From The Melbourne Engineer:

Researchers at the University of Melbourne are partnering with Rubicon Water on a wide-ranging project with the aim of improving the operation of large-scale networks of open canals for the distribution of irrigation water, such as those in Australia’s vitally important food bowl, the Goulburn Valley.

Researchers in the Melbourne School of Engineering have been engaged with engineers at Rubicon Water for more than a decade, in a collaborative effort, originally led by Professor Iven Mareels and Associate Professor Erik Weyer, to devise techniques for automating irrigation networks. World benchmark installations of technologies arising from this partnership are operational in Victoria and New South Wales in Australia, Washington, Colorado and California in the USA, China, Chile and Mexico.

Members of the team (Cantoni, Weyer, Dower, Alpcan) recently won a third round of funding from the Australian Research Council Linkage Project Scheme. The $615,000 awarded will allow them to continue work aimed at achieving reductions in the environmental footprint and positive impacts on the productivity of irrigation districts.

Lead Chief Investigator for the new project, Associate Professor Michael Cantoni from the Department of Electrical and Electronic Engineering, said that irrigation canal networks have been used to support agriculture for thousands of years.

“Water is fundamental to everything we do. Future food production pressures call for improvements on many fronts, including the efficiency with which we distribute the water needed to grow our food.”

He said that under manual operation there is a tendency to oversupply gravity-powered irrigation networks in order to satisfy the quality of service requirements of end users.

“The more water you store in an irrigation canal, the bigger the flow available on demand at delivery points without pumping. But once water is committed to the canal network, from a reservoir or river for example, it is either consumed by irrigators or it is wasted due the limited storage capacity of the network and the lack of opportunities to recover spillage at the end. Postponing the release of water from rivers and reservoirs until it is needed, in a demand driven fashion, keeps it available for other purposes such as maintaining environmental flows. Of course, this must be traded-off against the typically long transport delays associated with gravity powered channel networks.”

Assoc. Prof. Cantoni said the team was working on ways to optimize the efficiency of open-canal irrigation networks through automation.

“Now we’re looking at how to augment the automatic control systems we have devised in earlier work so as to overcome the limited performance these can deliver. The main idea is to introduce two higher layers of control to dynamically adjust the levels of water stored throughout the network on the basis of load schedules, and to automate aspects of the ordering process used to produce such schedules, including demand-side management.”

Rubicon Water is an important partner in the project, and an employer of many former University of Melbourne engineering students. Assoc. Prof. Cantoni said that the long-standing project has been invaluable as a training ground for a number of research fellows and students, who have gone on to apply their knowledge in industry, with Rubicon and other companies, as well as academia.

“At the end of the day, whether or not all research outcomes are adopted in practice, the training of people along the way is always a very positive impact.”

He said that thanks to the latest grant, the team was now looking for two or three Ph.D. students and two post-doctoral researchers. Potential researchers with interests in control systems, signal processing, and game theory, are advised to contact Assoc. Prof. Cantoni for further information.