Steel For Fuel N° 010 of 56 1 Oct 2023

What will we do with all the clean hydrogen?

by Andy Lubershane, Partner and Head of Research, Energy Impact Partners

In this note

01The question
02The answer
03The argument
04What you need to know first
05Details worth keeping
06Claims worth citing
07Where it’s contested

The question

Governments have committed enormous sums to producing clean hydrogen. Who is going to consume it, and for what?

The answer

Do the things that only hydrogen can do, which is a shorter list than the subsidies imply. His bet is that most clean hydrogen ends up replacing the hydrogen already used to make ammonia, methanol and petrochemicals, with secondary roles moving and storing renewable energy the grid cannot reach, and synthetic fuels as a wildcard that could dwarf the rest. He offers it as a current hypothesis rather than a forecast.

03The argument

The post starts from an asymmetry rather than from chemistry. Governments almost everywhere have committed to clean hydrogen production, and the United States answered Europe with a subsidy that pays several times more for a joule of zero-carbon hydrogen than for a joule of zero-carbon renewable electricity, and lets an electrolyzer running on wind power claim both. Lubershane expects the production credit to exceed the entire cost of making hydrogen before the decade is out, which would mean a producer could vent the gas into the air and still turn a profit. Supply is being taken care of. Nobody has settled who buys it.

His warning about the demand side runs through the metaphor hydrogen’s advocates like best. A swiss army knife is versatile, which is exactly why nobody pulls one out on a job site. Hydrogen holds very little energy per litre, embrittles the steel that tanks and pipes are made from, and escapes through gaps that would hold any other gas. Making it from electricity compounds the problem: electrolysis throws away nearly a third of the energy going in, and the equipment costs roughly what the wind or solar plant feeding it costs. He is confident the equipment gets much cheaper, partly on the strength of a company his firm has backed, but the direction of the inequality never changes. A joule of clean hydrogen will always cost more than a joule of clean electricity, electricity can usually be put to work more efficiently at the far end, and the grid that delivers it already exists. The burden of proof sits on hydrogen in every application.

Three uses clear that bar, and he numbers them. First, moving energy: gas pipelines are far cheaper per unit of energy than transmission lines, so hydrogen may be the only practical way to tap renewable resources the grid cannot reach, which matters because interconnection queues now hold more capacity than the entire installed generating fleet. Second, storing it: salt caverns and depleted gas fields hold hydrogen for a tiny fraction of what batteries cost, and he argues nothing else can balance supply between one year and the next, or insure against shocks the way a month of fuel underground does. Both come heavily qualified. The pipeline case rests on a reasonable assumption rather than built examples, and the storage case works only at enormous scale and only where the geology cooperates. The third he treats as settled, and it is chemistry rather than energy: ammonia and methanol need hydrogen atoms and nothing else will substitute. The world currently gets those atoms by stripping them out of hydrocarbons and venting the carbon, and he calls displacing that “grey” hydrogen the only truly mission-critical role for the clean kind, and the one most likely to take off first. Synthetic hydrocarbons made from clean hydrogen and captured carbon he rates no more than a maybe, while noting they could end up the largest source of demand of all.

What the post rules out is nearly everything at retail scale. Blending hydrogen into existing gas networks stalls at a small share by energy, and going higher would mean guaranteeing that every appliance on a shared network can handle it, which he judges politically unrealistic outside a handful of jurisdictions. Anything at truck-stop scale fails on delivery rather than production: under assumptions he calls charitable, getting the hydrogen to the pump costs more than making it, and the efficiency gap between fuel cells and batteries widens the loss again at the wheels. Intercontinental shipping of liquefied hydrogen he does not argue against in his own voice; he quotes Michael Liebreich’s verdict that the underlying physics forbids it. Ammonia is the carrier that might work, because the world already ships and stores it in bulk. And the structure carries a stated expiry date: naturally occurring underground hydrogen is a resource nobody has sized, and he says plainly that if it turns out to be large he will scrap the theses in this post.

04What you need to know first

Electrolysis, and “green” hydrogen: Splitting water into hydrogen and oxygen using electricity. Green means the electricity was zero-carbon; he notes the word is sometimes used to mean renewable specifically.
“Grey” hydrogen: The hydrogen industry uses today, stripped out of hydrocarbons with the leftover carbon released to the atmosphere. Most of it goes into fertilizer and chemicals rather than into anything that looks like fuel.
Volumetric versus gravimetric energy density: Energy per litre versus energy per kilogram. Hydrogen is excellent on the second and poor on the first; in many applications, he says, volume constrains as much as weight or more.
Hydrogen embrittlement: Contact with hydrogen weakens common metals, steel especially. This is why existing pipelines, tanks and valves cannot simply be handed a new molecule.

05Details worth keeping

He opens by crediting Michael Liebreich’s public writing on hydrogen and says he will count the post a success if it builds on it rather than restating it.
Two long passages quote his own earlier posts: one on gas turbines being built retrofit-ready for hydrogen, one on the safety limit for blending into gas networks. He carries both forward rather than revising them.
The blending objection is physical: hydrogen flows wherever the gas flows, so raising the blend means knocking on every door, and missing one house puts someone at risk.
The truck stop thought experiment prices three ways to get 30 tons a day to a station: compressed in trucks, liquefied in trucks, or a dedicated pipeline.
Ammonia’s attraction is that it is already traded worldwide and far easier to liquefy than hydrogen; its problem is that it is both highly toxic and explosive. Shipping moves first, with Maersk ordering methanol-capable vessels and MAN aiming at a first ammonia-fired marine engine in 2024.
Naturally occurring hydrogen seeps to the surface in circular patterns behind “fairy ring” folklore. Michael Webber, his own firm’s chief technology officer, was among the first people to take the resource seriously, and the firm invested early in Koloma on the strength of that. Webber’s summary, quoted in the post: “It’s big enough to be like, ‘Okay, this is exciting.’”
Three subjects are deferred to future posts: hydrogen made from natural gas, hydrogen from biomass, and heavy transport.
Several figures carry evidence the prose does not restate, notably the comparison of delivered hydrogen against grid electricity at a truck stop and the share of local gas distribution pipe that is plastic rather than steel.

