Ian Edmonds, Solartran Pty Ltd, Brisbane, Australia.



A scheme to transport water 2500 km from the Burdekin River in Queensland to the Murray River Basin is outlined. Favorable economy, a payback period of three years, is obtained by floating water in fabric containers on the East Australian Current for 93 % of the distance with a pipeline utilized for only 7 % of the distance. The scheme, as outlined, could deliver 1 GL/day of water into the Murray River at a cost of $160/ML or 1 GL/day into the Snowy River at a cost of $70/ML. The running cost of the scheme is reduced by utilizing the East Australian Current for low cost sea transport over 2300 km. Further cost reduction is obtained by offsetting the energy to lift the water up the Divide with electrical energy regained as the water falls through the Snowy Hydroelectric Scheme down the western side of the Divide into the Murray or Snowy. Two indirect advantages of the scheme are (1), a water supply system that can be called upon in times of drought to supply water to the cities on the east coast of Australia and (2), the provision of a high capacity (120 MW) pumped storage facility that would assist in the efficient implementation of renewable technologies such as wind power.        



The food basket of Australia, the Murray River Basin, is dying of thirst as climate change moves the weather systems further south into the Southern Ocean. At the same time the northern tropics are flooded with rain that flows unused into the Coral Sea. In the 40 day period between January 5 and February 17 this year 20,000 GL of water flowed over the Burdekin Dam wall, an average flow rate of 500 GL/day.  This is about 100 times the average daily flow rate of the Murray River.



Figure 1.  The Burdekin River in flood in 2009.  The Burdekin Falls Dam capacity is 1,806 GL.


Logic demands the transfer of Australia’s abundant fresh water from the North to the increasingly drier food bowl in the South. But how can this be economically achieved? A pipeline running from the northern rivers to the Murray River Basin appears to be an obvious solution. Unfortunately, piping water over long distances is extremely expensive in infrastructure and operating cost. A 900 km pipeline to supply about 1 GL/day from the Burdekin Dam to Brisbane was recently estimated to cost $14B, a cost that precluded building the pipeline in the foreseeable future [1 ]. Based on this estimate a pipeline from the Burdekin River to the Murray River, a direct distance of 1800 km would cost $28 B. The cost of pumping water over mountainous terrain is the major impediment to implementing overland water pipelines.  


Australia is an island continent and the northern rivers flow into the sea. Fresh water floats on sea water and provided the fresh water is enclosed in an impermeable membrane or water bag it continues to float and remain pure indefinitely. A water bag containing many thousands of tonnes of water may be towed over large distances with the only energy cost being the energy to overcome water friction on the bag. Delivering fresh water from mainland sources to islands by towing water bags is common practice in the Mediterranean. If an ocean current flows between the water source and the point of supply the water bag can be simply floated on the current with zero energy expenditure.


The East Australian Current (EAC) flows from the Coral Sea down the Australian coast to the Victorian border. A clear delineation of the course of the EAC [2] and daily velocity maps of the EAC can be downloaded from the CSIRO [3]. The EAC flows at about 3 km/hr and takes about 32 days to flow the 2300 km from the Burdekin River in North Queensland to the sea off Tathra in NSW.  An article [4] in 2007 described a scheme to float water from the Tully River to Brisbane to provide water to the, then, drought stricken S E Queensland region. This article extends the above concept to the utilization of the EAC to deliver northern river water to the Murray River Basin.


The scheme.

Water bags, large cigar shaped containers made from plastic fabric, with a capacity 250 ML are filled with fresh water at the Burdekin River.  The water bags, are towed by tug boat out through the Barrier Reef and released into the EAC. Each bag is accompanied by an unmanned power pod that provides up to 250 km correction to the EAC path if necessary.  The progress of the water bags South is monitored by beacons on the power pods until the bags have arrived offshore of Tathra. Tathra has a nearby natural water storage reservoir, the Wallagoot Lake. Further, Tathra is in close proximity (80 km) to the headwaters of both the Murrumbidgee and Snowy River systems and is within 160 km of Lake Jindabyne in the Snowy Mountains Hydroelectric Scheme. The headwaters mentioned lie on the Dividing Range about 1000 m above sea level. Thus water discharged at Tathra can be pumped over a relatively short pipeline into the headwaters of the Murrumbidgee or Snowy Rivers and in another short pipeline pumped into a reservoir of the Snowy Mountains Scheme.  The broad detail of the proposed scheme is shown in Fig 2.



