CompanyGraph
Follow hydrogen from water or hydrocarbon feedstock through production, purification, compression, transport, and use in ammonia, refining, steel, or a fuel cell. The molecule may be the same at the destination, but its energy source, pressure, purity, and route determine what the chain has actually delivered.
A hydrogen user does not need hydrogen in the abstract. An ammonia plant needs hydrogen atoms to combine with nitrogen. A refinery uses hydrogen in reactions that remove sulfur and alter hydrocarbon molecules. A steelmaker can use it to remove oxygen from iron ore. A fuel cell needs a clean, controlled stream that can react electrochemically with oxygen. These are different physical services, even when each is described as hydrogen demand.
Almost all hydrogen supplied today is manufactured from fossil fuels or water. Naturally occurring geologic hydrogen is being explored, but it does not yet provide a material share of commercial supply. For manufactured hydrogen, the molecule is an energy carrier rather than a primary energy source: it is separated from a compound using heat or electricity before a user can consume it. The IEA reports that global hydrogen demand surpassed 100 million tonnes in 2025, while low-emissions production grew by 20% to almost 1 million tonnes. Most current demand remains in established refining and industrial uses, not in the future applications often placed beside them in policy plans.
The molecule must first be separated
Hydrogen is usually bound to carbon in methane or to oxygen in water. A production plant supplies heat, electricity, catalysts, membranes, vessels, controls, and feedstock to separate it and remove unwanted gases or water.
In steam-methane reforming, high-temperature steam reacts with methane to produce a hydrogen-rich gas. Shift reactors convert carbon monoxide and steam into additional hydrogen and carbon dioxide, and purification removes the gases that the customer does not want. The process is mature and uses the existing natural-gas system, but its carbon dioxide stream is part of the product route rather than an external detail. The U.S. Department of Energy describes natural-gas reforming as a high-temperature process typically operating with steam at 700—“1,000—”°C.
Hydrogen is also produced by gasifying coal. The resulting synthesis gas is shifted and purified, while the route produces a substantial carbon-dioxide stream unless capture and storage are added.
Adding carbon capture changes the plant but does not make the carbon dioxide disappear. The gas must be separated, compressed, transported, injected into a permitted storage site, and monitored. Methane entering the plant also has a production and pipeline history. A capture rate measured at the reformer therefore cannot by itself describe the climate result of the whole gas route.
Electrolysis sends electricity through water in an electrolyzer, producing hydrogen at one electrode and oxygen at the other. Alkaline, proton-exchange-membrane, and solid-oxide machines use different materials, temperatures, and operating controls. DOE describes electrolysis as splitting water into hydrogen and oxygen, with the electricity source determining the emissions associated with the route. The electricity connection, water treatment, stack lifetime, and the timing of renewable generation are therefore part of the hydrogen plant.
Purity and pressure are part of the product
Fresh hydrogen is not automatically ready for every user. Water vapour, oxygen, nitrogen, carbon monoxide, sulfur compounds, and particles can matter differently to an ammonia catalyst, a refinery unit, a steel reactor, or a fuel-cell membrane. Drying, purification, filtration, compression, and sometimes cooling create the condition that a contract specifies.
A delivery can therefore be on time and still unusable. A fuel-cell vehicle needs a pressure and contaminant profile compatible with its tanks and stack. An industrial process may accept a different pressure or impurity range because it has its own purification step. A mass meter, a purity certificate, and a pressure reading observe different parts of the delivery; none alone establishes that the receiving equipment produced its intended service.
On-site production avoids long-distance transport
Most established hydrogen is made close to the unit that consumes it. A refinery or ammonia complex can bring natural gas and water to a reformer, move hydrogen through short internal pipes, and use it immediately. The molecule does not need a merchant pipeline, a tube trailer, a cryogenic tank, or a port. This is why the existing chain can be large without resembling a public distribution network.
