2.1.3 Hydrogeology and groundwater quality



Hydrochemistry data were available for three regional aquifers (Cadna-owie – Hooray Sandstone, Hutton Sandstone and Clematis Sandstone), three regional partial aquifers (Winton-Mackunda formations, upper Permian coal measures and Joe Joe Group) and three regional aquitards (Rolling Downs Group, Injune Creek Group and Moolayember Formation). All of the hydrogeologic units in the Galilee subregion show high variability in solute concentrations, ion:chloride (Cl) ratios, and sample depth. Based on major ion chemistry, hydrogeologic units in the subregion can be grouped into three hydrochemical systems: two in the Eromanga Basin and one in the Galilee Basin. In stratigraphic order these are: a strongly Na-Cl dominated system with minor SO4, consisting of the Winton-Mackunda formations partial aquifer and the Rolling Downs Group aquitard; a Na-HCO3-Cl dominated system, consisting of the Cadna-owie – Hooray Sandstone aquifer, the Injune Creek Group aquitard and the Hutton Sandstone aquifer; and a Na-Cl system with minor to significant HCO3, consisting of the Clematis Group aquifer, the upper Permian coal measures and the Joe Joe Group.

The distinctive hydrochemical signature, evident within each hydrogeologic unit, suggests that each is hydraulically separated at a regional scale by aquitards. However, at a local scale there may be some mixing of waters between aquifers. For instance, available data suggests mixing of groundwater may be occurring where the Hutton Sandstone aquifer is in contact with Galilee Basin aquifers in areas adjacent to the Maneroo Platform.

Salinity values are highest in Eromanga Basin units on the Maneroo Platform, coinciding with where the Moolayember Formation and Clematis Group pinch out.

Water levels

Potentiometric surfaces developed from water level observations in the Cenozoic aquifers (Quaternary alluvium and Cenozoic sediments) indicate that these units comprise local flow systems which are highly influenced by topography. The curvature of potentials around certain streams indicate that Jordan Creek, the Alice River and the Belyando River are potentially gaining streams and Dunda Creek, Tallarenha Creek and Lagoon Creek are potentially losing streams.

In the Eromanga Basin, the potentiometric surface of the Winton-Mackunda formations indicates a potential for regional flow from east to west, with some local groundwater mounding in areas of higher topography indicating a local component to flow. The underlying confined aquifers (Hooray Sandstone and correlatives, and Hutton Sandstone and correlatives) show a much smoother regional flow pattern from the intake beds in the east towards the west. Hydraulic gradients are much higher in these units in the area of the intake beds than further west, and this is believed to be due to reduced permeability in the outcrop areas due to weathering.

The potentiometric surfaces of the Galilee Basin units (Clematis Group, upper Permian coal measures, and Joe Joe Group) all show a basin centred groundwater divide in the vicinity of the Galilee and Jericho 1:250,000 scale geological map sheets, with potential for groundwater flow on one side going east toward the Belyando river basin and outcrop areas, and on the other side flowing to the west. Also, significant variability in vertical hydraulic gradients in the upper Permian coal measures partial aquifer was found. To further understand the complexity, water level data were further split between three hydrogeological sub-units, which from top to bottom are informally designated BC1, BC2 and BC3. The partitioning was done on the basis of approximately similar groundwater pressures existing in each sub-unit.

Water level trends

Time series groundwater level data for a number of observation wells were collected from the Queensland groundwater database.

The distribution of observation wells is weighted slightly to the east of the subregion, with a number of clusters of bores located around proposed areas of coal resource development.

The majority of bores in the observation dataset are non-artesian but a number of artesian bores were located in the Jurassic and Cretaceous units in the western part of the subregion.

Time series data from all bores in the dataset were analysed by the Theil-Sen regression method to identify whether there were statistically significant trends in the time series of the observation bores.

