Hydrogen is an energy carrier that must be manufactured using other energy sources, then compressed, liquefied, or chemically converted for transport — with significant energy losses at every conversion step — through infrastructure that largely does not yet exist at scale.
How the energy-carrier paradox, storage physics, and infrastructure absence interact to produce a supply chain where the molecule with the highest energy density by mass becomes one of the most difficult to move, where production method determines environmental meaning, and where the gap between industrial reality and energy-transition ambition is measured in missing infrastructure.
Introduction
The hydrogen supply chain produces, moves, and delivers the simplest and lightest element in the universe for use as a chemical feedstock, a reducing agent in metallurgy, and — increasingly in policy ambition if not yet in physical reality — a fuel for transport, power generation, and heating. Hydrogen is already an industrial commodity at enormous scale: global production exceeds seventy million tonnes per year, consumed almost entirely in oil refining and ammonia manufacturing. This is not a speculative supply chain. It is an established industrial system with well-understood production methods, existing customers, and decades of operational history.
What makes the hydrogen supply chain structurally distinctive is the gap between what it currently does and what energy transition plans require it to do. The existing system produces hydrogen from natural gas at the point of consumption, avoiding the transport problem entirely. The envisioned system would produce hydrogen from renewable electricity at locations with abundant wind or solar resources, then transport it hundreds or thousands of miles to industrial consumers, fuel cell vehicles, power plants, and potentially buildings. These are fundamentally different supply chains that share only a molecule.
The distinction matters because nearly every public discussion of hydrogen conflates the existing industrial system with the aspirational energy system. The existing system works because it avoids the hardest problems: hydrogen is produced where it is consumed, from cheap natural gas, through well-understood chemistry. The aspirational system must solve all three root constraints simultaneously — energy-negative production, physically punishing storage and transport, and absent infrastructure — at costs competitive with alternatives that already work. Understanding this supply chain requires separating what hydrogen does today from what it is being asked to do tomorrow, and examining the physical constraints that determine whether the gap between them can be closed.
Root Constraints
Why is hydrogen an energy carrier rather than an energy source?
Hydrogen does not exist in useful concentrations in nature. Unlike coal, oil, or natural gas, which can be extracted and burned, hydrogen must be manufactured by breaking it out of compounds — water or hydrocarbons — using energy from another source. This makes hydrogen an energy carrier, like a battery, rather than an energy source, like a fuel. The distinction is fundamental because it means producing hydrogen always consumes more energy than the hydrogen contains. The laws of thermodynamics guarantee a net energy loss in every production method.
The dominant production method is steam methane reforming (SMR), which reacts natural gas with steam at high temperatures to produce hydrogen and carbon dioxide. SMR accounts for roughly ninety-five percent of global hydrogen production. The process converts about seventy to seventy-five percent of the energy in the input natural gas into energy contained in the output hydrogen. The remaining twenty-five to thirty percent is lost as heat during the reaction. This is a mature, optimized industrial process operating near its thermodynamic limits. It produces cheap hydrogen — roughly one to two dollars per kilogram — but emits nine to twelve tonnes of carbon dioxide per tonne of hydrogen produced.
Electrolysis, the alternative production method, splits water into hydrogen and oxygen using electricity. The theoretical energy requirement is about thirty-nine kilowatt-hours per kilogram of hydrogen. In practice, current electrolyzers consume fifty to fifty-five kilowatt-hours per kilogram, achieving sixty to seventy percent efficiency. If the electricity comes from renewable sources — wind or solar — the hydrogen is classified as green, with no direct carbon emissions. If it comes from the grid, the hydrogen carries whatever carbon intensity the grid carries, which in most countries makes it worse than producing hydrogen from natural gas directly.
