Australia's clean energy experiment

As I've been arguing for over five years in previous opinion pieces, Australia's official power generation statistics make it obvious that we have no chance of meeting the clean energy needs of an all-electric fossil-fuel-free modern industrialised economy. Governments crow about progress of "renewable energy targets". But they are irrelevant, applying only to present electricity grids, not to those ultimate, much greater, national energy requirements.

Now AEMO, the Australian Energy Market Operator has confirmed my pessimistic claims. Under this recent headline from The Australian, 26 June 2024 "Put rocket under renewables for net zero" AEMO makes the extraordinary admission that "Australia will need to install 16 times its current capacity of batteries and pumped hydro by 2050, while large-scale wind and solar generation will have to jump six fold, if the country is to deliver its transition to net-zero emissions by 2050".

After many years of overexcited reporting of progress of the energy transition the quango in charge is not happy. How has this suddenly happened?

Let's start with the basics. Clean energy almost always comes as electrical energy, electricity generated with little or no accompanying carbon dioxide emissions. Concerns about climate change have led many nations to decide on progressive closure of their fossil-fuel based electricity generation. Australia is one of them. Power companies must switch to clean energy sources. The choices generally are the ones we call renewables – solar, wind and hydroelectricity. Here we rely on solar and wind for growth as most hydroelectricity sites are already exploited. Nuclear is excluded, by law.

Progress is monitored against clean energy targets. Two kinds of target with different time horizons have emerged. In the shorter term, all existing generation will be clean; that's the target. The longer term aim is to eliminate all uses of fossil fuels. That requires a suite of new technologies running on clean electricity. Such "electrification of everything" is a much more ambitious exercise. It calls for much more clean electricity and many technology innovations.

These are big programs. They need plans with timelines and milestones, even an initial feasibility study to make sure it's all achievable.

Australia has done some of those things. Clean energy preferences are in place. Progressive rollouts of solar and wind installations occur. The two targets have names, "100% renewables" for present electricity grids and "Net Zero Emissions" for future retirement of all fossil fuels by around 2050.

Specific milestones for progress of rollouts against targets seem lacking. And now it appears that AEMO has suddenly discovered the rollouts are proceeding much too slowly. Bad decisions? Bad planning? Bad management? Hard to tell.

Let's look at the plans. Where? They're not easy to spot. AEMO does publish something it calls the Integrated System Plan, which it says is a "roadmap". I look at it later. The present evidence is that until very recently AEMO and the government seem not to have noticed signs of the shortfalls just announced.

I've been looking at slow growth rates and progress of solar and wind generation for several years. Actual clean energy outputs offer the clearest way to see progress and shortfalls. Does anyone take notice? Maybe the general atmosphere of self-congratulation from the industry and excitement about "renewable energy superpower, world class renewables" etc. drowns out any critical analysis.

My latest update of energy data is in the table below, which shows Australia's total annual electricity generation, by energy source. The numbers are units of energy. Note that AEMO tends to communicate in terms of installed capacity (which measures the limit of what can happen), and power (which measures a rate of energy generation, transmission or consumption). Those are all important network parameters. But the ultimate determinant of adequacy of an electricity system for the nation is its electrical energy output. Energy is what must be measured. That's one thing that AEMO needs to fix.

The numbers in the table are energy totals for all Australia. One often sees somewhat smaller numbers for the National Electricity Market, essentially the East Coast grid, which comprises about 75% of national total generation.

The energy units used here are petajoules, PJ. They may be less familiar than traditional electrical units like kilowatt-hours or terawatt-hours but they now have a firm place in Australia's official statistics. Conversion to other units is simple, e.g. 1 PJ = 0.278 TWh. 1 TWh = 3.6 PJ.

The differences between the first two rows measure remaining contributions from fossil fuels (mainly coal and natural gas).

The critical numbers for annual clean energy growth from solar plus wind are highlighted in the last row. The table shows that prior to 2019 (only two points included here) growth rate of solar and wind output rose steeply. Then growth looks to have stabilised at around 35 PJ per annum for the next five years (to date). That is the true measure of growth under current plans.

How does it look in relation to estimated targets? Are we nearly there yet?

As I said above there should be two targets, one for Australia's existing power supplies, the other for future clean electricity requirement for replacing all fossil fuels.

