Linke, G.K. and Kendall, M.B. 1995. Groundwater Interception from a Multi Well Point Scheme Using Airlift Pumping. Murray Darling Workshops - extended abstracts. Wagga Wagga. 11 - 13 September 1995. AGSO record No. 1995/61.

Highly saline groundwater inflows into Pyramid Creek result in economic losses to downstream irrigation areas and urban and industrial centres. The Pyramid Creek Interception Scheme aims to reduce these groundwater inflows in the upper section of the creek (13 kilometres in length) by eliminating the hydraulic gradient through groundwater pumping.

Pyramid Creek is an enlarged natural stream that is now used as a major irrigation carrier from Kow Swamp to the Kerang and Swan Hill Irrigation Districts, northern Victoria. Typical flows in the creek are 1,200 ML/day during the irrigation season reducing to 100 ML/day in the winter months. Water that is not diverted for irrigation eventually outfalls to the Murray River.
Approximately 50,000 tonnes of salt enters Pyramid Creek each year from highly saline regional groundwater discharge. The distribution of groundwater inflow is non-uniform, with approximately 50% of the salt inflow discharging in the upper 13 km of the creek. Agricultural productivity and the high environmental value associated with many of the wetlands in the area are suffering severely from the effects of salinity.

The factors which contribute to the high watertables and seasonal fluctuations in the area include both local and regional changes to the water balance. In particular, recharge to the groundwater system at both local and regional scales has increased due to a combination of land clearing and hence less water use in up-basin dryland areas and the application of irrigation water, generally in excess of crop requirements, in the Kerang area.

The aquifer of importance in the study area is known as the Shepparton Formation. At the trial site, the Shepparton Formation is approximately 80 metres thick and consists of clay, silt and thin fluvial beds of sand. These sand layers take the form of discontinuous sand lenses, commonly referred to as shoe-string sands formed from prior streams. From groundwater pumping tests the hydraulic conductivity of the Shepparton Formation is estimated to be in the order of 0.5 m/day, while the transmissivity over a 1 km length determined from short term pumping tests varied from 20 to 200 m3/d.

Pyramid Creek is approximately 3 metres deep and 15 metres in width. The creek is incised into a landscape with high watertables, and as such, the creek acts to discharge groundwater on both sides. Monitoring of observation bores has demonstrated that groundwater levels remain 0.5 to 2.5 metres above creek levels, creating a permanent hydraulic gradient towards the creek.

A pilot scheme was constructed over a 1.2 km reach of the Pyramid Creek. A total of thirteen bores were constructed, six on the northern side of the creek and seven on the southern side. The bores were constructed 200 metres apart, with bores on one side of the creek offset by 100 metres from those on the opposite side. Airlift pumping was chosen due to its relatively low capital and maintenance costs. The pilot scheme was constructed with air supply from one side of the creek only, with airlines running across the bed of the creek. .
For the pilot scheme, short life span reciprocating compressors were utilised. However in the completed scheme, the compressor station or stations will be far larger and probably located elsewhere, at which time industrial quality rotary screw compressors will be selected.
At the time the article was written, the pilot scheme had been operating for a period of two to three months.

The discharge per bore to achieve drawdown in groundwater level to below the summer creek running levels is approximately 0.5 L/s. The corresponding flows in winter are significantly higher at approximately 1 L/s. The total summer and winter system flows over the 1.2 km pilot scheme are 7 L/s and 14 L/s, respectively, or 0.6 ML/day and 1.2 ML/day.
The capital costs for the pilot scheme, including bore construction, airlift components, temporary disposal line and monitoring requirements, were approximately $80,000 (1995 prices). The major proportion of the works for the pilot scheme is to be used in the completed scheme. The typical air: water ratio achieved for the pilot scheme was 1.8:1 which corresponded to an electric power running cost of $21 / ML (1995 prices) for 24 hour, 7 day pumping.

Total flow generated for the full scheme is expected to be in the order of 1,000 ML/y. While the final form of the groundwater control along the full length of the scheme and the disposal scheme is uncertain, capital costs are estimated to be in the order of $4.5 million (1995 prices).
The compressor, typically electrically powered, can be located remote from the wells and the air conveyed to the pump sites via relatively inexpensive airlines. This is expected to be far less expensive than supplying electric power to every site in a system. In a multiple well system such as at Pyramid Creek, pressure regulation and airflow control at each bore allows for optimisation of water flow and drawdown to match the variable aquifer characteristics.

Reducing the salt load along Pyramid Creek by 20,000 tonnes per year would result in an increase in the gross margin for the Kerang and Swan Hill areas of $390,000 per year (1995 prices). There would also be significant benefits to the salinity in the River Murray. The salt load passing through the Kerang system would be reduced by 16,000 tonnes per year. This would generate an economic benefit to the users of the River Murray water, in particular the urban population of Adelaide, of approximately $340,000 per year (1995 prices). Taking into account these benefits, and first order estimates of capital and ongoing costs of interception and disposal, the scheme is estimated to have a benefit cost ratio of 2.6 and a Net Present Value of $8.64 million (1995).

The trial interception scheme undertaken at Pyramid Creek is shown to be both technically feasible and economically viable for the area investigated. The primary benefit derived from the interception scheme is the protection of the River Murray from highly saline creek inflows, and as a result significant economic returns are generated.
The economic viability and technical practicalities of such an interception scheme in a dryland context would require the value and degree of protection provided to a receiving water body to be determined on a site by site basis.

The following are key determining factors for the successful implementation of groundwater interception schemes in dryland areas:

• adequate discharge from the groundwater bores to achieve a drawdown below the water level of the receiving water body;
• a suitable disposal strategy; and
• a perceived or actual benefit from the protection of the surface water body.