We Connect Asia

Hydrogen

Green Hydrogen Price Outlook: What I Put in the Model and Why

10 min read  ·  Sep 9, 2025  ·  By Alex

hydrogen price outlook

The price I write into a model for a hydrogen plant is a cost-based range anchored on what that same plant costs to build and run, not a figure I lift from a headline. Green hydrogen cost two to three times the cost of blue hydrogen in 2020. That gap widened through 2024 and 2025. Cost competitiveness by 2030 remains visible only in China. That is my green hydrogen price outlook. Price outlook means the selling price assumption I carry from one model version to the next. It is a band with a date, a boundary and a named buyer attached. The plant I am valuing is a green hydrogen plant, and before I pay for feasibility work I want that assumption to survive the tests I set out below.

What this article covers

This article orders three things: what the cost of a kilogram is made of, what moves that cost up or down, and the assumptions I refuse to model at all. I keep cost and price apart on purpose, because a cost is not a price. In addition, a scenario is not a market.

In short

  • Cost base: Two to three times the cost of blue hydrogen in 2020. A gap I date rather than carry forward.
  • Composition: Electricity is the larger share of production cost and the electrolyser. The machine that splits water with electricity, is the second.
  • Gap today: The gap with fossil production widened through 2024 and 2025 as gas prices fell and electrolyser costs rose.
  • Geography: Only China looks cost competitive by 2030, so other regions keep a cost premium in my number.
  • Refusal: I model no price that lacks a date, a boundary and a buyer behind it.
Facebook
Twitter
LinkedIn

In this article

The price question I am actually answering

A selling price in a model has to survive three tests, and I write all three down before I choose a number.

First, it must cover the levelised cost of hydrogen, which is the all-in cost of producing one kilogram once capex. The money spent to build the plant. The model also spreads running costs over that output. Second, it must be what a buyer in the target market can actually pay, based on offtake evidence I date to 2023 and 2025. Third, it must hold long enough for the debt to amortise.

A headline price with no date and no condition is worth nothing to a model. The same kilogram under a different contract, in a different year, to a different buyer is a different number.

What this means for you: ask me for the date, the cost boundary and the buyer behind any price before it enters the model, because those three lines are what make the number defensible.

What the cost is made of

IRENA put green hydrogen at two to three times the cost of blue hydrogen in 2020, then split the cost into its parts. Electricity is the larger share of production cost. In addition, the electrolyser is the second.

The same study reported electrolyser costs of 650 USD/kW and 1 000 USD/kW in 2020. Those fall to 307 USD/kW at 1 TW of installed capacity and 130 USD/kW at 5 TW in 2050. In that 2020 study the cost curves ran on electricity at 65 USD/MWh and at 20 USD/MWh.

Cheap power moves the kilogram cost more than the machine does. The study found two cost cuts: 40 percent in the short term and up to 80 percent in the long term. Either could take the price below 2 USD per kilogram within a decade. That is a 2020 projection, not a price I can buy today.

For the same cost stack on a real site, see green hydrogen cost in Southeast Asia in 2026.

What this means for you: before you accept an equipment quote, check the electricity price behind it, because that line moves the cost of a kilogram more than the electrolyser does.

What the gap looks like now

The IEA executive summary of the Global Hydrogen Review 2025 records a wider cost gap with unabated fossil production. Natural gas prices fell from their 2022 and 2023 levels while electrolyser costs rose. As a result, support schemes have to stay in place for longer than planners assumed.

Also, the same summary puts the cost of making and installing an electrolyser at 2 000 to 2 600 USD/kW outside China in 2024, against 600 to 1 200 USD/kW inside China, and notes that a Chinese electrolyser installed outside China in 2024 still lands at 1 500 to 2 400 USD/kW once transport and tariffs are counted.

More than half of that total is engineering, procurement, construction and contingency. Therefore, most of my installed cost sits in the work around the machine. The same summary expects renewable hydrogen to become cost competitive in China by the end of the decade.

In addition, the build-up sits in how plant economics break down line by line.

What this means for you: treat installed cost as a range carrying a year, and treat any single point estimate as unfinished work until engineering and contingency are priced.

What would move it

The 2026 Global Hydrogen Review executive summary says low-emissions hydrogen stays more costly than fossil-based hydrogen in most of the world. China is the exception, where renewable hydrogen could become cost competitive by 2030.

Elsewhere adoption carries a cost premium that someone has to pay. Higher fossil fuel prices can significantly reduce the gap. However, they would have to stay high for extended periods to trigger sustained investment decisions. That is why I never treat an energy spike as a permanent fix.

Urea prices doubled between January and May 2026. That shows what an energy shock does to the price of the products hydrogen helps make. It also shows what a buyer can afford to pay.

That premium has to be reflected in how I structure project finance in Southeast Asia.

What this means for you: run the model twice, once with fossil prices that stay higher for extended periods and once without, and if the project only works in the first case. The premium is yours to carry.

