Why electrolyzer cells and stacks perform differently: what to check when moving from cell results to a stack

New electrodes and membranes are usually tested in a single cell first. A single cell is easy to control and gives results quickly. When the same materials go into a stack, however, the voltage per cell is often higher and degradation is often faster. The materials are the same; the environment each cell sits in is not.

This article explains why cell and stack results differ, for performance and for durability, and lists what to check when you review a stack from single-cell results.

• Do not compare the average cell voltage of a stack directly with a single-cell voltage. A stack adds variation in flow, temperature and clamping pressure.
• Check the spread in cell voltage as well as the average. A stack with the same average but a wider spread will have a shorter life.
• Judge by the weakest cell. A stack reaches its operating limit at the cell that degrades first.
• Ask for a stack-level durability record. A single-cell record does not stand in for stack life.

How a single cell and a stack differ

A single-cell test measures the materials. A stack adds the variation that comes from stacking many cells. Most of the gap between cell and stack results comes from that variation.

Why performance differs

Flow distribution. Electrolyte enters through a manifold and is divided between the cells. Cells near the inlet and far from it can receive different flow. In a starved cell, bubbles do not clear and cover the electrode surface, and the cell voltage rises.

Cell temperature. Cells in the middle of a stack receive heat from their neighbours; cells near the end plates lose heat to the outside. At the same current, cells run at different temperatures, and a cooler cell runs at higher voltage. A single-cell test uses an external heater, so this difference does not appear.

Clamping pressure. A stack is clamped with end plates and bolts in one go. The larger the area and the more cells, the harder it is to apply pressure evenly. Where pressure is low, contact resistance rises; where it is too high, the membrane or gasket can be damaged.

Shunt current. The cells share the same electrolyte manifold, and the electrolyte conducts, so part of the current bypasses cells through the electrolyte. This shunt current produces no hydrogen and lowers stack efficiency. A single cell has no bypass path, so the loss does not appear.

Active area. Test cells are usually a few to a few tens of square centimetres; stack cells are much larger. Over a larger area, current, bubbles and temperature are less evenly distributed, and at the same current density a large cell often runs at a higher average voltage than a small one.

Why durability differs

The weakest cell sets the stack's life. The cells are in series, so every cell carries the same current. Conditions differ slightly from cell to cell, and so does the rate of degradation. As operating time builds up, the spread in cell voltage widens, and the cell that degrades fastest reaches the voltage limit first. A stack can have to stop because of one cell while its average voltage still has margin.

System events are recorded in the stack. A stack runs with a power supply, pumps, a heater and a make-up water supply. Temperature changes when make-up water enters, changes in circulation flow, and stops and restarts all affect the stack. A single-cell test bench is managed so that such events hardly occur, which is one reason the same materials often give a shorter durability result in a stack.

Starts and stops act differently on each cell. Each stop and restart brings a large swing in cell voltage and temperature. Cells recover temperature and refill with electrolyte at different rates depending on position, so the same stop puts a different load on each cell.

Telling the causes apart in the lab

If you test stacks or short stacks yourself, the following methods separate most of the factors that raise cell voltage.

Plot cell voltage by position. Measure each cell individually and plot the voltages in stack order. Different causes leave different patterns.

Estimate the temperature effect at a few millivolts per degree. In our HXS-2 long-term record, a 1 °C drop in stack inlet temperature raised cell voltage by about 6–8 mV. On that basis, end cells running 3 °C cooler than the middle cells would sit about 20 mV higher. If the spread between cells is of that order, check the temperature distribution before blaming the material.

Check shunt current and crossover through hydrogen output. The theoretical hydrogen output of an N-cell stack at current I is N × I ÷ 2F, where F is the Faraday constant, 96,485 C/mol. A 23-cell stack at 50 A should produce about 480 NL/h (0 °C, 1 atm). If the measured dry hydrogen flow is lower, the difference comes from shunt current, crossover and leaks. These losses take a larger share at lower current, so measuring at several currents helps narrow down the cause.

Put a short stack between the single cell and the full stack. Rather than moving straight from single-cell results to a full stack, check first in a short stack of a few cells. You need at least three cells to tell the end cells from a middle cell. Measure each cell voltage and, if you can, the temperature by position. Distribution, temperature and clamping effects show up at much lower cost than in a full stack.

What to check when reviewing a stack from cell results

Why we publish stack data separately

We sell electrodes (HXP-an, HXP-ca) and stacks (HXS-2). Electrode data come from single-cell and short-stack tests; stack data come from full stacks running inside a system. We measure them separately because cell results do not carry over to a stack unchanged.

If you are reviewing a stack, you can check the stack-level polarisation curve and cumulative operating record for the HXS-2. If you would like to compare results from your own cells or stacks, contact us with the test conditions.