This is a pre-print article from my “Cool Fuel” column published bi-monthly in Cold Facts — The Magazine of the Cryogenic Society of America.
The worst safety event we’ve had in the HYPER Center occurred back on a sunny morning in August of 2016 when a bird flew into a substation and knocked out power to campus. Normally this wouldn’t have been an issue but on this particular morning we were prepared to do a liquid hydrogen test and had one of our experiments full. When the power shutoff, so did our cryocooler allowing heat to reach liquid and boil, rapidly pressurizing our storage tanks. Per standard, all of our pressure vessels have dedicated, redundant, pressure relief devices (PRDs) connected to hydrogen vent stacks in case of this situation. However, when one of our PRDs opened, a jet of hydrogen of unknown origin streamed up the back of some of our electronics. Even though none of our experiments contain enough hydrogen to reach the lower flammability limit in the room, the uncontrolled hydrogen release prompted me to pull the fire alarm and get everyone out of the building. In the end, the hydrogen dissipated quickly and the firefighters couldn’t detect any, no fires started, and nobody got hurt. The question burning inside me was: How could we have missed this leak during our semesterly safety testing?
The cause of the leak remained elusive. There was nothing visually wrong and we couldn’t find any leaks with our mass-spectrometer leak detector. So we recreated the event by testing our PRDs with cold helium – and found our leak to be a small hole in the side of a PRD body. The hole, known as a ‘weep port’, is designed to drain water from the downstream side of a PRD to limit ice buildup that could prevent a PRD from properly functioning. Weep port leakage is not an issue with most cryogens, but certainly is with hydrogen. We couldn’t detect this leak with the valve closed during normal safety testing, we had no need for such a hole as the experiment wasn’t outdoors, and the hole was so small (see the figure) that nobody noticed it. How did nobody notice it? It’s the size of a ball-point pen-head and I train people new to plumbing and fittings. To prevent this from happening again, I banned PRDs with weep ports from the lab, switched PRD suppliers, and changed our safety testing procedures to require PRD testing every semester to be signed off by another person. That should’ve been the end of this story.
Seven years later, a supply crunch forced us to order PRDs from different suppliers. Sure enough, during the commissioning of some new PRDs, we found leaky weep ports when operated. However, the students had followed lab policy and ordered valves without weep ports. What happened? The company that makes the PRDs had directly conflicting information on the ordering page — one place says include a P in the part number for the weep port, another says include a P for no weep port. We notified the company right away.
Facing this same problem seven years later sadly means there are likely many PRDs in service unknowingly with unnecessary weep ports that can result in significant leaks when operated. These leaks could be a significant safety concern depending on the type of cryogen used. Since PRDs are not designed for repeated operation, our tests indicate significant leakage through closed valves after less than 5 cycles, they are likely not tested routinely for this type of leakage.
We all know that if you’re in a situation when you are relying on a PRD to operate, the last thing you need is another safety concern, like a hydrogen leak where one shouldn’t be. If you’re noticing periodic leakage from your cryogenic system, this could be a culprit. Please check for unnecessary weep ports on cryogenic PRDs, you may be glad you did.
