The ever-shrinking role of gas in Australia’s power system

There’s a lot riding on predictions of the future, when it comes to climate. In Australia, the future of coal is mostly a debate about whether its demise is fast, or slow. For the country’s gas-fired power generation, the debate is slightly more complicated.

The government, for instance, believes that Australia ought to be building new gas-fired power stations, in order to buttress grid reliability and “affordability”. 1,000 megawatts of new gas in the Hunter Valley is erroneously justified on the grounds that it is “required” to ensure reliable provision of electricity. This view is supported by Australia’s Labor party, with their draft policy platform declaring that “Labor recognises that gas-power generation has a critical role to play in firming the National Electricity Market (NEM) to ensure reliability and price affordability as it transitions to net zero emissions and as other technologies emerge”.

The other key player in this debate is former Chief Scientist Alan Finkel, who’s strongly maintained that gas ought to retain a role in Australia’s future power grids. “Maybe 20 or 30 years from now we’ll have new kinds of batteries, vastly powerful more extensive batteries and we can do it with batteries,” Finkel said last year. “But gosh, the quickest way to develop our renewable electricity system is to support it with gas”.

The Sydney Morning Herald reports today that Finkel, now “Special advisor on low emissions technology”, has been tasked with convincing world governments that Australia “has a plan to achieve net zero emissions”. He has reiterated his support for fossil gas. “I like to say that if you were to remove public health from society, you’d be back to the Renaissance. If you were to remove education from our society you’d be back to the Middle Ages. And if you remove manageable energy from society you’d be back to the Stone Age”.

The ‘gas and renewables’ partnership narrative has been championed by the gas industry, too. Instead of framing themselves as at odds with zero carbon sources, they present themselves as complementary, with a liberal usage of “low emissions” (how low? Don’t think about it) and cute visual metaphors ramming the message home through paid Facebook advertising:

The underlying story here is that climate change and emissions have both been deprioritised, and fossil fuel technology, economics and system security have come to take up the bulk of the conversation about Australia’s power system. This is pretty clear in a recent report from Australia’s gas pipeline industry group.

AEMO’s ISP isn’t as anti-gas as you might think

The 2020 iteration of the Australian Energy Market Operator (AEMO)’s Integrated System Plan (ISP) models a least-cost bundle of technological changes that meets cost and reliability demands into the coming decades.

In most of their scenarios, including those which outline a “slow” change, gas comprises only a small fraction of either total electricity generation, or total installed capacity of generation. Their modelling leans on new interconnection between regions, battery storage, pumped hydro, demand response and new planning schemes for renewable energy (“renewable energy zones”) to do the majority of the work in integrating wind and solar into the grid.

The installed capacity of gas remains relatively stable in each of AEMO’s scenarios, but the actual generation is relatively low. This is because for the most part, gas is only called upon to generate during extended periods of relatively low wind.

AEMO’s 2020 ISP is, if anything, slightly favourable about how much gas there will be in the future grid. “GPG can provide the synchronous generation needed to balance variable renewable supply, and so is a potential complement to storage”, write AEMO in their report. “Existing combined-cycle gas turbines (CCGTs) and open-cycle gas turbines (OCGTs) are forecast to play critical complementary roles when significant coal generation retires in the 2030s”.

The “need for gas powered generation” is even detailed in a special section highlighting gas output during a couple of low wind periods …. in 2035. ‘Baseload’ CCGTs run during wind droughts, and the fast-response OCGTs run during times when low winds coincide with failures in other generation and low solar output.

AEMO even includes an additional section detailing new gas pipelines and new gas mining that’ll have to occur to sustain this level of gas-fired power in the NEM, alongside demand from other uses. “To ensure this level of supply, let alone enable a greater reliance on gas, a policy intervention may be required”, they write.

This depends on a fine balance between the costs of gas, and the costs of batteries, and AEMO highlight this too. “In the 2030s when significant investment in new dispatchable capacity is needed, [the cost] advantage could shift to batteries, especially to provide dispatchable supply during 2- and 4-hour periods. Based on the cost assumptions in the ISP, new batteries are more cost-effective than GPG in the 2030s”.

As energy analyst Simon Holmes à Court points out above, an optimal simply doesn’t require as much gas is required today. And this is before we consider pricing the emissions impact of extracting the gas used as fuel, or the carbon impact of burning it to make electricity. The ISP isn’t anti-gas – it carves out as big a space for it as it possibly can. But even that process ends up with a tiny sliver among a wide range of zero carbon options for integrating VRE.

Australia’s gas industry is also pro-gas

Recently, the Australian Pipelines and Gas Association, APGA (an industry group that has members like Shell, AGL and Jemena) published a study that it claims supports the need for a gas in high VRE systems. They commissioned Frontier Economics to complete the analysis, which asserts that gas is required to “insure” the system against stretches of low wind and solar output.

