Monitoring regional groundwater extraction: the problem.

As hydraulic disturbances (signals) are propagated through a groundwater system two things happen: (1) the higher frequencies in the disturbance are filtered out by the physics of the system and (2) the disturbance takes time to propagate through the system. The filtering and time delays depend on the aquifer diffusivity. This means, for example, if one is observing a water table aquifer at some distance from where annual recharge is occurring, only the long-term average effect of the recharge will be transmitted to the observation point--the system filters out annual variations. These facts have profound impacts on what is feasible to monitor. For example, if one is concerned about the impact of pumping on a spring in a water table aquifer, where the pumping is more than 20 miles or so from the spring, there will be a long delay before the pumping impacts the spring and there will be an equally long delay before a long-term reduction in the pumping regime will restore the spring. The filtering by lower diffusivity groundwater systems makes it impossible to discriminate between the impacts of several major pumpers in the system and/or long-term climate changes.

Strategy for radioactive waste disposal in crystalline rocks.

A strategy for waste disposal is proposed in which the repository would be situated in a crystalline rock mass beneath a blanket of sedimentary rocks whose ground-water flow characteristics are well understood. Such an approach exemplifies the concept of multiple barriers to the isolation of radioactive wastes from the biosphere. This strategy has the advantages that (i) ground-water flow within the sedimentary rocks can be investigated and modeled by conventional, well-under-stood theory and technology; (ii) under favorable circumstances, the flow system operates as an active barrier, so that a long migration path and extremely low flow rate to the biosphere can be assured; and (iii) since, in many locations, ground water is nonpotable, the possibility of future human intrusion can be minimized.

An experiment in earthquake control at rangely, colorado.

An experiment in an oil field at Rangely, Colorado, has demonstrated the feasibility of earthquake control. Variations in seismicity were produced by controlled variations in the fluid pressure in a seismically active zone. Precise earthquake locations revealed that the earthquakes clustered about a fault trending through a zone of high pore pressure produced by secondary recovery operations. Laboratory measurements of the frictional properties of the reservoir rocks and an in situ stress measurement made near the earthquake zone were used to predict the fluid pressure required to trigger earthquakes on preexisting fractures. Fluid pressure was controlled by alternately injecting and recovering water from wells that penetrated the seismic zone. Fluid pressure was monitored in observation wells, and a computer model of the reservoir was used to infer the fluid pressure distributions in the vicinity of the injection wells. The results of this experiment confirm the predicted effect of fluid pressure on earthquake activity and indicate that earthquakes can be controlled wherever we can control the fluid pressure in a fault zone.