Low-cost sensors to support hydrogen for heating
1.0 Background
Hydrogen for heat has been the subject of extensive research in the UK and across Europe as gas transmission and distribution network operators look to reduce carbon emissions. Notwithstanding this, one of the major barriers to developments has been public reluctance to trial the use of hydrogen at any meaningful scale.
DNV’s UK Energy Transition Outlook for 20251 forecasts that electricity is set to provide the majority of domestic heat in the long term with gas use set to decrease, but there is some uncertainty in this with the imminent decision from the UK government. Nevertheless, natural gas will be needed in the interim, with nations across Europe looking to introduce low levels of hydrogen to domestic supply.
Hydrogen has a wider flammable range than natural gas – i.e. a greater range of concentrations in air that will sustain combustion. A build-up of flammable gas to between these limits has the potential to ignite and may lead to explosion. The consequence of an explosion correlates to the laminar burning velocity, which can be compared to other typical fuel gases.
Figure 1 (Left) The relationship between fuel concentration and laminar burning velocity and (Right). The relationship between burning velocity and hydrogen concentration in a hydrogen/methane blend.
Figure 1 (Left) shows that hydrogen will burn over a much broader range of fuel content. The higher the burning velocity of the fuel, the greater pressure produced during combustion and hence the greater severity of a confined explosion. Figure 1 also indicates at low hydrogen concentrations the burning velocity of hydrogen is similar to natural gas. Figure 1 (Right) shows the effect on burning velocity of blending hydrogen into natural gas, illustrating at low hydrogen concentrations there is little change to burning velocity. The challenge therefore is to prevent the accumulation of hydrogen within confined spaces, to avoid the build up to higher hydrogen concentrations.
Harm to residents generally occurs when leaks go undetected and a flammable atmosphere can develop. Early gas detection and warning systems would provide an additional safety measure alerting residents and helping to reduce the number of incidents involving undetected gas leaks. The properties of hydrogen potentially allow for the re-use of readily available gas detection equipment with little to no additional cost.
2.0 Hydrogen Detection Systems
There are several options to reduce the likelihood of hydrogen accumulation within properties. The first of these is by increasing ventilation. This might be suitable for some external applications but will meet greater resistance when applied to people’s homes due to the inevitable cold draughts.
A second option would be gas detection. No such system is currently recommended practice for natural gas in residential properties, with detection relying upon the occupier smelling the odorant added to natural gas. Gas detection might reduce flammable gas incidents beyond the reduction achieved through odorisation alone. Atmospheric natural gas or hydrogen monitors are available to purchase but their cost limits their use to industrial applications. Carbon monoxide (CO) alarms however are low cost, and in some regions such as Scotland mandatory in domestic properties.
CO alarms are often based on electrochemical reactions happening when carbon monoxide encounters the sensor. This creates an electric current within the detector which triggers the alarm. The same reaction occurs when hydrogen comes into contact with the sensor. The question is, how well do these sensors detect atmospheric hydrogen?
3.0 Carbon Monoxide Alarm Testing
To provide some confidence in the application of this method, a series of tests were carried out by DNV to assess the response of common CO alarms to a range of hydrogen concentrations in air. The test set-up mimicked the installation of a CO alarm in a domestic property. The sensors were placed inside a test chamber and supplied with a controlled blend of hydrogen in air. Several tests were repeated to determine an effective “calibration curve” showing the equivalent carbon monoxide concentration for increasing hydrogen concentration.
CO alarms are designed with three alarm set points and must alarm faster in high concentrations than when low concentrations are detected2. Sensors from three leading brands, as packaged for normal operation for CO detection, were exposed to varying hydrogen concentrations and the preliminary results are shown in Figure 2.
Figure 2 Preliminary hydrogen calibration curve results. Figure 2 shows the response from three leading manufacturer’s devices to increasing hydrogen concentration. Two of the devices responded rapidly to increasing hydrogen concentration showing a high cross-sensitivity. The third was slower but was still predicted to meet the high alarm setting when exposed to a hydrogen concentration of approximately 2800 ppmv, or 0.28% hydrogen in air by volume. This equates to 7% of hydrogen’s lower flammable limit.
5.0 Discussion
Initially, it is anticipated that hydrogen blends in natural gas will be a step on the road to a transition to full hydrogen systems. While the tests were performed on pure hydrogen leaks, the detection of low concentrations means that these detectors would still respond rapidly to a concentration well below the flammable limit of a blend of 20% hydrogen in natural gas. Further work to modify the alarm set points would improve the case for using CO alarms for domestic hydrogen, or blended hydrogen leak detection.
It is important that further research be conducted to assess the suitability of all available alarms for this purpose and whether long term or repeated exposure to hydrogen could reduce sensor lifetime and help develop standards for these devices to conform to.
6.0 Conclusion
The study provides confidence that a low-cost and reliable means of detecting pure hydrogen or blended natural gas leaks is possible and could contribute to reducing the risks associated with hydrogen for heating. The use of CO alarms as early warning systems for the presence of hydrogen does require further study to broaden the range of detectors tested and understand the overall spread of responses. It is also important to develop more information on measurement errors and uncertainties and any long-term effects of repeated or continuous exposure.
This study could also lead to development of combined safety systems, where an audible alarm or other signal type could be used to automatically shut off the supply of gas to domestic properties. This type of system would minimise the inventory of a gas leak and help to prevent further accumulation.
The functionality of carbon monoxide alarms as early hydrogen leak warning systems could help to ease the public concern around the use of hydrogen. The fact that a device they already own (or similar variation to) might reduce the risk associate with the gas use might alleviate the concerns they have raised.
Further work is required to verify and certify these sensors for hydrogen use but the preliminary results suggest an affordable solution to a significant and complex decarbonisation issue.
Authored by Adam Robinson, Consultant, Low Carbon Fuels.
References
29/04/2025 10:34:00