Energy reliability: risks so small “they are barely visible”

Modelling done for the AEMC’s Reliability Panel has put a lie to the scare campaign on Australia’s energy supply – led by the federal government – saying the risks of power supply not being met are “so small, they are generally not visible on the chart.”

The assessment of Australia’s power supply includes the expected closure of the Liddell coal-fired generator in 2022, and contrasts with the Coalition government’s scare campaign about potential blackouts if the closure goes ahead.

In its “base” scenario – modelled on the current settings of the RET, the Victoria renewable energy target and no particular action from the federal government on climate – the risk of supply shortfalls in NSW even after the closure of Liddell is 0.0000010 per cent.

That’s right – 0.0000010 per cent (five zeros after the decimal point) – which is to say, one-millionth of one per cent.

Contrast this to what the government, including prime minister Malcolm Turnbull and energy minister Josh Frydenberg have been telling us about the “greater chance of blackouts” if the ageing and increasingly unreliable Liddell coal generator in NSW is closed.

This is what conservatives have been telling us: “This not only puts the state at risk of blackouts but more importantly it drives up the cost of electricity,” says Craig Kelly, the chair of the backbench committee on energy.

The government has been using the scare campaign about blackouts to try to force AGL to either keep the plant open, or sell it to a company such as the rival retailer and generator Alinta.

And Alinta, which on Monday reportedly tendered a formal but non-binding offer to AGL to buy Liddell,  has been lapping it up and also warning of a “greater risk of blackouts”, in the hope of getting a forced sale and a cheap entry into the NSW market.

But the modelling done for the Reliability Panel by EY is emphatic. The reliability standards would not be breached. It says: “EY’s modelling found that NSW would require a further 1,300 MW of NSW coal capacity to be retired to exceed the reliability standard in NSW.”

Its modelling also showed that the greatest risk of any supply shortfalls was in Victoria, but the chance of this was also so remote as to be barely visible – in fact, the risk was 0.0000065 (six and a half times in a million – see top graph).

In only three out of 200 simulations trying to create a shortfall did the modelling actually succeed, and this would be on a single January day in Victoria in 2021, and would rely on a sequence of unlikely events.

This would require half the brown coal capacity being unavailable at the same time, only 5 per cent of wind, limitations on the network links to NSW, and extremely high demand.

In fact, the only scenarios that might lead to a breach of the reliability standard would be in the case where one-third of the gas-fired capacity in South Australia was retired within the next couple of years, or if one half of all brown coal capacity in Victoria was retired before 2024.

Both outcomes are highly improbable, and if they did occur – contrary to the new rules that require significant notice of retirements, then presumably the market operators could have time to respond – most likely with demand management and behind the meter resources.

And try as it might, it couldn’t model any supply shortfalls in Queensland, even if it forced the retirement of 1,500GW of relatively youthful coal generation. It said this would simply send an economic signal for the construction of 2,000MW of wind and solar, with no threat to supply.

The EY modelling also dismissed concerns that a move to five-minute settlement periods – rather than the current 30 minute settlement period – would have any substantial impact on reliability settings.

The modelling was used by the Reliability Panel to recommend that Australia’s existing reliability settings, which include delivering energy 99.998 per cent of the time, remain in place.

This is one of the tightest reliability settings in the world – and equates to the allowance of just 11 minutes of outages per customer per year.

The Reliability Panel dismissed suggestions that the reliability standard should be lifted to 100 per cent, saying that would require billions of dollars of additional investment.

“Setting the reliability standard involves a trade-off between the prices consumers pay for electricity and the cost to consumers of not having electricity there when it’s needed,” said panel chairman Dr Brian Spalding.

“In doing this review, the Panel was extremely cognisant of getting the balance right to avoid what some have called ‘gold plating’ with excess capacity built but not required for years.”

Many argue that gold plating has existed for years, particularly with network costs. Yet 98 per cent of all outages experienced by consumers comes from equipment failures, bushfires and storms – and not from unserved energy.

The Reliability Panel also said the settings on the market price cap ($14,200/MWh), the administrative price cap (in case of interventions), the cumulative price threshold, and the market floor should all stay unchanged.

Spalding said the panel had modelled a range of scenarios that included a world with more household battery storage systems and more wind generation. It also factored in the current state of uncertainty in the market and how the market is transforming.

RenewEconomy asked minister Frydenberg’s office for a response to the Reliability Panel modelling, and whether warnings of blackouts had been exaggerated, but had received no response by the time of publication.

Update: AGL said in a statement it had received a “highly conditional” and non binding offer of $250 million cash for Liddell from Alinta and its Chinese owner. It said it would think about it.