The Scottish Home Hydrogen Trial And The Ethics Of Delay

Support CleanTechnica’s work through a Substack subscription or on Stripe.

The Fife, Scotland, hydrogen trial is moving forward at Easter 2026, and that timing matters more than it might seem. Easter is a season associated with renewal, honesty, sacrifice, and the choice to leave behind what no longer serves the common good. Against that backdrop, SGN is advancing a project that does the opposite. H100 Fife is not a bridge to the future of home heating. It is a mechanism for delaying the end of the gas distribution business model, wrapped in the language of innovation and decarbonization. By 2026, that is not an innocent mistake. The economics are known. The safety evidence is known. The policy direction is known. The end-of-trial reversion to natural gas is known. What remains is not uncertainty, but a deliberate choice to preserve a legacy network at the expense of ratepayers and, in practice, at the expense of some of the more economically vulnerable households in Scotland.

H100 Fife is a purpose-built local hydrogen system in Levenmouth, serving homes in Buckhaven and Denbeath with hydrogen produced by electrolysis, stored on site, and distributed through a new local network. SGN presents it as the United Kingdom’s first end-to-end hydrogen system for homes. Nel’s public material says the electrolyser can produce up to 2,093 kg of hydrogen per day and that the system was sized for up to 900 homes, even though the initial project is commonly described as serving around 300 homes. The project presents this as a low-carbon heating pathway. What it really demonstrates is that SGN cannot make the case for hydrogen using existing gas assets and ordinary household economics, so it has had to build a bespoke, heavily supported, highly managed demonstration instead.

The economic case collapses even before the safety case does. SGN’s public-facing number for the project is £32 million. The Ofgem project direction and amended schedule show a lower formal project budget of about £20.93 million in that stage of the process, but the larger number is more useful for understanding the real-world system as built and discussed publicly. A typical home in the project materials is assumed to use 11,500 kWh of gas per year. At the 900-home design point, that means 10.35 GWh of delivered household energy demand each year. Converting that into hydrogen demand at 33.33 kWh/kg gives about 310,531 kg of hydrogen per year. Spread £32 million over 20 years and the capital burden alone is about £1.6 million per year, or about £5.15/kg before paying for electricity, maintenance, stack replacement, compressors, storage, metering, operations, insurance, and all the other balance-of-plant realities of a working hydrogen system.

The electricity math is the part SGN and its allies never want to sit with for long. Nel’s stated performance for the H100 Fife electrolyser implies about 57.3 kWh of electricity per kg of hydrogen, so at 10 p/kWh the electricity alone costs about £5.73/kg. That brings the combined cost to about £10.88/kg. The scaled 900-home case points to hydrogen costing about 32.7 p/kWh before VAT, profit, and retail overhead. Add the same 5% VAT already embedded in Ofgem’s gas price cap and the figure rises to about 34.3 p/kWh. Add supplier overhead, financing risk, and margin, and a plausible customer tariff would likely land somewhere in the mid-30s to around 40 p/kWh, before any standing charge.

Compare that with Ofgem’s April to June 2026 natural gas unit rate of 5.74 p/kWh including VAT. On a real retail basis, hydrogen would likely be around 6x to 7x as expensive as natural gas for the same delivered energy. And that is the strong version of the hydrogen case, the version where the project is stretched to its 900-home design point and the capital is amortized over 20 years. The actual 300-home trial economics are much worse. There is no realistic political pathway in which ratepayers would knowingly accept heating fuel that costs seven times as much as gas, especially when the alternative is not merely natural gas, but electrification, which removes combustion from homes and makes much better use of electricity.

The inefficiency becomes more obvious when compared to electrification. Hydrogen takes electricity, runs it through an electrolyser, compresses and stores the gas, moves it through a specialized network, and then burns it in a boiler at home. A heat pump simply moves heat into the house. Even before cost, one pathway is a chain of losses and the other is a chain of gains. SGN knows this. Any gas utility with an engineering staff, a regulatory team, and access to market analysis knows this. There is no credible claim of ignorance available in 2026. The company can know less than it claims about politics, but not less than it knows about arithmetic. A fuel that requires about 57.3 kWh of electricity per kg to produce and then delivers only 33.33 kWh of fuel energy into a home is already at a disadvantage before combustion losses at the appliance are counted. A heat pump using that same electricity does not need to fight physics at every step.