The efficiency chain reveals the core challenge. To deliver one unit of useful energy as hydrogen from renewable electricity: the renewable source generates electricity (with its own capacity factor limitations), the electrolyzer converts electricity to hydrogen (sixty to seventy percent efficiency), the hydrogen must be compressed or liquefied for transport (consuming ten to thirty-five percent of the hydrogen’s energy content), transported to the point of use, and then converted back to electricity in a fuel cell (fifty to sixty percent efficiency) or burned in a turbine (thirty to forty percent efficiency). The round-trip efficiency from renewable electricity to useful energy delivered via hydrogen is roughly twenty-five to thirty-five percent. Using that same renewable electricity directly — through the grid, to charge a battery, to drive a motor — delivers seventy to ninety percent of the original energy.
This does not make hydrogen useless. It means hydrogen makes thermodynamic sense only in applications where direct electrification is physically impossible or impractical — not merely inconvenient or currently lacking infrastructure, but genuinely infeasible. Steelmaking requires a reducing agent that strips oxygen from iron ore; hydrogen can do this where electrons alone cannot. Ammonia synthesis requires hydrogen atoms as molecular building blocks. Long-distance shipping and aviation may require energy-dense fuels that batteries cannot practically provide. These are the applications where hydrogen’s role is defined by chemistry, not by choice.
The color spectrum of hydrogen production reflects these energy realities. Grey hydrogen (SMR without carbon capture) is cheap but carbon-intensive. Blue hydrogen (SMR with carbon capture and storage) reduces emissions by sixty to ninety percent depending on capture rates, but requires carbon capture infrastructure, long-term geological storage, and continued natural gas supply. Green hydrogen (electrolysis powered by dedicated renewables) eliminates direct emissions but costs three to six times more than grey hydrogen at current prices and depends on abundant, cheap renewable electricity that many regions cannot yet provide. Each color represents not just a production method but a fundamentally different supply chain with different infrastructure requirements, different costs, and different dependencies.
Why is hydrogen hard to store and transport despite its high energy per mass?
Hydrogen has the highest energy content per unit mass of any fuel: one kilogram of hydrogen contains roughly thirty-three kilowatt-hours of energy, nearly three times the energy per kilogram of natural gas and more than twice that of gasoline. This property makes hydrogen attractive on paper. It is also misleading, because energy systems move volume, not mass, and hydrogen’s volumetric energy density is among the lowest of any fuel.
At atmospheric pressure and ambient temperature, one kilogram of hydrogen occupies roughly eleven cubic meters — a volume the size of a small room. To make hydrogen practically transportable, it must be compressed to seven hundred bar (roughly seven hundred times atmospheric pressure) for vehicle fuel tanks, or liquefied by cooling it to minus 253 degrees Celsius — just twenty degrees above absolute zero — for bulk transport. Each approach imposes substantial energy penalties and engineering challenges.
Compressed hydrogen transport by tube trailer — the current method for most merchant hydrogen delivery — carries roughly three hundred to five hundred kilograms of hydrogen per truck. A gasoline tanker carries the energy equivalent of roughly four thousand kilograms of hydrogen. Moving equivalent energy by hydrogen requires eight to thirteen truck trips for every one gasoline delivery. The economics of trucking compressed gas mean that delivery costs increase rapidly with distance, making centralized hydrogen production with long-distance truck delivery uneconomical beyond roughly two hundred miles.
Liquid hydrogen tankers carry more — roughly four thousand kilograms per truck — but liquefaction’s energy penalty and the continuous boil-off of liquid hydrogen during storage and transport (roughly one to three percent per day lost to evaporation through even the best-insulated vessels) impose their own costs. Liquid hydrogen infrastructure operates in a temperature regime colder than liquid natural gas by nearly one hundred degrees, requiring specialized materials, seals, and handling procedures that add cost at every point.
