GSA Annual Meeting in Seattle, Washington, USA - 2017

Paper No. 121-8
Presentation Time: 3:15 PM


TIMMS, Wendy1, DAVID, Katarina1 and BARBOUR, Lee2, (1)School of Mining Engineering, UNSW Australia, High Street, Kensington, NSW, Sydney, 2033, Australia, (2)Civil and Geological Engineering, University of Saskatchewan, Saskatoon, SK S7N 5A9, Canada,

Realistic values of specific storage (Ss) for groundwater systems are important to determine the spatial extent and timing of changes in groundwater heads when the groundwater system is stressed. However, numerical groundwater models of underground excavations typically assume constant literature values of Ss. One part of our research program utilised high frequency pore pressure data to evaluate variability and changes in Ss within sedimentary strata overlying a longwall coal mine. Pore pressure data from a vertical series of 6 vibrating wire piezometers (50 to ~278 m depth) recording at hourly intervals over a period of several years (including data before and during mining) were used in the analysis along with barometric pressure data recorded at the same frequency. The site was located near the centre of a longwall panel that extracted ~3 m of coal at a depth of ~330 m. The data was processed to calculate loading efficiency and Ss values by multi-method analyses of hydraulic head response to barometric and earth tide loading. In situ Ss results varied over one to two orders of magnitude and indicated that Ss was changing before and after excavation of underlying coal seams. The vertical leakage of groundwater within the constrained zone (~10 to ~150 m depth) was found to be limited, although some degree of vertical hydraulic connectivity was observed. Depressurization was evident in the fractured zone directly overlying the coal seam, and Ss changes at ~250 m depth indicated this confined aquifer may have became unconfined. Our results demonstrate that high frequency pore pressure data can provide realistic Ss values. In situ Ss values were an order of magnitude lower than Ss measured by geomechnical tests of cores, and were significantly different to textbook values set in majority of local groundwater models. The timing and extent of groundwater level drawdown predicted by models may therefore be underestimated. We have shown, for the first time, that variability of Ss can be significant, and that these changes can provide important insights into how shallow and deep groundwater systems respond to underground mining.