A few years back I sat through a pitch deck in a Fremantle boardroom that had hydrogen doing everything: trucks, trains, home heating, peaking power, export cargoes to Japan, the lot. Every slide had a different end use and every end use had the same green arrow pointing at it. Nobody in the room asked the question that actually matters, which is: how much electricity does it take to make this stuff, and is there a cheaper way to do the job it’s being asked to do? That question still isn’t being asked often enough, and it’s the one this piece is built around.
Green hydrogen, hydrogen split from water using electrolysis powered by renewable electricity, is a real technology doing real work in a handful of places. It is not, on current economics, a general-purpose decarbonisation tool. The gap between those two statements is where most of the public argument about it gets lost.
Agree on the units before you agree on anything else #
Electrolysis takes electricity in and makes hydrogen out, and every step of that conversion loses energy as heat. Run the hydrogen through a fuel cell or burn it to make electricity again, and you lose more. Depending on the pathway, round-trip efficiency from renewable electrons back to useful electrons can sit somewhere in the vicinity of 30 to 40 per cent – compare that to a big battery, which the AEMO and various grid studies put at roughly 85 to 90 per cent round trip. That single number is why hydrogen is a poor fit for short-duration grid firming and a perfectly sensible fit for jobs where you were never going to use electricity directly anyway.
This isn’t a criticism of the chemistry. Water electrolysis has been understood since the early 1800s and the physics hasn’t changed. What’s changed is that renewable electricity got cheap enough, in patches, for the economics to start working on some use cases. The CEFC and ARENA have both been candid in various published materials that the technology’s value proposition is about displacing fossil-derived hydrogen and fossil fuel inputs in processes that already need a molecule, not a general substitute for wires.
Ammonia and existing hydrogen demand are the obvious starting point #
Australia already makes and uses a fair amount of hydrogen, mostly as “grey” hydrogen derived from natural gas, feeding into ammonia production for fertiliser and industrial processes. That’s roughly 500,000 tonnes a year of existing demand, give or take, sitting at fixed sites with existing infrastructure, existing customers and an existing price benchmark to beat. Swap the grey hydrogen for green at an ammonia plant and you’ve displaced real emissions without needing a single new truck, train or appliance to change behaviour.
This is the single least glamorous and most defensible use case in the whole hydrogen conversation, and it’s underweighted in the public debate relative to how much it actually matters. Yara’s ammonia plant at the Pilbara and various proposed green ammonia projects around Gladstone and the Eyre Peninsula are chasing exactly this market, not new demand, but decarbonising demand that already exists and already pays a known price. When a developer tells you they’re targeting existing ammonia offtake rather than an entirely new hydrogen economy, that’s usually the more credible pitch.
Steelmaking is the genuinely hard problem worth solving #
Hydrogen-based direct reduced iron is one of the few plausible pathways to decarbonise primary steelmaking, because you need a reducing agent to pull oxygen out of iron ore, and hydrogen can do that job where coking coal currently does it. This isn’t a case where electricity could do the job directly instead. You need the molecule. Fortescue has poured real money into hydrogen-based iron and steel trials, and BHP has run its own direct reduction pilot work, both of them chasing the fact that global steel buyers, particularly in Europe and increasingly in parts of Asia, are starting to price carbon content into long-term offtake contracts.
The catch is scale and cost. Replacing a blast furnace’s coking coal with green hydrogen at the volumes Australia’s iron ore exports imply would need an amount of new renewable generation and electrolyser capacity that dwarfs anything built in the NEM to date. The CSIRO’s GenCost work and various AEMO Integrated System Plan scenarios treat large-scale green steel as a 2030s-and-beyond proposition at best, contingent on electrolyser costs falling the way solar panel costs fell over the past fifteen years. That’s a plausible bet. It is not yet a sure one, and the honest framing is “promising pilot stage”, not “arriving”.
Heavy transport, not your driveway #
Hydrogen fuel cell vehicles keep getting raised as a competitor to battery electric cars, and for passenger vehicles I think that argument is largely settled the wrong way for hydrogen. Battery costs have fallen far enough, and charging infrastructure has spread far enough, that the case for a parallel hydrogen refuelling network for everyday cars doesn’t stack up against the round-trip efficiency penalty described above. You’d be building two infrastructure systems to do one job, and the less efficient one needs more renewable generation behind it for the same number of kilometres driven.
