Induced Seismicity

A Ceiling on the Shaking

Sections 9.1 to 9.3 said whether a fault slips and when. This one asks how big. The cleanest answer is a bound, not a forecast. McGarr (2014) showed that the total seismic moment an injection can release is limited by the volume of fluid injected: $M_0 \le G\,\Delta V$, with $G$ the rock's shear modulus and $\Delta V$ the injected (or extracted) volume. Convert that moment to magnitude with the Hanks-Kanamori scale and you have the largest earthquake the injection could possibly drive.

The canon: inject 10,000 cubic meters into rock of shear modulus 30 GPa and the bound is a moment magnitude of 3.6, a small but felt earthquake. The curve is nearly straight against log volume, because each ten-fold rise in injected volume lifts the magnitude ceiling by about two-thirds of a unit. Inject a million cubic meters and the ceiling climbs toward magnitude 5, into the damaging range. This is why the largest injection-induced earthquakes have come from the largest operations, and why volume, not pressure alone, is a control operators watch.

A Ceiling, Not a Forecast

The honesty this section insists on: the McGarr bound is a ceiling, not a prediction. Most injections produce events far below it, or none at all, because the bound assumes every bit of injected volume feeds seismic slip on a single fault, which rarely happens. What sets where the real event falls below the ceiling is the geology: whether a large, well-oriented, critically-stressed fault exists to receive the pressure (Sections 9.2 and 9.3), whether that fault ruptures seismically or creeps, and how the pressure front spreads. The bound tells you the worst case; the stress model and the fault map tell you how close to it you are likely to get. The same physics runs for depletion, since extracting fluid also changes the stress and can trigger earthquakes (Segall 1989), and the bound applies to the produced volume too. Producing fields, not only injectors, induce seismicity.

The Verdict for Ogbon-1

For Ogbon-1 the pieces assemble into a risk statement. The stress model sits at mobilized friction 0.58, a fraction of a MPa from reactivating its optimally-oriented fault; if such a fault exists and a large volume is injected or produced, the McGarr ceiling says the resulting event could reach magnitude 3 to 4, felt at surface. The mitigation is the one the whole part has pointed to: know the fault population, keep the pressure change small and slow, and stay below the reactivation limit of the most dangerous fault. The last section of Part 9 turns all of this on a single decision an operator actually faces, whether a caprock will seal or leak under injection.

References

• McGarr, A. (2014). Maximum magnitude earthquakes induced by fluid injection. Journal of Geophysical Research: Solid Earth, 119(2), 1008-1019.
• Hanks, T. C., & Kanamori, H. (1979). A moment magnitude scale. Journal of Geophysical Research, 84(B5), 2348-2350.
• Ellsworth, W. L. (2013). Injection-induced earthquakes. Science, 341(6142), 1225942.