What if making green hydrogen cheaper to produce is not enough to make it commercially viable?

As governments and businesses explore hydrogen's role in the energy transition, much of the conversation focuses on reducing production costs, developing more efficient electrolysers and expanding renewable electricity generation. These are important priorities, but they only tell part of the story.

In the first article of this series, I explored why green hydrogen has yet to achieve widespread adoption despite its considerable potential. One of the central conclusions was that commercial viability depends on an interconnected ecosystem of energy generation, production, infrastructure, storage and demand.

Through our research with Imperial Business School, we examined how decisions across this ecosystem influence costs and where opportunities exist to improve the business case. In this article, I look at the physical side of that ecosystem. Using actual hourly 2025 weather data for Great Britain, we modelled what it really takes to supply an industrial customer with a constant 1,000 kg of hydrogen every hour from wind and solar power. What becomes apparent is that making green hydrogen competitive requires us to look beyond individual technologies and consider how the entire supply chain balances energy over time.

The price of hydrogen is only the beginning - hidden costs of the hydrogen supply chain

Green hydrogen remains significantly more expensive to produce than conventional alternatives. While technological advances and economies of scale may help narrow this gap, focusing exclusively on production costs risks overlooking other factors that determine commercial viability.

Renewable electricity must be sourced, electrolysers installed and operated, and, once hydrogen has been generated, it may need to be transported and stored before reaching the end user. Each stage introduces costs: transport distances, infrastructure availability and storage requirements all affect commercial viability. This means that two organisations purchasing hydrogen at the same production price could face very different overall economics.

Figure 1 shows five common ways of connecting renewable power to an industrial user. In a fully co-located system (A), renewables, electrolyser, storage and user share one site. The electrolyser can instead sit at the user's site and draw power from the grid, bought on the wholesale market or through a PPA (B). Production can also happen where renewable resources are strongest, with hydrogen delivered by road as compressed gas (C), by a dedicated pipeline (D) or, for intercontinental trade, by ship after liquefaction or conversion into ammonia (E).

Figure 1. Hydrogen value-chain configurations, from renewable electricity to industrial end use. Electricity flows in blue, hydrogen in green, hydrogen carriers (liquid hydrogen or ammonia) in violet. Configurations A and B are modelled in this article.

Each step in Figure 1 adds equipment, energy losses or both: compression for trucking, a pipeline that must be built and filled, or the conversion and reconversion steps that make shipping possible.

The impact of such supply chain decisions is clear if we consider two industrial organisations purchasing hydrogen produced at the same cost. One is located close to production facilities, with suitable infrastructure and predictable demand. The other operates further away, requires more complex transportation arrangements and needs greater flexibility in supply. Although their initial hydrogen costs may be identical, the economics of delivering and using that hydrogen could be substantially different.

This illustrates why improvements in individual technologies do not necessarily translate into equivalent savings across the supply chain. A reduction in production costs could be offset by infrastructure or transportation requirements elsewhere. Our research with Imperial reinforces the importance of understanding these relationships, rather than evaluating each stage independently.

Why cheaper electricity doesn't automatically mean cheaper hydrogen

Renewable electricity is fundamental to green hydrogen production, but its availability fluctuates with weather, time of day and season. Industrial demand, meanwhile, is often flat: a refinery, an ammonia plant or a steelworks typically needs the same amount of hydrogen every hour of the year. This creates a mismatch between when hydrogen can be produced and when it is required.

Figure 2 makes that mismatch visible. It follows a plant built to supply 1,000 kg/h (about 33 MW of hydrogen energy) entirely from wind and solar, through three weeks of real GB weather in October 2025. Making exactly 1,000 kg/h takes around 53 MW of electricity. Because renewable output swings between almost nothing and well above that level, the plant needs an 84 MW electrolyser, 1.6 times the continuous requirement, so that it can catch the windy and sunny hours.

The first ten days are windy. The electrolyser runs at full load for long stretches, producing about 1.6 t/h, and the surplus 0.6 t/h flows into H₂ storage, which gains around 67 tonnes, while output above the electrolyser rating is curtailed. Then the wind drops to almost nothing for six days. The electrolyser falls to a fraction of its rating, storage supplies up to 90% of the customer's demand, and H₂ storage loses more than 130 tonnes, ending the period almost empty.

Figure 2. Hourly simulation over 1–21 October 2025 for a wind & solar + electrolyser + hydrogen storage plant supplying a constant 1,000 kg/h. (a) Renewable generation against the electrolyser rating; (b) electrolyser load against the 53 MW needed to match demand, with the surplus sent to storage and the shortfall drawn from it; (c) hydrogen storage inventory.

These calm spells, rather than the gap between day and night, are what set the size of H₂ storage. Figure 3 shows the same H₂ storage over the whole year: it fills in windy and sunny weeks and is drawn down hard in calm ones, several times a year. To keep the customer supplied with 100% renewable hydrogen, this particular system needs around 150 tonnes of storage: more than six days of demand, even with a deliberately smooth, fleet-wide wind profile. A single wind farm would need more. Accepting 10% grid electricity shrinks H₂ storage to two days of demand, but only because the grid steps in every time it runs empty.

Figure 3. Hydrogen storage inventory across 2025 for the 100% renewable design (152 t) and the 90% renewable design with grid back-up (48 t). Storage is dimensioned by multi-day low-wind periods rather than by the daily solar cycle.

So rather than focusing exclusively on electricity prices, organisations must consider how production patterns, equipment sizing and storage interact. Cheap renewable electricity is only cheap if the system can use it when it is available and still deliver hydrogen when it is needed.

Can batteries close the gap?

A natural question is whether a battery could do some of this work. A battery energy storage system (BESS) can absorb surplus electricity and release it when renewable output drops, smoothing the electrolyser's operation and allowing a smaller unit.