06Claims worth citing

All figures as stated on 2023-10-01. Subsidy design, electrolyzer costs and hydrogen project economics were all moving quickly when this was written, and several figures are his firm’s own analysis.

42 countries had promulgated national hydrogen strategies as of July 2023. Bloomberg New Energy Finance, cited by Lubershane
The European Commission has estimated the cost of meeting its 2030 clean hydrogen production goals at on the order of €100 billion; a footnote splits this into €24-42 billion for electrolyzers and €65 billion for transport, storage and delivery. European Hydrogen Strategy investment agenda, cited by Lubershane
The US will pay more than three times as much to produce a joule of zero-carbon hydrogen as it pays for a joule of zero-carbon electricity from renewables. Lubershane
The clean hydrogen production credit, 45V, is $3 per kilogram, and he is confident it will exceed the total cost of production by the end of the decade. Lubershane, with Electric Hydrogen’s cost trajectory as the basis
Pressurized to 700 bar, hydrogen holds under 15% of the energy per litre of diesel, and a little over twice that of a lithium-ion battery. Lubershane
Electrolysis consumes nearly a third of the energy input, leaving around 70% embodied in the hydrogen; about 5% of electricity is lost in the electrical grid. Lubershane; the grid figure a US Energy Information Administration average for 2017-2021
Fully installed electrolysis systems cost roughly $1,000-1,500 per kilowatt, the same ballpark as large-scale wind or solar generation. Lubershane
Electric heat pumps are over three times more efficient than burning hydrogen for heat, in applications below about 400 degrees Fahrenheit; a battery and motor is about twice as efficient as a fuel cell and motor. Lubershane
US interconnection queues hold about twice as much wind, solar and battery capacity as the country’s total installed generating capacity of all kinds. Lubershane
A gas pipeline capable of carrying the same energy over the same distance costs about six times less to build than an electric transmission line, on average in the US. Lubershane
Large geological hydrogen storage can cost as little as 1% of the capital cost of lithium-ion battery storage, a few dollars per kilowatt-hour, but only at very large scale, because a site costs about the same whether it holds thousands or millions of tons. Lubershane
The US keeps about 47 days of primary energy supply on hand as coal piles, oil tanks and gas reservoirs, more than all the renewable power it generates in a year. Lubershane, quoting his earlier post “Four Ways to Store Sunlight”
Gas turbines installed in the next five years will nearly all accept 10-18% hydrogen by energy content blended into the fuel, and the three big turbine manufacturers are promoting designs retrofit-ready for up to 100% by 2030. Lubershane, quoting the same earlier post
The gas industry’s tentative consensus limit for blending into existing pipelines is about 20% hydrogen by volume, which is 6-7% by energy content. Lubershane
US interstate gas pipelines can carry about 1.3 terawatts, nearly twice the country’s peak electricity demand; a footnote gives that peak as 720 GW in August 2021. Lubershane
Delivering hydrogen to a truck stop, under charitable assumptions and with production at $2 per kilogram, costs at least 50% more than producing it in the first place. Lubershane
Shipping liquid hydrogen will cost four to six times more than liquefied natural gas, and the obstacle is physics rather than scale, technology or cost of capital. Liebreich, “The Unbearable Lightness of Hydrogen”, quoted by Lubershane
Grey hydrogen production is responsible for about 2% of global greenhouse gas emissions. Lubershane
Fully decarbonizing ammonia and methanol production would take about two thirds of all the wind and solar power deployed globally to date; about 10% of ammonia production already crosses borders. Lubershane
Producing enough hydrogen for synthetic jet fuel to substitute for all the jet fuel the US consumes annually would require about three times as much wind power as the country produced last year. Lubershane

07Where it’s contested

There is no second voice here, and the post argues against hydrogen maximalism rather than against a named opponent. What it carries instead is a lot of marked confidence and several stated unknowns.

The subsidy claim rests on an interested case. His confidence that the credit will exceed production cost is grounded in Electric Hydrogen’s trajectory, and he states that his firm is an investor. It is conditional in his own words: it holds if the dynamic holds up.
The pipeline case is assumption, not evidence. He lists the open questions and answers none: whether hydrogen pipelines will be easier to permit because activists do not oppose them or harder because the public thinks of the Hindenburg, and whether hydrogen’s low viscosity makes them cheaper to build or its handling makes them dearer.
He grades his own conclusions. Ammonia gets “chalk me up as ammonia-curious”; synthetic fuels get “maybe a good reason”; small-scale distributed hydrogen gets an explicit general pessimism, with the truck stop numbers described as charitable to hydrogen.
The case against blending is a political judgment carried over from earlier writing. He is, in his own words, extremely skeptical that the public would tolerate the intervention required, and that skepticism rather than a technical finding is what makes blending niche in his telling.
He says the post may not survive. The size of the naturally occurring hydrogen resource is unknown, with error bars running from a modest complement to green hydrogen up to the world’s next great primary energy resource, and he is mentally preparing to scrap his theses if it is the latter.
What goes unexamined. Whether the subsidies persist long enough to matter; whether the clean power his preferred end state implies can actually be built; and the competing routes to hydrogen from natural gas and from biomass, named and then deferred.