Figure 2. Outline of the scheme to float water on the EAC from the Burdekin River to Tathra thence, by pipeline, to the headwaters of the Murray River.



At Sydney the EAC divides into huge eddies that may swing in close to the coast or flow further off shore. In this region the power pods can be used to shepherd bags out of the eddies. As a result of eddies the towing distance from the EAC to Tathra is likely to average about 200 km. After discharge the water bags are collapsed and rolled into a form which may be shipped or railed back to Northern Queensland along with the power pods. The technology of water transport in large water bags is reasonably advanced with companies in Australia [5] and the USA [6] developing different forms of this technology.




Figure 3. Detail of the pipelines transferring water from Tathra to the headwaters of the Murrumbidgee and Snowy rivers or to Lake Jindabyne in the Snowy Hydro Scheme.


Figure 3 illustrates the two pipelines that lift the water up the Dividing Range. The pipeline from Tathra follows the Snowy Mountain Highway to its junction with the Monaro Highway. Near the junction, about 80 km from Wallagoot Lake, the pipe branches to the headwaters of the Numeralla River and the McLaughlin River. Water discharged into the Numeralla flows to the Murrumbidgee River and services Canberra and eventually flows to the Murray River. Water diverted to the McLaughlin River flows into the Snowy River. The pipeline then extends from the Monaro Highway near Nimmitabel a further 80 km to Lake Jindabyne at 900 m altitude. Water discharged into Lake Jindabyne becomes a resource to the Snowy Mountain Hydro Scheme. The water can be lifted by the Jindabyne Pumping Station 230 m into Lake Eucumbene. From there the water can be transferred through the Eucumbene - Snowy Tunnel or the Eucumbene – Tumut Tunnel to the west side of the Dividing Range. As the water falls down through the power stations of the Snowy Mountain Hydro to the Murray or Murrumbidgee rivers about 2/3 of the pumping energy can be regained. A very simplified schematic of the Snowy Scheme is shown (with red lines) in Figure 4. The pipelines of the present scheme are shown as purple lines in Figure 4.



Fig. 4. Simplified schematic of this scheme (purple) combined with the Snowy Scheme (red).  Altitude is exaggerated relative to distance in this diagram.


The advantage of the scheme outlined in this article is that water from the Burdekin is transported by the EAC a distance of about 2300 km at near zero running cost to a location in close proximity to headwaters of the Murray and Snowy Rivers and to the reservoirs and power stations of the Snowy Mountains Hydroelectric Scheme. The effective running cost relates to sea towing cost over distances of about 300 km and water pumping costs over distances of about 160 km and a lift of about 1200 m. Two thirds of the pumping energy is regained as the water falls 800m through the power stations to the Murray.  The infrastructure cost relates to the cost of a moderately large number of water bags, one charging facility in Queensland, one discharging facility in NSW, two pipelines each of  length about 80 km and one water pumping station near Tathra.


Infrastructure cost.


(A) The sea going link.

The sea going or EAC link comprises water bags, power pods, means of charging and discharging the water bags, tug boats to tow the water bags and pods into and out of the EAC and gantries to load empty bags and power pods onto and off freighters, Figure 5.



Figure 5. Components of a sea going water transport scheme include water bag, power pod, tugboat, discharge pipes, loading gantry and barge for freighting water bags.   


The costing of this scheme will be based on a 1 GL/day or 365 GL/annum supply to the Murray River Basin. Assuming the water bags are 250 ML capacity (500 m length, 10 m draft and 50 m beam) similar to that illustrated in Figure 5 the 1 GL/day supply requires, each day, the charging of four bags in North Queensland and the discharging of four bags in NSW.  The sea distance is 2300 km and the travel time in the 3 km/hr EAC is 32 days. Thus 128 water bags are required for the sea link and at any one time 128 bags would be floating south on the EAC with a average distance between bags of 18 km. Each bag is accompanied by an radio controlled power pod that is essentially a two tonne petrol tank and a 115 HP motor.  At any one time about 22 pods and empty bags would be being shipped north by freighter. Thus the total number of water bags and pods required to deliver one GL/day would be 150.  Based on an earlier cost estimate for a 60 ML bag [4] of $60,000, a 250 ML bag would cost $125,000 based on the increased area of fabric required. Each power pod costs about $50,000. The cost of 150 bags and pods would be $26 M.   Six tug boats, two at the Burdekin and four at Tathra, cost $12 M.  One charge facility and one discharge facility as in Figure 5 cost about $4 M.  One unloading gantry in Queensland and one loading gantry in NSW, each of 40 tonne capacity, would cost $4 M. Two 2000 tonne capacity coastal trader ships each costing $2 M to return the pods and empty bags to North Queensland cost $4 M. The radio control link to the pods is estimated to cost $10 M.  This gives a total infrastructure cost for the sea link of $60 M.


