional stimulation might otherwise be abandoned based on site-specific seismic hazard estimates.