A new, large-scale, water-cooled nuclear reactor that has just come online in the US state of Georgia is estimated to have cost about $US35 billion ($53 billion).
The potential for large taxpayer subsidies is reason to be circumspect about nuclear, despite its appeal in providing reliable, zero emissions power over a very long time.
But regrettably, the CSIRO’sGenCostreport has several shortcomings.
It fails a fair apples-with-apples comparison betweenPeter Dutton’s nuclear ambitionsand the Albanese government’s renewables.
Nuclear lasts 80 years, not 30
GenCostunderplays the potential use of nuclear and underestimates the total system costs of unreliable renewables.
First,GenCostcalculates the cost of various energy projects over 30 years.
Nuclear power plants provide always-on power for households and businesses for more than 80 years. Solar panels typically last up to about 30 years, after which they are old and inefficient. Wind turbines last up to about 25 years. Batteries for energy storage last up to about 15 years and may in the future reach 20 years.
After these renewables expire, billions of dollars must be spent – again – on replacements.
The environmentally damaging materials, such as huge amounts of scrap metals, must be replaced. Hopefully, they are recycled.
Repeated investment cycles are required.
For solar panels, replacements are needed 2.7 times over 80 years. For wind turbines, replacements are required 3.2 times over 80 years. For batteries, replacements are needed four to five times over 80 years.
Hence, the cost of energy sources should be benchmarked over a “like for like” whole-of-life period.
If not 80 years, then at least halfway between the 20-30-year life of renewables and the much longer life of a large nuclear power plant. Let’s call it, say, 50 years.
The CSIRO says it uses 30 years because it is the time commercial finance is usually available before repayment is required.
Yet, after 30 years a nuclear power plant should have a residual economic value or resale price, which is ignored by the CSIRO.
A second shortcoming of theGenCostreport highlighted by Centre for Independent Studies energy research director Aidan Morrison is the assumption that the capacity range of nuclear power plants is between 53 per cent and 89 per cent.
Modern nuclear plants overseas can run for above 90 per cent of their capacity. The ongoing costs of nuclear power plants are very low after the upfront build, so they typically run near full throttle.
Finally, the CSIRO uses the levelised cost of electricity (LCOE) to compare the competitiveness of power generation technologies.
‘Apples v oranges’ comparison
But it underplays the advantage of always-on, dispatchable power, such as fossil fuels and nuclear, over mainly off, intermittent wind and solar generation.
Former Treasury economist Geoff Carmody says this “apples versus oranges approach is very misleading”.
To fix this, the LCOE should be calculated on a reliability-equal, whole-of-life-asset basis.
Solar power, on average, is fully on about 20-25 per cent of the 24-hour day and off about 75-80 per cent of the day. Wind power is fully on about 30 per cent of the day and off 70 per cent.
The CSIRO attempts to adjust for the unreliability of renewables in its LCOE estimates by including costs of storage, transmission and firming.
But the assumptions underpinning the renewable integration costs are murky because the CSIRO says it doesn’t want to release modelling that could contradict the Australian Energy Market Operator’s integrated system plan.
The CSIRO admits that LCOE is a “simple screening tool” and is “not a substitute for detailed project cash-flow analysis or electricity system modelling which both provide more realistic representations of electricity generation project operational costs and performance”.
David Pearl, a former Treasury assistant secretary and adviser to former Labor prime minister Kevin Rudd, says the LCOE is an accounting metric, not an economic one.
“It measures the total unit costs a generator must recover to meet all expenses – plant, equipment, land, raw materials and labour – including a return on investment,” Pearl writes inThe Spectator.
“It says nothing about the revenue side of the commercial equation: what prices can the generator earn on the wholesale market and, given their costs, what profits can be earned?”
Pearl likens it to an unreliable car that costs less to make than a reliable vehicle but is of less value to consumers.
“By the same logic, an inherently unreliable source of power, like solar or wind, cannot be said to be cheaper in an economic sense than a reliable source of power, regardless of how much it costs to supply when the sun is shining and wind is blowing, which as we know is only 20 to 40 per cent of the time,” he writes.
“For the 60 to 80 per cent of the time when intermittent power cannot be supplied at any price, its economic cost can be said to be infinite.”
Pearl calls for the Productivity Commission to conduct a renewable energy audit that includes: the direct cost of subsidies; the system-wide costs – including transmission, storage and back-up dispatchable power including the cost of subsidies to keep coal-fired power stations operating; the economic, social and environmental losses wind, solar and new transmission lines are inflicting on farmers and others living in regional communities; and the welfare costs of a more volatile and less reliable grid.
“This renewable energy audit should be embraced by all in the community, regardless of their views on climate change, net zero and the merits of nuclear energy,” Pearl argues.
CSIRO updated previous report
The CSIRO has previously fixed shortcomings in its analysis to give a more realistic estimated cost of the net zero energy transition.
Morrison last year exposed theGenCostreport for assuming that any pre-2030 cost of building renewables was “sunk”, or effectively zero.
This artificially lowered the estimated cost of renewables, such as the six-fold blowout in the Snowy Hydro 2.0 project to $12 billion and tens of billions being spent on batteries and transmission. These capital costs will be paid for by energy users and/or taxpayers.
To the CSIRO’s credit in response to Morrison’s pressure, it updated theGenCostreport to reflect the previously ignored pre-2030 costs.
The independent CSIRO must maintain its credibility and give a realistic “apples with apples” comparison of nuclear and renewables to inform the energy transition debate.
Australia needs evidence-based policy, not policy-based evidence.
