Compressing Green Hydrogen: The Pressure the Electrolyser Does Not Deliver

A green-hydrogen project does not buy one compressor. It buys a compression step whose inlet pressure is set by the electrolyser, and that inlet is low. In a 22,000 Nm³/h project now out to tender in Gansu, the hydrogen leaves electrolysis below the 3.0 MPa of the sphere tanks that receive it, and the scope calls for two 22,000 Nm³/h two-stage reciprocating machines to close the gap (1).

The ceiling is on the electrolyser, not on the compressor

The output pressure of an electrolyser is a design parameter with a practical ceiling, and it is not high. A Chinese provincial open-call specification for a current-generation alkaline stack of at least 1,000 Nm³/h fixes the hydrogen production pressure at 1.6 MPa, alongside a rated output of at least 1,000 Nm³/h, a DC power consumption of no more than 4.1 kWh/Nm³, an operating range of 50–120% of rating and a 25-year design life (2). A commercial Chinese PEM system is offered with an H₂ outlet pressure of either 1.6 MPa or 3 MPa, at 99.999% purity (3). International industry practice puts typical output pressure for both alkaline and PEM at 2–3 MPa, with development work under way to raise it — and states directly that because alkaline output pressure is not expected to exceed about 6 MPa, an additional compression step will always be needed for alkaline plants injecting into a transmission system (4).

That sentence is the reason this article exists. An electrolyser is an electrochemical cell with a diaphragm or a membrane between the hydrogen and oxygen spaces; raising the gas outlet pressure means loading that separator, the gaskets and the cell frame. The compressor carries the consequence instead of the cell.

Where the hydrogen has to arrive

The green-hydrogen ladder has more rungs than the refuelling ladder, because most green hydrogen is an industrial feedstock rather than a transport fuel.

The Gansu project puts the first two rungs in one tender. Its hydrogen production system is 22,000 Nm³/h of alkaline water electrolysis; it stores into four 1,000 m³ sphere tanks at 3.0 MPa; and it also builds a pilot platform with a 1,000 Nm³/h alkaline unit, a 500 Nm³/h PEM unit and a 20 MPa storage system holding 800 Nm³. The compressor package in the tender is two 22,000 Nm³/h two-stage reciprocating machines (1).

Green hydrogen broke the old selection rule twice

Hydrogen was already a difficult gas to compress for reasons unrelated to being green. Its molecular weight is 2.02, so it leaks through dynamic seals that hold heavier gases. It embrittles many steels, so the wetted materials have to be selected for it. And where the gas is destined for a fuel cell, contamination is a hard limit, which is why oil-free construction is standard rather than an option.

Green hydrogen added two constraints to that.

The inlet moves. A project fed by wind and solar has an electrolyser whose output follows the weather. A machine sized for one inlet pressure and one flow rate does not fit that duty: the compression step has to hold its discharge pressure while its inlet pressure and flow move. This is why variable-inlet capability, more than discharge pressure, has become the specification that decides the machine at the production end.

The flow is large and the pressure is low. A 22,000 Nm³/h electrolysis system is not a laboratory. Its compressor is a large machine, and it has to be oil-free — which removes the lubricated reciprocating machine that would otherwise be the cheapest way to move that much gas.

Who is building what

Reciprocating, oil-free. The Chinese reference for the production-end duty is the two-stage oil-free reciprocating machine — the class the Gansu tender specifies at 22,000 Nm³/h (1). It gives flow and ratio at the same time. The costs are pulsation and the purity ceiling of a piston-ring seal.

Dry screw. Screw machines serve the duties where the flow is very large and the pressure rise is modest. A Chinese builder supplies single oil-free screw machines with a volumetric flow of 1,266 m³/min, which the company describes as a record it has repeatedly raised (5). The same company won the core equipment package for the world's first 1.2-million-tonne green-electricity / green-hydrogen / pure-hydrogen metallurgy line: a hydrogen dry screw compressor, for a duty with a maximum suction volume of 83,430 Nm³/h (6). A European builder's hydrogen screw range starts at atmospheric pressure, reaching 6 bar in a single stage and 16 bar in two, with a water-buffered sealing system instead of a gas barrier (7).

