Fuel Cell and Electrolyzer Testing: Polarization, Gas Utilization, Efficiency, and Durability
Electrochemical testing converts a fabricated cell into measurable performance data. A polarization curve can reveal voltage losses, impedance spectroscopy can help separate contributions from different processes, and gas analysis can establish whether the measured current corresponds to the intended electrochemical reaction.
These measurements are only meaningful when operating conditions are controlled and reported. Temperature, gas composition, flow rate, pressure, humidity, current-collection geometry, sealing, and thermal history can all change the apparent performance. This guide presents a reproducible approach for testing fuel-cell and electrolysis cells, with particular attention to ceramic electrochemical devices.
1. Define the operating mode and sign convention
A reversible electrochemical cell can operate in two directions:
- Fuel-cell mode: a spontaneous chemical reaction generates electrical power. For hydrogen oxidation, hydrogen is consumed and water is produced.
- Electrolysis mode: an external power supply drives the reverse reaction. Steam is consumed and hydrogen is produced.
For the hydrogen/oxygen reaction,
\[ \mathrm{H_2 + \tfrac{1}{2}O_2 \rightleftharpoons H_2O} \]
the same cell architecture can therefore be tested in either mode, provided the electrodes, gas supply, current collectors, and seals are suitable for both environments.
Always state the sign convention used for current and voltage. In this article, current density magnitude is reported as a positive value in either operating mode; the mode itself determines whether the cell produces or consumes electrical power.
2. Establish a controlled test configuration
Before heating the cell, document the complete test configuration.
| Parameter | What to record |
|---|---|
| Cell | Composition, architecture, active area, thickness and fabrication history |
| Sealing | Sealant type, curing procedure, compression and leak-check method |
| Current collection | Mesh, paste, wire, contact area and applied load |
| Temperature | Set point, measured location, ramp rate and stabilization criterion |
| Fuel-side gas | Composition, flow rate, humidification and pressure |
| Oxygen-side gas | Composition, flow rate, humidity and pressure |
| Electrical setup | Instrument, wiring, polarity, current range and voltage-sensing points |
| Test history | Reduction, conditioning, thermal cycles and prior polarization |
Active area and current density
The geometric active area must be defined consistently. For a cell with measured current \(I\) and active area \(A\),
\[ j=\frac{I}{A} \]
where \(j\) is current density. Report the area in cm² if current density is reported in A cm\(^{-2}\).
Do not substitute the total pellet area for the electrochemically active area unless the entire area is exposed to the intended gas and electrically contacted in the same way. State how the active area was determined.
Gas flow and humidification
Report whether flow rates are specified at standard or actual conditions, and identify the reference temperature and pressure used by the flow controller. For humidified hydrogen, report the humidifier temperature or measured water-vapour partial pressure. A nominal hydrogen flow without humidity information is not sufficient to reproduce the fuel-side chemical potential.
For each gas stream, distinguish between:
- the dry-gas flow set by the mass-flow controller;
- the water vapour added by the humidifier;
- the total wet-gas flow reaching the cell.
3. Open-circuit voltage (OCV)
At open circuit, the external current is nominally zero. The measured voltage can be compared with the reversible Nernst potential calculated from the gas activities at the electrodes.
For the hydrogen/oxygen cell reaction, assuming ideal-gas behaviour and unit activity for liquid water,
\[ E_{\mathrm{N}}=E^\circ(T)+\frac{RT}{2F} \ln\left( \frac{p_{\mathrm{H_2}}\,p_{\mathrm{O_2}}^{1/2}} {p_{\mathrm{H_2O}}} \right) \]
where \(R\) is the gas constant, \(T\) is absolute temperature, \(F\) is the Faraday constant, and the partial pressures must use a consistent reference standard state. For a different product phase or non-ideal gas mixture, the appropriate activities must be used.
The standard potential \(E^\circ(T)\) is temperature-dependent. Do not insert the room-temperature standard potential into a high-temperature calculation without correction.
