8 replies to this topic

[Zauberberg]

What is the purpose of water seal at the bottom of dry gas flare? I have never seen this anywhere before.

It's more hazardous than it appears useful - especially when you think of e.g. flare overpressuring and blowing the flare gas out together with the water seal. Also, you have to check it from time to time and see if water is dripping out from the ground connection or not - and to refill the system as required? It looks quite stupid and hazardous in my opinion, but perhaps I'm wrong.

 Dry_Flare.JPG   70.38KB   130 downloads

[riven]

I too am intrigued.

The only question I could ask is what is the water seal sealing against? And if its purpose is to seal against something, why not use a mechanical/different seal?

[djack77494]

The water seal looks to be a configuration that permits any liquid that may condense within the flare stack to freely drain while still containing the potentially flammable gases within the stack. (I've never designed a flare system, so this is my "guess" based on other seals I have seen.) The seal prevents gases from escaping out the base of the stack because the gas pressure would need to build enough to overcome the height of liquid contained in the seal leg in order to escape.

[JoeWong]

Besides what Doug proposed...

rain water is one of the concern especially in tropical country...

[chemtan]

I have designed Flare systems and evaluated the one under operation...

Emathis has already answered this in detail... Would just add that use of water seal is advised by API, see APIs 520 and 521 to find where\why water seal could be used.

Insurance guys would not want to work with you unless they believe your Flaring and De-pressuring system is correctly designed using recommended practices. And for this, they rely on international practices like APIs (atleast here in Pakistan)

[Zauberberg]

If there is no contact with air, there is no ingress of air - that was my question from the beginning.
This is a dry flare, and I believe an ordinary drain valve at the bottom of the flare (including the level transmitter) would serve the purpose in a more efficient and safe way.

[djack77494]

I do not disagree with what you are saying here, and I am not recommending a water seal leg. Nonetheless, I liked JoeWong's comment of what about rain? I assume that the operations group would prefer to not have to check for liquid in the base of the flare and would definitely prefer to avoid the manual operation of draining off any liquid accumulation. So, I see no problem with the water seal. I do not share your concern with blowing the water from the seal, since this should not occur in a well designed seal leg. Your concern about the need to check and fill the seal leg is definitely valid and should be balanced against the effort required for the alternative. Ultimately, I think it's a matter of personal preference.

[astro]

What is the purpose of water seal at the bottom of dry gas flare? I have never seen this anywhere before.

It's more hazardous than it appears useful - especially when you think of e.g. flare overpressuring and blowing the flare gas out together with the water seal. Also, you have to check it from time to time and see if water is dripping out from the ground connection or not - and to refill the system as required? It looks quite stupid and hazardous in my opinion, but perhaps I'm wrong.

 Dry_Flare.JPG   70.38KB   130 downloads

I'm used to seeing this kind of flare / vent configuration and don't find it a disagreeable design on the surface. That said I would make improvements (see bottom of the post) depending on the risks and finding out further information.

I went searching in my 5th edition copy of API STD 521 and found section 7.3.4.2, 6th paragraph:

In every vent stack installation, careful consideration should be given to two potential problems:
a) accumulation of liquid in lines that terminate at the vent stack

The section goes on further to say:

Accumulation of liquid in lines to the vent stack can result from leakage into the system of high-relative-molecularmass vapours that condense at ambient temperature. If appreciable quantities of liquid collect, they will subsequently be discharged to the atmosphere when vapours are released into the system.

To avoid liquid accumulation, pockets should be prevented from occurring in the lines and the system should be sloped to a low-point drain. These drains can be installed to function automatically by using a properly designed seal. The height of the seal should provide a head equivalent to at least 1,75 times the back pressure under the maximum relief load to avoid release of vapour through the seal. As an alternative to a sealed drain, a small disengaging drum may be installed at the base of the vent stack. This type of installation is recommended where significant quantities of liquid can occur.

My comment to improve the design of the P&ID presented would be to include a local level gauge and consider instrumentation to address high level events due to rain or other causes (which need to be determined from evaluation of the normal & non-steady state operation and emergency scenarios via HAZOP or some equivalent technique). The action of the instrumentation being commensurate to the assessed process and safety risks.

There are downsides to any design selection. By adding some mechanism to prevent one problem, the result is that often another problem has been engineered into the design. Seal legs make a lot of sense because they offer a low ops/maintenance design but because of this they can suffer neglect and degrade to a latent failure mode (e.g. blocked seal leg). Other more "active" designs provide a counter position. Ultimately it's a balance of risk and your ops/maintenance philosophy, which the Engineer at the work face is in the best position to assess and understand.