Where hydrogen has a more credible shot is heavy, long-haul freight and certain mining and rail applications, where payload and refuelling time matter more than they do for a commuter sedan, and where battery weight starts eating into the thing you’re actually trying to carry. Anglo American and Fortescue have both trialled hydrogen in heavy haul mining equipment. It’s early, the fuelling infrastructure is thin on the ground, and nobody should mistake a demonstration fleet for a solved problem, but the physics argument for this niche is sturdier than it is for family cars.
The export dream has run into the price reality #
A good chunk of the hydrogen enthusiasm in Australia over the past five years was about exporting the stuff to Japan and South Korea as a shipped fuel, either as liquid hydrogen or converted to ammonia for the voyage. Several flagship projects have been scaled back, delayed or had partners withdraw as the hard number came in: landed green hydrogen in Asia, after production, conversion, shipping and reconversion losses, has consistently priced above what buyers have been willing to commit to in binding, long-term contracts at the volumes originally pitched.
That doesn’t mean export is dead, but it means the timeline has stretched and the volumes have shrunk from the headline gigawatt-scale announcements of a few years ago to more modest, bankable first stages. It’s a pattern this desk has seen before in other corners of the transition: the ambitious first number is a marketing number, and the bankable number comes two or three years later once someone has to sign a contract. Readers who followed how SunCable’s export ambitions compressed after Sun Cable’s initial ownership shake-up will recognise the shape of this story.
Why hydrogen keeps losing to batteries and transmission for grid firming #
This is worth separating out because it’s where the policy conversation gets muddiest. Green hydrogen is sometimes pitched as a long-duration storage solution for the grid: make hydrogen when renewables are abundant, store it, burn it in a turbine or fuel cell when the wind drops. The chemistry allows it. The economics, at current electrolyser and storage costs, generally don’t, because you’re paying the round-trip efficiency penalty on every cycle, on top of capital costs for electrolysers, compression or liquefaction, storage vessels, and the generation plant on the other end.
Batteries win on short-duration firming by a wide margin, which is why the the firming maths comparing gas peakers and big batteries has increasingly tilted toward batteries for anything under about four hours. For the genuinely long, multi-day firming gap, pumped hydro and transmission interconnection are doing more of the heavy lifting in AEMO’s planning than hydrogen storage is – see how pumped hydro versus batteries actually plays out for that comparison. Hydrogen for grid firming isn’t impossible, it’s just currently the most expensive way to solve a problem that transmission and pumped hydro solve more cheaply, which is roughly the same conclusion the AEMO Integrated System Plan keeps landing on when it models the technology mix out to 2050.
Where the next few years actually point #
My honest read, for what it’s worth, is that Australia’s hydrogen industry over the rest of this decade will look a lot less like the glossy 2021-era renders of liquefaction trains at the Pilbara, and a lot more like unglamorous electrolysers bolted onto existing ammonia and fertiliser plants, plus slow, careful progress on green iron pilots backed by Fortescue, BHP and a handful of international steelmakers with carbon-exposed customers. That’s a smaller story than the one that got pitched a few years ago, but it’s a more honest one, and frankly a more investable one: the Capacity Investment Scheme and various state hydrogen strategies are increasingly being written with that narrower scope in mind rather than the everything-pitch I sat through in Fremantle.
None of this makes hydrogen a bit player in the broader transition. Steel and ammonia are enormous, hard-to-abate sectors and decarbonising even a fraction of them matters more for global emissions than another gigawatt of grid-scale batteries. But it does mean the test for any hydrogen project crossing your desk should be the same one that boardroom in Fremantle skipped: what’s the existing molecule demand, what’s the realistic cost of green hydrogen landed at that demand, and is there a cheaper way, probably involving a wire instead of a pipe, to do the same job. Most projects that can’t answer that cleanly are worth a long second look before the cheque gets signed.
Whether the industry settles into that narrower, harder-edged shape or keeps chasing the broader vision is still an open question, and one I’ll keep watching, probably from somewhere with a decent view of the Swan and a glass of local cider, which is about as far from a Pilbara electrolyser as this correspondent gets most weeks.
– Tom Fitzgerald, Baseload & Fuels Correspondent
Photo by Jackery Power Station on Unsplash