Even at Fife’s high domestic electricity prices, a modern, well installed heat pump is already cost competitive with natural gas. While UK doesn’t currently carbon price residential gas, the EU ETS coming into effect does, and the UK has a policy of convergence with EU carbon pricing policies. Further, it’s committed to closing the spark gap, lowering electricity prices. Hydrogen will end up 8-10 times more expensive than the same heating delivered by heat pump.

The trial economics are even harder to defend. If 300 homes use 11,500 kWh each, that is 3.45 GWh per year or 6.9 GWh over a two-year trial. That works out to roughly 207,000 kg of hydrogen over the entire demonstration period. If the £32 million lifetime project figure is charged against that demonstration output, the implied cost is on the order of £155/kg, or about 100 times the cost of heating with natural gas if those prices were passed to consumers. If the lower Ofgem-stage budget is used, the implied number falls, but only to a level that still has no relationship to a sane consumer proposition. The point is not whether the trial cost is £80, £100, or £150/kg. The point is that it is so far from plausible consumer economics that the only way to present hydrogen as affordable is to shield participants from the actual cost and spread it through public and regulated funding. That is not a market test. It is an accounting disguise.

After two years, the H100 Fife trial will end because no plausible retail customer base would ever accept the real cost of the hydrogen once the subsidy shield is removed. Even on the generous 900-home scaled case, the hydrogen points to about 32.7 p/kWh of fuel energy delivered, versus 5.74 p/kWh for natural gas, so there is no serious prospect of households voluntarily staying on full-price hydrogen, especially in a community with elevated fuel-poverty risk. The residents will almost certainly take the free return to new natural gas appliances when offered, because that is the cheaper and administratively easier option built into the project design.

At the same time, there is no obvious nearby industrial or chemical offtaker around Methil able to absorb roughly 2 tons a day of hydrogen from the electrolyser as a standing commercial demand, which means the production asset is going to end up stranded: the wrong size for local demonstration uses, the wrong location for industrial feedstock demand, and detached from any durable heating market. SGN will not meaningfully be on the hook in the way an ordinary commercial investor would be, because the project sits inside an approved regulatory funding framework, allowing the costs to be socialized across the wider customer base rather than borne as a failed private bet by the utility itself.

And then there’s the safety aspect. The cleanest continuum is from fully electrified home to natural gas home to hydrogen home. A fully electrified home with heat pump space heating, electric or heat pump hot water, and electric cooking still has electrical fault risk, as all homes do, but it removes in-home combustion, removes fixed-appliance carbon monoxide risk, and removes household fuel-gas leak and explosion pathways. Natural gas adds those hazards back. Hydrogen then increases the risk picture further. The Health and Safety Executive’s March 2026 assessment says the Potential Loss of Life for houses in the hydrogen base case is about 2.6 times higher than the natural gas base case. Even after planned mitigations, HSE says the evidence still points to the potential for greater damage from 100% hydrogen incidents. This is not a small technical wrinkle buried in an appendix. It is the official safety authority saying that hydrogen heating is, on current evidence, less safe than the fuel it is supposed to replace.

Potential Loss of Life is a quantitative risk assessment metric used for systems where leaks, ignition, fire, explosion, and mass-casualty consequences have to be modeled across networks, buildings, and end-use equipment. It is the language of hazardous fuel systems, which is why HSE uses it for comparing natural gas and 100% hydrogen heating. The revealing point is that there is no equivalent public debate about the Potential Loss of Life of heat pumps, electric hot water, and induction stoves, not because electrified homes are risk free, but because they do not create the same distributed combustion, gas leak, and deflagration pathways that require this kind of major-hazard analysis in the first place. Electrification still carries ordinary electrical fire risk, as any modern building does, but it removes the entire class of household fuel-gas scenarios for which PLL becomes a central policy and regulatory metric. In that sense, the absence of a comparable electrification PLL debate is not a gap in analysis. It is evidence that moving heat out of combustible gases and into wires removes a category of danger rather than merely managing it.

The physical reasons for hydrogen’s higher risk rating are familiar. Hydrogen is not just somewhat different from natural gas. Its flammability range in air is about 4% to 75%, compared with methane’s 5% to 15%. Its minimum ignition energy is about 0.02 mJ, compared with roughly 0.3 mJ for methane, so it can ignite with around one fifteenth of the spark energy. And it diffuses through air at roughly three times methane’s rate, while being about eight times less dense, making it harder to contain and more prone to collecting at high points in enclosed spaces.