Pipeline transport, which works efficiently for natural gas, faces specific obstacles with hydrogen. Hydrogen molecules are the smallest in existence, leaking through seals and joints that contain larger molecules. Hydrogen causes embrittlement in many steel alloys used in existing natural gas pipelines, degrading the steel’s structural integrity over time and potentially causing catastrophic failure. Repurposing existing natural gas pipelines for pure hydrogen requires either replacing pipeline sections built with susceptible steel grades or limiting hydrogen concentrations to five to twenty percent blended with natural gas — a compromise that reduces the transport benefit proportionally. Purpose-built hydrogen pipelines exist but are rare: roughly five thousand miles globally, versus millions of miles of natural gas pipeline, almost entirely serving clusters of industrial facilities like refineries and chemical plants where hydrogen moves short distances between adjacent operations.
Chemical carriers offer a third approach: converting hydrogen into ammonia, methanol, or liquid organic hydrogen carriers (LOHCs) for transport, then extracting the hydrogen at the destination. Ammonia — three hydrogen atoms bonded to one nitrogen atom — is already traded globally at scale and can be shipped in conventional chemical tankers. But converting hydrogen to ammonia and back consumes roughly thirty percent of the hydrogen’s energy content, and the reconversion process adds infrastructure and cost at the receiving end. Ammonia itself is toxic and corrosive, requiring handling infrastructure distinct from hydrogen infrastructure. The chemical carrier approach solves the transport problem by introducing a conversion problem, trading one set of losses for another.
Why does hydrogen have no distribution infrastructure at scale?
Unlike electricity, which has grids spanning continents, or natural gas, which has pipeline networks built over a century, hydrogen has no distribution infrastructure at scale. The roughly five thousand miles of dedicated hydrogen pipeline worldwide serve a handful of industrial clusters — the Gulf Coast of the United States, parts of northwest Europe, and scattered chemical complexes elsewhere. These are point-to-point connections between adjacent industrial facilities, not a distribution network.
The absence of infrastructure creates a coordination problem that is structural, not merely financial. Building hydrogen production facilities requires customers who will purchase the output. Building distribution infrastructure requires production volumes to justify the capital investment. Industrial users cannot switch to hydrogen without reliable supply and delivery. Each element of the supply chain requires the other elements to exist first, and none can proceed economically without them.
This chicken-and-egg problem is not unique to hydrogen — it appeared in the early days of electricity grids, natural gas networks, and more recently electric vehicle charging networks. But hydrogen faces a version that is unusually severe because the competing alternative — direct electrification — is already being built. Electricity grids already exist. Battery electric vehicles already have a growing charging network. Heat pumps already connect to the existing electrical infrastructure. Hydrogen must build its distribution system from near-zero while competing with an alternative that plugs into infrastructure already in place.
The geographic mismatch compounds the problem. The best renewable energy resources — strong, consistent wind and intense solar radiation — exist in locations that are often far from major industrial demand centers. Patagonia, the Australian outback, the North Sea, and the Sahara have excellent renewable resources. The steel mills, refineries, and chemical plants that would consume green hydrogen are in the Ruhr Valley, the US Gulf Coast, coastal China, and South Korea. Connecting abundant renewable generation to industrial hydrogen demand requires either building long-distance hydrogen transport infrastructure or building long-distance electrical transmission to co-locate electrolysis with industrial demand. Neither option is free, fast, or straightforward.
Government policy has attempted to resolve the coordination problem through subsidies, mandates, and hydrogen strategies. Over forty countries have published national hydrogen strategies. The US Inflation Reduction Act provides production tax credits of up to three dollars per kilogram for green hydrogen. The European Union targets ten million tonnes of domestic green hydrogen production by 2030. These interventions aim to break the chicken-and-egg cycle by guaranteeing returns to early investors. Whether they succeed depends on whether the physical constraints — energy losses, storage penalties, and infrastructure requirements — can be managed at costs that remain within the subsidy envelope as projects scale from demonstration to industrial volumes.
How Constraints Shape the System
The three root constraints interact to create a supply chain challenge more severe than any single constraint implies.
