Purge velocities in Horizontal flares 3

[Bala0610]

Hi

Is there any publication / guideline / industry practise for purge rates to be maintained in Horizontal flares (always lit)? Are any provisions made to address change in direction of wind /gusts in such conditions ?

Thanks

Bala

 

[Flareman]

Bala016
Most horizontal flares are associated with a liquid Burn Pit to allow wet gases and liquids to be discharged at the same point.
Unfortunately, horizontal flares are not really addressed too well in any "official" publication. API-537 has 2 lines which only acknowledge they exist.
Part of the problem is that the standards are essentially tailored to US refineries which do not generally use this design approach.
I have a draft of a "standard" approach [BURN PIT CRITERIA] which I will post on the website

http://www.geocities.com/flareman_xs

(navigate to Main Index | Downloads) within a few of days of this post. This paper will contain some formulae to facilitate approximation of the radiant loads and flame positions over the liquid pool.
The horizontal tip itself, when burning gas, produces a wind-blown flame which is not covered in any of the API flame models. However, another paper [..COMPREHENSIVE MODEL FOR FLAMES AND PLUMES] has formulae to approximate flame position based on various wind speeds and directions for flares at an angle to the vertical (such as horizontal). It’s a bit of a slog but only needs a systematic approach. You may find these useful.

If there is any component of the wind which applies a pressure to the end of the discharge pipe, it will influence the outgoing rate of flow. A 30 mph wind has a velocity pressure of 0.016 psi. Applied suddenly, directly and squarely to the end of the flare line, it could drive the air/gas interface into the header, compressing the contained volume. Assuming a header length of 3000 ft, the interface could move back by between 3 – 4 ft independently of the gust duration. If there is a concurrent purge rate of (say) 0.1 fps it will take roughly ½ minute to restore the interface to the end of the pipe. In the grand scheme of things, this is not a disaster. Winds are usually not so commonly high at grade level, nor always in the worst direction, and the burn pit should have some kind of containment wall (bund or berm) which can mitigate the effects of the wind on the pipe.
In a vertical flare, you always have to monitor the minimum rates for burn-back (formula on the stated website) and horizontal tips are no different. My best recommendation would be to use the rate for the equivalent vertical tip and apply a small FOS (say 1.5), and don’t go below 0.1 fps for flammable gases or a recovery time of about 1/2 minute.
In any event, at low rates, the flame will burn upwards against the top of the horizontal pipe and that will affect its longevity according to how well protected it is (design concept again !) Any sporadic wind gusts and direction changes will simply distribute the damage over different parts of the tip. In this respect, slightly larger rates are probably worse than minimum, because the flame lick on the leeward side or the top may be more widespread.
The other similar consideration is volume shrinkage of the contained gases due to temperature changes or condensation in the header. These produce much greater interface movements and usually need a small study to determine the probabilities of occurrence, and the appropriate volume replacement rate in each case. Often, instrumented solutions are needed where this is a high probability.
One way to assist is to isolate the tip discharge from the rest of the header by using some sort of seal. Most commonly, this is a water seal or a relief valve although I realize that water is sometimes an inconvenient choice in those parts of the World where horizontal tips are most common.

If, at some time, you switch to an inert purge, you can probably go as low as 0.05 fps on the basis that the purge provides a cushion between the air and the flammable gas. However, be aware that some pilots are easily extinguished by a surge of inert gas, so be sure to get a design which is proven to be resistant to inert atmospheres.

Regards
David