Lessons Learned: Hydrogen Release Due to Laboratory Gas Header Manifold Mix-up

Summary

On April 18, 2022 the hydrogen alarms were activated in lab C323, the Fuel Cell Development and Test Lab (FCDTL) in NREL’s Energy Systems Integration Facility (ESIF). Researchers were attempting to use house air to dry out a piece of laboratory equipment, however they had accidentally connected a line to hydrogen (H2) supply instead of house air. The H2 and house air valve drops are labeled and located next to each other. The lab area is not a intended as a continuous hydrogen environment.

It was estimated that the H2 was flowing into the room through 1⁄4” tubing with some valve restriction for 2-3 minutes at approximately 70 psi before the hydrogen alarms activated.The lab went to Emergency Power Off (EPO) status approximately 30 seconds later. There were no injuries or equipment damaged associated with this event.

Event Details

At approximately 4:00 pm on April 18, 2022, a hydrogen (H2) release occurred in the ESIF FCDTL, lab C323. A researcher (R1) and a research technician (RT1) were preparing to use the lab’s compressed air to dry a variable area flow meter (Rotameter) that would be used for controlling the flow of nitrogen (N2) into the gas slipstream of the Crossover Flow Cart that they were constructing. RT1 intended to connect the input of the rotameter to a compressed air connection on one of the lab’s equipment gas distribution manifolds.

The compressed air line had previously been disconnected and left in place for an undetermined future use. RT1 accidentally connected the compressed air tubing to an H2 supply line immediately adjacent to the compressed air connection. The gas lines on the manifold were labeled, but the height and location of the labels made them difficult to read and the similarity of the labeling itself made the connections difficult to distinguish.

When the RT1 opened what they thought was the compressed air supply valve, H2 flowed through the process piping and rotameter and exhausted into the lab space. Approximately three minutes after the H2 began flowing, the lab H2 High alarm was activated for the laboratory. A few seconds later, the H2 High-High alarm activated. The lab emergency power off (EPO) activated approximately 30 seconds after the H2 High-High alarm, shutting off all electrical power and hazardous gas flow to the lab, including the house hydrogen. When the alarm sounded, personnel immediately evacuated the lab. RT1 stopped, turned off the H2 flow valve, and evacuated the lab.

Readings obtained from the room’s gas detection system (VESDA ECO Aspirating Gas Detector) indicate that the H2 levels reached a maximum level of 2.1% by volume of air, which is below the LEL for Hydrogen of 4% by volume of air. H2 levels in the lab started to drop as soon as the EPO was activated. The VESDA system acting in concert with the automatic EPO of the laboratory, functioned as designed and prevented an explosive atmosphere from developing within the lab space.

After a safe level of H2 was indicated by the VESDA system, the area work supervisor (AWS) and a research technician entered the lab and used a four-gas monitor to confirm hydrogen had stopped flowing and verify levels within the lab had reached a safe level. Once the researchers were able to enter the lab after the event, the abandoned gas line (a combination of 1⁄4” stainless tubing and flex tubing), previously in use to connect a stationary piece of equipment, was disconnected from the H2 connection on the manifold and the H2 line was capped at the gas manifold. Work on the Crossover Flow cart was paused until a safe path forward could be determined.

Lessons Learned

The lessons learned below are aligned with one of NREL’s eight safety principles:
Hazards are identified and evaluated for every task, every time. This incident highlighted an opportunity to improve the visual and engineering controls related to the house gas system.

The lessons learned are as follows:

1) The gas line was connected to the wrong supply without verification of connection. This was an error that would have been detectable and correctable if a check of the completed, or partially completed, work was performed.

2) The Gas Line Labeling did not comply with ANSI/ASME A13.1 or SEMI-S2, making identifying the gas in each line and their hazards hard to determine.

3) Valve handle color is the same for all gases on the manifold, making identifying the hazard associated with opening the valve challenging to determine. Similar controls were indistinguishable from one another.

4) No devices on the ESIF house hydrogen supply are present to limit flow in case of a catastrophic leak or valve misalignment/mix-up.

Actions Taken

Actions taken in response to the lessons learned include:

1) Labeling the gas lines on compact equipment manifolds in the service aisles of ESIF Lab C323, consistent with the ANSI/ASME A13.1 standard.

2) Changing the valve handle color on valves to comply with the ANSI/ASME A13.1 standard. See before and after below.

3) Engineering and installing a restrictive flow orifice (RFO) and/or an excess flow valve on the house hydrogen system feeding ESIF Labs. It is very unlikely a flammable atmosphere can be achieved utilizing an RFO due to the normal ventilation in the lab.

4) Developing Lessons Learned Article

SME for additional information: Emily Gogel, Emily.Gogel**At_Symbol_Here**nrel.gov

Please email the NREL Lessons Learned Program with questions or for additional information: lessons.learned@nrel.gov.

Audience: Public

NREL-FY22-S-005-SEPT

---
For more information about the DCHAS-L e-mail list, contact the Divisional membership chair at membership**At_Symbol_Here**dchas.org
Follow us on Twitter @acsdchas