Figure 4 compares the two approaches over three days in March. With a battery of 13 MW / 58 MWh, the electrolyser shrinks from 84 MW to 79 MW and hydrogen storage falls from 152 to 148 tonnes. The battery cycles at least once a day and does what it is designed to do: it moves energy by a few hours.

Figure 4. Configuration 1 (renewables + electrolyser + hydrogen storage) versus Configuration 2 (with a BESS) over 72 hours (19–21 March 2025). (a) Electrolyser load; (b) battery state of charge; (c) hydrogen storage inventory. Both designs are sized for a 100% renewable supply.

But the problem that storage has to solve lasts days, not hours. Figure 5 shows what happens as the battery is made progressively larger. A 400 MWh battery would cut the electrolyser by almost a fifth and more than halve the hydrogen storage, yet the plant would still need 21 tonnes of H₂ storage.

The scale gap explains why. The 150 tonnes needed for a fully renewable supply hold about 5,000 MWh of chemical energy, nearly 90 times the 58 MWh battery in Figure 4. Batteries are excellent at shifting energy within a day; bridging a windless week is a different order of magnitude. Even when batteries improve flexibility, a significant volume of hydrogen storage is still required.

How much storage does a renewable supply need?

To see how these requirements build up, we sized the same plant for different shares of renewable electricity, from a grid-powered electrolyser to a fully renewable supply. Figure 6 shows the result.

Figure 6. Design requirements for a constant 1,000 kg/h as the renewable share of electricity rises: (a) hydrogen storage in days of demand; (b) electrolyser capacity relative to the 53 MW continuous load; (c) electrolyser utilisation and curtailed renewable output.

A grid-powered electrolyser at the customer's site is simple: it runs almost continuously (97% utilisation), needs only a one-hour buffer and an electrolyser barely larger than the load. It is also not green: on the 2025 GB grid mix, its electricity embeds around 6.7 kg of CO₂ per kilogram of hydrogen.

Every step towards renewable supply moves the problem into the plant. At 80% renewable electricity H₂ storage already must hold more than a day of demand; at 100% it holds more than six. The electrolyser grows to 1.6 times the continuous load, yet runs at only 63% of its capacity over the year, and around 14% of the renewable output cannot be used and is curtailed. The last few percentage points are the hardest: going from 95% to 100% renewable almost doubles the storage required.

Could storage actually make hydrogen cheaper?

Storage provides an interesting example of how infrastructure investment can create economic value. While often viewed as an additional expense, our research highlights how storage can help organisations manage renewable energy intermittency, improve supply stability and potentially reduce costs.

By allowing hydrogen to be produced when renewable electricity is abundant and made available when required, storage turns a variable resource into a reliable supply. The larger and more flexible that storage is, the more renewable output the system can capture instead of curtailing, and the longer it can ride through calm spells without falling back on the grid.

Different business models can also affect these costs. Our research identifies opportunities for longer-term savings when storage is integrated more closely into the hydrogen supply chain, where third-party mark-ups would otherwise affect margins. These benefits must however be balanced against upfront investment. This illustrates an important distinction: the lowest-cost individual component does not necessarily deliver the most economical overall solution. Sometimes, strategic investment in one part of the supply chain can improve the economics of the wider system.

For organisations considering green hydrogen, the price per kilogram therefore remains an important measure, but it cannot tell the whole commercial story. The reliability of supply, infrastructure requirements, operational flexibility and commercial arrangements all influence long-term economics.

No approach is inherently preferable. The most appropriate model depends on the organisation's circumstances and how its production, storage and supply arrangements interact.

Making hydrogen competitive requires smarter decisions

There is understandable optimism that technological progress, expanding renewable capacity and economies of scale will improve green hydrogen's competitiveness. However, waiting for production costs to fall is not the only route towards a stronger business case.

As explained, our research with Imperial suggests that opportunities also exist through smarter decisions about how hydrogen is supplied, stored and commercially managed: how renewables are connected, how the electrolyser is sized relative to demand, and how much flexibility is built into the system and in what form.

It is equally important to recognise that hydrogen will not necessarily be the most cost-effective decarbonisation solution in every application. Businesses must first understand where hydrogen makes commercial and operational sense before determining the infrastructure and arrangements required to support it.

Storage: the missing piece

Taken together, the results point in one direction. Renewable hydrogen is intrinsically variable; industrial customers need it to be constant. Electrolysers, batteries and storage can be combined to bridge that gap, but each comes with trade-offs. Batteries handle hours. Hydrogen storage has to handle days, and the closer we get to a fully renewable supply, the more of it we need.

That makes the ability to store hydrogen at scale one of the decisive variables for renewable hydrogen. Without it, system designers are pushed towards oversized electrolysers, curtailed renewables and grid back-up. A storage solution able to hold weeks rather than days of demand would change those trade-offs.

Perhaps the more valuable question is not simply how we make green hydrogen cheaper to produce, but how we make the entire hydrogen ecosystem more cost-effective, and where the hydrogen can wait between production and use.

This leads me to the final article of this series, where we explore one of the most important decisions within that system: hydrogen storage. With multiple technologies available, how should organisations determine which solution offers the greatest value?

Methodology note: results come from an hourly model of capacities and operation over 8,760 hours of 2025 GB data (Elexon onshore wind generation and Sheffield Solar PV_Live, scaled to new-build capacity factors), with a constant demand of 1,000 kg/h and an electrolyser consumption of 52.7 kWh/kg. Each configuration is sized with an optimisation model; the economics of these configurations are discussed in the final article.

Download the full BIP and Imperial research paper, Fuelling Sustainability and Savings: The Business Case for Green Hydrogen Storage, to explore the findings in more detail.