Merchant hydrogen does travel between sites, usually in industrial clusters or to customers without their own plant. DOE reports that about 1,600 miles of hydrogen pipeline operate in the United States, concentrated near refineries and chemical plants. Those lines demonstrate a functioning industrial route, but they do not amount to a general-purpose network connecting every potential user.
Moving production away from the user can make renewable electricity or another feedstock available, but it replaces a short internal pipe with compression, storage, transport, delivery equipment, and new interfaces. The physical distance is not a neutral change: each additional transfer creates another place where pressure, purity, leakage, custody, and payment must be managed.
Making hydrogen movable changes its condition
Compressed gas is the simplest transport form, but hydrogen is light, so a tube trailer carries much less mass than a tanker carrying a comparable amount of liquid fuel. Compression also needs equipment and electricity, and the receiving site must safely reduce or raise pressure as the use requires.
For larger quantities without a pipeline, hydrogen can be liquefied below —253—°C and carried in insulated tanks. DOE notes that present liquefaction consumes more than 30% of hydrogen’s energy content and that stored liquid can be lost through boil-off. The route adds cryogenic materials, refrigeration, loading, unloading, and a way to use or manage evaporated gas.
For storage over days or seasons, surface tanks are not the only option. Salt caverns and other underground formations can hold bulk hydrogen, but their usefulness depends on suitable geology, cavern integrity, compression, injection and withdrawal rates, contamination, leakage, and proximity to the pipeline or user. DOE describes geologic storage as a possible way to handle seasonal demand and reports four existing salt caverns used for hydrogen storage. Underground storage is therefore a site-specific facility, not simply a larger version of a surface tank.
Hydrogen can also be converted into ammonia or another carrier. Ammonia is already shipped in chemical tankers and can be the final product for fertilizer, but it is not hydrogen. If the destination needs hydrogen, a cracking plant must split ammonia again, with its own heat, catalyst, separation, emissions, and residual-ammonia controls. A carrier can solve a transport-volume problem while creating a second conversion plant at the receiving end.
Use changes the molecule again
In ammonia synthesis, hydrogen combines with nitrogen to form NH—. The ammonia may become fertilizer or a chemical feedstock; in that case hydrogen has supplied atoms to a new product rather than remained an energy carrier.
In refining, hydrogen reacts with sulfur-containing and unsaturated hydrocarbons. Sulfur leaves the process in streams that require further treatment, and the hydrogen is no longer available as a separate gas. The useful result is a changed petroleum feedstock, not a quantity of hydrogen delivered to a tank.
In direct-reduced iron, hydrogen removes oxygen from iron oxide and produces water vapour while leaving reduced iron for later melting and shaping. The reduction reactor, hydrogen quality, heat, iron-ore form, and electric arc furnace all matter. Hydrogen availability alone does not establish that a steel plant can make the required grade.
A fuel cell combines hydrogen with oxygen to produce electricity, water, and heat. The useful service depends on stack condition, cooling, power electronics, controls, and the load being served. A kilogram of hydrogen delivered to a fuel-cell system is not the same observation as kilowatt-hours available at the vehicle wheels or to a building.
New demand is a different chain
Replacing fossil-based hydrogen in an existing refinery or ammonia plant can preserve much of the downstream equipment while changing the production and delivery route. Supplying a new steel plant, shipping fuel system, seasonal store, or power station can require different pressure, purity, storage duration, safety distances, and conversion equipment.
The proposed locations also differ. Electrolyzers may be placed beside renewable generation, while ammonia, steel, and refining demand may be elsewhere. A project must then connect electricity, water, hydrogen equipment, storage, transport, and an offtaker. The IEA reports that global hydrogen demand surpassed 100 million tonnes in 2025, while low-emissions production grew by 20% to almost 1 million tonnes. Its announced low-emissions project pipeline for 2030 has contracted to 27 million tonnes, and firm offtake remains insufficient to unlock large-scale investment. These figures describe current project development, not proof that any one use will never work.
