PEM vs Alkaline Electrolyzer: How to Choose for a Hydrogen Project

For most wind and solar projects, PEM's ability to run at very low load is worth less than it sounds. In the worked 100 MW example below, that flexibility recovers about 5% more hydrogen, worth roughly $0.10/kg. The bigger levers are the installed price gap and how often the stacks must be replaced.

That does not make alkaline the default. It means the choice should come from a model of your own power profile, quotes and service terms, not from a rule of thumb. This guide gives you a worked example, then six checks that settle the decision.

This is general pre-feasibility guidance, not engineering or investment advice. Confirm all figures with vendor quotations and qualified engineers.

How the two technologies differ

Both are low-temperature technologies that split water using electricity. The practical differences that matter for a project are summarized below.

Two things in that table surprise many developers. First, the U.S. Department of Energy puts both technologies at the same system efficiency today, so efficiency alone rarely settles the choice. Second, DOE's cost figures exclude installation. Engineering, construction and other site costs are excluded, and that is where most of a project's money goes.

The IEA's Global Hydrogen Review 2026 assumptions put the 2025 installed cost of water electrolysis at $1,920–2,500/kW as a global average and $550–1,100/kW in China, without splitting by technology. The IEA uses a 50,000-hour stack life as an economic optimum, while noting stacks can reach up to 95,000 hours. Treat any single "PEM costs X" figure with caution.

Worked example: what is low-load flexibility worth?

Consider a hypothetical 100 MW plant on a hybrid wind and solar supply. All inputs are illustrative assumptions, not quotes, project data or H2Hub outputs.

Shared assumptions: 55 kWh/kg for both technologies, power at $45/MWh paid only on electricity used, fixed operating cost of 3% of capital cost a year, and an 8% discount rate over 20 years (capital recovery factor 0.10185).

Power profile: the supply can deliver the following share of the plant's 100 MW across the year's 8,760 hours.

Technology assumptions:

• Alkaline: $2,000/kW installed, minimum load 20%, stack replacement at $250/kW every 60,000 operating hours
• PEM: $2,300/kW installed, minimum load 5%, stack replacement at $450/kW every 40,000 operating hours

Stack costs and lifetimes use DOE's 2022 status values. Replacement is charged in proportion to operating hours.

Result on the variable supply

The alkaline plant cannot use the 1,900 hours below its 20% minimum load, so it loses 18,500 MWh, or 4.9% of the available energy. PEM captures that energy and makes about 336 more tonnes a year.

Yet PEM still costs $1.26/kg more. Most of the gap comes from stack replacement: PEM runs 7,560 hours a year, so at 40,000 hours its stacks last about 5.3 years, against about 10.6 years for alkaline.

To isolate the flexibility, give the alkaline plant PEM's 5% minimum load and change nothing else. Its cost falls from $6.89/kg to $6.79/kg. In this profile, low-load flexibility is worth about $0.10/kg, or about $50/kW of extra capital cost. If PEM's installed premium over alkaline is larger than that, flexibility alone does not pay for it.

How the result changes

• Longer PEM stack life: at 80,000 hours, DOE's 2026 target, PEM falls to $7.53/kg. It would then need an installed cost of about $1,960/kW to match alkaline, or slightly below alkaline's price.
• Lower alkaline minimum load: at 10%, the alkaline plant recovers the 15% hours and falls to $6.79/kg.
• Baseload power: at 8,000 full-load hours on firm supply, both use the same electricity, and alkaline comes to $4.52/kg against PEM's $5.18/kg.

The lesson is not "alkaline wins." It is that the price gap and the stack service terms in your actual quotes decide this, and the flexibility benefit is smaller than many pitch decks imply. See our LCOH explainer for how these cost lines combine.

What the example leaves out

This is a screening model built from bins, not an hourly simulation. It ignores several effects that can favor PEM:

• Delivery pressure: many PEM systems deliver hydrogen at higher pressure, which can reduce downstream compression.
• Standby and restart: alkaline plants may need standby power to stay warm and may take longer to restart, reducing usable hours further.
• Ramping and gas purity: fast swings and long periods near minimum load can raise impurity limits and trip alkaline plants.
• Footprint: PEM is usually more compact, which matters at tight brownfield sites.

It also ignores effects that can favor alkaline, such as iridium supply risk and longer degradation records. Degradation under heavy cycling is not well established for either technology. DOE notes its degradation status values are based on continuous operation.