This is a modest infrastructure cost for the transfer of one GL/day or one million tonnes of water per day over a distance of about 2300 km. Comparison with the estimated infrastructure cost of $14 B for transporting about 1 GL/day by a 900 km pipeline between Brisbane and the Burdekin [1] illustrates the superior economics of water transport. However, this present scheme also includes pipeline links.


(B) The pipeline links.

The two pipeline links, each about 80 km in length, form only about 5% of the length of this approximately 2500 km long water transport scheme. However, these pipeline links are the most expensive part of the scheme due to the high cost of large diameter, high pressure, steel pipe. The cost estimate for the pipelines is based on the cost, $750 M, of the North-South pipeline that will pump Murray River water 70 km over the Dividing Range between Goulburn and Melbourne.  Based on this cost each 80 km pipeline link in the present scheme would cost $860 M. For both links the cost would be $1,720 M. Thus the pipeline part of the scheme, representing only about 5% of the transport distance, is about 30 times more expensive than the estimated infrastructure cost ($60 M) of the sea-going part of the scheme. The total infrastructure cost of the scheme to provide 1 GL/day of northern river water to the Murray River Basin is estimated to be $1,780 M, about two billion dollars.


Operating cost of the scheme.


The principal operating costs are associated with (1), the energy cost for towing the water bags a distance of about 100 km into the EAC and 200 km out of the EAC and (2), the cost of electricity to pump the water 1200 m up the Dividing Range and how much of that energy is regained when the water falls 800 m down through the Snowy Hydro Scheme.


(A) Towing energy cost.

The work done in towing a water bag is the product of the drag force and the distance towed. The profile of water bags is designed to minimise friction and a typical drag coefficient is about 0.1. Using this coefficient and estimates of tug boat efficiency the fuel cost to tow two 60 ML bags of water 90 km was estimated to be about $5000, [4]. The cross section area of the 60 ML bags was 300 m2. The cross section area of the 250 ML bags is 500 m2. With a simple extrapolation to the towing of four 250 ML bags 300 km we obtain a daily fuel energy cost of $28,000 per day.


(B) Pumping energy cost


(1) Pipeline Tathra – Nimmitabel.  The energy required to lift 1 GL of water 1000 m is 1013 Joules. If the electrical pumping station is 50% efficient 2 x 1013 Joules are required. This is equivalent to 5.4 million kWHr. Assuming electricity is available at 6 cents per kWHr the pumping energy cost is $324,000 per day.


(2) Pipeline Nimmitabel – Lake Jindabyne. Nimmitabel and Lake Jindabyne are at about the same altitude. Thus the pumping energy between the two locations is expected to be relatively small.


(3) Pumping from Lake Jindabyne to Lake Eucumbene. Again, assuming electrical power at 6 cents per kWHr the energy consumed by the existing pumping station at Lake Jindabyne to pump 1 GL/day through the 231 m lift to Lake Eucumbene would cost $75,000 per day.    


The total energy cost for pumping is $399,000 per day. The total running cost of the scheme is $427,000 per day. However, as discussed below a large fraction, (67%), of the energy cost for pumping can be regained by generating electricity as the water falls down through the hydro electric scheme to the Murray River.