A lower-than-expected OCV can arise from several causes:
- gas leakage or crossover;
- electronic leakage through the electrolyte;
- mixed ionic-electronic conduction;
- incorrect gas composition or humidification;
- incomplete electrode activation;
- unstable temperature or electrical contacts.
OCV alone does not identify which mechanism is responsible. A leak test, gas analysis, thickness dependence, or independent transport measurement may be needed.
4. Current–voltage and power-density curves
A polarization curve records cell voltage as a function of current density. In fuel-cell mode, voltage generally decreases as current density increases. In electrolysis mode, the required cell voltage generally increases with current density.
The area-specific power density in fuel-cell mode is
\[ P_A=jV \]
where \(P_A\) is in W cm\(^{-2}\) when \(j\) is in A cm\(^{-2}\) and \(V\) is in volts.
A useful test sequence is:
- Stabilize the cell at the target temperature and gas composition.
- Record OCV after the voltage has reached a defined stability criterion.
- Apply a sequence of current steps or a controlled current sweep.
- Allow the voltage to stabilize at each point, using the same criterion throughout the experiment.
- Record voltage, current, temperature, gas flows and elapsed time.
- Repeat selected points to assess drift and hysteresis.
- Return to OCV and check whether the initial state has been recovered.
A current sweep can be faster, but the sweep rate must be reported and should be slow enough to avoid confusing transient behaviour with steady-state performance. Current-step measurements are often easier to interpret when gas conversion and electrode relaxation are significant.
Interpreting polarization losses
A practical decomposition is:
\[ V_{\mathrm{FC}}=E_{\mathrm{N}}-\eta_{\mathrm{act}} -\eta_{\mathrm{ohm}}-\eta_{\mathrm{conc}} \]
for fuel-cell operation, where the terms represent activation, ohmic and concentration losses. This is a conceptual decomposition; the individual terms are not necessarily separable from a single polarization curve.
For electrolysis, the required voltage can be expressed schematically as
\[ V_{\mathrm{EC}}=E_{\mathrm{rev}}+\eta_{\mathrm{act}} +\eta_{\mathrm{ohm}}+\eta_{\mathrm{conc}} \]
The reversible voltage and overpotentials depend on temperature and gas composition. In particular, local gas compositions at the electrode surfaces can differ from inlet compositions at high conversion.
5. Ohmic resistance and area-specific resistance
If the high-frequency intercept or another well-justified method provides the ohmic resistance \(R_{\mathrm{ohm}}\), the area-specific resistance is
\[ \mathrm{ASR}_{\mathrm{ohm}}=R_{\mathrm{ohm}}A \]
with units of Ω cm² when resistance is in Ω and area is in cm².
For a simple linear ohmic contribution,
\[ \Delta V_{\mathrm{ohm}}=j\,\mathrm{ASR}_{\mathrm{ohm}} \]
The total cell ASR should not automatically be equated with electrolyte resistance. It may include electrode, contact, current-collector and interfacial contributions, depending on the measurement arrangement.
A frequently used estimate,
\[ \mathrm{ASR}_{\mathrm{app}}=\frac{E_{\mathrm{OCV}}-V}{j} \]
is an apparent polarization resistance relative to the chosen reference voltage. It includes non-ohmic losses and should not be labelled electrolyte ASR or purely ohmic ASR.
For a more complete analysis, combine polarization curves with impedance measurements at OCV and, where appropriate, under operating load. State the DC bias, AC amplitude and frequency range used for each spectrum.
6. Fuel utilization and steam conversion
Gas utilization links the electrical current to the amount of reactant supplied. For hydrogen oxidation, two electrons are transferred per mole of hydrogen consumed. The ideal hydrogen consumption rate is therefore
\[ \dot n_{\mathrm{H_2,cons}}=\frac{I}{2F} \]
For a fuel stream containing hydrogen at inlet molar flow rate \(\dot n_{\mathrm{H_2,in}}\),
\[ U_{\mathrm{H_2}}= \frac{I}{2F\dot n_{\mathrm{H_2,in}}} \]
For steam electrolysis, the ideal steam consumption rate is also
\[ \dot n_{\mathrm{H_2O,cons}}=\frac{I}{2F} \]
and the inlet-based steam conversion is
\[ X_{\mathrm{H_2O}}= \frac{I}{2F\dot n_{\mathrm{H_2O,in}}} \]
These expressions assume that the measured current is entirely Faradaic for the stated reaction and that the inlet molar flow is known on the same wet/dry basis as the composition used in the calculation.
