Stress and subsidence resulting from subsurface fluid withdrawal in the epicentral region of the 1983 Coalinga Earthquake

Abstract

The proximity of the May 2, 1983, Coalinga earthquake to active oil fields on Anticline Ridge led to speculation that the earthquake might have been triggered by oil field operations. Elsewhere, earthquakes have been associated with pore pressure increases resulting from fluid injection and also with subsidence resulting from fluid extraction. Simple calculations show that shales, which underlie the oil producing strata, hydraulically isolate the oil field from the earthquake focal region. The large volumes of fluid extracted from the oil fields caused a 50% decline in reservoir pressures from 1938 to 1983. These observations independently rule out substantial increases in pore pressure at focal depths due to fluid injection. A theoretical method, based on Biot's constitutive theory for fluid‐infiltrated elastic media, is used to evaluate the change in stresses acting in the focal region resulting from fluid extraction in the overlying oil fields. As an independent check on the method, the subsidence of the earth's surface in response to fluid withdrawal is calculated and compared with measured elevation changes of Anticline Ridge. The producing horizons are taken to be horizontal permeable layers, bounded above and below by impermeable horizons. Strains within the producing layers are related to extraction‐induced changes in pore fluid mass. Contraction of the producing layers causes the free surface to subside and strains the elastic surroundings. The calculated subsidence rate of Anticline Ridge between 1933 and 1972 is 3 mm/yr, in good agreement with the measured subsidence rate of 3.3±0.7 mm/yr. Calculated pore pressure changes in the deepest producing zone also compare well with observed changes in reservoir pressure. Although the shear stresses induced by extraction favor reverse slip on either the northeast or southwest dipping nodal plane, the induced normal stresses are compressive, inhibiting fault slip. The driving stress (shear stress minus frictional resistance) acting across the northeast dipping plane increased by 0.01 MPa (0.1 bar) between 4 and 9 km depth, weakly favoring slip, and decreased by half that amount at depths of 9 to 11 km, weakly inhibiting slip. The driving stress on the southwest dipping plane increased by 0.02 MPa (0.2 bar) at 10 km, slightly favoring slip. The sign and magnitude of the pore pressure and stress changes at hypocentral depths do not support the hypothesis that the earthquake was induced, although knowledge of the rate of tectonic stress accumulation is required to assess properly the significance of the extraction‐induced stresses.