The first is easy. The numbers for "% renewables" are in the third row of the table. If we assume clean generation keeps rising at the same 35 PJ per annum then that number will reach 100% in around 18 years, in 2042.

The second is tougher. There is as yet no agreed target, no agreed method for calculating it. That's partly because the "clean technologies" for many major industrial products now reliant on fossil fuels, like steel, fertiliser, explosives, plastics and cement, are yet to be developed (despite politicians always sounding confident – they have no idea). Without commercial trials and application there is no way of knowing the electricity requirements for those hypothetical processes.

However several other ways of making informed estimates have been published and they seem to be converging. These "educated guesses" suggest a figure at least 2.5 times today's electrical energy usage. Let's use that now as a first approximation, remembering it's rough and probably a bare minimum.

In 2023 Australia used 983 PJ of electricity. On that basis we would need 2.5 X 983, say 2460 PJ, to displace all fossil fuel use. Latest data for renewables generation is 343 PJ for 2023, which means another 2460 – 343 PJ to go. That is, we still need another 2120 PJ (rounded) solar and wind to reach the target.

Let's assume the solar/wind growth rate stays the same as for the "% renewables" target, 35 PJ per annum. At that rate it would take 60 years. Certainly not high enough to it the tentative 2050 deadline, which is only 26 years away. To do that we'd need to raise the average annual increase to about 80 PJ per annum, 2.3 times faster than now. AEMO calls for a sixfold leap to reach net zero. Why the difference? It's hard to say on what has been published.

The conclusion we are stuck with is that total solar/wind energy output growth is now too slow to meet a conservative target.

AEMO's Integrated System Plan is its "roadmap". It presents a set of scenarios comprising various assumed mixes of built generation, transmission and storage that could complete the clean energy transition. How, when and where those projects will happen, who manages them and what they'll cost are not given. And no-one appears to be in charge of implementation.

The ISP is not a normal construction plan. It's "a plan for investment in the NEM to ensure a reliable and secure power system through Australia's transition to a net zero economy".

So AEMO's role is to urge investment in renewables and offer investment signals to the market. In other words, with governments foreshadowing closure of fossil fuel generator, the energy market operator is encouraging investors to enable the energy industry to replace them with renewables. Sounds like crowd funding. Is it working? The recent "rocket" comment suggests not.

Which brings me to the question, is an Australian net-zero clean energy future based on the trio solar, wind and storage actually feasible? There was no feasibility study. It was just assumed. The last few years comprise a preliminary experiment that so far suggests the renewables trio is not adequate.

That's precisely the reason why nuclear energy must be added to the means of generating clean energy. Solar and wind are not doing the job.

What improvement would nuclear energy make to the growth rate? The arithmetic is easy. Like coal-fired power stations, nuclear power plants are often built to total capacities of around 1 gigawatt, much like the coal-fired power stations they must replace. Often they comprise several smaller units on the one site. Power engineers decide these things around what's commercially available or feasible and the expected demand.

Annual energy output is a simple function of rated power and capacity factor (fraction of the year when the unit works at rated capacity). A generator with rated output of 1 GW and capacity factor 80% has an annual output of 25 PJ.

As we have seen Australia's total solar/wind output is presently increasing by 35 PJ per annum, well below what energy policy needs. A typical nuclear power plant can supply 25 PJ per annum. That's not far behind what's lost when a coal-fired plant shuts down. The current shortfall facing us is easily made up with nuclear energy.

Cost and construction rates always come up in nuclear energy debates. On cost the current government has a clear position; "everyone knows renewables are cheapest". Both the Federal government and AEMO rely on one main source of cost information for deciding policy, CSIRO's annual GenCost report. This now includes nuclear costings, even though AEMO itself has explicitly rejected nuclear energy.

It has been suggested that GenCost omits extra system costs like new transmission and storage ("firming") in its renewables costings. Conversely others claim that GenCost's nuclear costings are too high. Nuclear is consistently dismissed by Australian politicians on the basis of its cost. The slogan "nuclear too slow, too expensive" caught on about a decade ago and is still influential. Cost debates will no doubt remain vigorous while the relevant evidence is in short supply.

I enjoy recounting my involvement long ago in a peer-reviewed scientific study of nuclear costs (https://www.sciencedirect.com/science/article/abs/pii/S036054421000602X.