The energy carrier constraint combined with storage difficulty means that hydrogen suffers compounding losses. The energy lost in production (thirty to forty percent for electrolysis) is followed by energy lost in compression or liquefaction (ten to thirty-five percent), followed by energy lost in any chemical conversion for transport, followed by energy lost in reconversion to useful work at the point of use. Each step takes its fraction from an already-reduced total. The round-trip efficiency problem is not a single loss but a cascade of losses, each applied to what remains after the previous step. Improving any single step helps, but the multiplicative nature of the losses means that even significant improvements at one stage yield modest improvements in total system efficiency.
The energy carrier constraint combined with infrastructure absence creates a competitive disadvantage that grows with distance. Near the point of production, hydrogen competes on the cost of production alone. But as distance increases, the cost of storage, compression, transport, and reconversion accumulates. At short distances — within an industrial complex — hydrogen is cost-competitive for applications that require it. At medium distances requiring truck transport, hydrogen costs escalate rapidly. At long distances requiring pipelines, ships, or chemical carriers, the accumulated penalties make hydrogen more expensive per unit of delivered energy than alternatives for most applications. The infrastructure absence means these costs cannot be reduced through network effects or shared infrastructure as they can for electricity or natural gas.
The storage difficulty combined with infrastructure absence means that the transition from grey to green hydrogen requires building not one supply chain but two simultaneously: the production infrastructure (electrolyzers, renewable generation) and the delivery infrastructure (pipelines, compression stations, storage facilities, fueling stations). Grey hydrogen avoids the delivery problem because it is produced on-site at refineries and ammonia plants using natural gas delivered through existing pipeline networks. Green hydrogen, produced where renewable energy is cheapest rather than where hydrogen is consumed, must solve the delivery problem that grey hydrogen sidesteps. The green transition does not simply swap one production method for another. It introduces an entirely new logistics challenge that the current system does not face.
System Context
The existing hydrogen system operates almost invisibly within the industrial economy. Roughly half of global hydrogen production feeds oil refineries, where it is used to remove sulfur from fuels and to crack heavy hydrocarbons into lighter products. Roughly another forty percent goes to ammonia production — the primary feedstock for nitrogen fertilizers that support global agriculture. The remainder serves niche applications in methanol production, electronics manufacturing, food processing, and metallurgy.
In this existing system, hydrogen is produced captively — within the facility that consumes it, from natural gas piped to the site through conventional gas infrastructure. The hydrogen never enters a distribution network. It is produced and consumed in the same industrial complex, sometimes within the same building. This captive production model works precisely because it avoids the storage and transport constraints. The molecules travel feet or yards, not miles. No compression, liquefaction, or chemical conversion is needed.
The merchant hydrogen market — hydrogen produced at one location and delivered to another — accounts for roughly five to ten percent of total production. This market is served by industrial gas companies operating fleets of tube trailers that deliver compressed hydrogen to customers within a regional radius. The economics of tube trailer delivery limit the practical market to customers within roughly one hundred to two hundred miles of a production facility. The merchant market is small not because demand is limited but because delivery costs escalate so steeply with distance that only customers with no alternative and relatively low volume requirements find it economical.
The proposed transformation asks this system to expand by an order of magnitude or more: from seventy million tonnes per year serving a handful of industrial applications to potentially hundreds of millions of tonnes serving steel mills, shipping fleets, power generation, and potentially buildings and vehicles. This expansion would require not just more hydrogen production but fundamentally different production (electrolysis instead of SMR), fundamentally different logistics (long-distance transport instead of captive on-site production), and infrastructure that does not currently exist at any meaningful scale.
Flows and Visibility
The hydrogen supply chain’s flow characteristics differ radically depending on which system — existing industrial or proposed energy transition — is being described.
In the existing system, flows are short and visible. Natural gas enters a refinery or chemical plant. Steam methane reformers within the facility convert it to hydrogen. The hydrogen moves through internal piping to the process unit that consumes it. The entire flow occurs within a single facility’s boundary fence. Input and output volumes are metered and monitored in real time as part of standard industrial process control. There is no distribution network to track because there is no distribution.
