GAST Gas Flow Test Preliminary Interpretation

pdf Nagra (2024): GAST – Gas Flow Test Preliminary Interpretation. Nagra Arbeitsbericht NAB 23-40 Rev. 1(45.75 MB)

Gas-permeable seal sections can provide an engineering solution for limiting gas overpressures by gas release along the backfilled structures of an L/ILW repository. The Gas-permeable Seal Test (GAST) is a large-scale experiment at the Grimsel Test Site (GTS) to demonstrate the effective functioning of gas-permeable seals at repository scale. For that purpose, a seal made of a mixture of 80% sand and 20% bentonite was constructed in 2011-2012 and saturated over about a decade prior to performing a series of gas tests.

This report documents the main results obtained from the long-term multi-phase Gas Flow Test (GFT) that was carried out May 2022 – November 2023. The test successfully demonstrated end-to-end gas flow along a flow path within the highly saturated sand/bentonite seal. The path developed very quickly (within a week) and at a small gas overpressure, as indicated by early time responses of pore pressure sensors. The end-to-end gas flow path along the 8 m-long seal section was fully established about one month later and was kept at near constant pressure for six months. The gas injection rates increased during this period indicating a significant reduction in flow resistance over time. The highest sustained gas injection rate achieved during the various test phases of GFT was 1380 mlN/min, equivalent to 725 m3N/yr. Almost all injected gas was recovered from the extraction point. The injected Helium tracer was also largely recovered at the extraction point, and detected both within the sand/bentonite and at the sand/bentonite granular bentonite interface.

Mid-term shut-in and tracer-flushing phases were followed by the injection of a further gas tracer mix using conservative gas tracers (Helium, Argon, Xenon) with differences in solubility and diffusivity. Tracer breakthroughs and recovery were similar to those for Helium in the first gas tracer phase. Final tracer-flushing and characterisation phases concluded the GFT with an overall gas tracer test duration of 16 months.

Interpretation of the observed flow and pressure, tracer and other sensor data indicates:

  • The speed of gas flow development indicates a localised discrete flow path that is likely to have exploited heterogeneity in porosity and bentonite content within the seal section.
  • The gas flow path is interpreted as being of low volume with little desaturation of the bulk of the sand/bentonite although the resistance to gas flow significantly decreased over time.
  • GFT has demonstrated the ability of the sand/bentonite to transport large volumes of gas at repository-relevant flowrates across an 8m-long sand/bentonite seal at low entry pressure and relatively small gradients, from initial saturated conditions in the sand/bentonite.
  • Injected tracer gases were unambiguously detected within the sand/bentonite.
  • The GFT dataset is of high quality and suitable for the development of detailed two-phase flow models of the test.

In conclusion, the key objectives of the GAST and specifically of the GFT have been achieved and the enhanced gas permeability of a large-scale sand/bentonite seal was successfully demonstrated. Water permeability before and after the gas tests remained low. Detailed aspects of the observed responses may usefully be addressed by further interpretation and modelling.