12 replies to this topic

[gazepdapi1]

I wanted to know what you guys do to prevent water from collecting next to the relief from the discharge piping? Usually we put the caps with the shock cords but this is a pain to do especially with reliefs that discharge above the building limit. I think just a weep hole (1/2") would be just fine to prevent water from collecting but this wouldn't work as well during hail storm or a good snowfall.

[Technical Bard]

In a plant I worked in 20 years ago we had some large PSVs (10-12" discharge lines) that were protected against rain/snow by placing a simply plastic pail (5 gal / 20 litre variety) upside down over the stack. In the event of relief the pail would "fly". Not sure if that would pass modern plant safety practices, but it was simply and effective.

[gazepdapi1]

Well this is what we have on some of our systems now and they work well when the relief is short. When it does relieve, someone has to go back and put the pail back on but on reliefs have vent above the building height, this is a pain. This is why I'm looking at some other possibilities.

[paulhorth]

Nertil,
I don't really understand what kind of reliefs you are talking about - presumably non-flammable and non-toxic? But to keep rain out, why not fit a 180 deg pipe bend on the end, so that the outlet faces downwards? This feature is seen on many storage tank vents, such as filling station underground petrol tanks, etc. This will depend on the relief discharge not creating a hazard below.

Paul

[gazepdapi1]

The reliefs are relieving natural gas the a compressor goes above the MAOP. THe 180 degree pipe would would create a hazard if there were people right below it.

[Dacs]

Since I have yet to get involved in a location that involves snowfall, I'll just take a jab on this issue.

Can a rain cap does the job (similar to this one)?

Attached Files

•  raincap.JPG   8.65KB   13 downloads

[gazepdapi1]

I though of that but unfortunately the force of the gas is so high that it blow the cap right off.

[paulhorth]

Nertil,
Natural gas from a compressor discharge? Rain is the least of your worries. The discharge of a natural gas PSV direct to atmosphere, in the vicinity of a building, would be prohibited by the standards of all the international oil companies that I have worked for,.
How far away is your nearest source of ignition? What would prevent a flammable mixture being drawn into the building by the HVAC?

Paul

Edited by paulhorth, 07 September 2012 - 10:14 AM.

[Lowflo]

Paul - This is a case where you can't trust your intuition. PSV releases of flammable vapor pose very little risk if the tailpipe is pointed upward into an open area. The VCE risk is extremely small, almost nonexistant. Your comments about flammability hazards are just like those i used to say, until someone turned on the light for me. API 521 has been saying this for years, but I wasn't a believer until I ran a series of test cases in PHAST. As long as the velocity is 40 ft/sec or more, you'll induce enough air into the stream to keep it safely diluted below the LFL. Like API says, the flammable envelop will have a predictable characteristic pattern (like what you get from a burner tip). As long as it's released vertically into an open area, the only risk is radiation. I'm not saying radiation hazards are trivial. That can be quite severe, depending on how much gas is burning. But, unlike being blown up, one can usually escape from radiation exposure, or mitigate the hazard by elevating the tailpipe.

As for preventing ice plugs in PSV tailpipes, this is a topic I'm keenly interested in. I've heard nothing but complaints from folks who used tailpipe covers/socks. Regardless of the brand, they all seem to get dislodged and require repeated reinstallation. We put a weep-hole in all tailpipes, and I think most of our cold-weather sites rely on just that weep-hole. That's been a continuous source of worry for me, especially for the plants in Canada and northern Europe. Fortunately, we haven't had any incidents. Probably the most effective solution is to heat trace the tailpipes, but that's a costly solution.

So, I don't know a good solution. I'd like to hear one. In the meantime, I continue to recommend the use of a tailpipe cover or heat tracing, while telling them honestly that tailpipe covers are likely to be a nusiance.

[gazepdapi1]

Lowflo, I agree with what you're saying. Honestly, my recommendation will be to cut the stack to a height they can reach to put the covers back on but also have the weep hole. Another idea is to just have a tee installed on top of the tailpipe with pups on either side cut at a 45 degree angle to prevent water/snow from getting in.