Filtering of the dataset was required to ensure that interpretations were made on reliable information. Many bores had time series of less than two years and were analysed separately from bores with longer records.

The majority of non-artesian bores with a statistically significant trend and records longer than two years showed a decreasing trend in water levels over the recording period.

The majority of artesian bores did not show any statistically significant trend in the data. For some bores this is likely due to the small number of observations available for statistical analysis (only three or four points for some bores), but for many bores a change in the hydrological regime caused by the Great Artesian Basin Sustainability Initiative (GABSI) program may be the reason. The hydrographs of several bores in which a statistically significant trend could not be identified show stable water levels, or a decline in water levels, until sometime in the 1990s, after which water levels begin to rise.

Both increasing and decreasing trends in water level were seen in most hydrostratigraphic units.

Head difference

When head differences between adjacent aquifers are in the range –10 m to +10 m, this is considered to be a necessary but not sufficient condition for the aquitard to be leaky and for inter-aquifer leakage to be occurring. Conversely, where the head difference is less than –10 m or greater than +10 m, such areas may be interpreted as indicating where the aquitard forms a tight seal and that negligible inter-aquifer leakage occurs.

Regions where the head in the Hutton Sandstone is higher than the head in the Hooray Sandstone occur near the intake beds in the eastern zone (where the outcrop of Hutton Sandstone is topographically higher than the Hooray Sandstone) and in the western artesian areas on the Winton and Mackunda 1:250,000 sheets. These are areas where it appears that the intervening aquitard, the Injune Creek Group, forms a tight seal. Elsewhere, in the central eastern, and in parts of the central western and western zones, the Injune Creek Group aquitard appears to be leaky.

About 40% of the Injune Creek Group aquitard forms a tight seal and the remaining 60% is leaky. The notion of a significant component of the Injune Creek Group being leaky is supported by the hydrochemistry data. The dominant vertical flow direction through the leaky aquitard would be upwards from the Hutton Sandstone because chemically the Injune Creek Group is closer to the Hutton Sandstone than the Hooray Sandstone.

Analysis of the head difference between the Hutton Sandstone and Clematis Group indicates that the Moolayember Formation forms a tight seal near the intake beds of the Hutton Sandstone in the eastern zone and parts of the central eastern zone, and is leaky over the majority of the central eastern zone and for all of the Manuka 1:250,000 sheet in the central western zone.

Examination of head difference between the Clematis Group aquifer and the BC1 partial aquifer (of the upper Permian coal measures) indicates that in most places the intervening aquitard, the Dunda beds and Rewan Formation, forms a tight seal.

Analysis of the head differences between the BC1 and BC2 partial aquifers indicates that on the Muttaburra and Jericho 1:250,000 sheets the aquitard at the top of BC2 (the BC interburden sandstone) forms a tight seal to exclude vertical hydraulic connection between BC1 and BC2 and is approximately collinear with the groundwater divide. The tight aquitard occurs across about 40% of the mapped area but is leaky elsewhere.

Examination of the head difference between the BC2 and BC3 partial aquifers indicates that the intervening aquitard (DE interburden sandstone) forms a tight seal on the Tangorin, Muttaburra, Galilee, Jericho and Springsure 1:250,000 sheets. Elsewhere, on the eastern and western margins the DE interburden sandstone aquitard appears to be leaky.

Investigation of the head difference between the BC3 partial aquifer of the upper Permian coal measures and the Joe Joe Group aquitard reveals the head in the Joe Joe Group is higher than the head in BC3 over about 75% of the mapped area. While there is a potential for vertical upwards flow, leakage is excluded by the tight seal afforded by the Joe Joe Group aquitard (Jochmus Formation). The head difference map shows an area on the western margin, and a smaller one on the eastern margin, where the Joe Joe Group aquitard appears to be leaky.