Illustrative hydrogen-utilization calculation
Consider a cell operating at \(I=0.50\) A with a dry hydrogen inlet flow of 100 sccm. At a stated reference condition of 0 °C and 1 atm, the molar volume is approximately 22.414 L mol\(^{-1}\). Thus,
\[ \dot n_{\mathrm{H_2,in}}= \frac{0.100\;\mathrm{L\,min^{-1}}} {22.414\;\mathrm{L\,mol^{-1}}\times60\;\mathrm{s\,min^{-1}}} \]
\[ \dot n_{\mathrm{H_2,in}}\approx7.44\times10^{-5}\;\mathrm{mol\,s^{-1}} \]
Using \(F=96485\) C mol\(^{-1}\),
\[ U_{\mathrm{H_2}}= \frac{0.50} {2(96485)(7.44\times10^{-5})} \approx0.0348 \]
The calculated inlet-based hydrogen utilization is approximately 3.5%. This example assumes pure dry hydrogen and does not include humidification or gas crossover. If the inlet contains a hydrogen fraction \(y_{\mathrm{H_2}}\), use the hydrogen molar flow, not the total flow.
Why utilization matters
High utilization can improve reactant efficiency but may also create steep concentration gradients, local fuel starvation, voltage instability and accelerated degradation. Report the inlet flow and calculated utilization at every operating point when current changes substantially.
For co-flow, counter-flow or cross-flow cells, the local gas composition varies along the active area. A single inlet-based utilization is a useful overall metric, not a direct measurement of the local electrode-surface composition.
7. Faradaic efficiency and product quantification
Faradaic efficiency compares the measured amount of product with the amount predicted from the charge passed.
For hydrogen production,
\[ \mathrm{FE}_{\mathrm{H_2}}= \frac{2F\,n_{\mathrm{H_2,measured}}} {\int I(t)\,dt}\times100\% \]
For constant current over a time interval \(t\),
\[ \mathrm{FE}_{\mathrm{H_2}}= \frac{2F\,n_{\mathrm{H_2,measured}}}{It}\times100\% \]
The measured hydrogen amount should be determined using a calibrated method such as gas chromatography or a calibrated flow measurement with suitable composition analysis. Correct for dead volume, leaks, water condensation, gas sampling losses and background hydrogen.
A Faradaic efficiency below 100% can indicate electronic leakage, gas crossover, side reactions, unaccounted product retention or measurement error. A value apparently above 100% usually signals a calibration, baseline, flow-reference or charge-integration problem; it should not be interpreted as enhanced electrochemical efficiency.
8. Thermoneutral voltage and electrolysis efficiency
For water splitting, the reversible voltage is related to the Gibbs free-energy change:
\[ E_{\mathrm{rev}}=\frac{\Delta G}{2F} \]
The thermoneutral voltage is related to the enthalpy change:
\[ E_{\mathrm{tn}}=\frac{\Delta H}{2F} \]
Both quantities depend on temperature and on the thermodynamic reference state, including whether water is treated as liquid or vapour.
At the thermoneutral voltage, the electrical energy supplied per mole of hydrogen equals the reaction enthalpy under the chosen reference conditions. Operation below or above this voltage changes the net heat requirement of the cell; it does not by itself establish the system-level efficiency.
A voltage-based electrolysis efficiency can be defined relative to a specified heating-value convention:
\[ \eta_{\mathrm{voltage}}= \frac{E_{\mathrm{reference}}}{V_{\mathrm{cell}}} \times100\% \]
where \(E_{\mathrm{reference}}\) must be explicitly identified. For an HHV-based convention, use the corresponding thermodynamic voltage; for an LHV-based convention, use the LHV-equivalent voltage. Report Faradaic efficiency separately. A cell-level voltage metric is not the same as system efficiency because it excludes balance-of-plant energy, heating, gas processing and power-electronics losses.
