Hydrogen energy: a technical primer

Hydrogen energy is a set of industrial processes rather than a single fuel: electricity splits water into hydrogen and oxygen, the gas is stored or moved, and it is burned or recombined in a fuel cell to return electricity and heat. The technical case rests on where that electricity comes from, what the round trip costs, and whether the gas can reach a user through existing pipes. A reference site covering the field from production to end use, hydrogen systems explained, sets out the same chain in operational terms for project holders and plant operators.

What does green hydrogen production actually cost per kWh?

Electrolyser efficiency is quoted in kWh per kilogram of hydrogen, and the published range for commercial alkaline and PEM units sits between roughly 48 and 55 kWh/kg at stack level, rising when the balance of plant, compression and drying are counted. A plant running at 60 per cent load factor therefore consumes more electricity per kilogram than its nameplate figure suggests, because part-load operation lowers stack efficiency and increases the share of parasitic load.

The levelised cost of hydrogen follows from three inputs: the price of the electricity, the capital cost of the stack amortised over its service life, and the load factor. Where renewable generation is intermittent, the electrolyser is sized against a capacity factor that may fall below 40 per cent, and the stack is cycled rather than run flat out. Cycling raises degradation rates, and manufacturers publish different figures for steady operation and for variable operation, which is why two projects with the same nameplate capacity can report levelised costs that differ by a factor of two.

Coupling an electrolyser to a wind or solar farm is a question of matching curves. Curtailment is avoided by allowing the stack to follow the generation profile, but the stack then operates far from its design point for much of the year. The alternative, a grid connection with a power purchase agreement, trades renewable attribution for a steadier load. Both arrangements appear in the operational literature, and the choice is usually settled by the contract structure rather than by the physics.

How do fuel cell vehicles and stationary stacks differ in duty cycle?

A fuel cell vehicle is a hybrid: the stack provides baseload power and a battery buffers acceleration and regenerative braking. The duty cycle is dominated by transient load, with the stack spending much of its time between 10 and 40 per cent of rated power. Refuelling at 350 or 700 bar takes minutes, and the usable payload penalty against a diesel equivalent is set by the mass of the storage vessels, typically several hundred kilograms for a heavy truck.

Stationary stacks run a different profile. Backup power and microgrid units are sized for continuous or long-duration output, and their degradation is driven by hours at steady load rather than by cycling. Where a stationary unit is used for peak shaving, the duty cycle resembles the vehicle case; where it is used for resilience, the stack may sit idle for months and then run for days. Idle periods introduce their own failure modes, including membrane drying and catalyst oxidation, and operators schedule conditioning runs accordingly.

Which infrastructure questions decide whether a project proceeds?

Blending hydrogen into an existing gas pipeline is the cheapest way to move molecules over distance, and the published limits are set by the pipeline material, the pressure rating and the end use. Steel transmission lines tolerate single-digit percentages by volume without modification in many cases, while distribution networks with polyethylene or older steel and a high share of domestic connections face tighter limits. The blend percentage is therefore a network-specific figure, not a universal one.

Safety practice around hydrogen rests on three properties: it is buoyant, it has a wide flammability range, and it burns with a nearly invisible flame. Ventilation at the highest point of an enclosure, detection at the points where leaks accumulate, and avoidance of confined spaces are the standard measures. Detection systems are calibrated for hydrogen specifically, since catalytic sensors designed for methane respond differently.

Offtake is the other gate. A project without a contracted buyer is a project without a financier, and the instruments used are offtake agreements with a fixed volume and a price formula, sometimes indexed to the price of the fossil alternative. Industrial clusters, where steel, fertiliser and refining plants sit within pipeline distance of each other, allow a single production unit to serve several buyers and to share the infrastructure cost. The cluster model also concentrates the safety case and the permitting effort in one place.

Where does the round trip lose the most energy?

Every conversion step costs. Electrolysis converts electricity to hydrogen at 60 to 70 per cent efficiency on a higher heating value basis. Compression to 700 bar consumes a further 10 to 15 per cent of the energy content of the gas. A fuel cell returns electricity at 50 to 60 per cent efficiency. The round trip from electricity back to electricity therefore lands in the range of 25 to 35 per cent, which is why hydrogen is used where direct electrification is impractical: long-haul transport, high-temperature industrial heat, and seasonal storage.

Storage itself is a loss term. Compressed gas in salt caverns has a low standing loss but a high capital cost per unit of working gas. Liquefaction consumes roughly 30 per cent of the energy content of the hydrogen, which restricts liquid hydrogen to applications where volume and mass matter more than efficiency, such as aviation research and some road transport trials.

What figures should a project holder verify first?

Four numbers determine whether a scheme is credible. The first is the electricity price and its profile, because it dominates the levelised cost. The second is the stack's degradation rate under the actual duty cycle, not the datasheet figure for steady operation. The third is the pipeline or transport limit at the point of injection, which may be lower than the national blend target. The fourth is the offtake volume and its price formula, since a project with a firm buyer can absorb a higher production cost than one selling into a merchant market.

A fifth figure, often overlooked, is the water supply. Electrolysis consumes roughly nine litres of ultrapure water per kilogram of hydrogen, and the water treatment train is part of the capital cost. Sites with limited water access need to account for it before the electrolyser is ordered.

How should safety and permitting be sequenced?

Permitting follows the hazard, and the hazard is defined by the inventory. A small blending injection point is a different permitting case from a 100 MW production unit with storage. Ventilation studies, dispersion modelling and detection layout are usually required before construction, and the results feed back into the site layout. Operators who treat safety documentation as a post-design exercise tend to rebuild the layout later.

The sequence that appears in operational practice is: define the inventory, model the dispersion, place the detection and ventilation, then fix the layout. Contracting follows the same logic, with the offtake agreement and the grid connection agreement signed before the electrolyser order, since both determine the load profile the stack will actually see.