Two 200 MW projects, two different choices

Two refinery-supply projects of the same size in northwest Europe show that experienced buyers reach different answers.

Air Liquide Normand'Hy, France — PEM. The 200 MW plant uses Siemens Energy PEM technology. Air Liquide said in 2023 that TotalEnergies would generate up to 100 MW of renewable power for it, with the rest from long-term renewable PPAs and low-carbon grid power. The company's 2025 annual report describes commissioning as planned for 2026 and output of around 28,000 tonnes a year.

Shell Holland Hydrogen 1, Netherlands — alkaline. thyssenkrupp nucera is supplying a 200 MW plant built from its 20 MW alkaline modules. Shell's 2022 investment decision said power would come from the Hollandse Kust (noord) offshore wind farm, with output of up to 60,000 kg a day.

Neither company has published the comparison behind its choice, so no conclusion about cost can be drawn from them. They show only that both technologies are being used at scale with renewable supply.

Six checks before you choose

1. Build the hourly profile first. Count the hours your supply sits below each vendor's minimum load. If there are few, flexibility is worth little. Firming, grid top-up or a battery can also shrink that count.
2. Compare installed cost, not equipment price. Ask both vendors for the same scope, including rectifiers, gas treatment, water treatment, compression and construction. The IEA found that installing Chinese electrolyzers outside China cost $1,500–2,400/kW once transport and tariffs were included (IEA 2025), close to the cost of other suppliers.
3. Get the stack terms in writing. Ask for the guaranteed lifetime and degradation rate under your duty cycle, the replacement price and lead time, and what the warranty excludes.
4. Check the minimum load in practice. Ask for the guaranteed minimum load, standby power, cold-start and warm-start times, and ramp rates for your plant size.
5. Price the whole outlet. Include delivery pressure, purity and the compression your plant design needs, not only the stack.
6. Test the supplier, not just the chemistry. Check service presence near your site, delivered reference plants at a similar size and the supplier's balance sheet. A 20-year plant needs a 20-year service partner.

Compare the cases in H2Hub

Start your first H2Hub project free to test electrolyzer cost, operating hours and power price for a candidate site. The LCOH Lab explains the inputs behind each result.

H2Hub is a screening and learning tool. Its outputs are estimates, not vendor quotations or engineering opinions.

Frequently asked questions

Is PEM or alkaline more efficient?

On DOE's 2022 status values, both use about 55 kWh per kilogram at system level. Specific products differ, so compare guaranteed values at beginning and end of life for your operating range.

Is PEM better for renewable energy?

PEM can run at lower load and respond faster, which helps on variable supply. Whether that is worth paying for depends on how many low-power hours your profile has and the price gap between quotes. In our example it was worth about $50/kW.

Why does PEM cost more?

PEM stacks use platinum-group metal catalysts, including iridium, and DOE's status values show higher stack and system costs. Installed project cost is also affected by scope, location and supplier.

How long do electrolyzer stacks last?

DOE's 2022 status values are 60,000 operating hours for alkaline and 40,000 for PEM, with 80,000 as the 2026 target for both. Guaranteed lifetimes vary by vendor and duty cycle.

Primary sources and further reading

1. U.S. Department of Energy — Technical Targets for Proton Exchange Membrane Electrolysis, page updated 29 May 2026.
2. U.S. Department of Energy — Technical Targets for Liquid Alkaline Electrolysis, page updated 29 May 2026.
3. International Energy Agency — Global Hydrogen Review 2026: Assumptions Annex. Installed cost, efficiency and stack life assumptions.
4. International Energy Agency — Global Hydrogen Review 2025: Executive summary. Installed cost of Chinese and non-Chinese electrolyzers.
5. Electric Power Research Institute — Technology Brief: Water Electrolyzer Stack Degradation. Load ranges and degradation mechanisms.
6. Air Liquide — Air Liquide takes a further step in developing the hydrogen sector in France, 14 September 2023, and Integrated Annual Report 2025.
7. thyssenkrupp nucera — thyssenkrupp to install 200 MW green hydrogen facility for Shell in port of Rotterdam, 10 January 2022.
8. Shell Nederland — Holland Hydrogen 1 final investment decision, 6 July 2022 (Dutch).

Editorial verification: 6 November 2026. Worked examples are calculations from stated assumptions, not forecasts, market prices or H2Hub-generated results.