Why Hydrogen Quality Matters for Industrial Heating

Hydrogen is increasingly being considered as a key fuel for decarbonising energy-intensive industrial processes. However, ensuring its reliable use involves more than simply supplying hydrogen to a burner: its quality and composition can directly influence flow measurements, process control, combustion performance and emissions.

As part of the HyInHeat project, research has explored how hydrogen impurities affect fuel-gas characteristics and flow measurement, as well as the potential of ultrasonic technology to monitor hydrogen purity directly in the pipeline. In this interview, Elsa Busson (RWTH Aachen) shares the hey findings, technical challenges and practical implications of this work, highlighting how it can support the reliable use of hydrogen in industrial heating applications.

Why is accurately measuring hydrogen quality so important for industrial applications?

Accurately measuring hydrogen quality is important because industrial processes are controlled through measurable quantities such as gas flow, energy input and the air-fuel ratio. If the hydrogen composition changes but the measurement system assumes pure hydrogen, the process may operate with more or less fuel than intended. In our tests, only 2 vol% impurities caused errors of up to 20% on some flow measuring devices, mainly because the gas density changed (see Fig.1). In an industrial furnace, an error of this size can affect temperature control, product quality, efficiency and emissions. I think hydrogen quality measurement is not simply about checking whether a specification is met. It is needed to translate the measured gas flow into the actual energy entering the process and to operate the plant consistently, even when the hydrogen supply changes. 

Many people assume that hydrogen is always ‘pure’. Why is it important to consider impurities?

Hydrogen is often referred to as if it were one uniform gas, but its exact composition depends on how it is produced, purified, stored and transported. A remaining fraction in the gas may contain nitrogen, oxygen, water, methane, carbon monoxide or carbon dioxide (or even other gases), and the type of impurity can matter as much as the total amount. In our study, mixtures containing only 2 vol% impurity had densities up to around 42% higher than pure hydrogen, depending on the density of the impurity. We also observed different effects of these impurities during combustion: oxygen increased NOx emissions, whereas methane reduced them under the tested conditions. This means that saying “98% hydrogen” is not enough to fully describe how the gas will behave. It’s better to know what makes up the other 25 or to at least know the density of the hydrogen fuel with impurities, particularly when operating flow meters and assessing emissions.

What was the biggest technical challenge you encountered during this work?

The biggest technical challenge was establishing a trustworthy “impure” gas mixture that differed from pure hydrogen by only very small amounts. We had to distinguish a real response of the flow meter from uncertainties in the gas mixture we were producing , pressure, temperature and in our calculations. Water was particularly difficult because condensation or droplets in the pipe could not be fully excluded. During the ultrasonic tests, even changes in ambient temperature caused pressure drift and had to be corrected before the refence composition could be calculated. So, a detailed uncertainty quantification was essential because without a reliable gas mixture as reference, a sensor or algorithm can appear accurate even when the comparison itself is not. 

One of the project’s innovations is using ultrasonic technology to estimate hydrogen purity. Why is this approach particularly promising?

The ultrasonic approach is promising because it can potentially provide two important pieces of information with one inline instrument (cf. Fig.3): the hydrogen flow and an estimate of its purity. Hydrogen has a much higher speed of sound (SOS) than typical impurity gases, so even a small impurity fraction produces a measurable change (cf. Fig. 2). The algorithms developed in the project compared the measured SOS with the value expected for pure hydrogen under the same pressure and temperature. In the simulations, the estimated uncertainty was approximately 0.03 to 0.09 mol%, and the tests with nitrogen and methane agreed well with those predictions. The main advantage for industry is that quality changes could be detected continuously and directly in the pipeline, without relying only on occasional samples or a separate analyser. It is not a universal composition measurement, but it is a very practical monitoring concept when the likely impurities are known. 

How does this research contribute to making hydrogen a more reliable industrial fuel?

This research contributes to reliability by connecting hydrogen quality directly with the way industrial plants measure and control their fuel. We identified which flow-meter principles were comparatively robust and which were affected by changes in gas density. We also showed that the error of density-sensitive flow meters can be reduce when the mixture density is known and an appropriate correction is applied. In parallel, the ultrasonic work demonstrated a possibility to detect quality changes directly in the pipeline. Together, these results can help operators maintain the intended fuel input, air-to-fuel ratio and process conditions even when hydrogen comes from different production or supply routes. In my view, this is an important step from simply proving that a burner can operate with hydrogen to ensuring that the complete fuel-supply and control system can operate predictably over time, independently of hydrogen impurities. 

Were any of the results particularly surprising to you?

What surprise me most was the scale of the effect caused by such a small impurity fraction. I expected density-sensitive sensors to be influenced, but a mixture containing 3 vol% impurity led to deviations of up to 20% for the orifice plate and 16% for the area flow meter. This shows how unusual hydrogen is compared with many conventional fuel gases. Because its density is so low, even a small amount of a heavier gas can change the mixture properties substantially. At the same time, it was encouraging to see that the thermal mass meter remained within about 2% of deviation and that a relatively simple density correction reduced the deviations of the other flow meters to a maximum of 4%. The results were therefore both a warning and an indication that these effects can be managed. 

If you could leave readers with one message from this work, that would it be?

The main message I would leave is that hydrogen quality should be treated as a process variable, not as a fixed label. Knowing that a gas is “hydrogen” or even “98% hydrogen”  does not automatically tell us how accurately it can be measured or how it will behave in a burner. The identity of the remaining components can influence the measured flow the actual energy input, the combustion-air requirement and pollutant emissions. This does not mean that industrial hydrogen must always have extremely high purity. It means that the expected quality range must be understood and considered when the measurement and control system is designed. If industry includes quality monitoring and suitable correction strategies from the beginning, hydrogen can be used much more predictably without unnecessarily demanding the highest purity for every thermal process.

Fig.1:  Deviation from the set point for the Mass Flow Meter (MFM), Area Flow Meter (AFM) and Orifice Plate at three power outputs (16kW, 20kW,  and 24 kW) in relation to the density ratio of hydrogen mixtures with impurities to pure hydrogen.

is “hydrogen” or even “98% hydrogen” e same 

Fig.2: Effect of different impurity gases on the SOS of hydrogen.

is “hydrogen” or even “98% hydrogen” e same 

Fig.3: 4″ Ultrasonic Flow Meter (USM) prototype for H2 measurements from Endress+Hauser SICK