9. Impedance measurements under operating conditions
EIS can help determine how the cell response changes with temperature, gas composition and current load. A useful minimum dataset may include:
- an OCV spectrum after equilibration;
- spectra at selected current densities;
- spectra at relevant fuel or steam utilization levels;
- repeated spectra after durability testing.
Report whether the impedance is measured with a two-electrode or reference-electrode configuration. In a two-electrode cell, the measured spectrum contains contributions from both electrodes, electrolyte, contacts and interfaces. Do not assign an arc to one electrode without additional evidence.
Check that the perturbation amplitude is sufficiently small for the operating point, and inspect the raw spectrum before equivalent-circuit fitting. A fitted circuit is a model, not a unique microscopic description.
10. Durability and degradation testing
A durability test should distinguish initial conditioning from long-term degradation. Record the stabilization period, operating mode, current or voltage control, gas composition, flow, temperature, and any interruptions.
For constant-current operation, voltage degradation rate can be calculated from a defined time interval:
\[ r_V=\frac{V(t_2)-V(t_1)}{t_2-t_1} \]
For fuel-cell operation, a negative \(r_V\) indicates voltage loss under the chosen convention. For electrolysis at constant current, an increasing required voltage indicates deterioration. Report the absolute rate and the normalization used, for example mV kh\(^{-1}\).
A simple linear rate should only be reported if the data support approximately linear behaviour over the stated interval. If degradation is nonlinear, show the full time series and describe the fitting interval rather than extrapolating a single slope.
Useful diagnostic checkpoints include:
- OCV and leak integrity before and after the test;
- polarization curves at identical operating conditions;
- impedance spectra at matched temperature and gas composition;
- gas composition and flow verification;
- post-test phase, microstructure and interface characterization.
A change in voltage is not, by itself, proof of a specific degradation mechanism. Correlate electrochemical trends with post-test structural and chemical evidence.
11. Common experimental errors
- Unreported active area: makes current-density and power-density comparisons unreliable.
- Unspecified flow reference conditions: prevents reliable conversion from sccm to molar flow.
- Ignoring humidification: changes reactant partial pressures and calculated utilization.
- Comparing cells at different utilization: confounds intrinsic cell performance with concentration polarization.
- Using OCV as proof of gas-tightness: electronic leakage and mixed conduction can also reduce OCV.
- Calling apparent ASR an electrolyte resistance: the measured value may include several cell components.
- Reporting Faradaic efficiency without calibrated gas analysis: product quantification must be traceable.
- Mixing conditioning with degradation: initial activation or relaxation should not be included blindly in a long-term degradation slope.
- Comparing different thermal histories: heating, reduction and redox cycling can change the cell state.
12. Minimum reporting checklist
A reproducible report should include:
- Cell composition, architecture, fabrication, dimensions and active area.
- Electrode and current-collector materials, contact configuration and sealing method.
- Temperature, measurement location, stabilization criterion and thermal history.
- Gas composition, humidification, flow rates, pressure and flow-reference conditions.
- OCV, current–voltage data, current density and power density where applicable.
- Fuel utilization or steam conversion, with the molar-flow calculation shown.
- Faradaic-efficiency method, calibration and product-analysis details.
- EIS conditions, bias, amplitude, frequency range, fitting model and uncertainty.
- Durability protocol, operating time, interruptions and degradation-rate definition.
- Replicate measurements and uncertainty estimates.
Further reading
- R. O’Hayre, S.-W. Cha, W. Colella and F. B. Prinz, Fuel Cell Fundamentals, 3rd ed., Wiley (2016).
- S. C. Singhal and K. Kendall (eds.), High-Temperature Solid Oxide Fuel Cells: Fundamentals, Design and Applications, Elsevier (2003).
- International Electrotechnical Commission, IEC 62282 series, Fuel cell technologies (consult the applicable part and current edition for the test being performed).