Hydrogen Purity Comparison Testing: Standards, Methods, and Technical Applications

Hydrogen purity comparison testing is a systematic interlaboratory or inter-method verification procedure used to confirm the accuracy, consistency, and traceability of hydrogen purity and impurity measurements against recognized frameworks such as ISO 14687, ISO 21087, ISO/IEC 17043, and ISO 13528. This technical article explains the sample types, tested parameters, analytical methods, statistical evaluation criteria, and practical applications of hydrogen purity comparison testing for fuel cell, industrial, and laboratory quality assurance programs.

Technical Overview

Hydrogen purity comparison testing (also referred to as interlaboratory comparison or proficiency testing for hydrogen purity) is a structured exercise in which homogeneous hydrogen samples are distributed to two or more laboratories, or analyzed by two or more analytical methods, and the resulting purity and impurity data are compared using defined statistical criteria. The process is governed by ISO/IEC 17043 for proficiency testing schemes and ISO 13528 for statistical evaluation, while the underlying analytical work is performed under ISO/IEC 17025 laboratory management systems.

The primary objectives of a hydrogen purity comparison test are to verify laboratory competence, validate analytical methods, quantify measurement bias and uncertainty, and demonstrate metrological traceability to certified reference materials. Comparison testing is particularly critical for hydrogen because the required detection limits for certain impurities reach the parts-per-billion (ppb) range, where even minor differences in sampling, calibration, or detector performance can produce significant discrepancies between laboratories.

Hydrogen purity is reported either as a direct measurement or, more commonly, calculated "by difference," where the sum of all quantified impurities is subtracted from 100%. Representative grade thresholds include 99.99% for pure hydrogen, 99.999% for high-purity hydrogen, and 99.9999% for ultra-pure hydrogen as specified in GB/T 3634.2, and 99.97% minimum purity for PEM fuel cell applications as specified in ISO 14687 Grade D and GB/T 37244. For fuel cell grade hydrogen, individual impurity limits are extremely stringent, for example carbon monoxide at 0.2 µmol/mol and total sulfur at 0.004 µmol/mol, which is why a complete impurity suite must be measured rather than relying on a single purity value.

Testing Samples

Cylinder Gas Samples

Compressed hydrogen samples from industrial production sources such as water electrolyzers, steam methane reforming (SMR) plants, pressure swing adsorption (PSA) units, and by-product recovery systems are collected into high-pressure cylinders. Sample vessels are typically electropolished or inert-surface-coated stainless steel cylinders that have been evacuated, purged, and preconditioned with hydrogen to minimize adsorption, desorption, and memory effects. Source pressures commonly range from 5 MPa to 20 MPa, and cylinder preparation procedures are documented as part of the comparison protocol.

Hydrogen Refueling Station Samples

Hydrogen dispensed at refueling stations is sampled on-site using passivated sampling vessels connected through the fueling nozzle in accordance with ISO 19880-1 and the sampling practice described in ASTM D7606. Because station hydrogen may be pre-cooled and delivered at pressures up to the 70 MPa service class, sampling protocols must control pressure letdown rates and thermal effects to prevent condensation or loss of trace contaminants. Field sampling records, including temperature, pressure, and handling conditions, accompany each vessel in the chain of custody.

Certified Reference Materials and Synthetic Mixtures

Certified reference gas mixtures containing trace impurities in a hydrogen matrix, prepared gravimetrically in accordance with ISO 6142 and certified with documented expanded uncertainties, are used as anchor points in comparison exercises. These materials allow each participating laboratory to be assessed not only against peer results but also against an independently certified value, strengthening the traceability chain of the entire comparison.

Homogeneity and Stability Verification

Prior to distribution, the comparison samples are subjected to homogeneity testing to confirm that between-unit variation does not bias the statistical evaluation, and stability testing to confirm that impurity concentrations remain unchanged during storage and transport. Both requirements are specified in ISO 13528 and form a mandatory part of a technically valid comparison scheme.

Testing Items

A hydrogen purity comparison test covers the full impurity profile needed to establish the purity figure, with the specific parameter list selected according to the applicable product specification. Typical tested items include:

• Hydrogen purity (reported directly or calculated by difference from 100%)
• Oxygen (O2)
• Nitrogen (N2) and argon (Ar), reported individually or combined
• Methane (CH4) and total hydrocarbons (as CH4)
• Carbon monoxide (CO) and carbon dioxide (CO2)
• Helium (He) and other noble gases
• Moisture (H2O) and water dew point
• Total sulfur and sulfur species (H2S, COS, CS2, mercaptans)
• Ammonia (NH3)
• Formaldehyde (HCHO) and formic acid (HCOOH)
• Halogenated compounds (HCl, HF, and organic halides)
• Particulates and oil mist in compressed hydrogen systems

For fuel cell applications per ISO 14687 and GB/T 37244, all listed impurities relevant to the specification must be quantified, since fuel cell catalyst poisoning by CO, sulfur, and ammonia can occur even at ppb levels despite an otherwise acceptable aggregate purity value.

