Rule of thumb for determining how much purge gas or sweet gas need to maintain flare header pressure 1

[Vivaldi M. Smith]

Good day ladies and gents.
I was wondering if there is rule of thumb to determine how much purge gas is needed or required to maintain the Flare System Header Pressure . Anyone knows what are the steps to make those calculations? is there ny API standard to refer to?
Muchas gracias

 

[LittleInch]

Depends on your flare tip.

I've seen 10% of max flow mentioned before but if you have a pilot flare then maybe 5.

Typo corrected.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

You mean how much purge gas to prevent flashback ?
Flame velocity in oil/ gas installations is highest for methane at about 0.45m/sec, and about 0.3m/sec for natural gas. See table 27-18 in Perry Chem Engg Handbook 7th edn for other hydrocarbons. So minimum purge gas velocity has to be higher than this at the flare burner tip to prevent flashback. Suggest adding some 20-30% more to ensure no flashback.

 

[shvet]

how much purge gas is needed or required to maintain the Flare System Header Pressure

What is a purpose to maintain a certain pressure? Explain please.

 

[Vivaldi M. Smith]

thanks for the inputs. littleInch, you are correct I meant sweet gas. georgeverghese, yesh I meant flashback to ensure that the gas will overcome the flare tip to avoid air egress. to asnwer the request from shvet, you have to maintain a certain pressure in your flare header to prevent air egress to your flare header.

 

[LittleInch]

It's all about the flare tip design and what its min flow is.

This could be anything from 10% of max flow upwards.

Flame speed of natural gas is around 0.35m/sec so if you know the flare tip nozzle area have that as a minimum flow to prevent flash back.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Vivaldi M. Smith]

We have a ground flare with multiple burners tips. I have found a formula where minimum purge gas to flare per year is calculated by multi[lying flare diameter*total time flare in operation*(7.85*10^-4). But since we have a ground flare with multiple burners, I dont know if to take the diameter I should consider each flare header or just the two common headers heading to the burners tips?

 

[LittleInch]

That formula will only give you a length, not a volume?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Vivaldi M. Smith]

check document attached on page 24

 

[LittleInch]

Nothing attached

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Vivaldi M. Smith]

Done now.

 

[LittleInch]

It's multiply flare diameter SQUARED.... x 7.85 x 10-4 x no of hours of operation.

This formula assume a min velocity in the flare header of 0.04ft/sec. Note that this is related to the min flow required by a flare stack seal.

If your flare does not have one of these then the min flow is, as it says on P24 " the amount of sweep gas required can be estimated by the gas flow rates
needed to prevent oxygen ingress at the flare tip."

Hence you need to know if you have a seal or not and if not you need details of the flare tips.


Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

Can we see sketch of the 2 common lines going to the flare burners.
Are there 2 sets of flare burners, one for each common flare header? Or are the flare burners operating off a single common manifold as a single set for both flare headers ?
Also, are the flare burners operating as staged burners, where more burners are placed into operation depending on the built up backpressure at each common header?

 

[Vivaldi M. Smith]

Georgeverghase, Yes there are HP and LP flare headers heading each individually to four set of stages. Each stage opens depending of the built up backpressure. First stage is for for normal plant operations. Rest are for upsets.

 

[georgeverghese]

This sketch tells us a lot more.

From what you have disclosed so far, the 0.3m/sec min velocity ( assuming this is natural gas) at the flare burners tips applies only to the stage 1 burners for the HP ground flare and the stage 1 burners at the LP ground flare, since you have ample pressure in the flare headers to ensure forward flow for stages 2-4.

Since these burners are staged, min flow purge must be injected just upstream of the stage 1 burners so that this purge gas does not flow into the stage 2,3 and 4 burner sets. This approach saves you the additional OPEX and environmental impact of additional min flow purge gas if min flow purge gas were to be alternatively injected upstream of the flare KOD.

To keep the flare headers free of air, additional sweep gas should then be applied at the upstream ends of each flare subheader. You can use the formula in that page 24 of your design note from the EPA (with v = 0.04fps applied over the cross section of the flare subheader)for sweep gas since it appears you have a liquid seal drum for each HP flare / the LP flare header (ie drum upstream of flare KOD).

You could take credit for the upstream sweep gas rate to minimise min flow purge gas rate at the stage 1 burners.

 

[shvet]

@all
What is a purpose to purge flare header for staggered ground flare? How can air come into pipe? Or purging relates to another issue? Can anyone explain please?

 

[LittleInch]

Are those drums liquid seal?

Or a secondary knock out drum?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

Since this is a staged flare, you can safely relocate the min flow purge meant for stage 1 burners back to the upstream end of the flare headers in the plant. This is because stages 2-4 open only when there is ample pressure building up in the flare headers (ie. plenty of forward flow). This upstream location would then account for sweep gas requirements also on the upstream plant flare stream collection headers.

 

[shvet]

@all
What will be not ok if this header is not purged? Fire, explosion, intoxication, corrosion, freezing, undesired deposits buildup? What is a problem purging is intended to solve/avoid?
Purging for purging. Topicstarter purges header to provide purging itself, am I right? And forummembers discuss with topicstarter how to do it better? Just you know... to provide purging because a header needs to be purged. Anyone, correct me please.

 

[georgeverghese]

@shvet,
Flare / vent header purge is
a)to prevent reverse migration of air into flare headers, which will happen if there is zero forward flow out of flare tips. Once air gets in, there is a risk of flashback or worse still, detonation in the FKOD / flare headers.
b)maintain flame stability even in cases when forward flare stream flow is below the turndown of the flare burners.