215 facilities. That’s the current population of the Safeguard Mechanism, the scheme that’s supposed to be the sharp end of Australia’s industrial decarbonisation. Steelworks, cement kilns, LNG trains, alumina refineries, coal mines with big fugitive emissions. Between them they account for something like a third of the country’s total emissions, and the mechanism is meant to bring that number down on a schedule. The honest read is that it’s working, slowly, unevenly, and mostly on paper.
I’ve been going back through the Clean Energy Regulator’s published safeguard data for the past few weeks, partly because the scheme is due for another look under its statutory review process, and partly because the debate around it has gone quiet in a way that usually means someone’s about to make noise. Heavy industry lobbyists aren’t loud right now. That’s not the same as content.
What the Safeguard Mechanism is actually meant to do #
The mechanism sets a declining emissions baseline for each covered facility. Miss the baseline and you buy Australian Carbon Credit Units to cover the gap. Beat it and you generate Safeguard Mechanism Credits you can sell. It’s a baseline-and-credit scheme, not a cap-and-trade one, which matters more than it sounds like it should. Facilities aren’t competing for a fixed pool of permits. Each one gets its own declining line, roughly 4.9 per cent a year, and is judged against itself.
That design was the 2023 reform’s big call, and it was defensible. A flat industry-wide cap would have handed windfall advantages to whoever got there first. But it also means the scheme’s total ambition is only as good as the sum of 215 individual baselines, and some of those baselines were set generously.
Who’s actually exposed and who isn’t #
Not every covered facility feels this the same way. A gas-fired power station competing in the NEM passes costs through to wholesale prices reasonably easily. An export-facing LNG train or an aluminium smelter selling into a global commodity market can’t do that, not without losing the contract to a producer in a country with no equivalent scheme. That’s the trade exposure argument, and it’s a real one, not an industry talking point invented for effect.
We’ve written before about how that exposure plays out on the ground — Zen Energy Australia’s position propping up Whyalla is a case study in what happens when a trade-exposed, emissions-heavy asset also happens to be the only large employer in town. Steel doesn’t get a carve-out from physics, but it does get carve-outs from policy, and the two aren’t always aligned.
New gas fields and the credibility test #
The sharpest argument right now isn’t about the existing 215 facilities. It’s about the ones not yet built. Every new LNG project, every Beetaloo-style gas development, has to enter the Safeguard Mechanism on day one with a baseline calculated off international best-practice emissions intensity. Santos has been through this exact fight over Narrabri and the Moomba carbon capture project it’s using partly to manage its own compliance position. Whether CCS at that scale delivers what’s modelled is still an open question, and I’d treat the injected-tonnage figures with some scepticism until a few years of actual monitoring data are in.
The gas industry’s defenders argue the sector has been written off prematurely by people who haven’t looked closely at where east-coast supply actually comes from. There’s something to that — we’ve made the case ourselves that the industry gets dismissed too early in some of these debates. But the Safeguard Mechanism is precisely the test of whether new gas can genuinely operate at a credible emissions intensity, not just claim one on a compliance filing.
Follow the money: where the credits actually go #
Here’s where it gets interesting, and where most of the coverage stops short. Safeguard Mechanism Credits and ACCUs aren’t burned once used. They’re traded. A facility that beats its baseline generates a saleable asset; one that misses it buys someone else’s headroom. Follow the money and you find that a chunk of the heaviest-emitting facilities in the country are, in effect, paying a modest carbon price to a handful of landfill-gas projects, native forest regeneration schemes and a smaller number of genuine industrial abatement projects that happen to be ahead of schedule.
ACCU prices have sat in the low-to-mid thirties per tonne for a while now, nowhere near high enough to force a genuine technology switch at a cement kiln or a blast furnace. That’s the honest read on why heavy industry hasn’t rushed to electrify or install carbon capture at scale. It’s cheaper to buy the credit than fix the process. The mechanism is doing what a market is meant to do — find the lowest-cost abatement — but the lowest-cost abatement in this market often isn’t inside the covered facility at all. It’s a forestry offset three states away.
The reliability side nobody mentions #
There’s a second-order effect worth flagging. As coal generators close faster than new firmed capacity comes on — a trend we’ve tracked closely and one that keeps outpacing the replacement build — some of the heavy industrial load that the Safeguard Mechanism is meant to be squeezing is also becoming critical demand-response capacity for AEMO. A smelter that can shed load during a tight afternoon is doing the grid a genuine favour, and the Capacity Investment Scheme has started to recognise adjacent flexibility even if the Safeguard baseline doesn’t give it much credit directly. The two schemes were designed by different parts of government at different times, and it shows.
What the review needs to actually fix #
The scheme is due another look under its legislated review cycle, and the sensible items on the list are not exotic. Baseline-setting methodology for new entrants needs tightening so a new gas project can’t lock in a generous intensity figure that ages badly. The interaction between Safeguard credits and the voluntary carbon market needs clearer rules so buyers aren’t double-counting abatement. And someone needs to have an honest conversation about whether trade exposure assistance is protecting genuinely uncompetitive-without-it industries, or just delaying an adjustment that was coming anyway.
My own view, and it’s a contrarian one in this sector, is that the Safeguard Mechanism gets too much credit for ambition and too little scrutiny for mechanism design. It looks tough on paper — a declining cap, a hard 100 million tonne ceiling by 2030 for covered facilities. But a scheme where compliance can be satisfied almost entirely through purchased credits rather than site-level abatement isn’t really squeezing heavy industry. It’s squeezing their accounts payable. That’s not nothing — money changing hands does eventually change behaviour — but it’s a slower, blunter instrument than the headline figures suggest, and I think the government’s own modelling probably knows it.
What to watch next #
Two things worth tracking through the rest of this year. First, whether the Department of Climate Change, Energy, the Environment and Water’s consultation on the review timetable produces any tightening of new-entrant baselines — that’s the number that will tell you whether Beetaloo-scale gas gets an easy run or a genuinely hard one. Second, whether ACCU and Safeguard credit prices move meaningfully off their current range. If they stay flat, the honest read is that the mechanism has settled into a compliance cost rather than a decarbonisation lever, and heavy industry will keep paying it the way a club side pays a fine for a slow over rate — annoyed, but not changing anything about how they actually play.
Whether that’s good enough for a scheme meant to carry a third of the country’s emissions to net zero by 2050 is the question the review should actually be asking. So far, nobody in Canberra seems keen to ask it out loud.
— Marcus Wren, Editor
Photo by Compagnons on Unsplash