The groundwater at Rome was notoriously unpalatable, and water from the Tiber was unsafe to drink. Aqua Appia, Rome’s first aqueduct (312 BC) was commissioned by the censor Appius Claudius Caecus as a publicly funded major project.

By the late 3rd century AD, Roman aqueducts supplied Rome with water by a combined conduit length of 800 kilometres, of which only 47 km were carried above ground level by the familiar masonry supports. They supplied around 1 million cubic metres (ie 10,000 GigaLitres) a day; enough to supply a modern city of population of 10 million. The longest was the Constantinople (Turkey) around 500 km. Gradients for Roman Aqueduct were 1:4800 and typically around a metre wide and deep with a flow rate estimated to be about 35,000 m3 /day (or 350GL/day) depending on the season .

With the fall of the Roman Empire, some aqueducts were deliberately cut by enemies but many more fell into disuse through lack of organized maintenance. Their failure had an impact on the population of cities; Rome declined from its high of over 1 million people in the Imperial era to as low as 30,000 in the medieval era. 

The New Bradfield Scheme is very similar length, capacity and function to the Roman aqueducts being totally gravity fed. With an estimated fall of 1:5000, 100m from Hell’s Gate over a distance of 500 km to the Lakes Buchanan and Galilee Storages, a wider, lined channel could easily provide the 20,000GL per annum. Harvesting the flow from streams along would be possible, or desirable helping to reduce flooding lower in the catchments and leading to potentially greater harvest rates. Spillways would dump the excess if the water level got too high.

As Sir Humphrey Appleby said about public projects (‘Yes, Minister’): “Anything is possible for government, so long as it isn’t the first time.” The Roman Empire has done a similar-sized gravity-fed aqueduct system 2000 years ago.

In Queensland, CSIRO have identified the Mitchell catchment (yellow area above) as the greatest potential to increase greenfield development opportunities in Northern Australia, and the coastal regions of the Fitzroy and Burdekin River catchments for enhancements. However, developments in these would do nothing for the water poor, rich soil areas of Clermont, Richmond, Longreach, Barcaldine, Muttaburra that already have established agricultural infrastructure that could be expanded, established human settlements (and importantly voters), and further west potentially channel water into the Murray Darling Basin.

The CSIRO studies have not, so far, evaluated the feasibility of aqueducts between catchments. Therefore, the locations they identify rely on the coincidence of good soils, dam locations and rainfall. They do not consider connected water networks with harvesting levees and high catchment dams, or aqueducts that could transfer water between high rainfall poor soil, to low rainfall good soil locations. Not considered, also, are potential higher returns from enhanced town water supplies.

The problem – very few people live in the Mitchell catchment or want to go there (see map). I am concerned the CSIRO are going down the wrong path again, potentially creating another massive ‘white elephant’ like the Ord River Irrigation Scheme that may take 100’s of years to reach the projected potential. A system of gravity-fed aqueducts from the high rainfall northern rivers of Tully and the Walsh Rivers into the existing populated outback areas with irrigation potential – aka the new Bradfield Scheme – needs to be included in the CSIRO feasibility studies.

CSIRO have investigated the potential of northern Australia’s water resources to support increased regional development as part of our engagement in delivery of the Australian Government’s White Paper on Developing Northern Australia, for which one of the key initiatives is the development of northern Australia’s water resources. They have delivered water resource assessments for three priority regions in northern Australia for the Australian Government after working with northern jurisdictions, research partners and communities over 2.5 years. An orphaned dam in the Mitchell catchment is all we get? Disappointing.

Let’s look at similar water storage and distribution systems to the New Bradfield Scheme, in California.

California is 400 km wide and 1200 km long. It has a land area of 423,970 km² of which 36,421 km² is irrigated. Agricultural production is $47 billion per year, a large proportion of which is irrigated or irrigation related.

California is largely desert and its population centres and agriculture watered by a number of long aqueducts exceeding 500 km in length.

For example, the California State Water Project, commonly known as the SWP, collects water from rivers in Northern California and redistributes it to the water-scarce but populous south through a 650 km length aqueduct, with pumping stations and power plants. About 70% of the water provided by the project is used for urban areas and industry in Southern California and the San Francisco Bay Area, and 30% is used for irrigation in the Central Valley.

By comparison, Northern Queensland has a land area of approximately 500,000 km². An extended Bradfield Scheme may provide irrigation for 30,000 km2 of land. This would be achieved by a system of aqueducts and dams, of similar length to similar to those in California. In return, we would expect agricultural production of $50 billion per year, similar to California.

In some ways Queensland is more suited than California, as due to good luck or blessing, the new Bradfield Scheme may entirely gravity fed through tunnels and aqueducts. The cost of water may therefore be considerably less than the cost of water in the SWP in California.

The construction of the aqueducts in the new Bradfield Scheme could be staged, developing first those areas that are closer to population centres and prepared for irrigation, and releasing the excess water into parched river systems. However, the irrigation of large areas black soil plains in the central state would open up the largest areas to new agriculture.

The new Bradfield Scheme could provide 30 thousand km2 of new irrigated land in inland Queensland, more than doubling the land under irrigation, and catapulting Australia from 23rd place by irrigated land area (between Peru and Japan) to 8th place (between Brazil and Thailand).