How to qualify an alkaline electrolyser separator

Most separator evaluations are not wrong, they are just not comparable: different cells, different electrodes, different days. The fix is simple and costs nothing: change one thing at a time, and agree in advance what “better” means.

1. Agree the conditions and the acceptance criteria first

Before any sample is cut, write down the test conditions and keep them identical for both separators. Then agree the acceptance criteria with everyone who will sign off the result, including your licensor if you have one. A criterion agreed after the data arrives is a negotiation, not a test.

Cell and active areaThe same cell, frames and gaskets for both separators.

ElectrodesSame anode and cathode, ideally from the same batch.

ElectrolyteKOH concentration, and how it is circulated.

TemperaturesRoom temperature and your operating temperature.

PressureAtmospheric or your operating pressure.

Current-density rangeUp to and beyond your nominal operating point.

CompressionThe same torque or gap for both builds.

Reference separatorThe one you use today, from a known lot.

Typical acceptance criteria cover cell voltage at nominal current density against the reference, hydrogen in oxygen at minimum load, voltage drift over the durability run, and the condition of the separator after the test.

2. Prepare the samples the same way

Record the lot of every piece you cut, for both separators, and keep the certificates of analysis with the test record.
Wet before assembly in KOH at the test concentration, following each supplier's instructions, so the pores are completely filled. A partly wetted separator reads as a high-resistance one.
Keep it wet from wetting to assembly; a separator that dries out may not recover.
Condition the cell at nominal current density until the voltage is stable before recording any data.

3. Polarisation: compare at fixed voltage and at fixed current

Run galvanostatic steps from low to high current density and record the voltage once it has settled at each step, at room temperature and at your operating temperature. Then read the two curves in two directions:

At the same voltageHow much more current, and so hydrogen, the candidate gives. This is the stack-size view: more current per cell means fewer cells for the same output.

At the same current densityHow many millivolts lower the cell runs. This is the power-bill view. At 100% Faradaic efficiency, 1 mV lower cell voltage saves about 26.6 Wh per kilogram of hydrogen, so 50 mV is about 1.3 kWh/kg.

Compare only inside the current range both curves cover; extrapolating one curve beyond its data is the most common way to overstate a difference. The Separator Evaluator does both readings from your instrument files, in your browser.

4. High-frequency resistance: find out where the difference comes from

Polarisation tells you how much better a cell is; high-frequency resistance (HFR, from impedance at the operating point) tells you why. The separator's contribution is ohmic, so the HFR difference between the two builds, multiplied by current density, is the part of the voltage saving you can attribute to the separator and contacts:

Worked exampleHFR of 0.32 Ω·cm² with the reference and 0.22 Ω·cm² with the candidate, at 0.5 A/cm²: (0.32 − 0.22) × 0.5 = 0.05 V, so 50 mV of the saving is ohmic. If the total saving is 70 mV, the other 20 mV comes from kinetics, bubbles or wetting, and is worth understanding before you scale up.

5. Gas purity: test where it is hardest, at minimum load

Hydrogen crossing into the oxygen side is diluted by less oxygen when the cell runs at low current, so hydrogen-in-oxygen is highest at minimum load. The lower flammability limit of hydrogen in oxygen is about 4% by volume, and plants trip well below it, commonly around half. Measure hydrogen in oxygen from minimum to nominal load for both separators: the margin to your trip point at minimum load sets how far down you can turn the electrolyser, which matters when it follows renewable power.

6. Durability: hold, measure the drift, agree the duration

Hold the cell at nominal current density and record voltage over time. Report the drift as microvolts per hour from a fitted line, not two readings, and agree the duration up front: a shorter screening run first, a longer one for design-in. Temperature and pressure cycling, if your plant will see them, belong in the plan too.

7. Look at the separator after the test

Open the cell and inspect both separators: visual condition, edges and the gasket area, thickness and, if you can, bubble point against the certificate. A separator that performed well but came out damaged is not qualified.

8. Turn the data into a decision

Collect conditions, lots, curves and the metrics above in one data pack, compared against the criteria you agreed in step 1. The outcome is one of three: qualified for a trial lot, more testing with a clear question, or not qualified. All three are useful; an unclear result is not.

Checklist

Same cell, electrodes, KOH, temperature, pressure and compression for both separators
Acceptance criteria written down before testing
Lots recorded; samples fully wetted; cell conditioned
Polarisation at room and operating temperature, read at fixed voltage and fixed current
HFR at the operating point, to separate the ohmic saving
Hydrogen in oxygen from minimum to nominal load
Durability with drift in µV/h; post-test inspection
One data pack, one decision

Notes. Energy per kilogram from 2F / M(H₂) at 100% Faradaic efficiency (about 26.6 kWh/kg per volt). Hydrogen-oxygen flammability limit from standard safety data; trip settings are site-specific. Worked examples are illustrative.

Back to the resource centre