3 replies to this topic

[Once lost now found]

Hi everyone,

 

We have an interesting scenario on site (we're busy commissioning). We have a flare stack that must be able to handle a flow of between 400Nm³ (Min Design) and 10000Nm³ (Max Design for which the flare will still burn). Usual operation should be around 400Nm³ (Operating Point), however, this could experience very rapid (step) changes of around 2000Nm³.

 

At a flow rate of 5000Nm³ the water seal is blown completely out of the flare. The flame is not put out, but neither is the water completely evaporated. Quite a bit of water ends up running down the side of the flare stack and showering the surrounding area. The water travelling up the inside of the flare doesn't seem to do the dust monitor and equipment in the nitrogen purge line any good either.

 

According to the flare vendor this is quite normal (??). The idea, it seems, is that at high flows, the water seal will be constantly topped up (and blown out the top of the flare) so that in the event of the flow decreasing, the water seal will remain intact. This is important since the primary function of the water seal is as a flame arrestor (which of course is only relevant at low flows).

 

The problem is that we cannot afford to waste that amount of water out the top of the flare. The plant will not be in constant operation and so at each start-up the water seal will be lost during the initial high flows.

 

So, my question is this: is it actually normal for a flare seal to be blown out? What can we do to prevent this happening? If there really isn't much that can be done, any suggestions on a control philosophy for keeping the liquid seal intact without blowing the water into the atmosphere?

 

BTW I haven't looked at API 521 yet. There is a height of around 1m above the water level before the stack narrows. If there is a thread that deals with this topic, please do direct me to it - I couldn't find anything when I searched.

 

Best regards,

 

William

 

 

 

 

 

[Bobby Strain]

Looks like you got a bad design.

 

Bobby

[Once lost now found]

Thanks Bobby,

 

That is our opinion as well. However, none of us are specialists in flare design. We need to show this conclusively. The flare vendor currently denies that this is a problem.

 

William

[flarenuf]

William

I’m guessing this is quite a small flare judging by the quantities.  What your vendor says is partially true, that is to say it is normal that for larger size flares when high flows are put to the flare the water in the seal drum will be entrained & “lost” up the stack.  This may be acceptable for a large flare where achieving flow rates high enough to do this is a rare event & of quite short duration, the water is evaporated in the flame & no one worries about it.  Some flares will only manage  10% of their capacity before the water is lost, this isn’t an issue if that 10% is say 100tph & still represents a major (rare) flaring event!

 

However, an experienced vendor knows that you cannot apply the same logic to a small flare particularly one associated with a plant that frequently starts up & flares close to its design rate. 

In this case the WSD needs to be rated (no water loss) for as high a rate as feasible otherwise you lose water all the time.  Thus, as flares get smaller, the water seal diameter may only reduce slightly or not at all.  Basically the  vendor should be aware of the mode of operation of the plant and design accordingly. The vendor should probably not made the WSD so small,

 

imho

brian