Review of groundwater model for Rocky Hill Coal Project

Australian Groundwater and Environmental Consultants Pty Ltd (2013) developed a groundwater model using MODHMS modelling package (HydroGeologic Inc, 2001). The model was developed to define the impact of open-cut coal mining on groundwater systems, which also included a few water supply bores that are currently in use, and groundwater levels in alluvial aquifers. It also was used for cumulative impact analysis, when in addition to the proposed Rocky Hill Coal Project, the model accounted for AGL CSG activities and the Stratford Coal Mine extension.

The model extent is 6 km × 14.5 km comprising 188 rows by 95 columns, and about 178,600 rectilinear cells. The model has ten layers: alluvial (one layer), colluvium/weathered Permian (one layer), Permian interburden and minor coal seams (four layers), and major coal seams and minor interburden (four layers). However, there is no information on spatial distribution of these layers or their thickness. Only hydraulic properties of individual layers are presented. The majority of the model boundaries were set as no-flow boundaries with limited cells in Layer 1, as constant head boundaries, to represent inflow and outflow from the alluvium. In the model, ephemeral creeks were represented as drains with bed elevation of 4 m below the topographic surface elevation.

The hydraulic conductivity of materials was discretised into six zones along the horizontal plane. Hydraulic conductivity varied from 0.002 to 150 m/day. Recharge from rainfall was imposed as a fixed percentage of rainfall in three zones. Evapotranspiration was applied to the entire model domain as the mean annual rate of 1059 mm/year, with an extinction depth of 2 m below ground surface using the evapotranspiration package. Irrigation wells were not represented in the model, as the amount of produced water was considered to be too small in volume. The modelling was undertaken in three steps:

  1. The steady-state model was designed within MODFLOW, and was used to define initial groundwater level conditions for the second step.
  2. Transient calibration was undertaken using MODHMS’s full capacity (overland flow, channel flow, unsaturated zone modules were included) using a daily stress period from March 2011 to February 2012.
  3. The transient predictive model was undertaken using MODHMS’s full capacity (overland flow, channel flow, unsaturated zone modules were included) using a quarterly stress period.

The calibration of the steady-state model resulted in hydraulic conductivities of 0.5 m/day for the alluvium, 5 x 10-3 m/day for the weathered zone, 2.64 x 10-2 m/day for the coal seams, 4 x 10-3 m/day for the interburden and 1 x 10-6 m/day for the Alum Mountain Volcanics. Vertical hydraulic conductivity is a factor 10 less than horizontal. For the transient calibration, a daily time step was used for rainfall, and the transient prediction model adopted a quarterly time step over a 14-year modelling period. The water balance was not given either for transient calibration or prediction models and comparison between the two calibration models was not proposed.

Model verification was done using the groundwater level/hydraulic heads monitoring data, which was collected for 12 monitoring bores. Sensitivity analysis was undertaken for the steady-state calibration model, when the sensitivity of simulated heads to a series of model parameters was based on the relative composite sensitivity (RCS) approach as defined in PEST. Two parameters were identified: hydraulic conductivity of overburden and the interburden layers, to which model calibration was most sensitive. Uncertainty analysis was undertaken for one water balance component: water inflow to the pits. The model parameters were assigned the range ±50% and the resulting inflow to the pit ranged from 9% reduction to a 2% increase of the base case model. The parameters which the inflow to the pit was most sensitive to were: specific yield in Layer 1 (alluvial sediments, colluvium and regolith) and hydraulic conductivity of interburden layers and minor coal seams.

The groundwater model was used to project mining impact on groundwater users and impact on the alluvial aquifer:

  • groundwater users: zero drawdown was identified for the privately owned bores, which are located within the boundary of the model domain
  • alluvial aquifer: the cumulative impacts were estimated as (i) from the surrounding operations (AGL CSG and Stratford Coal Mine) (i.e. without Rocky Hill Coal Project); and (ii) Rocky Hill Coal Project and the surrounding operations (AGL CSG and Stratford Mining Complex). The impact was presented as the changes in inflow from the Permian units to the alluvial aquifer. The inflow is 0.4 to 0.6 ML/day (without the Rocky Hill Coal Project). The Rocky Hill Coal Project operation projected to reduce this inflow from the Permian units to the alluvial aquifer to 0.1 to 0.4 ML/day.
Last updated:
5 November 2018
Thumbnail of the Gloucester subregion

Product Finalisation date