Left: A closeup of a sensor element. Hydrogen sensor prototypes were installed at this refueling station in Burbank (right) and others to test how well the sensors detect hydrogen concentrations. Click to enlarge.

The SCAQMD project has been led by engineer Amanda Wu at LLNL and by Brosha at LANL, and together they have seen the long effort come to fruition by field demonstration.

Hydrogen is the smallest of molecules, making it notoriously difficult to contain, and prone to leakage. The point of field trials is to demonstrate the ability of our sensors to detect hydrogen consistently and reliably. We’re seeing if they can outperform current, commercially available sensors, with improved hydrogen detection at levels of interest, robustness and durability.

—Amanda Wu

Researchers want to incorporate the sensors into the refueling stations as a safety device that would trigger an emergency shutoff or prompt an alert to when detection levels reach 2%. Eventually, Glass said, the sensors also will be adapted for installation within the passenger compartment and in critical areas near the fuel supply line in vehicles to detect hydrogen leaks.

Current commercial sensors, the researchers said, need frequent recalibrations and can generate false alarms, which lead to shutdowns and unnecessary, costly response from fire departments. The advantage of the new sensors, Glass said, is that they can be made low-cost, with higher durability and reliability than current commercial sensors, which have proven to be error-prone in practice.

Instability, as indicated by baseline drift, has been a long-term problem with commercial sensors. This technology is more stable and is designed to show less drift over time. Ultimately, we want to use these on a hydrogen fuel cell car, with even more demanding performance requirements, and as a first step we're using them at these refueling stations.

—Bob Glass

So far, according to Brosha, the sensors have shown they are sensitive enough to detect even minute levels of hydrogen released during normal refueling operations while screening out airborne pollutants, hydrocarbons and combustion byproducts, and smog and smog-forming agents present in the atmosphere.

We’ve been collecting data continuously and we’ve proven everything we’ve claimed our sensors could do with regard to baseline stability. We’ve never had a single false alarm. We’re able to detect the day-to-day activity, and we’ve never seen a peak that wasn’t the result of a fill activity. We have a strong indication they’re performing very well.

—Eric Brosha

SCAQD and the two national labs are funding the trials. Dan Poppe of Hydrogen Frontier Inc. owns and manages the Chino and Burbank stations and is facilitating the on-site demonstrations.

The project, an outgrowth of the second developmental phase of a Department of Energy project that ran from 2008-2014, ends in January. Following the test run, Brosha said the sensors would be brought back to LANL to test their durability and performance and attempt to show experimental data with little to no drift in sensor calibration resulting from extended operation in the field.

Currently, there are about two dozen retail hydrogen refueling stations operating in California, with about 20 more in development and expected to come online in the future. The researchers hope to expand to other stations in the next year, and to explore augmentations on the current design and electronics platform, according to Wu.