In order to combat human-caused climate change, the emission of greenhouse gases from the transportation sector must be addressed. However, to make a meaningful impact on emissions in a market with such a large number of distributed point sources poses a daunting challenge. Battery electric vehicles are a viable commercial option, but consumer demand for range and convenience have prevented their widespread adoption. Hydrogen fuel cell vehicles (FCV’s) offer the high performance of battery EV’s, but with the driving range and refueling times that most consumers demand. Presently, California leads the nation in the introduction of hydrogen filling stations and sales/leases of FCV’s by major automobile manufacturers. The California Energy Commission has been tasked with implementing 100 hydrogen fueling stations within the state with a near term goal of 50 hydrogen fueling stations open to the public by the end of 2017.

Hydrogen is highly flammable, stored at high pressures, and quickly propagates if released into the air. The introduction of odorants to provide an automatic mechanism for human detection of leaks is not possible because these chemicals will damage the fuel cell stack. Yet hydrogen must not accumulate to concentrations greater than the lower flammability limit (LFL) of 4 vol%. Because the risk of a hazardous event involving hydrogen can be mitigated through use of hydrogen safety sensors, these devices are recognized as a critical component in the safety design for any hydrogen system whether production, distribution as in filling stations, or as part of the consumer product itself.

The number of available sensor technologies is quite great and most mature hydrogen sensor technologies can be categorized into a relatively small number of platforms. The US DOE has published target specifications for hydrogen safety sensors for stationary and automotive systems. In order to understand the capabilities and limitations of the existing sensor platforms, the DOE created the Hydrogen Safety Sensor Testing Laboratory at the NREL. The findings from extensive sensor testing show that there are gaps in performance in all presently available sensor platforms and that none of the commercially available hydrogen sensors meet DOE requirements for hydrogen safety sensors and the needs of real-world applications. The principle issues are that (1) no one technology appears to meet all of the DOE requirements and (2) there is a clear absence of technology offering superior performance in the 1-4vol% H₂concentration range which does not suffer from drift that may lead to false positives (a costly nuisance) or false negatives (potential for loss of life and property). The frequency for the occurrences of false responses can be mitigated by frequent testing and calibration, which adds operational expenses. To address sensor limitations, the US DOE began sensor research programs at Los Alamos and Lawrence Livermore National Laboratories in 2008 to develop a new safety sensor platform to precisely fill the critical performance gaps identified in the existing commercial platforms.

In this work, zirconia mixed potential-based electrochemical hydrogen safety sensors were field-tested at hydrogen filling stations in the Los Angeles area over a two-year period. Two versions of the sensor – one with a tin-doped indium oxide working electrode and one with a strontium doped lanthanum chromite working electrode – were tested in four rounds of field trials experiments conducted at hydrogen stations located in Burbank and Chino CA. The sensor systems detected numerous hydrogen releases over two years of testing. These events were localized in time and always easily distinguished from the sensor baseline behavior. No evidence of sensor baseline drift was detected during extended baseline stability testing at the Chino station. No false alarms or false positives were encountered during the testing program at either station. We found excellent correlation of recorded H₂ release events with customers filling their FCVs or related to normal station maintenance activities. Multiple fill events were recorded in filling station logs and sensor resolution was sufficient to identify fuel cell vehicles queued to refill. Multiple refueling events could be identified in the sensor data even when these actions took place in less than 15-minute intervals. Moreover, sensor sensitivity was sufficient to detect hydrogen permeating from dispenser components at the Burbank station during the refueling activities. There was confirmed exposure of the sensor enclosure to significant and sometimes severe weather events, with no discernable, deleterious effects to the sensor. The installation of two field trials units at the Burbank station together with the addition of a weather station permitted correlations to be made vis-à-vis hydrogen detection and prevailing wind direction at the outdoor station.