Nicholson, Biegler, Brook, Energy, January 2011). We used published technical cost literature from 2000 to 2010; the study is certainly dated. But it remains relevant and of special interest in its conclusion that just 13 years ago we claimed nuclear was the cheapest clean electrical energy technology for Australia, at least under the conditions explained below.

At the time, 2010, it looked inevitable that Australia would be introducing a price on carbon emissions. As it turned out, it didn't, but it would have hit hard the competitiveness of fossil fuels in power generation and improve prospects for the low emissions technologies. Also we took into account a set of "fit-for-service" criteria. Candidates for future low emission power generation had to produce reliable, continuous, controllable power (often called baseload power), just as electricity consumers have always expected.

The result was unsurprising. Wind and solar energy were scratched before the race started. Without additional technologies they couldn't meet that condition.

Was our fit-for-service exclusion list contentious at the time? Not that I'm aware. Nor do I recall we ever considered including costings for standalone solar and wind power. They were simply unacceptable given their uncontrolled variation with weather and time of day.

Now there are questions about whether the "renewables are cheapest" claim of today might be omitting proper costs of storage. It's expensive and usually site-specific, which makes generalised cost estimates impractical at the least.

If debate on lifting the nuclear ban is coming, cost will undoubtedly come up as a central issue. This raises an awkward question. Can a future nuclear contribution to Australia's electricity supply be properly costed while nuclear power remains illegal?

I doubt it. Arguably, while nuclear energy is illegal the proper process for obtaining reliable cost estimates from the global nuclear construction industry can't be used. Quotations are expensive to prepare. Why would a business bother when they know the job can never proceed?

It's worth recalling how Australia's nuclear ban came about. Former Prime Minister John Howard recently reminded us. He says he was 'blackmailed' into the ban as the price for getting minority (Greens Party) support for a modernised nuclear reactor at Lucas Heights (in Sydney). That facility run by Australia's Nuclear Science and Technology Organisation (ANSTO) is a small nuclear reactor dedicated to manufacturing radioisotopes for medical and industrial applications and to research in materials science. It does not generate nuclear energy. The ban was a piece of political deal-making totally unconnected with clean electricity generation for reducing carbon emissions. Now 26 years later it is a legal barrier to Australia using a clean energy technology common in some 30 industrialised nations (the USA has 93 nuclear reactors generating electricity, France 56, Australia zero).

AEMO says regular updates of its Integrated System Plan cannot include consideration of nuclear energy because it's illegal. But AEMO commissions from CSIRO an annual electricity technology cost survey GenCost that does include nuclear costings. How a technology deemed illegal by government can be costed impartially and reliably by one government authority for another government authority is surely puzzling.

And why doesn't AEMO choose ANSTO, Australia's Nuclear Science and Technology Organisation, as its preferred provider of information on nuclear energy, including cost? Another puzzle.

Nuclear may well be expensive. Its cost has definitely been rising in recent years. There are no obvious reasons other than reduced growth and reduced competition in the construction industry. The global nuclear industry has been shrinking or stagnant since the boom period of the 1970s. Rising interest in clean energy might reverse the trend. It's hard to say. Competition is always a significant economic factor.

For the time being there's no doubt that recent high nuclear construction costs abroad will be milked in any political conflict on lifting the ban in Australia. My view is to take the cost of nuclear energy in Australia as indeterminate while it's not legal and wait until proper cost estimates can be provided by the nuclear construction industry. Costings from other avenues should be treated with caution.

Indeed all Australia's energy policy decisions should be based on sound evidence. Growth rates should always be based on electrical energy outputs. With the correct data it is easily concluded that Australia's clean energy growth from solar and wind alone will not meet needs. Growth planning and financing strategies need improvement. Nuclear energy may well be needed to supplement and even replace renewables. Nuclear costs will be debated but are still largely speculative. Furthermore while nuclear power is illegal Australia cannot seriously engage in establishing its true cost. For that reason alone any government contemplating nuclear energy must first remove the ban. Also the electorate must be given the full costs of a renewable energy supply fully equipped (firmed) with the added technologies needed to make it "fit-for-service" as reliable, continuous, controllable power. Until then an informed choice cannot be made.