In the bioregional assessment (BA) programme only the rock outcrop areas of the aquifers have been used to estimate groundwater recharge rates using the chloride mass balance method. The sub-crop areas are assumed to be blanketed by dense, plastic Cenozoic clay which greatly impedes recharge. Consequently, an assumed recharge rate of 0.2 mm/year has been applied to such areas, irrespective of the substrate.

The estimated recharge flux for the Hutton Sandstone and Hooray Sandstone are 18,672 ML/year and 12,252 ML/year. This contrasts with previous estimates by Kellett et al. (2003) of 25,710 ML/year and 21,360 ML/year. The significant decreases in these revised recharge estimates are due to only mapped areas of outcrop being considered as applicable for receiving direct recharge from rainfall due to the potential for Cenozoic cover to impede recharge. Whereas, these effects of Cenozoic cover on recharge were not taken into account in the original estimates by Kellet et al. (2003).

A total recharge flux of approximately 101,000 ML/year was estimated across the subregion. Recharge to the Winton-Mackunda formations occurs over the entire area of occurrence, but the recharge rate is not uniform. In many places, the Winton-Mackunda formations are blanketed by a thick layer of saprolite and the lower horizon of the weathered profile greatly impedes downward infiltration of the wetting front and therefore also recharge. Groundwater (recharge) mounds occur in those places where the saprolite has been eroded exposing relatively unaltered rock. Recharge rates in such areas are about 1 mm/year, but in places where the saprolite cover has been preserved, recharge rates are of the order of 0.1 mm/year.

The recharge fluxes for the Rewan Group, upper Permian coal measures and the Joe Joe Group are particularly low (<700 ML/year).


Artificial discharge of groundwater by pumping from bores, or discharge from free-flowing artesian wells, is a component of groundwater discharge for every formation in the Galilee subregion. For the Hooray and Hutton sandstones, flow from controlled or uncontrolled artesian water wells is by far the largest proportion of discharge from these aquifers. The remainder of the groundwater flux in these two aquifers, except for a minor component of flow to rejected recharge springs in the Barcaldine Springs complex, is ultimately naturally discharged in springs, salt lakes or vertical leakage in the south-west Eromanga Basin.

Natural groundwater discharge occurs from several groups of springs in the Galilee subregion. The source aquifer for the Barcaldine Springs complex is the Ronlow beds (mainly Hooray Sandstone equivalent). The source aquifer for the Doongmabulla Springs complex (about 10 km west of the proposed Carmichael Coal Mine) is primarily the Clematis Group aquifer. The Colinlea Sandstone is the source aquifer of the Mellaluka Springs complex and Albro Springs in the east of the subregion. The Dunda beds are source aquifer for some small springs including Hector Springs.

With the notable exception of the Clematis Group aquifer, artificial groundwater discharge by pumping from wells is negligible for the Galilee Basin formations. However this is set to change dramatically when dewatering of the upper Permian coal measures begins in 2018. Initially it is proposed to pump 6,000 ML/year from the upper Permian coal measures, ramping up to 11,350 ML/year over 30 years.

A component of flow in the upper Permian coal measures partial aquifer and the Joe Joe Group aquitard discharges eastwards towards the Belyando River valley, but the majority of the groundwater flux in these formations is towards the west. It appears that the groundwater discharge from these formations is dominantly vertical upwards leakage into the overlying formations at the western margins of the Galilee Basin where the strata pinch out. This also appears to be the case for the Triassic units of the Galilee Basin sequence, with the exception of the Clematis Group.


A significant amount of data could not be used due to insufficient information being available to adequately determine which geological unit the data were obtained from. Additional hydrochemical data collection and analysis (e.g. cluster analysis and isotopic data) would help in identifying inter-aquifer mixing and regions where different chemical processes are dominant.

Water level and water level monitoring data are sparse and unevenly distributed in the subregion. Few nested piezometers exist to enable direct comparison between aquifers at a single location.

Last updated:
6 December 2018
Thumbnail of the Galilee subregion

Product Finalisation date