Long-term creep tests have been performed on rock-salt and argillite samples under very small uniaxial loadings 0.02 to 0.1 MPa. V To minimize the effects of temperature variations, testing devices were set in a mine where temperature fluctuations are of the order of one-hundredth of a degree Celsius. The mechanical loading was provided by dead weights. The deformations were measured through special displacement sensors with a resolution of 8 10. 'H Strain rates as small as 13 1 7 10 s H u were measured. These tests allow rock-sample creep to be investigated at very small strain rates. The tests also prove that extrapolation of constitutive laws at very small rates is often incorrect.

This report documents research performed to develop a new stress-based criterion for predicting the onset of damage in salt formations surrounding natural gas storage caverns. Laboratory tests were conducted to investigate the effects of shear stress, mean stress, pore pressure, temperature, and Lode angle on the strength and creep characteristics of salt. The laboratory test data were used in the development of the new criterion. The laboratory results indicate that the strength of salt strongly depends on the mean stress and Lode angle. The strength of the salt does not appear to be sensitive to temperature. Pore pressure effects were not readily apparent until a significant level of damage was induced and the permeability was increased to allow penetration of the liquid permeant.Utilizing the new criterion, numerical simulations were used to estimate the minimum allowable gas pressure for hypothetical storage caverns located in a bedded salt formation. The simulations performed illustrate the influence that cavern roof span, depth, roof salt thickness, shale thickness, and shale stiffness have on the allowable operating pressure range. Interestingly, comparison of predictions using the new criterion with that of a commonly used criterion indicate that lower minimum gas pressures may be allowed for caverns at shallow depths. However, as cavern depth is increased, less conservative estimates for minimum gas pressure were determined by the new criterion.ii EXECUTIVE SUMMARYThe main objective of the research discussed in this report is to improve the predictive technology used to evaluate the structural stability of natural gas storage caverns in bedded salt deposits. The structural stability of caverns in bedded salt depends on many interrelated factors, including local hydrology, local geology and rock properties, cavern operating conditions, cavern depth, cavern geometry, and cavern location with respect to other caverns. Cavern design entails avoidance of conditions known to be adverse for cavern stability. For caverns sited in salt deposits, integrity of the salt is crucial for long-term cavern stability. Rock salt is a viscoplastic material that is difficult to fail under moderate levels of confining pressure, which is one of the reasons salt is a favored storage medium. To maintain the integrity of a host salt formation, cavern design philosophy involves circumventing states of stress that cause the salt to dilate. Dilation manifests as a volumetric expansion resulting from microfracturing of the material. Therefore, structural stability is maintained by avoiding or limiting microfracturing in the salt.

The multimechanism deformation model for the creep deformation of salt is extended to treat the response of salt to imposed stress drops. Stress drop tests produce a very distinctive behavior where both reversible elastic strain and reversible time dependent strain occur. These transient strains are negative compared to the positive transient strains produced by the normal creep workhardening and recovery processes. A simple micromechanical evolutionary process is defined to account for the accumulation of these reversible strains, and their subsequent release with decreases in stress. A number of experimental stress drop tests for various stress drop magnitudes and temperatures are adequately simulated with the model.

No abstract

No abstract