A hulking tower, 50 metres high, rises above Australia’s newest power station on the shores of Lake Illawarra. Since starting up in June, the turbine inside Tallawarra B has shown it can ramp to its maximum output in under 30 minutes and run for as long as it’s needed at any time of the day.

This makes it a valuable asset in a grid shifting rapidly from coal to more fluctuating renewable energy. But its owner, EnergyAustralia, intends to use it as little as possible. That’s because its turbine is powered by natural gas – an expensive fossil fuel that releases planet-heating emissions.

For the most part, and with good reason, the world’s hastening efforts to combat climate change centre on moving away from burning fossil fuels like coal, oil and gas – the key culprits of global warming.

But the green energy transition has a dirty secret: as things currently stand, officials believe Australia has little choice but to urgently lift investment in a whole new generation of gas-powered turbines, like Tallawarra B. These turbines will have an infrequent but important role in backing up renewables and keeping the lights on at times of critical need.

Deploying large-scale renewable energy generation and storage projects and more transmission lines to add them to the grid will be enough to get us most of the way to banishing coal from the grid. But not the whole way, warns the Australian Energy Market Operator (AEMO) in its 25-year transition road map, released last month.

With most coal-fired power plants slated to shut by 2040, energy authorities are pushing hard for a speedier renewable rollout amid concerns that not enough is being built in time to compensate for coal closures. But increasingly, they are also worried about how to fill gaps when the wind disappears for long stretches and clouds blanket the skies – events known in the industry by the German term dunkelflaute, or “dark doldrums”.

According to AEMO and executives at utility giants AGL, Origin Energy and EnergyAustralia, it’s for this reason that gas – a fuel less polluting than coal but still a source of harmful emissions – must remain in the mix for now. While the overall use of gas-fired power is not likely to grow in the future, they say, new infrastructure, including flexible, open-cycle gas-powered turbines, must be added to bolster the grid’s access to rare but enormous injections of electricity when needed.

The ‘Judi Dench’ model

Unlike traditional gas generators that run regularly, so-called gas “peaking” plants (or “peakers”) won’t run often at all. A typical plant may generate just 5 per cent of its annual potential, amounting to a few days, even hours, across a whole year. But the role it plays will be critical.

“I sometimes describe it as the ‘Judi Dench’ model – a cameo role,” says Tony Wood, energy director at the independent Grattan Institute. “If you want to fill a theatre, you get Judi Dench to walk on for five minutes and walk off again – she didn’t do very much, but, boy, was she important. That’s what gas is going to do.”

However, for gas to play that small but significant role, there are high hurdles to overcome. The first is the scale of the build-out required: AEMO’s modelling suggests that Australia needs an extra 13 gigawatts of back-up gas generation to ensure the grid will remain resilient under power supply squeezes and extreme weather.

Assuming each plant can generate about 0.5 gigawatts, approximately 26 plants would be needed. And if recent experience is anything to go by, this will be no small ask: EnergyAustralia’s Tallawarra B is the only peaking plant added to the NSW grid in the past decade, while just one more remains under construction at Kurri Kurri.

At the same time, domestic gas supplies are running perilously low and are forecast to diminish further as legacy gas fields in Bass Strait rapidly dry up without new production to replace them.

Australia is a major exporter of liquefied natural gas, but massive volumes of production in Queensland are locked into overseas export deals, and west coast supplies cannot be sent east because there are no pipelines or gas-import terminals.

“So how do you get the gas to those plants that are only going to run 5 per cent of the time?” asks Wood. “Where is it coming from?”

The risk of supply shortages for gas-powered electricity is so acute that the AEMO plan stipulates that future plants must be built with onsite storage for extra diesel or hydrogen fuel to keep the turbines running if the gas runs out.

Where’s the money?

Ahead of the next election, the climate wars have flared up again, as Labor and the Coalition pitch competing visions for future of energy. Opposition Leader Peter Dutton is pushing for nuclear power plants. The Albanese government is doubling down on efforts to drive the transition by expanding renewables to 82 per cent of the grid this decade.

Neither the government nor the opposition has yet outlined concrete plans to tackle the gas problem, though.

The government’s Capacity Investment Scheme (CIS) – a federally funded program to underwrite projects capable of dispatching on-demand power to the grid when it’s needed – is limited to wind and solar farms backed up by grid-scale batteries or hydroelectric dams.

Going against the advice of the now-disbanded Energy Security Board, which called for a technology-neutral scheme, the CIS’ final design barred gas. That’s because some states, including Victoria, the most gas-dependent in the country, refused to support taxpayer money going to fossil fuels. Developers warn it will be difficult to invest in peakers without some form of government financial support, given the plants will run sparingly, limiting the chance to generate revenue to support the investment.

