The Current You Are Buying

Three routes split water at industrial scale, and the published parameters tell you which one you are actually buying. Proton exchange membrane runs at technology readiness level 8 to 9 and returns 60 to 68 per cent efficiency on a lower heating value basis. That is 50 to 55 kilowatt hours of electricity for one kilogram of hydrogen, delivered at 30 to 70 bar, with a stack that lasts 60,000 to 80,000 hours. Alkaline is the mature one. Readiness level 9, 63 to 73 per cent efficiency, the same 50 to 55 kilowatt hours per kilogram, 10 to 30 bar at the outlet and a stack rated for 60,000 to 90,000 hours. It is the cheapest hardware in the room and it has the longest track record. Solid oxide is the efficient one. Readiness level 6 to 7, 74 to 84 per cent efficiency and up to 90 per cent when the waste heat is recovered, running at 700 to 850 degrees Celsius with a stack rated for 20,000 to 40,000 hours. The International Energy Agency, IRENA and the United States Department of Energy published all three sets of figures. They do not agree on which is best, because they are not answering the same question. One route is compact. One is cheap. One is efficient. You get to pick one.

The headline efficiency is quoted at a single operating point, and that point is not where the plant spends its life. Proton exchange membrane holds a dynamic range of 30 to 100 per cent load and ramps faster than the alternatives. Alkaline responds more slowly and carries a corrosive liquid electrolyte. Solid oxide is tied to steady state operation and dislikes thermal cycling. So 60 to 68 per cent and 74 to 84 per cent are both true, and neither one describes a machine following a renewable supply curve for a year. A plant that runs at 90 per cent of its rated efficiency for 2,000 hours and at 60 per cent for the other 6,760 has an average that the data sheet does not print. The load range is the number nobody puts on the front page.

Capital cost separates the three routes by a factor of four. Proton exchange membrane sells for 1,000 to 1,250 dollars per kilowatt in Western markets and 600 to 900 dollars for Chinese equipment. Alkaline is 750 to 1,000 dollars and 300 to 500 dollars respectively. Solid oxide is 2,000 to 4,000 dollars per kilowatt today, with a target near 800 dollars by 2030 and a projected reduction of 50 per cent to between 720 and 810 dollars. Now take the levelised cost of hydrogen for 2025 and 2026. Membrane comes in at 3.50 to 6.00 dollars per kilogram on grid power and 2.50 to 4.00 dollars on renewables. Alkaline is 3.20 to 5.50 dollars. Solid oxide is 3.00 to 5.00 dollars where there is waste heat to use. Read that again. The expensive machine produces the cheapest kilogram, if you can find the heat. The capital is paid once. The efficiency is charged every hour for twenty years.

Two specifications decide the cost of the stack, and neither of them is the efficiency. The first is current density. Membrane runs at 1 to 3 amps per square centimetre. Alkaline runs at 0.2 to 0.8. Solid oxide runs at 0.3 to 1.0. A higher density means fewer cells for the same output, which means less steel, less floor space and less balance of plant. The second is what the catalyst is made of. A membrane stack loads 1.3 grams of iridium per kilowatt today, against a target of 0.4 grams. The metal is scarce. At current loading the published assessment puts a ceiling on membrane scaling at roughly 50 gigawatts a year. Alkaline uses no precious metal at all. Solid oxide uses none either. So the trade is not efficiency against cost. It is current density against material risk. One route carries a supply chain that ends at a mine. Two do not.

One route is compact. One is cheap. One is efficient. You get to pick one.

Once the hydrogen exists, the product line is a menu with published prices, and every item starts at the same meter. Ammonia is the Haber-Bosch loop. 400 to 500 degrees Celsius, 150 to 250 bar, 15 to 25 per cent yield per pass, and 8 to 10 megawatt hours per tonne of ammonia when the hydrogen comes from natural gas. That loop fed the world and accounts for 1.6 to 2.0 per cent of anthropogenic carbon dioxide. Green ammonia lands at 600 to 1,200 euros per tonne. Grey ammonia is 300 to 400 euros. Methanol is carbon dioxide plus three hydrogen at around 250 degrees Celsius and 50 to 100 bar, at 56 per cent system efficiency, consuming 1.4 kilograms of carbon dioxide per kilogram of methanol. Biogenic methanol on green hydrogen is 900 to 1,200 euros per tonne. Fully renewable is 4,000 to 4,510 euros. Grey methanol is 300 to 400 euros. Synthetic aviation fuel through Fischer-Tropsch returns 35 to 45 per cent of the electricity as liquid, at 3,000 to 6,000 euros per tonne. Synthetic methane converts above 95 per cent per pass at 200 to 600 degrees Celsius, at 50 to 65 per cent power-to-gas efficiency. Every one of those numbers starts with a kilogram of hydrogen and a price for electricity. The molecule at the end is packaging.

Electrochemical ammonia synthesis sits at readiness level 2, on the bench. The lithium-mediated route carries a theoretical faradaic efficiency near 28 per cent and has demonstrated under 10 per cent in practice. The binding energy of ammonia on transition metals is the barrier, and the literature has not removed it. Solid oxide degradation is unresolved. Twenty thousand hours of stack life against ninety thousand for alkaline is a factor of four and a half, and thermal cycling shortens it further. Synthetic aviation fuel is expected to stay near three times the cost of fossil kerosene even in the 2050 case, because the conversion efficiency caps the arithmetic before any learning curve applies. The honest sections of the record are the useful sections. Each gap here is a research question with a budget attached.

The mandates are the buyer, and they are written in years. Renewable Energy Directive III sets a 29 per cent transport energy share, or a 14.5 per cent greenhouse gas intensity reduction, by 2030, with a 5.5 per cent subtarget for advanced biofuels and renewable fuels of non-biological origin, and a maritime target of 1.2 per cent from 2030 in member states with ports. ReFuelEU Aviation mandates 2 per cent sustainable aviation fuel at European airports in 2025, rising to 6 per cent in 2030 of which 1.2 per cent must be synthetic, 20 per cent in 2035 with 5 per cent synthetic, 34 per cent in 2040, 42 per cent in 2045 and 70 per cent in 2050 with 35 per cent synthetic. FuelEU Maritime sets a greenhouse gas intensity reduction of 2 per cent in 2025, 6 per cent in 2030, 14.5 per cent in 2035, 31 per cent in 2040, 62 per cent in 2045 and 80 per cent in 2050. Not one of those quantities is a preference. Each is a volume that has to be bought from somebody. When a market is created by a regulation, the regulation is the business model.

I restructured a industrial group and built companies across twelve countries. I have deployed 210 energy systems across Africa and Asia, and every one of them was a bet on the electricity price and on the running hours. Entrade built 200 wood gasification machines across Britain and Europe. Those machines were sold on a fuel number, and the fuel number was the whole sale. The same trade is open again, and it is the same arithmetic. On one side, capital per kilowatt. On the other, kilowatt hours per kilogram, multiplied by a power price you do not control and by a number of hours you do not choose. Efficiency is a figure on a data sheet. Current is the thing you buy. The probability that the cheapest kilogram in 2035 comes from the efficient machine rather than the cheap one is not zero, and most of the market is pricing it at zero. The electricity is the product. Everything else is packaging.