That would matter in any setting, but it matters much more in housing because homes are full of imperfectly ventilated, partly enclosed, and irregular spaces. Ceiling voids, service cupboards, lofts, boxed-in risers, and meter enclosures are not idealized laboratory geometries. The recent deflagration literature is not reassuring on this point. It keeps showing that geometry, confinement, venting, and structural response dominate outcomes. That is why HSE’s current stance is so much more qualified than the old hydrogen-heating talking points. The system is not merely asking whether a boiler can burn hydrogen. It is asking whether an existing housing stock with all of its ordinary leaks, awkward spaces, and mixed construction can safely absorb a fuel that is less forgiving. The answer, from the latest official review, is not yes. It is not yet proven, and the base case is worse than gas.

The natural gas leak burden makes this even more troubling. A 2024 Energy Policy study found 2.4 million gas leak-coded fire department responses in the United States from 2003 to 2018 and concluded that uncombusted leak responses were about 600 times more frequent than gas-leak-caused fires. The catastrophic events are the visible tip of a much larger background of ordinary failures, small leaks, emergency callouts, and near-misses. The right lesson from that is not that gas leaks are manageable and routine. The right lesson is that fuel-gas systems impose a steady burden of failure and response even when the most dramatic incidents are rare. Replacing methane with hydrogen in that environment does not make the system more robust. It makes the ordinary leak burden more dangerous. The risk is not only in the rare explosion. It is in the much larger field of events that do not quite become one.

To spell this out, the 2.4 million gas leak-coded fire department responses would be higher for hydrogen in any mature system because it’s a lot harder to contain, and a lot more of them than 1/600th would have seen ignition because the energy required to ignite hydrogen is so much lower and because it will ignite across a much broader range of air-hydrogen ratios. When ignition occurred, the result would be much more explosive than for natural gas, causing a lot more damage. Hydrogen is a risk multiplier along several dimensions. We use it in industrial settings when it is required, and industrial engineers treat it very carefully as befits its hazardous nature.

SGN cannot plausibly say it is learning these things in real time. A company of this scale, with regulated obligations, engineers, commercial analysts, and direct access to project economics and government safety processes, knows the difference between a promising pathway and a delaying tactic. It knows hydrogen is materially more expensive than gas on an energy basis and that it will remain more expensive. It knows the latest safety assessment is worse than gas, not better. It knows how how hard it is to keep even methane inside a retail distribution pipeline system and home appliances, and that hydrogen will be harder to contain as the system ages. It knows electrification is the likely long-term destination for low-temperature heat. It knows that sending households back to natural gas at the end of the trial is not a transition strategy, but a stall. This is the heart of the moral problem. The defense of ignorance is not available. What remains is the defense of self-interest.

That end-of-trial design is especially revealing. The H100 Fife documentation and related project materials make clear that participants are intended to return to natural gas at the end of the trial at no cost to them. That means the project does not leave households electrified, safer, or structurally better prepared for the future. It takes a community off of natural gas temporarily for a hydrogen demonstration and then locks them back into gas appliances and gas supply. In a rational heat transition, the off-ramp from natural gas is electrification. In H100 Fife, the off-ramp is a detour that returns to the motorway it was supposed to leave. That is not a bridge to the future. It is a loop designed to consume time.

The socioeconomic context removes any remaining moral comfort. Area-level data for Levenmouth show that 54% of data zones are in the 20% most deprived nationally, compared with 20% across Fife. Fife’s own local profile highlights Buckhaven Central and Denbeath South among the areas of highest fuel poverty risk. That does not mean every participant is poor, and it should not be phrased that way. It does mean the trial is being conducted in communities with elevated deprivation and fuel vulnerability. It is being conducted in an area with many households who will find the £1,000 signing bonus for the trial and free natural gas appliances when it’s over very fiscally attractive, blinding them to the risks. The households with the least margin for error are being used as the proving ground for a fuel pathway that would be rejected immediately if it were presented honestly to consumers on unsubsidized terms. That is not a side issue. It is part of the indictment.

This is where Jonathan Haidt’s moral foundations become useful. His theory argues that human moral judgment is built on a small set of intuitive foundations, commonly framed as care, fairness, loyalty, authority, sanctity, and later liberty. He arrived at them through years of cross-cultural psychology research, drawing on survey data, anthropological evidence, evolutionary reasoning, and experiments suggesting that people usually feel moral intuitions first and then construct rational explanations afterward. They provide a structured way to test whether SGN’s conduct is merely controversial or whether it is wrong from every major moral angle at once.