However, as Energy Minister Chris Bowen points out, the scheme contains a workaround allowing states that want more gas in the power mix to forge ahead with it. To be eligible for funding, states must meet specific reliability standards. That means ensuring sufficient back-up for renewables, which can come in the form of batteries, hydropower or gas.

“As part of these commitments, where necessary, states will commit to strategic reserves comprising technologies of their choosing, including possible gas power generation,” Bowen says.

Dutton, meanwhile, has seized on the need to back up weather-reliant renewables to promote his campaign of adding seven nuclear generators to the grid from 2037. Experts, including the CSIRO, however, say such a timeline is overly ambitious, and warn turning to nuclear technology could lock Australia into the highest-cost form of electricity.

While the Coalition proclaims it also “strongly supports” the need for more gas alongside nuclear and renewables, it has so far only offered broad support for speeding up permitting for gas projects and is yet to detail how it would expand the role of gas-powered generators.

“We have already announced that a returned Coalition government will speed up approvals, unlock gas in new basins, defund the Environmental Defenders Office, reinstate the National Gas Infrastructure Plan and commit to an annual release of offshore acreage for exploration and development in the Northern Territory and Western Australia, and there will be more to come,” opposition energy spokesman Ted O’Brien says.

Climate campaigners have typically argued that gas is not needed as a “bridge” fuel from coal to clean power. But there is increasing acceptance in the green movement that regulators see a limited role for gas in “firming” renewables. The hope now is for that to be as small – and as temporary – as possible.

“The [AEMO] plan’s acceptance of ongoing gas use is disappointing, as gas is a highly polluting fuel,” Australian Conservation Foundation campaigner Ella Factor says.

“But the plan advocates for flexible gas generation only – peaking power, not baseload – so we don’t waste our carbon budget on needlessly burning fossil fuels.”

In addition to gas, other technologies can help plug renewable power supply gaps without the same emissions. Big batteries – mostly lithium-ion – are being deployed at speed to soak up surplus energy when it’s sunny or windy and store it for later use. They also have near-zero start-up time and can inject critical power discharges in response to frequency changes to keep the system stable.

Today’s batteries, though, have key limitations. Their typical size means they will often exhaust their stored energy in two to four hours of maximum output, minimising their ability to plug longer-lasting solar or wind shortfalls.

Pumped hydro is another option. Motors pump water uphill to a higher reservoir, then release it downhill to spin turbines connected to generators. The technology has the advantage of being able to run for many hours, even days. The Snowy 2.0 project is being designed to last for up to a week.

However, these projects are proving difficult to deploy on time and on budget. Pumped hydro needs significant height differentials over short distances, making it challenging to secure the right locations and carry out construction. When it was announced in 2017, Snowy 2.0 had an expected completion date of 2021 and a price tag of $2 billion. By last year, it had blown out to $12 billion and a deadline of 2029.

The last 10 per cent

In the brief time since Tallawarra B opened in NSW, its role in supporting the grid and restraining prices has already been apparent. Last month, a wind drought caused “unusually low” output from renewables, just as cold weather was hiking power and gas demand from home heaters. In the two weeks from June 18, Tallawarra B was fired up 19 times, often at very short notice.

“Whenever demand is high, we turn on Tallawarra B, and whenever supply is low from other sources, Tallawarra B will turn on as required to make sure there is power supply to match demand,” says EnergyAustralia chief executive Mark Collette.

But critically, Collette adds, fast-response plants like Tallawarra B can “just turn off” when there is plentiful supply, unlike coal, which takes hours to wind down.

“The real beauty of something like Tallawarra B is it complements batteries – which are there for an hour or two or maybe up to four hours – with something that can start quickly, run for an indefinite period of time and then get out of the way when there are lots of renewables available,” Collette says.

According to the Grattan Institute, the lowest-cost road to net zero for Australia’s grid is for renewables to supply 90 per cent and gas the last 10 per cent.

However, it says that clean-burning “green hydrogen”—hydrogen made using renewable energy—could one day displace gas in that final 10 per cent, depending on how quickly it can be scaled up and become commercial and competitive.

Tallawarra B’s turbine has been built to incorporate an initial 5 per cent green hydrogen once it’s available. This year, EnergyAustralia also begins a $90 million upgrade of neighbouring Tallawarra A to enable up to 37 per cent hydrogen.

The Business Briefing newsletter delivers major stories, exclusive coverage and expert opinion. Sign up to get it every weekday morning.