Market prices for Murray River Basin water entitlements can be downloaded from the Federal Department of Environment web site [7]. For high reliability water the volume weighted average price in 2008-2009 was $2,200/ML. Assuming water supplied by the present scheme would receive a high reliability rating the market price of 1 GL delivered into the Murray River Basin would be $2.2 M. The  running cost of the scheme to supply 1 GL/day was estimated above to be $427,000/day or about $0.4 M/day. Thus the nett earning rate of the scheme is $ 1.8 M/day or $ 657 M/annum.  As the estimated infrastructure cost of the scheme is $ 1,780 M the simple payback time of the scheme is 2.7 years. This payback period is short as compared with the payback period for other major infrastructure projects such roads, tunnels, dams, power stations and airports which may be 10 years or more.


Dividing the costing into a sea transport part and a pipeline part provides the useful information that the total infrastructure cost of the scheme is dominated by the cost of the pipeline link. This is useful as the sea transport part involving water bag transport is a new technology that has not been used in Australia. Thus there is a large uncertainty in the estimation of infrastructure cost of this part of the scheme. However, due to the domination of the overall cost by the pipeline link, the estimated cost of the sea transport link ($60 M) could be ten times higher ($600 M) without substantially altering the overall cost of the scheme (an increase from $1,780 M to $ 2,320 M). This would increase the estimated simple payback period from 2.7 years to 3.5 years.


This article has assumed that the water bags could be charged with free water running into the ocean at the Burdekin River mouth. The situation is less simple than this. A continuous supply from the North requires storage reservoirs as the rivers do not run continuously.  The resource data for the Burdekin catchment, [8], indicates a catchment area of 47,000 km2, a runoff of about 3,500 GL/annum, storage for 5000 GL and surface water use of 500 GL/annum. The Burdekin Dam has a capacity of 1,860 GL. Of the entitlements150 GL are unallocated and 300 GL are significantly underutilized [9]. Thus the Burdekin system has a surplus storage capacity of 450 GL/annum and the water asset pricing at close to zero ($2/ML) reflects that [10]. Purchasing water at $2/ML in the Burdekin catchment and selling it at $2,200/ML in the Murray catchment is the basis of a viable business.


There are three alternative delivery points in the proposed scheme. (A), to the Numeralla River, (B), to the McLaughlin River or (C), to Lake Jindabyne.


(A) Water discharged into the Numeralla River flows to the Upper Murrumbidgee River and can service the needs of Canberra. In 2006 the ACT had approximately 26 GL inflow into reservoirs yet 62 GL was used, [11]. The ACTEW is constructing the Murrumbidgee to Googong water transfer to provide 10 GL/annum of  water into the Googong Reservoir [12]. Canberra uses about 100 GL/annum. The scheme presented here could be scaled down by a factor of 3.6 i.e. from 365 GL/annum to 100 GL/annum to supply Canberra via a Burdekin to Tathra sea link and an 80 km Tathra to Numeralla River pipeline.  


(B). Discharge into the McLaughlin River. The Snowy Mountains Scheme [13] diverts water from the Snowy River through the Snowy Mountains via several power stations to the Murray River. Originally, all the Snowy River water was diverted and the Snowy ran dry. Today, environmental concerns require some water to be released from the Snowy Scheme to replenish flow in the Snowy River. When water is released from Lake Jindabyne into the Snowy River it is no longer available for power generation or for irrigation in the Murray River system. The scheme outlined here provides for water from Queensland to be delivered to the Snowy River via a Tathra to McLaughlin River pipeline, (purple line in Figure 4). Reference to Figure 4 suggests that water to the Snowy River could be delivered at an even lower running cost than estimated above by compensating pumping energy with generation energy.  Suppose 1 GL of water from Queensland is delivered via the Tathra pipeline into the McLaughlin and Snowy Rivers. This makes available 1 GL of water from Lake Jindabyne or Lake Eucumbene for electricity generation and irrigation in the Murray River system. This 1 GL of water falls through 800 m of tunnels and power stations of the Snowy Scheme and produces electrical energy equal, in principle, to 4/5 of the electrical energy required to pump the 1 GL of water from Tathra 1000 m up to the McLaughlin – Snowy Rivers.  Thus 4/5 of the pumping energy for water delivered into the Snowy River can be regained from water made available to the Snowy Hydroelectric Scheme. This would reduce the running cost of this alternative to about $70,000/GL ($70/ML) for water delivered to the Snowy River.     


