Supplemental Purge Gas Requirement

[keddo]

I am trying to calculate the supplemental purge gas rate required following a major hot release. The additional purge gas is required to prevent air entering the system due to thermal contraction.

The method I was using was to determine the heat loss from the insulated flare header at minimum ambient conditions and then calculate the corresponding shrinkage rate. The flare stack is not insulated and thus the heat loss would be greater than that of the header. Do I need to take this into account or would the addition of the purge gas at the base of the flare stack be enough to push out the hot gas?

 

[TD2K]

"Do I need to take this into account or would the addition of the purge gas at the base of the flare stack be enough to push out the hot gas?"

Not unless the volume being added is greater than the skrinkage in the uninsulated flare stack. You're back to do the calculations I think on the flare stack.

 

[Homayun]

I porpose 2 options:

1: consider installing a seal drum. that way you are always sure that air will not flow into your flare header

or

2: as a rule of thump have a purge rate of 1 ft/s in your header..

done..

 

[Flareman]

keddo
The entire system out to the end (where the air can enter) has to figure in the contraction calculations.
You might have any one of the following conditions
- Starting with a cold system, you get a short term release of hot vapor which fills or partially fills the header. The rate of contraction is probably controlled by the vapor cooling back to system temperature. Match the greatest rate of contraction with the purge. Calculate how much purge you probably need as a straightforward replacement volume between hot and cold conditions.
- You have a long term release which heats up the piping. The rate of cooling depends largely on the external heat transfer. Match the greatest rate of contraction with the purge. The total quantity you need depends on the thermal mass of the headers and how long they take to get down to ambient.
- You may have a high concentration of hot vapor which can reach dew point. Purge it out quickly and completely with light vapor before the cooling gets down to dew point. Then treat any residual temperature difference as above.

A water seal is certainly beneficial in case of contractions but not always practical.
How do you trigger your supplementary purge? Think carefully about the instrumentation and make sure that it measures what you think it will. Remember that the height of the stack contributes (or subtracts) pressure when you use a PT in the header. You can sometimes measure a negative or a positive simply as a function of the gas density in the stack (against atmospheric ref) and that can be a very misleading signal. You need to use two PT's with a vertical separation so that you can "weight" the gas in the stack. Also the instrument deadband can sometimes be greater than the pressure you are measuring. Oh boy these things are never easy!

David