Jump to content

Featured Articles

Check out the latest featured articles.

File Library

Check out the latest downloads available in the File Library.

New Article

Product Viscosity vs. Shear

Featured File

3-Stage Propane Ref Performance

New Blog Entry

Steam Tracing Design- posted in Ankur's blog


Inert Gas Blanketing For A Storage Tank - Flow Requirements And N2 Li

5 replies to this topic
Share this topic:
| More

#1 CabChemEngi


    Brand New Member

  • Members
  • 2 posts

Posted 06 September 2017 - 08:11 AM

Hello everybody, 


I am currently calculating the N2 flow requirements for a 5000m3 storage tank. Going through API 2000 Annex A, i see that the inert gas requirements are calculating according 2 parameters:

1. Liquid movement due to the pumpout rate of the tank pump

2. Thermal effects due to contractions and expansions for temperature variations. 


for 1, The product flash point that i have is >38°C so 0.94 Nm3.h-1/m3.h-1liq is being used

for 2, the Nm3/h for a 5000m3 tank volumen according to the table A.3 is 787Nm3/h air


I am assuming that the Nm3/h air calculated is equal to N2 due to the similar molecular weight. 


With the N2 supply conditions, i am obtaining approximately 170 m3N2/h. 


For the N2 line sizing calculation, i am considering single-vapour line hydraulic criteria where no velocity criteria is required, however a pressure drop between 4-7 Kpa/100m, with this criteria i have to go for a 4" line where comparing with other N2 blanketing lines of similar volume tank is pretty big as it is around 2" line. I do not know if the criteria i am following is too rigoruous, or i should take a different criteria for sizing the line,


It would be great if someone in this forum could confirm if the steps i am taking are correct or if there is something i should take into account.


Thanks in advance 

#2 Art Montemayor

Art Montemayor

    Gold Member

  • Admin
  • 5,481 posts

Posted 06 September 2017 - 03:59 PM



My copy of API 2000 is the 5th Edition and Appendix A - Basis of the Normal Venting for Tables 1 and 2 states the following:


“For liquids with a flash point below 100 oF (37.8 oC), this standard recommends a venting capacity of 12 Scfh of air for each barrel (2.02 Nm3/h per cubic meter) per hour of filling rate.  Of this quantity, one half, or 6 Scfh (1.01 Nm3/h per cubic meter) of air, represents the vapor displacement caused by liquid movement.  The additional 6 Scfh (1.01 Nm3/h per cubic meter) of air was established on the basis of an evaporation rate of approximately 0.5 percent and to account for the conversion of dense vapors being vented to an air equivalent.”


Bear in mind that API 2000 was put together with the intent of an application for petroleum storage tanks.  Consequently, what the document refers to - and what Appendix A alludes to - is the change of phase in petroleum liquids when there is a temperature increase, a pressure decrease, or release of dissolved gases in the petroleum product.  This vapor generation requires additional capacity for the tanks venting capacity over and above that calculated for the pump-out capacity of the tank.  This Appendix is not meant to give you any indication of what is required for the nitrogen feed capacity required to maintain a blanket in the tank’s vapor space.  It has nothing to do with nitrogen blanketing.  In fact, API 2000 (at least my edition) has no mention of nitrogen blanketing.  You fail to state what API 2000 edition you have and what it states.


You fail to identify what liquid you are storing.  For such a large tank, I assume you are inerting a petroleum liquid.  Is that correct?


A 5,000 m3 (1.320 million gallons) is a very large tank.  I estimate it as 80 feet in diameter and 35 feet high.  Depending on your maximum pump-out rate, it will require a large nitrogen feed rate to maintain the inert blanket.  Depending on the supply pressure of the nitrogen gas and the location of the source, it might require a very large nitrogen supply line.  You have also failed to state these values.


I have designed and installed a fair number of nitrogen blanketing systems on storage tanks, so I can offer you the following advice:


The basic design scope for a blanketing system is to ensure that the liquid is kept inerted through the blanket while it is stored, pumped out, and pumped into the tank.  In keeping with this requirement, you must ensure that the existence and injection of nitrogen gas in the vapor space is kept constantly under controlled conditions of pressure.  This means you must avoid two basic tank hazards caused by a process upset or instrument failure:

  • A tank over pressure.  This can be caused by a nitrogen feed valve failed open, or a deficient vent valve failure or design.
  • A tank vacuum condition.  This can be caused by nitrogen feed valve failed closed or deficient capacity, or a deficient vacuum valve failure or design.

You must ensure that you calculate for the worst pump-out or level depletion condition and have sufficient nitrogen gas capacity to make up for the increase in gas blanket volume requirement.  This has nothing to do with air, as stated in Appendix A.  You must design for feeding sufficient nitrogen as is needed to maintain the design blanket pressure when there is a draw-down of liquid in the tank due to pump out or drain.  In the event your nitrogen make-up rate is not enough to prevent an approach to a vacuum condition, your tank's vacuum relief valve should save the tank.