(C) Discharge to Lake Jindabyne. If 1 GL of water is pumped through the second link of the pipeline from Nimmitabel to Lake Jindabyne and then up to Lake Eucumbene the total pumping height is about 1200 m. This water now becomes available to the Snowy Mountains Hydroelectric Scheme. The water flows via the Eucumbene Tunnel through the Snowy Mountains and flows down through the power stations on the western side of the Snowy Mountains into the Murray River. The fall through 800 m can, in principle, generate 2/3 of the energy required to lift the water the 1200 m from Tathra to Lake Eucumbene. This, in effect, reduces the pumping cost of this alternative, $399,000/GL, to 1/3 of this figure i.e. to $133,000/GL. The overall running cost, including the towing cost of $28,000/GL, is reduced to $161,000/GL, ($161/ML) for Queensland water delivered to the Murray River via Lake Jindabyne.


What are the main environmental concerns of the present scheme? The transfer of water from a tropical river to naturally snow fed rivers in the South may introduce new organisms into those rivers. However, the time of transfer, 32 days, in an impermeable container, would result in a major reduction of live organisms being transferred. The spillage of cargo from vessels at sea is a major concern for most cargoes. However, the spillage of 250,000 tonnes of fresh water into the sea from a water bag would have essentially the same effect as a heavy shower of rain at sea, i.e. no effect. Collisions with a large vessel such as an oil tanker would simply part the bag and would probably be imperceptible to crew on the vessel. In the event of a beacon indicating a water bag moving towards a reef or an island or being trapped in an eddy the accompanying power pod would be started and the bag towed up to 250 km to rectify the situation.  Many large water supply schemes are energy intensive with large greenhouse emissions. For example desalination plants, water recycling plants and pipelines such as the North - South pipeline. The present scheme is, relatively, much lower in energy intensity and greenhouse emission. This is due to most of the transport distance being accomplished, at zero energy input, by the East Australian Current and with water pumping energy being offset with the water generating energy afforded by transferring the water via the Snowy Mountains Hydro Scheme. The implementation of a pipeline that can transfer water from Lake Wallagoot 1000 m higher to Lake Jindabyne provides a high capacity (100 MW) pumped energy store. This has considerable environmental benefits in reducing peak energy generation from fossil fuel and allowing more efficient utilization of energy from intermittent renewable sources such as wind and solar.




This proposal utilizes three largely unused resources, water from the northern rivers, the East Australian Current and the excess generating capacity of the Snowy Mountains Hydroelectric Scheme. By floating the water for 95% of the distance and using a pipeline up the Divide over the last 5% of the distance the economies of transporting water 2500 km between the Burdekin and the Murray become favorable. Due to the low cost of northern water and the high cost of water in the Murray River Basin the simple payback time on this scheme to supply 1 GL/day is about three years. A further advantage of this present scheme is that, if implemented, 1 GL/day of fresh water would be floating down the EAC. All or part of this fresh water could be diverted when necessary to provide an emergency water supply to cities and towns on the east coast of Australia.     



Jock Wallace suggested extending the water bag concept to supplying the Murray River Basin. Andreas Luzzi suggested including, within the concept, electricity generation via the Snowy Hydro Scheme. Useful comments were made by Glen Johnston, David Griffin and Bob Swinton.


[1]  Direct connection pipeline-Burdekin to South East Queensland

[2] Renewable Energy Atlas of Australia “Highest near surface current power”.


[3] CSIRO.


[4]  Northern river water for Australian cities? Edmonds I., Water, Journal of the Australian Water Association, September 2007, 80 – 82.

Copy at :


[5]  Watertow Pty Ltd, Western Australia.  


[6]  Spragg and Associates.


[7]  Market prices for Murray-Darling water entitlements, Department of Environment, Water, Heritage and Arts 


[8]  Burdekin and Tully River resources. Australian Natural Resources Atlas


[9] Burdekin Shire Council Report


[10] Sunwater irrigation price review 2005-2006.


[11] ACTEW Water Security Program


[12] ACTEW Murrumbidgee to Googong Water Transfer,


[13] The Snowy Mountains Scheme – Australia.


[14] Endersbee L.A., Symposia Report, “The Snowy Vision”, Australian Academy of Technological Sciences and Engineering (1999).




The Author

Dr Ian Edmonds is a physicist who operates an R&D company ( that develops sustainable energy products for the building industry and new types of renewable energy systems.