Compressed air energy storage in salt caverns induces low-frequency cyclic loading effects in rock salt. An ongoing study is described which will furnish data for assessing associated stability of storage caverns. Some initial findings are presented from laboratory and in-situ mine tests.


The first compressed air energy storage (CAES) facility near Huntorf, West Germany, reportedly has been operating successfully for over two years. The facility employs two non-compensated (or dry) caverns in the Huntorf salt dome as a storage reservoir (Mattick and co-worker, 1975). Somewhat similar CAES facilities are under consideration in the U.S., and the long-term behavior of the associated storage reservoirs is of considerable interest (Loscutoff, 1979). The objective of this study is to obtain data from a generalized laboratory test program on behavior of rock salt for long-term cyclic loads typical of CAES operations. The data are currently being collected from relatively conventional laboratory tests and also from bench-scale in-situ tests in U.S. Gulf Coast salt mines. In addition, 1imited complementary numerical modeling is performed to aid in planning tests and analyzing resulting data. Figure 1 illustrates schematically the loading of a "typical" noncompensated or dry CAES Cavern. Loading is associated with the time and depth dependent "effective overburden" or cavern "pressure difference", ğ=g-p. Here g is the geostatic stress calculated from the depth and unit weight of overlying geologic materials, and p is the countering and time varying cavern pressure controlled by operators of the storage facility.

Conditions at specific locations in the salt around the cavern will depend on the effective overburden, cavern geometry, temperature, material properties, cavern loading history, and site-specific geologic effects (Thoms and Martinez, 1978 and Thoms, 1979). Rapid reductions in cavern pressure result in rapid increases in the effective overburden. Further-, gas filled caverns at large depths will be subjected to effects of large g values if the pressure p is reduced to atmospheric levels. Both of these situations reportedly have caused undesirable effects in natural gas storage caverns in salt domes. Röhr (1974) reported roof falls in a gas test cavern when p was decreased rapidly (approximately 14,500 kPa over 158 hours); and Dreyer (1974) noted significant closure in the Eminence dome gas storage caverns of the U.S. Gulf Coast, with around 40% reduction apparently due to large value of 9 over a period of a few months. The Eminence dome caverns were at depths from 1737 to 2042 m, thus ğ could range up to around 44,000 kPa. Because of this last example, Thoms (1979) has recommended that CAES salt caverns be limited to depths less than 1525 m. and, because of the well known sensitivity of salt to temperature, that temperatures of injected air be such that salt temperatures do not exceed 80°C. This is in contrast to a number of other studies which have been undertaken to determine salt response for static or slowly varying effective overburden loads and temperatures (Hansen, 1977; Wawersik, 1980).

This content is only available via PDF.
You can access this article if you purchase or spend a download.