Hi
I’ve been asked how far below the ceiling an Oxygen monitor should be in an NMR room to detect reduced Oxygen levels resulting from a release of Helium.

We have a room set aside for the NMR machine and nothing else. There is a low level Oxygen alarm to detect the release of Nitrogen at about 1M from the floor. Currently the low Oxygen alarm for detecting the release of Helium is at head height. However this is about 1.8M below the ceiling which is very high in this part of the building.

The detectors are attached to a high powered fan that will exhaust all the air/Nitrogen/Helium out of the room if they detect a decrease in the Oxygen level in the room. My view is that the low Oxygen level for detecting Helium release should be higher to give as much warning as possible of a reduction in the Oxygen level in the room.

I’m very cautious about this room since it is by definition a confined space. Does anyone know of any standards or guidance that I could consult to help inform my decision.

Thanks

Graham Watson

Seems to be:
Nuclear Magnetic Resonance (NMR) Room

Graham
You have a point about the height, but I think you need to look more deeply at the events that can happen in the room.

How much nitrogen is in the room at any time? Could it cause a serious problem? Under what circumstances - normal boil off? Dewar failure? Apparatus failure?

How much helium? What will be the source of helium at high level? If it is boil off, then how likely is it that it will fill the room at high level to cause a danger? Would this only happen if the machine were faulty and boil off was unusually high? If so, then the occupants would notice long before the alarm went off.

Is a quench the only event that could release this gas in a large quantity? If so, how fast does you apparatus release helium in a quench? This varies a great deal between machines. If you want to see a really scary quench look at this video http://mrisafetyvideo.co...nch_outside_shortSo.htm. It has been filmed from outside the building because this machine was correctly fitted with a quench pipe so that helium does not get released into the building. The quench starts at approx t=10s. A machine that quenches as violently as this should be vented directly to the outside air. It should NOT vent into the room and then wait for the fan to start up and suck out all the helium.

A quench is going to be noticed by the room occupants, even if it is a mild one. They should not stay in the room, but ventilate it and leave while the air balance recovers.

Have a look at the way the forced ventilation has been configured. I encountered one where the inlet for the air to be removed from the room was at knee height. This meant that if there had been a helium release, all the good air was removed from the room first, drawing down the helium last. Not good.

PM me if you want to discuss further.

I have been watching this matter discussed on another forum and have seen the following comment posted

I know Graham will have seen it but others may have not

“Also, be aware that if you are monitoring for oxygen depletion as a result of Helium escaping you shouldn’t use a standard O2 depletion sensor. I can’t recall the detail off the top of my head, but it has something to do with the very small size of the Helium molecule interfering with how the sensors detect Oxygen; the sensor somehow end up with artificially high levels of O2. Hence there are helium specific devices to use."