In the proposed green hydrogen system, flows would be long and complex. Renewable electricity from wind or solar installations would power electrolyzers, potentially at the generation site or at a central hub. The hydrogen would be compressed, liquefied, or converted to ammonia. It would move by pipeline, truck, rail, or ship to an import terminal or distribution point. It would be reconverted if necessary, then delivered to the end consumer. Each step introduces conversion losses, requires dedicated infrastructure, and adds cost. Tracking the carbon intensity of the delivered hydrogen — essential for regulatory compliance and market credibility — requires accounting through the entire chain, including the source and carbon intensity of the electricity used for electrolysis, the energy consumed in compression and transport, and any leakage losses along the route.
Hydrogen leakage introduces a specific visibility challenge. Hydrogen is an indirect greenhouse gas — it does not trap heat directly but extends the atmospheric lifetime of methane, a potent greenhouse gas, by competing for the same atmospheric oxidants. Leakage rates along the hydrogen supply chain are not well characterized because the infrastructure to measure them at scale does not exist. Estimates range from one to ten percent depending on assumptions about equipment, handling, and measurement methods. If leakage rates are at the higher end, the climate benefit of switching from grey to green hydrogen is substantially reduced, particularly over shorter time horizons where methane’s warming effect is most pronounced.
What This Reveals About Industrial Structure
- The hydrogen supply chain reveals that production method is not a technical detail but a structural determinant. Grey, blue, and green hydrogen are not interchangeable products with different carbon labels. They are different supply chains with different input dependencies, different infrastructure requirements, different cost structures, and different geographic logics. A company positioned in grey hydrogen occupies a fundamentally different industrial position than one positioned in green hydrogen, even if both produce the same molecule.
- The existing hydrogen system is large, mature, and functional precisely because it avoids the constraints that the proposed system must solve. Captive on-site production eliminates transport costs, SMR eliminates the efficiency penalty of electrolysis, and natural gas pipeline delivery eliminates the infrastructure gap. Every proposed change — from grey to green, from captive to distributed, from short-range to long-distance — reintroduces a constraint that the current system has structured itself to avoid.
- Direct electrification is not merely a competing technology. It is a competing supply chain that uses existing infrastructure, avoids the conversion losses inherent in the hydrogen pathway, and is scaling faster than hydrogen in nearly every application where both are technically feasible. Hydrogen’s viable domain is defined by the boundary where electrification stops being physically possible — steel reduction, certain chemical processes, potentially long-distance shipping and aviation — not by a general-purpose competition with electricity.
- The coordination problem is more severe for hydrogen than for historical infrastructure transitions because the competing alternative already exists. Electricity grids were built when there was no alternative for delivering power over distance. Natural gas pipelines were built when no substitute could deliver comparable energy density to buildings and power plants. Hydrogen infrastructure must be built while electricity grids and natural gas networks already serve most of the applications hydrogen targets, and while battery technology continues to extend electrification into applications previously thought to require chemical fuels.
- Government subsidies and mandates can break the coordination deadlock but cannot repeal the physics. Production tax credits reduce the cost of green hydrogen production but do not reduce the energy consumed by compression, the volume penalty of the molecule, or the distance between renewable resources and industrial demand. Policy can make hydrogen projects financially viable at current costs, but the physical constraints determine whether the system can scale to the levels that energy transition scenarios require without the cost of subsidy becoming unsustainable.
- Hydrogen embrittlement and leakage are not engineering problems to be solved in isolation. They are system-level constraints that determine which materials can be used in pipelines, storage vessels, and end-use equipment, which in turn determines whether existing natural gas infrastructure can be repurposed and at what cost. The answer shapes the entire economics of hydrogen distribution.
Inside CompanyGraph
View this perspective inside CompanyGraph.
CompanyGraph