Water restrictions during the drought cost households $1 billion a year, and a similar amount could have been saved if construction of the Sydney desalination plant had been delayed, according to the head of the Productivity Commission's inquiry into water, Wendy Craik.


Speaking at a water policy conference in Adelaide earlier this week, the Ms Craik said that state governments building six plants across the country had failed to take into account the financial burden of water saving measures on urban customers.


"The costs to the community have been high . . . higher than what they needed to be," she said.


"The sorts of things we're talking about are costs of putting in grey-water recycling systems, putting in rainwater tanks, rearranging your life so you can water your lawn when you're allowed to."


Dr Craik said repairing pipes in homes and unit blocks to stop leakage added to the toll.


"With all those sorts of costs . . . if you add it all up it is in excess of $1bn a year," she said.


She also criticised the billions of dollars wasted on building desalination plants, singling out Sydney's $1.9bn plant and Adelaide's $1.8bn facility.


"In Sydney, a review in 2006 of water augmentation options indicated that . . . if you delayed putting in a desal plant until the dams were at 30 per cent, you would save over $1bn a year," she said.


The government hade instead decided to go ahead with the plant in 2007 when dam levels were at 34 per cent of capacity, leading to higher building costs.


Furthermore, South Australia's Rann government should have saved money by buying water from upstream sellers of water licences, instead of building the $1.8bn desal plant.


SA Water chief John Ringham defended the decision as the state's "insurance policy" against future drought.


University of Adelaide's Mike Young criticised federal government spending on licence buybacks and infrastructure in the Murray-Darling Basin that cost $593,000 an irrigator.


Professor Young said the Murray-Darling Basin Authority "got it wrong" by determining environmental flows on the basis of average yearly flows instead of drought years.