Induced seismicity – or earthquakes caused by human activity – occurs in oil and gas shale fields, enhanced geothermal systems, carbon dioxides sequestration reservoirs and wastewater disposal sites. In all these cases, massive fluid injection into subsurface rocks alters pore fluid pressure and stress conditions, and it may eventually reactivate pre-existing critically-stressed faults – those prone to rupture. Laboratory experiments (triaxial rock deformation tests) and numerical simulations help us better Understand induced seismicity mechanics and mitigation. Let’s look specifically at new studies examining fluid injection direction and rate.
Direct versus Indirect Fluid Injection Effects
In 2017, an induced earthquake of 5.5 magnitude occurred near an enhanced geothermal system site in Pohang, South Korea. This event motivated a group of Chinese scientists to conduct a numerical simulation of fluid injection directly into a fault surface. Their results, reported in “Computers and Geotechnics,” suggest that seismic slip is determined by a sudden increase in shear stress at the boundaries of an aseismic slip zone. Maximum seismic stress events occur near the fluid injection site, and seismic events move along the pressure propagation front. For normal and reverse faults, aseismic slip along fault dip is greater, while for strike-slip faults, aseismic slip along fault strike is greater.
A laboratory study of granite samples published in the “Philosophical Transactions of the Royal Society” corroborated the above-mentioned computer simulation, revealing slow and continuous increase in fluid pressure eventually leads to reactivation of pre-existing faults. The study also found that fluid injection into an intact granite in high volumes and pressures could cause injection-induced seismic hazards; however, this process occurs in evolutionary stages of injection pressure buildup, crack initiation and breakdown.
Fast versus Slow Injection Rate Effects
Laboratory tests of granite samples subjected to varying fluid pressurization rates, confining pressures and fault stress states have looked at how injection rates impact seismicity. Published in “Applied Science,” the study found that at a low pressurization rate of 0.5 megapascal per minute, the injection pressure required for fault reactivations agreed with theoretical estimation: The fault slip was slow and prolonged. At higher pressurization rates, the injection pressure required for induced seismicity increased, the fault slip was rapid and the fault displayed a stick-slip behavior.
Poroelastic Effect
Possible mechanisms for fault reactivation induced by hydraulic fracturing may be placed into three categories: (1) Pore-fluids pressure diffusion across a pre-existing fault from a nearby injection source at relatively low injection rates; (2) tensile opening of the fault plane by direct fluid injection at high pressure rates; and (3) “poroelastic effect.” The third mechanism might not be favored by some researchers because of the very low permeabilities of granite or shale rocks; however, some recent studies support the poroelastic effect.
First developed by Maurice Biot in 1941, poroelastic stress theory considers the rock a porous and linear elastic material coupled with fluid diffusion. In other words, the overall response of the rock matrix to fluid injection is responsible for reactivation of pre-existing faults, even without the fracturing fluid invasion into the fault zone. A numerical simulation reported in “Interpretation” suggests that poroelastic stress changes in the rock formation impact reactivation of faults even at distances up to 200 meters away.
Another computer simulation reported in “Petroleum Science” considered how poroelastic effect relates to critical stiffness in a rock formation. Stiffness is the resistance of a given material to elastic deformation and is influenced by the rock’s mineral composition and porosity. Stiffness is expressed by Young’s (or elasticity) modulus or the ratio of stress to strain. The article in “Petroleum Science” was modeled on the Silurian-age Longmazi Formation, a major shale gas play in China’s Sichuan basin.
To estimate critical stiffness, researchers measured magnitude and rate of effective stress (normal vertical stress minus pore fluid pressure) in the rock. The study found a positive correlation between magnitude of effective stress and critical stiffness, and negative correlation between the change rate of effective stress and critical stiffness. Magnitude and rate both contribute to rock instability, which is prominent when the magnitude of the effective stress increases (constant injection at each fracturing stage) and the change rate of effective stress decreases (the injection process stops abruptly).
Safe and Efficient Injection Strategies
So, given all we just learned, how can the petroleum, geothermal and mining industries perform underground fluid injection more safely and efficiently, reducing the frequency, and avoiding large, induced earthquakes?
In short, mapping and characterization of pre-existing faults and their in-situ stress regimes as well as optimizing fluid rates and volumes should all be considered to mitigate induced seismicity (see sidebar).