No abstract

A seismological model is developed for earthquakes induced by subsurface reservoir volume changes. The approach is based on the work of Kostrov () and McGarr () linking total strain to the summed seismic moment in an earthquake catalog. We refer to the fraction of the total strain expressed as seismic moment as the strain partitioning function, α. A probability distribution for total seismic moment as a function of time is derived from an evolving earthquake catalog. The moment distribution is taken to be a Pareto Sum Distribution with confidence bounds estimated using approximations given by Zaliapin et al. (). In this way available seismic moment is expressed in terms of reservoir volume change and hence compaction in the case of a depleting reservoir. The Pareto Sum Distribution for moment and the Pareto Distribution underpinning the Gutenberg‐Richter Law are sampled using Monte Carlo methods to simulate synthetic earthquake catalogs for subsequent estimation of seismic ground motion hazard. We demonstrate the method by applying it to the Groningen gas field. A compaction model for the field calibrated using various geodetic data allows reservoir strain due to gas extraction to be expressed as a function of both spatial position and time since the start of production. Fitting with a generalized logistic function gives an empirical expression for the dependence of α on reservoir compaction. Probability density maps for earthquake event locations can then be calculated from the compaction maps. Predicted seismic moment is shown to be strongly dependent on planned gas production.

In 2003 hydraulic stimulations were carried out in a geothermal field in eastern El Salvador, Central America, as part of a project to explore the feasibility of commercial hot fractured rock energy generation. A key requisite for this environmentally-friendly energy source is that the fracturing of the hot rocks at depths of 1-2 km must not produce levels of ground shaking at the surface that would present a serious disturbance or threat to the local population. Thresholds of tolerable ground motion were inferred from guidelines and regulations on tolerable levels of vibration and from correlations between instrumental strong-motion parameters and intensity, considering the vulnerability of the exposed housing stock. The thresholds were defined in terms of peak ground velocity (PGV) and incorporated into a "traffic light" system that also took account of the frequency of occurrence of the induced earthquakes. The system was implemented through a dedicated seismograph array and locally derived predictive equations for PGV. The "traffic light" was used as a decision-making tool regarding the duration and intensity of pumping levels during the hydraulic stimulations. The system was supplemented by a small number of accelerographs and re-calibrated using records obtained during the rock fracturing.

Measurements of the strains and earthquakes induced by fluid extraction from a subsurface reservoir reveal a transient, exponential‐like increase in seismicity relative to the volume of fluids extracted. If the frictional strength of these reactivating faults is heterogeneously and randomly distributed, then progressive failures of the weakest fault patches account in a general manner for this initial exponential‐like trend. Allowing for the observable elastic and geometric heterogeneity of the reservoir, the spatiotemporal evolution of induced seismicity over 5 years is predictable without significant bias using a statistical physics model of poroelastic reservoir deformations inducing extreme threshold frictional failures of previously inactive faults. This model is used to forecast the temporal and spatial probability density of earthquakes within the Groningen natural gas reservoir, conditional on future gas production plans. Probabilistic seismic hazard and risk assessments based on these forecasts inform the current gas production policy and building strengthening plans.

SUMMARY Induced seismicity typically arises from the progressive activation of recently inactive geological faults by anthropogenic activity. Faults are mechanically and geometrically heterogeneous, so their extremes of stress and strength govern the initial evolution of induced seismicity. We derive a statistical model of Coulomb stress failures and associated aftershocks within the tail of the distribution of fault stress and strength variations to show initial induced seismicity rates will increase as an exponential function of induced stress. Our model provides operational forecasts consistent with the observed space–time–magnitude distribution of earthquakes induced by gas production from the Groningen field in the Netherlands. These probabilistic forecasts also match the observed changes in seismicity following a significant and sustained decrease in gas production rates designed to reduce seismic hazard and risk. This forecast capability allows reliable assessment of alternative control options to better inform future induced seismic risk management decisions.