Diaphragm. Where purity is absolute and the volume flow is smaller, the metal diaphragm machine is the reference: the process gas never touches the hydraulic fluid, and the containment is a static seal. Chinese diaphragm builders now cover the 45 MPa, 70 MPa, 90 MPa, 250 MPa and 320 MPa grades (8).

Hydraulic. The reciprocating machine driven by a hydraulic circuit rather than a crank is the newest of the four in this duty. With no crank forcing a fixed stroke relationship, the machine tolerates a moving inlet pressure and frequent starting. Chinese builders describe volumetric efficiencies above 95% and a footprint of 60–70% of a diaphragm machine, with products that accept inlet pressures down to 1.3 MPa; they have shipped 45 MPa ionic-liquid and 22 MPa hydraulic-piston hydrogen machines for a domestic energy station, and a 90 MPa hydraulic hydrogen compressor for refuelling duty (9)(10).

Centrifugal. Large flow, low ratio, no wearing parts in the gas path — a pressure rise that suits the front end of a liquefaction train rather than the storage end. That is exactly the duty a current Chinese open call separates out. A hydrogen liquefaction package is seeking both large and small oil-free reciprocating machines and a centrifugal machine, and its qualification threshold is a reference of more than 100,000 Nm³/h total flow in closed-loop high-purity dry oil-free hydrogen circulation, plus a reference above 30,000 Nm³/h for a centrifugal circulation machine (11).

A different drive. One line of work attacks the drive rather than the compressor. A university-and-company team in Shandong has developed a pump-and-motor integrated linear machine for a hydrogen compressor, on the argument that a rotary motor driving a compressor responds too slowly to the power swings of green-hydrogen production. The team reports a 60% smaller volume, 40% lower cost, 25% higher efficiency and a dynamic response more than twenty times faster than the conventional arrangement, with millisecond-level thrust control to follow production swings (12).

The boundary of this technical route

Four limits belong alongside the progress.

The performance figures above are makers' and projects' own. The 95% volumetric efficiency, the 60–70% footprint comparison, the 1,266 m³/min flow and the 90 MPa grade are company statements and product data, not independently measured field results. The linear-motor numbers come from a provincial entrepreneurship competition entry; they describe a developed machine, not a machine with an industrial service record.

The liquefaction package is a specification, not a machine. The 100,000 Nm³/h and 30,000 Nm³/h thresholds are qualification requirements in an open call issued in August 2026, with a seven-day response window that closed that month. As of the notice they describe a supplier to be selected, not equipment built for this duty.

No single machine family covers the ladder. The diaphragm machine is leak-tight and flow-limited. The hydraulic machine is flexible and carries a hydraulic system that needs maintenance. The reciprocating machine is efficient and pulsating, with a dynamic seal that leaves purity around 99.5% where a fuel cell wants more. The screw machine takes the largest flows at the lowest ratios. The centrifugal machine takes large flow only at low ratio. A project chooses a point on the ladder, not a machine for the whole of it.

And the constraint that opened the article has not moved. The electrolyser's outlet pressure is set by the cell, not by the compressor. Until stacks run reliably at storage or pipeline pressure, the compression step is not optional and cannot be designed away — only made to follow a variable supply.

What the trend looks like

Two movements are running at once. The production end is turning toward machines that tolerate a moving inlet — hydraulic drives and variable speed — because that is what a wind- or solar-fed electrolyser requires. The storage and liquefaction ends are pulling the largest flows onto screw and centrifugal machines, which is where the Chinese reference projects now sit. The machine that fits both ends has not been built; projects are buying two machines instead.

What this article does not cover

Note that this article does not select a compressor for any project, does not calculate stage count, power or intercooling, does not give prices, delivery times, purity specifications or a ranking of machine types, and does not compare manufacturers. It describes the pressure ceiling that electrolysis imposes, the destinations hydrogen has to reach, the machine families that serve each part of that ladder, and the limits of what has been demonstrated so far.

Related on this site

• High-Pressure Hydrogen Compression: The Pressure Ladder, the Sealing Breakthrough, and Where the Grades Now Stand
• Diaphragm vs Piston vs Screw Compressors: A Comparison Table

Sources