Hydrogen vents - Purge with N2 1

[jalvarez]

May anyone in the network recommend me a criteria to define the flow of N2 to effectively (and economically) purge a hydrogen vent? Obviously, I'm trying to minimize the probability of an explosion/detonation, due to lightning or occassional sparks.
I'm speaking of a couple of vents, 16 and a 20", 60 ft height, that occasionally handle hydrogen, 250 pounds/hr, and it is planned to add a Faraday sphere and a flame arrestor on top.
Best regards

 

[Montemayor]

jalvarez:

Knowing only what you've given us, I would not use a Nitrogen source to "dilute the H2 content down to the point where the mixture doesn't ignite. I would, instead, recommend that a small flare be used to ignite and burn the H2, producing a non-contaminating flue gas of pure water vapor.

The reasons I would avoid a "dilution" system are:
1) I suspect it will result as more expensive; the instrumentation requirements will, I believe, be more expensive than a flare;
2) You will need proven and reliable instruments to ensure that the resultant mixture is truly non-combustible; and these instruments will need maintenance and testing periodically;
3) H2 is a very flammable and potentially hazardous compound, but it has its benign side: it burns easily and very environmentally-friendly to produce only water vapor;
4) The controls on a small flare are much simpler and easier to operate and maintain - especially if it is only a vent that occurs from time-to-time and not on a steady basis.
5) When you hold your preliminary HazOp you will inevitably come to the situation where you have to address the scenario where you are venting the H2 and the supply of N2 gives out. This means you are going to have to ensure that the N2 supply is backed up or "fail-safe". You don't have this situation with a flare system.
6) Let's not forget the operating costs: N2 is an expensive utility to have on site; usually it has to be stored as a cryogen liquid with an "across-the-fence" supply contract with outside sources. A flare system doesn't need these type of supply costs - it only needs atmospheric air as an oxidizer.

A 16" and a 20" vent are a huge vent size! You don't state it, but I suspect the H2 is not pure. Otherwise, it represents a huge waste. I also suspect there is more basic data that you are holding back and for that reason I can't offer any other comments or advice. But based on the simple data you've furnished, a flare application is what I would opt for - not a dilution application.

I hope these comments are of some help.

Art Montemayor
Spring, TX

 

[unclesyd]

We use steam on all our process vents as we have variable, not enough to support a flare, flow of H2 at all times. The vents are 15 ft above the top floor of the processing building. All vents have a dedicated ground and are grounded to the supports. They are 4 or 5 vents of this type in a hydrogenation building. There is an occasional fire of short duration but additional steam can be injected on demand if needed.

 

[jalvarez]

Colleagues, as Art request more info, I'm adding the following:
H2 at present is being fired in a couple of small boilers. These vents are ocassionally used (fortunately very ocassionaly) when a control problem appears.
The vents already exist, they are equipped in the top with a temperature sensor and steam injection (triggered by an increase in the temperature) to blow out a flame in case of ignition.
A flare was considered, but the layout implies a very expensive solution. Another limitation for the flare is the existing very low presssure, less than 0.2 psig.
All this maybe explains why I'm asking for a criteria to define a safe flow of N2. Unclesyd suggests steam, but it seems not enough to me. Till now, I couldn't find any recommendation in the literature.

Have a safe day

 

[25362]

I've read of the practice of using fans to dilute possible hydrogen escapes to below a 4% LEL concentration. It appears to me it wouldn't be a totally fail-safe arrangement. There is a site that may be of interest:

http://www.hydrogensociety.net/hydrogen_safety_mesures.htm

 

[EGT01]

If you are insistent on the dilution approach, you may want to look at NFPA 69 covering explosion prevention systems and API RP-521 which give guidelines for flashback prevention in flare stacks.

Like Art said, 16" and 20" vents seem to be huge for the amount of H2 your talking about so these stacks are likely for something that you have not stated. That presents a lot of unknowns as to what other material may be in those stacks but I would suspect that the other material is not flammable if you aren't already concerned about flashback protection in the existing stacks.

It seems you could spend a lot of money trying to retrofit the existing system to handle something for which it wasn't originally intended compared to providing a much smaller, new system appropriately designed. Just a guess, but probably something about 4" diameter is all you need. If you try to retrofit the existing system, don't forget to consider what impact the modifications will have for the other functions of the stack.

 

[pmureiko]

The purpose of nitrogen dilution in your case is to bring the H2 conc. below the LEL. You would be going from 100% H2 to say 5% (I didn't look it up, but it is low). That means you will need 19 volumes of nitrogen for every volume of hydrogen, which works out to a ridculously high flow of 940,000 scfh.

Dilution could be used if the venting was for a very short time since the actual volume used would be small. A flame arrestor at the end of the vent may be sufficient to prevent the flash back you are concerned about.

Because the hydrogen vent rate is small, a separate, small hydrogen vent line could be used. I estimate a 3" line could be used in your situation. This would result in a more economically sized flame arrestor and allow the consideration of adding nitrogen to increase the vent velocity above the flame propogation speed.