Testing Methods

Gas Chromatographic Analysis

Gas chromatography is the core technique for permanent gas impurities. Thermal conductivity detection (GC-TCD) is used for higher-level components such as O2 and N2, while pulsed discharge helium ionization detection (GC-PDHID) in accordance with ASTM D7649 provides detection limits down to the ppb level for CO, CO2, CH4, N2, O2, and Ar in a hydrogen matrix. GC-FID coupled with a methanizer, consistent with the approach in GB/T 8984, is applied to carbon monoxide, carbon dioxide, and methane at trace levels. Calibration is performed using certified gas standards traceable to national metrology institutes.

Moisture Determination

Sulfur, Nitrogen, and Special Impurities

Total sulfur and individual sulfur species are determined by gas chromatography with sulfur chemiluminescence detection (GC-SCD) per ASTM D7652, or by flame photometric detection for screening purposes. Ammonia and acid halides are measured by impinger absorption followed by ion chromatography or spectrophotometry, while formaldehyde, formic acid, and ammonia can also be quantified by Fourier transform infrared spectroscopy (FTIR) at low ppm and sub-ppm levels.

Comparison Test Design and Execution

The comparison protocol defines a common sampling procedure, a unified or candidate analytical method, the number of replicate determinations, calibration requirements, and reporting formats. Samples are coded and distributed blind to participating laboratories, instrument runs are randomized, and each laboratory reports results with expanded measurement uncertainty (typically coverage factor k = 2) in accordance with ISO/IEC 17025 requirements for uncertainty estimation.

Data Evaluation and Acceptance Criteria

Comparison results are evaluated statistically rather than by simple numerical agreement. The most widely applied metrics are:

• Robust z-scores per ISO 13528, where |z| ≤ 2 is satisfactory, 2 < |z| < 3 is a warning signal, and |z| ≥ 3 is an action signal requiring investigation
• En numbers, where |En| ≤ 1 indicates agreement between the laboratory result and the reference value after accounting for measurement uncertainties
• Repeatability and reproducibility limits derived from ISO 5725 and its Chinese counterpart GB/T 6379 for method precision studies
• Bias assessment against certified reference values, confirming whether the certified value falls within the laboratory's reported uncertainty interval

Laboratories receiving action signals conduct root cause analysis covering calibration standards, detector contamination, carrier gas quality, sample handling, and data processing, followed by corrective action and, where necessary, a repeat comparison round.

Applications

Hydrogen purity comparison testing supports multiple quality assurance scenarios across the hydrogen value chain:

• Fuel cell vehicle hydrogen quality verification against ISO 14687 Grade D and GB/T 37244 requirements for dispensed fuel
• Fuel quality control and online analyzer validation at hydrogen refueling stations in line with ISO 19880-8
• Laboratory competence demonstration and proficiency testing for accredited third-party laboratories under ISO/IEC 17025, ISO/IEC 17043, and GB/T 15483
• Production quality control for electrolytic, reforming, PSA, and by-product hydrogen, including grade verification against GB/T 3634.1 and GB/T 3634.2
• Instrument and method validation, including calibration transfer between analyzers and qualification of new analytical procedures
• Research applications such as semiconductor process gases, hydrogen embrittlement studies, and analytical method development requiring verified ultra-high-purity hydrogen

FAQ

What is a hydrogen purity comparison test?

It is a controlled interlaboratory or inter-method exercise in which the same homogeneous hydrogen samples are analyzed by multiple participants, and the purity and impurity results are compared statistically per ISO/IEC 17043 and ISO 13528 to verify measurement accuracy and consistency.

Why is hydrogen purity often calculated by difference?

Because hydrogen is the matrix gas, its concentration is most accurately obtained by subtracting the sum of all measured impurities from 100%, provided the complete impurity suite specified in the applicable standard has been quantified with sufficient sensitivity.

Which hydrogen grades can be evaluated?

Comparison testing covers industrial hydrogen, pure hydrogen (≥99.99%), high-purity hydrogen (≥99.999%), ultra-pure hydrogen (≥99.9999%), and fuel cell grade hydrogen (≥99.97% with strict individual impurity limits).

What sample quantity and pressure are required?

Typical comparison programs require one or two high-pressure cylinders per laboratory, each containing several liters of hydrogen at pressures from a few MPa up to refueling station service pressures, with exact volumes determined by the impurity test list and instrument demands.

How are discrepancies between laboratories resolved?

Discrepancies flagged by z-scores or En numbers trigger root cause analysis of calibration standards, sampling and transfer procedures, instrument performance, and data treatment, followed by corrective action and repeat testing to confirm resolution.

How often should comparison testing be performed?

Accredited laboratories typically participate in relevant proficiency or comparison schemes on a periodic cycle defined by their quality system, while hydrogen producers and refueling station operators perform comparisons whenever methods, instruments, or supply sources change, and as part of ongoing quality surveillance.

What deliverables are provided at the conclusion?

A complete comparison report includes the assigned or certified reference values, individual participant results with uncertainties, statistical evaluations such as z-scores and En numbers, conclusions on comparability, and any recommended corrective actions.