Also, does API 521 state how far reliefs have to be from the compressor building or if the vent stack discharge has to be above the building?

Edited by nertil1, 07 September 2012 - 08:15 PM.

[paulhorth]

Lowflo,
Thanks for your response. I respect your views, your posts are always more clearly written and more accurate than my own. I have read the relevant section in API 521 and understand the technical justifications for relief to atmosphere. However, the practice (if not totally prohibited) is strictly limited, and is subject to stringent technical study and authorisation, by a number of international operating companies.
I quote extracts from three standards which I have at hand today.
Company A
Unless it can be demonstrated an alternative is safe beyond all reasonable doubt, all discharges shall be contained in a common gathering and disposal system. The aim here is ensure any discharge is routed to a safe location rather than local to a relieving device. Such containment is mandatory for systems containing potentially hazardous levels of H2S or where hazardous flammable mixtures could be produced.
In some circumstances it may be permissible for a relief valve in hydrocarbon service to relieve locally provided the requirements of API Standard 521 are met but this requires a dispensation.
Where hydrocarbons are released in this way the project / contractor / asset shall demonstrate the dispersion from the outlet point does not cause a hazard (e.g. no accumulation of flammable or toxic mixtures in locations which could cause harm). The possibility the release might ignite shall be taken into account in the design, particularly in terms of ensuring the point of release is at a safe location and any equipment or structures which could be affected by thermal radiation.
Company B
Wherever possible disposal streams shall be collected in a closed system and directed to a
flare or vent, except when they can be sent back into the process or stored.
(followed by a list of strict criteria for when atmospheric discharge is permitted, and verifications of dispersal, risk etc).
Company C
Atmospheric venting for flammable and/or toxic gases in Company C shall be eliminated or severely curtailed when practical. (followed by seven pages specifying the required dispersion analysis, consequence analysis, radiation analysis etc, to support the venting, and including these distances : minimum 30 m to any source of ignition, minimum 25 m above grade, minimum 3m above any platform or structure.) Any deviation from these requirements requires approval of the Company Director of Engineering.

These companies might well close down Nertil’s facility tomorrow, until the required demonstrations of safety have been prepared.

Relief valves do not always operate at full flow, they can leak, and pass a small flow of gas which might not be dispersed. Would you bet your life on this never happening?
Paul

Edited by paulhorth, 08 September 2012 - 08:33 AM.

[Lowflo]

Paul,

I understand the skepticism. Based on my experience, I find this to be counter-intuitive for most engineers, even very experienced ones. I was very experienced when I learned this. I was doing a project for an ethylene plant. The plant had a flare header but the C3 splitter PSVs were routed to the atm. The relief load for this column was huge, which I learned was typical for C3 splitters. I think the rated capacity was 300000 -400000 lb/hr. Based on that flowrate, and C3’s being heavier than air, I thought this was a disaster waiting to happen. But an engineer in the plant said that the risk was minimal due to the exit velocity.

We ended up routing this and other PSVs to the flare header, but I was intrigued by the comment that, due to the velocity, this was not a significant safety risk. I ended up doing a series of case studies using a rigorous gas dispersion analysis tool. It was an eye opener for me. Even when I reduced the exit velocity to 10% of the normal value, the LFL envelope never dropped below the discharge elevation and it maintained the characteristic shape (tall due to the mass flow, but very narrow in width). The bottom line was that this was an environmental risk but a trivial safety risk. Just for grins, I ran the model using even heavier flammable gases, and I found the same result. The take-away message for me was that we need to focus on preventing the release of flammable liquid – that’s always a risk of VCE. But, flammable vapor releases for PSVs are not a big risk, unless they’re poorly engineered (discharged too close to an elevated structure, or into a confined area, or in a direction other than vertically upward).

[paulhorth]

Lowflo,
Thank you, I appreciate your comments. I acccept what you say, as you have done the work to check the dispersion profiles. You have convinced me that an atmospheric discharge system can be designed which meets the various risk criteria. Modern practice ( as defined in the company standards) requires this to be proved, bit if it is demonstrated, then there is no objection..

Paul