In recent years interest in gas escape phenomenon from unlined mined rock caverns has increased considerably. This phenomenon is most critical to the feasibility of Compressed Air systems, closed surge chambers, and underground storage of hydrocarbons. It is of further interest in the long term sealing of radioactive waste repositories and in the potential siting of underground nuclear plants. All previous studies used extremely simplified analytical models and/or laboratory tests and attempted to define a "critical gradient" at which "bubble escape" would cease to occur. The predicted critical gradients ranged from 0.4 to 1.0. These ideas were sometimes translated into practice in the form of elaborate tunnel schemes with radiating pressurized boreholes in order to maintain sufficient gradients both during cavern construction and during use. This paper describes detailed laboratory experiments which were conducted to study two-phase countercurrent flow through simulated rock fractures. A modified Hele-Shaw parallel plate model was built which allowed variation of fracture aperture and orientation with respect to the pressurized cavity. Also, different entrance geometries were simulated to delineate their importance as far as bubble or slug initiation is concerned. An array of sensitive pressure transducers were used to obtain any small variation in the pressure field during bubble propagation. The deformation and volume of the bubbles were recorded via a high speed camera. The data is analysed in light of existing two-phase flow theories developed in the chemical and nuclear engineering fields. Engineering implications are discussed in detail.


Gas escape from unlined rock caverns involves two distinct problems:

  • the fundamental initiation and motion of gas "bubbles" through rock fractures, and,

  • the calculation of overall gas losses from storage caverns.

Solutions to the latter problem, at present, are best handled using an equivalent porous medium approach as suggested by Barton (1972) and Berg and Noren (1969). However, this provides no insight into the fundamental mechanism and parameters controlling the initiation of gas escape. It is this problem that the present research addresses. Interest in two-phase flow in the geotechnical area was originally restricted to the petroleum engineering field where drillstem and two-phase porous medium problems have received considerable attention. A phase is simply one of the states of matter and can be either a gas, a liquid, or a solid. The term two-component is sometimes used to describe flows in which two phases are not of the same chemical substance. Since the mathematics describing two-phase or two component flows are identical, it does not matter which definitions are chosen. Increasing use of underground space has aroused significant interest in developing a better understanding of single-phase and to a lesser extent of two-phase fracture flow. Recently Willet (1979a, 1979b) has given an excellent review of the history and present status of the economic use of the underground. Two-phase fracture flow may be a very important, if not critical, parameter in each of the following areas:

  • nuclear power plants sited underground,

  • crude oil and petroleum product storage.

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