Why location decides the number

The Global Hydrogen Review 2025 page of five key questions records that China can produce renewable hydrogen 40 to 45 percent cheaper than Europe or the United States, and that the electrolyser stack is only 15 to 20 percent of the investment cost of an installed electrolyser.

Buying a cheaper machine therefore changes less than buying cheaper power and cheaper money. Capital costs could reach 15 percent in emerging economies in 2025, three times the level in advanced economies. I define capital costs as the return lenders and equity holders require before money moves.

Location also decides how long a price stays reliable. In that 2025 comparison, Japan offers long-term certainty through 15-year contracts, not a spot sale with no buyer behind it.

The financing route moves with the same variables, as set out in how I finance a hydrogen project.

What this means for you: fix the country first, because power price, capital cost and contract length all move with it, and only then argue about the equipment.

What I refuse to model

I do not model a price that has no date, no boundary and no buyer behind it, and I never let announced capacity stand in for production.

The announced pipeline for 2030 stands at 27 Mt in 2026, against just above 6 Mt of committed and strong-potential projects. Firm offtake agreements were below 2 Mtpa in 2025, equal to 5 percent of announced 2030 production potential.

When firm contracts in 2025 covered 5 percent of announced 2030 output. A price quoted against the rest is a wish dressed as a forecast. In addition, a lender will not fund it.

Next to every price I write three assumptions: the year it applies to, the cost boundary it covers from the plant gate to the delivery point. In addition, the buyer class it is addressed to. A blank field kills the line.

What this means for you: refuse any price that arrives without those three fields completed, however attractive the number looks.

Why trade and certification decide the realisation price

A producer can hit a cost target and still miss the price, because a buyer pays for proof as well as for hydrogen.

The IRENA and RMI study of 2023 assessed eight voluntary and five mandatory schemes. It found that none of the existing hydrogen certification systems suits cross-border trade.

In that 2023 assessment, two of the eight voluntary mechanisms set no emissions threshold. The others set one between 1 and 4.9 kg CO2eq per kg for renewable pathways. Therefore, two shipments can carry the same label and different greenhouse gas information.

The study also identifies gaps in greenhouse gas information, common standards and ecolabelling that prevent fair comparison across borders. Until those gaps close, the price I model has to survive customs, not a producer’s spreadsheet.

Also, the trade side of the branch runs through my page on hydrogen, power to X and e-fuels.

What this means for you: ask which scheme a delivered kilogram will be certified under before you price it, because an uncertified kilogram. In addition, a certified one are not the same product.

How I sequence the pricing work

Between the 2025 and 2026 annual reviews, 300 ktpa of additional production reached a final investment decision. That is the point at which an owner commits the money. In addition, the schedule stops being a plan.

Low-emissions production grew 20 percent in 2025 to almost 1 Mt and passes 1 percent of global production for the first time in 2026, and, in the assessment published in 2025. The cost gap should narrow by 2030.

Each of those markers is a volume or a share rather than a price. None of them can set my selling price on its own. I read the two annual reviews side by side rather than trust one year alone. The direction of the gap matters more than its level.

What this means for you: I set the plant’s own levelised cost first, then test it against what a dated buyer in the target market can pay, and only then write the price into the model with its year, its boundary and its buyer class attached.

Sources

Frequently asked questions

It sits well above fossil hydrogen: two to three times the cost of blue hydrogen in 2020, with the gap widening through 2024 and 2025 as natural gas prices fell and electrolyser costs rose, and the cost of making and installing an electrolyser outside China stood at 2 000 to 2 600 USD/kW in 2024.

The gap is expected to narrow by 2030, with cost competitiveness expected in China by the end of the decade, while other regions keep a cost premium unless fossil fuel prices stay higher for extended periods.

China, where renewable hydrogen can be produced 40 to 45 percent cheaper than in Europe or the United States, and where cost competitiveness by 2030 is expected. Cheap power does more of the work than cheap equipment, since the stack is only 15 to 20 percent of the investment cost of an installed electrolyser.

A dated, conditional range built from the plant’s own levelised cost, the buyer’s ability to pay and the contract length, not a headline figure. Cost cuts of 40 percent in the short term and up to 80 percent in the long term could take the price below 2 USD per kilogram within a decade, as projected in 2020.

Not sure your project would survive a credit conversation?

A short call is usually enough to see where your project stands.

If you are at the point where the selling price assumption has to be fixed before you pay for feasibility work, I build that assumption with you, dated and conditional, inside a feasibility and financial model you can hand to a lender.

I do not provide investment advice or arrange investments. I act as a strategic advisor and facilitator.

Alex, Founder of WeConnect Asia

Alex, Founder of WeConnect Asia

I help low-carbon fuel and carbon removal project developers in Southeast Asia secure the commercial and financial partners needed to reach FID.

More about me·Get in touch

Facebook
Twitter
LinkedIn

Related articles