The investigation of the Rangely oil field has shown that artificially raised reservoir pressure in the neighborhood of a fault could initiate sliding and trigger earthquakes. The pore pressure along the fault at the site and depth where most earthquake loci were located has been monitored. Knowledge of the local stress configuration was needed, however, to confirm the type and direction of slip along the fault, establish the critical pore pressure necessary for an unstable equilibrium, and provide a basis for critical pore pressure estimates in future research of earthquake control. The hydraulic fracturing method was used to determine the in-situ stresses in the vicinity of the fault at a depth of more than 6,000 ft. below surface. A newly drilled and cased well was extended by diamond coring and an appropriate 12 ft. interval of solid rock was selected. The interval was packed-off and pressurized until fracture. Additional pumping extended the artificial fissure away from the well. The pressures recorded during this process together with impression packer results yielded the following principal stress configuration:
8550 psi (horizontal at N70°E) > 6300 psi (vertical) > > 4550 psi (horizontal at N20°W)
This stress distribution is in accord with the theoretically expected condition for the existing strike-slip fault and its right-lateral slip. Moreover, the results show that a serious danger of sliding along the fault would exist if the pore pressure were brought up to about 3500 psi. Pressure monitoring of the seismically active part of the fault confirms this result (3500-4000 psi). With some limitations, the method of hydraulic fracturing appears to be suitable for stress measurements along other active faults where critical pore pressures could thus be predicted and hopefully prevented.
The well publicized Denver earthquakes provided strong field evidence that a correlation exists between formation pore pressures and seismic activity. The mechanism suggested for earthquake triggering was the reduction of frictional resistance to faulting as a result of pore pressure increase (Healy et al, 1968). The Denver case prompted a much more detailed study of the Rangely, Colorado earthquakes which too appeared to be resulting from an artificial pore pressure intensification.
The Rangely oil field is located in the Rio Blanco County, close to the northwestern corner of Colorado. The reservoir, 12 miles long by 5 miles wide, is an anticlinal entrapment of the top few hundred feet of the Weber sandstone (about 6000 ft. below surface). A main fault crosses the long axis of the field. Fault-plane solutions indicate that the fault is strike-slip with an approximate strike of N50°E and a right-lateral slip (Fig. 1) (Raleigh et al, 1972). Water flooding of the oil field for secondary recovery, a process started in 1958, has resulted in an increase of the reservoir pressure. In particular, along the southern periphery pressure has risen from 2500 psi to 3500-4000 psi. Earthquakes have been recorded in the area ever since the installation of a nearby seismograph in 1962 (Munson, 1970).
A primary objective of the Geological Survey program was the correlation between earthquake initiation and pore pressure.