Current demand also should not be treated as a fixed measure of biological or technical necessity. Ammonia needs hydrogen chemistry; a refinery may need hydrogen under its present process and fuel standards; a vehicle or power plant may be designed around hydrogen because of a chosen system architecture. Whether additional hydrogen is required depends on the service, available alternatives, infrastructure, and the conditions created by the present organization.
Money has to arrive before the plant
A project cannot begin reliable delivery until production equipment, electricity or gas supply, water treatment, compression, storage, transport, and an offtake contract are ready together.
The financing gap is visible in current projects. The U.S. Department of Energy awarded the Gulf Coast Hydrogen Hub an initial $22 million tranche against a possible federal cost share of up to $1.2 billion, before the regional system can deliver its planned services. The IEA reports that only one of 31 low-emissions projects announced for 2030 had reached final investment decision, showing why announced capacity is not yet qualified supply.
An offtake agreement can help a producer borrow for an electrolyzer, but the buyer will not rely on the project until supply, pressure, purity, and delivery interruptions are covered. A pipeline developer needs enough committed volume to pay for construction, while a producer needs the pipeline before siting far from demand. The money arrangement changes which physical route is available; it does not change the molecule’s properties.
The current industrial chain reduces this financing problem by keeping production and use inside one complex. The proposed export chain separates renewable generation, hydrogen production, ammonia or liquid-hydrogen shipping, receiving terminals, cracking or distribution, and final use. More contracts can allocate responsibility, but they also create more places where an unpaid invoice or uncertain specification can stop a physical flow.
A production label is not a complete result
Terms such as grey, blue, and green describe a production route under a particular accounting scheme; they do not replace a physical description. A credible emissions claim must account for the feedstock or electricity, plant energy, capture and storage where relevant, methane leakage, transport, conversion, and the time period being compared.
Likewise, a production certificate may identify a facility and an approved calculation, while a meter records kilograms at a delivery point and a purity test samples a particular lot. A pipeline inspection can find defects in a defined section, but not prove that no leak occurred elsewhere. A carbon-intensity number can be useful for a procurement decision without establishing the condition of every upstream asset.
DOE notes that hydrogen is colourless, odourless, and highly flammable, and that pipelines and tanks need compatible materials, ventilation, leak detection, and controls for embrittlement. Hydrogen is not a direct greenhouse gas, but leakage can affect atmospheric methane and ozone, so leak measurements belong in an emissions assessment alongside the production route.
Outputs do not disappear at delivery
Reforming leaves carbon dioxide, process water, heat, and spent catalysts. Electrolysis leaves oxygen, heat, treated water streams, and worn stacks. Compression and liquefaction consume electricity and produce heat; liquid storage can release boil-off. Ammonia synthesis or cracking adds ammonia handling and residual-gas obligations. A useful hydrogen delivery therefore shifts material into new streams rather than ending the chain.
At retirement, reformers, electrolyzers, compressors, tanks, pipelines, cables, catalysts, membranes, and electrical equipment need separate routes for reuse, recycling, treatment, or disposal. Carbon dioxide captured from a reformer needs a monitored storage pathway for as long as the storage obligation requires. A hydrogen project is not physically complete because a customer has paid for a kilogram. Carbon dioxide, spent catalysts, membranes, wastewater, and retired equipment still need routes for storage, treatment, reuse, or disposal.
Inside CompanyGraph
Explore gas and electricity suppliers, water systems, reformer and electrolyzer makers, hydrogen producers, compression and storage operators, pipeline and port companies, ammonia and refining complexes, steelmakers, fuel-cell users, certification bodies, regulators, financiers, and closure contractors. CompanyGraph can map their handoffs and stated contracts; it cannot by itself prove a lot’s purity, a pipeline’s leak rate, a capture project’s storage integrity, or the service delivered at the receiving equipment.