11 replies to this topic

[KeJoSa2010]

There was an issue that was identified in a HAZOP meeting where operations at a facility are seeing a small vacuum pressure in the flare system when no venting is taking place (i.e. purge only)

 

There was a disagreement with a more senior engineer regarding vacuum pressures in flare systems. It was said that it is acceptable to have a vacuum pressure in the flare system "as long as the velocity of the purge gas is maintained at or above the flare stack manufacturers minimum flow rate at the velocity seal". It was explained that the reason for the vacuum in the header has to do with buoyancy effects and the fact that flare gas is less dense than air.

 

This is the first time I have ever heard this and it makes no sense. The only flow I see at the flare tip when there is a vacuum in the flare header is the air ingress back into the header. I see this as a major hazard. 

 

The answer to this question seems obvious to me but, I am would like to hear others thoughts on this.

 

Thanks,

Edited by KeJoSa2010, 10 March 2017 - 07:32 AM.

[Saml]

Your senior engineer is right.  In point of fact, senior engineers tends not to give an opinion if they are not sure. So I would say that there is some hubris on the way this was handled, and that this not belong to good decision making process. Please try to understand his point of view.

 

It works in a similar fashion than a natural draft heater stack.

 

At the tip, the pressure is atmospheric.

 

Let's say that the gas you are filling the stack with is methane. With MW of 16, the density is around 0.65-0,70 Kg/m3 depending on temperature (let's take 0.70 for this explanation)

 

Outside you have air, with a MW of 29, the density is around 1,2 to 1,3 (let's take 1.25 for simplicity)

 

Let's say that your stack tip is 50 meters above ground

 

Let's also say that you have a pressure of 100 000 Pa at 50 meters above ground (at tip's height)

 

The weight of air adds 1.25 Pa per meter. So at ground level your atmospheric pressure is 50 x 1.25 x g = 613 Pa higher than the pressure at 50 meters. That is 100613 Pa (*)

 

The weight of 50 meters of methane adds 0.7 Pa per meter. So at ground level, inside the stack you have 50x0.7xg = 343 Pa higher than the pressure at 50 meters. That is 100343 Pa. Actually it will be a little higher since you have to add the frictional pressure loss of the small purge flow.(*)

 

Both pressures are higher than the pressure at the tip, but the pressure inside is lower than outside. That difference is what you see as vacuum at ground level. It is about -270 Pa or a little more than an inch of water.

 

However, at the tip, the pressure is the same both for the gas and the air,  and you should be OK if you maintain the minimum velocity recommended by the flare vendor.

 

Anyway, be aware that if you have any opening in the flare header (like drain points kept open, or a PSV removed for maintenance with a leaking thru block valve) you will have air air going into the header, and that this leak is not easy to detect.

 

 

(*) this assumes that in 50 meters the density and temperature does not change significantly.

Edited by Saml, 10 March 2017 - 02:24 PM.

[Bobby Strain]

But, in fact, your senior engineer is wrong. You are correct in your thinking except the problem of air ingress is not at the tip, but rather via leaks in the flare header system. But you should seek some authoritative source to support your view. Engineering should not be conducted on an opinion basis. Or a survey via this site.

 

Bobby

Edited by Bobby Strain, 10 March 2017 - 02:24 PM.

[KeJoSa2010]

Thanks for both of your responses.

 

Bobby, you are absolutely right. It was not my intention to send out a survey of opinions on the topic to justify my own view. I want to get it right. If I am wrong, then I welcome the correction. I think others opinions are useful to at least point me in the right direction when I have questions (which you have done). Thanks.

 

Saml, I understand the argument I just don't think that it is quite correct. If you don't maintain a positive pressure in the header at ground level, what would prevent air ingress from leaks in the flare header system? The purpose of purge gas is to maintain a positive pressure in the flare header to prevent this scenario.

[Bobby Strain]

You probably can't afford the fuel gas required to maintain a positive pressure. A water seal provides a means to maintain a positive pressure.

 

Bobby

[Saml]

Bobby, you are right. A water seal is probably the best option.  We don't know what they are flareing, but if it is hot, or condensing, after the relief, the cooling down/condensation can produce enough vaccum to exceed the purge gas flow and air will come in (this is mentioned in API 521 with some detail)

 

I've personally seen an explosion caused by air ingress on a flare header that operated under slight vacuum. The sight of a rain of refractory bricks pieces coming from the flare tip is not nice. Nobody was hurt, luckily.  This was caused by a missunderstanding during a maintenance removal of a PSV.

 

That flare operated from 1968 to 1997 without any incident. There was a mistake in the maintenance procedure? yes.  Can be operated this way. Also yes, In fact, it operated more than 30 years in a pneumatic control, non PSM world. However, something that is one simple error away from an explosion is not normally a good design.

 

That is why I mentioned this fact at the end of my post.

 

However, the issue is  that you can indeed have a flare operating with vacuum without air coming back thru the tip. That was the original question. Obviously, at ground level, if you make a mistake or something does not close tightly you will have air inside, and will have a risk of explosion.

Edited by Saml, 11 March 2017 - 09:27 PM.

[Bobby Strain]

Often people ask the wrong question. It's up to seasoned engineers to respond correctly. As to the dangers of rapid condensation, there are control systems to dump large quantities of gas into the header to maintain positive pressure. Water seal alone won't provide proper protection.

 

Bobby

[paulhorth]

If you purge the flare header with nitrogen, you eliminate the chimney effect, and the risks that come with it, You also eliminate the adverse environmental impact of methane purging. The user would need to assess whether these benefits would be worth the higher cost of nitrogen purging.

 

Paul

[Fakhri]

I think this is because the default value for the atmospheric pressure in Flarenet is a small negative pressure. Please check and change it into 0.0.

 

Regards,

 

Fakhri

[Zubair Exclaim]

If you have a water seal drum or even better a molecular seal ,,, you should be safe ... 

as far as purge gas goes ,,,, its intended to create a positive back pressure higher than atmospheric  ... not essentially a velocity criteria 

[Bobby Strain]

There is some bad advice given. So, you need to research to make sure you don't do something dangerous.

 

Bobby

[KeepLearing]

how is the choice of an elevated flare over a ground flare ( or vice versa) affecting the built up back pressure on PSVs ? which type of flare is better to reduce the built up back pressure on PSVs?

Thanks .

Karim