Min. purge rate of storage tank

[timsonderby]

Hi,

Anyone who knows how to calculate the minumum purge rate of a tank with fuel gas.

I have been looking in API STD 2000 API 521,650 etc. but they only explain how to take thermal expansion/contraction into account.
What I'm looking for is the purge rate that will make sure that the gas/air in the tank is not within the flamable region.

The reason why I'm doing this is because I have a sump tank of 3400litre and is is currently being purged with 14.5Nm3/h (fuel gas) which I believe is way more than required.


any suggestions?

 

[TD2K]

I'm not sure what standard might address this, to me it's impossible to quantify how much air is leaking into the system and being purged by the natural gas to put in a standard.

I'd check the gas being vented from the sump and measure how much oxygen is in it and adjust the purge rate corresponding.
14.5 Nm3/hr for a 3.4 m3 sump means you are purging 3x per hour, that seems a lot for me but you need to estimate what the upper flammeability limit is and then how much O2 is in the vented purge gas.

 

[25362]

You state the rate but not the duration of the purging process.

Assuming this is sweep-through purging, with perfect mixing, starting with air as the gas originally filling the tank, and that the fuel gas contains no oxygen, it all depends on the final oxygen content required.

For example, starting with 21% and seeking a final O[sub]2[/sub] content of 1%, the required colume of the purging volume is:

V = 3.4 ln(21/1) = 3.4[×]3.05 [→] 10.4 m[sup]3[/sup], namely a minimum of three tank volumes.

Since perfect mixing is never truely attained, experience will tell you how many "tank volumes" should be used for purging to the desired O[sub]2[/sub] % level.

 

[timsonderby]

It is constantly being purged.


@TD2K: I agree with you that this is the normal way you would go in case that there is no standard within the area. However since it is an offshore facility it becomes more expensive to go and make this study on 10 tanks :-S

Thanks for the good feedback, and nice to know that you don't know of a standard either.

 

[25362]

Thread469-22878
Thread9798-66873
Thread798-168462