It’s odd, because it’s presented as something of a rebuttal to AEMO’s ISP. “[APGA] wants to ensure that the next version of AEMO’s Integrated Systems Plan, due in 2022, and the federal government’s National Gas Infrastructure Plan don’t set the scene for under-investment in gas generation and infrastructure such as pipelines”, writes the AFR’s Angela MacDonald-Smith.

Their modelling does, as you might expect, find that a system with some gas-fired power – up to 7% – is cheaper than a zero emissions grid. But strangely, it isn’t all that different from AEMO’s ISP. The results for South Australia’s installed capacity of gas-fired power in 2035 in their most pro-gas scenario – 93% renewables – is around 1,000 megawatts, for both 2030 and 2035. That’s actually even lower than what the ISP assumes, which is around 2,500 megawatts in 2030 and 2,000 in 2035.

Most significantly, it’s a very direct admission that South Australia is bloated with far too much gas capacity, and that even within ten short years, running a secure system while shutting down gas-fired power is completely feasible:

This model is sort of like a small slice of AEMO’s ISP, except with a bunch of features of that model removed. Demand response is excluded. It models South Australia’s grid, but with interconnection to other regions excluded. And it models only two years – 2030, and 2035, based on AEMO’s wind traces used for its ISP report.

If you remove two major zero emissions tools for integrating VRE into the NEM, demand response and greater interconnection, you would absolutely expect a modelled system to lean more heavily on gas, and for integration to be more expensive. It’s tough to figure out what value this analysis brings – it doesn’t even really refute the ISP – it reshuffles some variables with mixed results. Are they arguing we ought to build more gas-fired power stations? Or that we shouldn’t retire existing ones early? It really isn’t clear.

Like so many reports of this kind, what is framed as a “trilemma” – between security, costs and emissions – is treated like a “dilemma”, with the focus solely on the first two. Emissions are only mentioned once, and feature in none of the resulting data. Generally, the issue is shrugged off. The head of APGA told the AFR recently that gas “will ultimately probably be better offset than removed from the system” – the magic of carbon removal strikes again.

And on reliability, the claim that gas is key to serving large volumes of demanding during periods of low wind and solar output comes at a relatively awkward time. Last week, a cold snap froze the gas supply network in Texas, leading to a cascading failure of gas pipelines and gas-fired power stations, resulting in the largest single blackout event in American history, with incalculable costs to Texas.

American style gas in power systems – where huge plants provide constant, ‘baseload’ power is not a model any country is all that interested in. Europe has begun taking actions to place strict emissions control on gas-fired power, with Europe’s largest gas plant cancelled only days ago. That came paired with a Carbon Tracker report which shows that ‘clean energy portfolios’ of batteries, demand response, energy efficiency and renewables were far more cost effective than the UK’s planned gas fleet. And only a few months ago, the CSIRO’s draft GenCost report found that renewables in Australia paired with storage and transmission were far cheaper than gas – even at a 90% VRE share.

This report, which shows gas’ only remaining role in a power system is rare, infrequent, emergency backup, as confusing as the results are, seems at the very least to be a confirmation that gas’ role in power systems is shrinking fast. The next question is, of course, how long before even this small carve-out gets rubbed out by some cost-effective, long-term zero carbon storage option?

The underlying problem: Australia has no climate policy

The under-discussed problem around many of these debates is simply that Australia has no clear emissions control policies, and therefore it isn’t a factor in these reports, or corresponding coverage.

The “trilemma” is pure symbolism – you rarely find detailed climate metrics in the PDF outputs of modelling agencies and fossil industry groups. There is an Energy Security Board, but there is no Emissions Reduction Board. Pay attention to this, and you’ll notice how rarely climate is treated as a ‘must have’. Mostly, it’s ‘we’ll take whatever we can get’.

It is a relatively neat illustration of how a total absence of any clear climate policy from Australia’s government feeds all the way down into the lobbying of fossil industry groups and visions of future energy systems from planners. As I’ve written previously, even AEMO’s most ambitious scenario isn’t aligned with a 1.5C compatible trajectory; you’d never know that from reading these reports, or many of the articles that cover them.

“The Panel encourages the Australian Government to develop a national 2050 emissions reduction strategy by 2020, consistent with commitments under the Paris Agreement” was recommendation 3.1 of the 2017 Finkel Report. “This will set expectations and help to guide investment decisions in the electricity sector by providing an anchor point for Australia’s long-term emissions trajectory”.

Finkel may not be mentioning the government’s notable failure to take this on, four years on. But the consequences are clear: decision makers, lobbyists, industry groups and companies are free to forget about one third of the trilemma, and ignore the impacts of climate change in their messaging.

Update 05/03/2021 – Added links to advertisement, details on Gencost and minor wording changes