Large scale gas injection test (Lasgit): Results from two gas injection tests

Abstract

This paper describes the initial results from a large scale gas injection test (Lasgit) performed at the Äspö Hard Rock Laboratory, Sweden. Lasgit is a full-scale field-scale experiment based on the Swedish KBS-3V repository concept, examining the processes controlling gas and water flow in compact buffer bentonite. The first 2 years of the test focused on the artificial hydration of the bentonite buffer. This was followed by a programme of hydraulic and gas injection tests which ran from day 843 to 1110. A further period of artificial hydration occurred from day 1110 to 1385, followed by a more complex programme of gas injection testing which remains on going (day 1385+). After 2 years of hydration, hydraulic conductivity and specific storage values in the lower filter array were found to range from 9 × 10 −14 to 1.6 × 10 −13 m/s and 5.5 × 10 −5 to 4.4 × 10 −4 m −1 respectively, with the injection filter FL903 yielding values of 7.5 × 10 −14 m/s and 2.5 × 10 −5 m −1. A second set of hydraulic measurements were performed over 1 year and a half later yielding similar values, in the range 7.8 × 10 −14 m/s and 1.3 × 10 -13 m/s. The hydraulic conductivity of FL903 had reduced slightly to 5.3 × 10 −14 m/s while specific storage had increased to 4.0 × 10 −5 m −1. Both datasets agree with laboratory values performed on small-scale saturated samples. Two sets of gas injection tests were performed over a 3 year period. During the course of testing, gas entry pressure was found to increase from around 650 kPa to approximately 1.3 MPa, indicative of the maturation of the clay. The sequential reduction in volumetric flow rate and lack of correlation between the rate of gas inflow and the gas pressure gradient observed during constant pressure steps prior to major gas entry, is suggestive of a reduction in gas permeability of the buffer and indicates only limited quantities of gas can be injected into the clay without interacting with the continuum stress field. Major gas entry occurs at a gas pressure close to, or slightly in excess of, the local stress state and is associated with a rapid increase in flux, linked to a series of kicks in axial and radial stress and porewater pressure. Post peak gas flux exhibits dynamic behaviour, symptomatic of unstable gas flow. When gas injection is stopped, flux into the clay rapidly reduces before entering an extended period of very small flows. This is accompanied by a rapid reduction in pressure which decays to an asymptotic value close to that of the total stress locally within the deposition hole. Examination of the data shows considerable evidence for the development of a highly-dynamic, tortuous network of pressure induced pathways which evolve both temporally and geospatially within the clay, opening and closing due to local changes in gas pressure and/or effective stress. These detailed observations do not conform to standard concepts of two-phase flow. The bentonite response to the passage of gas suggests that pathway dilation is the primary mechanism governing gas migration within the Lasgit system. The important coupling between gas, stress and porewater pressure at the repository scale is clearly evident with the data. The importance and interdependencies of this coupling will be investigated in future experiments planned at the Lasgit experimental site.