Start with care. A company committed to care would steer households toward the safer long-term pathway. Electrification removes in-home combustion and the leak-and-deflagration pathway of fuel gases. HSE says hydrogen is worse than natural gas in the base case. Running a trial anyway in a vulnerable community is hard to square with a duty to minimize avoidable harm. On the care foundation, SGN fails because it is asking households to bear risk that better alternatives avoid.

Fairness is just as damaging. Participants are told they will pay the same as they would for natural gas. That sounds fair until the underlying economics are exposed. The real cost of the hydrogen system is multiple times the cost of natural gas, even in the best scaled case, and the residents are shielded from that only because the wider public and regulated customer base pick up the bill. The arrangement does not reveal hydrogen as affordable. It conceals hydrogen as expensive. It socializes the losses and privatizes the appearance of viability. On the fairness foundation, SGN fails because it is presenting an artificial bargain while concealing the real economics of the choice.

Loyalty is often framed as loyalty to one’s group or community. SGN can claim loyalty to its staff, its legacy assets, and the continuity of the gas system. But loyalty to a business model is not the same as loyalty to the public. In practice, H100 Fife looks like loyalty to gas distribution infrastructure at the expense of household interests and system decarbonization. The company is preserving the idea that gas pipes should remain central to domestic heat long after the logic for that position has weakened. On the loyalty foundation, SGN fails because it is more loyal to the survival of the network than to the people paying for it and living with its consequences.

Authority is no refuge either. SGN can say Ofgem approved the project and that it operates within a formal innovation framework. That is true. It is also insufficient. Authority does not absolve judgment. The UK government’s own hydrogen heating overview now says the country won’t be using hydrogen for home heating. HSE’s updated assessment makes the safety picture worse, not better. The regulator approved a demonstration and then failed to stop a project whose economics, safety profile, and strategic direction had already become poor fits with the public interest. Ofgem’s innovation framework and the Scottish and UK policy machinery should have recognized the point at which experimentation stopped being useful and became a way to prolong sunk-cost politics. The failure is not only SGN’s. It is also the regulator’s failure to say enough, stop, and redirect resources toward electrification.

Sanctity can sound out of place in an energy article until it is translated into ordinary terms. The home is supposed to become simpler, safer, cleaner, and easier to live in. That is what sanctity looks like in domestic infrastructure. A good heat transition removes hidden hazards, reduces combustion, and leaves households less exposed to the failures of fuels and flames. SGN’s project goes the other way. It keeps combustion in the home, swaps in a fuel with a tougher safety profile, and turns ordinary houses into the proving ground for a higher-consequence gas. On the sanctity foundation, SGN fails because it is polluting the idea of home improvement with a technology that increases risk rather than removes it.

Liberty is the last foundation and one of the most revealing. Participation in H100 Fife is voluntary. That matters, but liberty is not only about whether a household signs a form. It is about whether people are being led toward more durable freedom or back into dependency. A household electrified with a heat pump and electric hot water is being moved toward the future. A household moved from gas to hydrogen and then back to gas is being kept inside the old system. The set of choices offered is narrow and managed. The people most likely to join the trial are the ones least able to say no to £1,000. On the liberty foundation, SGN fails because it preserves dependency on gas infrastructure while presenting that dependency as optional experimentation.

The regulator’s failure deserves its own plain statement. Ofgem and the broader policy apparatus were not required to accept SGN’s framing forever. Innovation funding is supposed to reduce uncertainty in service of the public interest, not create new sunk costs around pathways whose core weaknesses are already visible. By 2025 the hydrogen-heating proposition had already been badly eroded by cost logic, by the government’s cooling policy stance, and by HSE’s own safety conclusions. Continuing to support H100 Fife at that point was not neutral oversight. It was a choice to permit a legacy utility to keep spending public and ratepayer-backed money on a detour. Regulators exist to prevent exactly this kind of lock-in, especially when the households involved are in areas with high deprivation and fuel poverty risk. They did not do their job.

That is why the moral judgment on SGN is so harsh in 2026. The company is not exploring the unknown. It is defending the known interests of a declining business model. The cost comparison does not work. The safety comparison does not work. The end-state logic does not work. The social justice case does not work. The moral case fails on care, fairness, loyalty, authority, sanctity, and liberty. No one looking at the full record should be in favor of this. The right future for domestic heat is fewer fuels, less combustion, lower long-run cost, and safer homes. SGN is asking the public to fund the resurrection of the wrong system at Easter, and in that there is a bitter kind of symbolism. Some things should be renewed. Household gas distribution is not one of them.