You don’t furnish copy of your nitrogen piping calculations, so our members can’t speculate on your calculation results.


The design of a nitrogen-blanketed tank is relatively simple in scope but it must be subjected to a hazop or similar study to ensure that the application is not only valid, but safe in all aspects - whether operating or not.  Do not neglect to include tank level and pressure safe guards in your overall design.


I hope this helps you out.

#3 CabChemEngi


    Brand New Member

  • Members
  • 2 posts

Posted 07 September 2017 - 02:30 AM

Art Montemayor, 


First of all thanks for your answer, i really appreciate it, 


A few things to expand my explanation:

  • I am using API 2000 7th Edition, i am not able to get the API 2000 5th edition, but in this edition there is a new Appendix F which state as "Guidance for inert-gas Blanketing of Tanks for Flashback Protection" where it describes three tank inert-gas-blanketing design levels.

            Level 1 has minimum inert-gas-blanketing requirements in combination with a specific flame-arrester classification.

            Level 2 has more stringent inert-gas-blanketing requirements with a different flame-arrester classification.

            Level 3 has the highest inert-gas-blanketing requirements with no flame arrester.

  • This tank is a multiproduct tank that store different petroleum liquids with different flash points (from 46°C, 65°C, 80°C and 130°C)
  • Nitrogen is being supply at 5kg/cm2g but i have no information about the location of the source

Going back again to API 2000 7th edition, for level 1 design level, the required flow rate for inert-gas blanketing is calculated as V = 0.1C ⋅ Ri ⋅ Vtk^0.7 +Vpe where:

       c is a factor that depends on vapor pressure, average storage temp and latitude (for my case is 6.5)

       R is insulation, in my case there is no insulation, therefore this value is 1

       Vtk is the tank volume (5000m3)

       Vpe is the maximun rate of liquid discharge (for my case approximately the pump out rate which is 250 m3/h)


With all these values i have a nitrogen flow rate of 500 m3/h (it does not say if this is calculated in Nm3/h, it only says that the V is expressed in cubit meters per hour, any idea at which conditions?


The other issue i am facing is regarding the line size of the N2 supply line i am designing, with the flow rate of nitrogen i came up with two approaches:

1. calculate the line regarding pressure drop criteria for single-phase vapour lines where i have to mantain approximately between 4-7 kpa/100m

2. calculate the sonic velocity of gas and calculate the minimun require area, (A=Q/Vs),

Any suggestion for this?


Thanks again in advance

#4 vjmromero


    Brand New Member

  • Members
  • 8 posts

Posted 05 February 2018 - 09:10 PM

Dear Sirs,


Are there any criteria for determining the operating pressure of the tank at blanket condition? Maybe the basis is the required head to deliver vent N2 to flare. I am currently designing a DEA tank with N2 blanket to avoid corrosion of tank roof.


My problem is that the tank to be injected with N2 has a design pressure of only 2 kPag i.e. atmospheric tanks... I am evaluating whether to install new tanks with higher design pressure or just vent-off N2 to atmosphere.


Thank you for your inputs.

#5 Art Montemayor

Art Montemayor

    Gold Member

  • Admin
  • 5,481 posts

Posted 05 February 2018 - 10:12 PM



Put your thinking cap on.  2 kPag is equal to 8 inches of water column.  That is sufficient to maintain a positive nitrogen blanket in the storage tank up to that pressure.


Just because your tank is designed for 2 kPag doesn't mean that it will not safely sustain a pressure higher than that.  The important criteria in any pressure vessel or tank is the Maximum Allowable Working Pressure (MAWP).   If you rate your storage tank (calculate the MAWP under the present conditions it is working under) you will probably find that it can safely contain more than 2 kPag.  If your tank is an old vessel, you should obtain the service of an experienced mechanical engineer to inspect and rate the tank as per the applicable API recommended procedures.


That is the reason I always have relied on the MAWP of a storage tank instead of the design pressure.  Like many purchasers of so-called "atmospheric" storage tanks, you (or someone besides you) did not want to have the tank calculated and furnished with a name plate stating the MAWP at fabrication.  It costs a little more to have the calculations made, but well worth it.  All that is saved by not having the calculations made is pennies compared to future expenses incurred - like you are facing at present.  In many cases involving "atmospheric" storage tanks, the tank's wall thickness is in excess of  the calculated design pressure.  This is so because of factors such as weldability, structural integrity, or simply because of available steel plates that are standard thickness in excess of the calculated.  Tank fabricators are not bound by the calculations and no one is going to produce an off-standard steel plate.  That is just common sense that is expected of any engineer.


I will never recommend you to inert a tank by continuously venting the nitrogen to atmosphere.  That is very expensive and would pay for hundreds of tank calculations and a name plate.  Additionally, if it is only DEA why would you have to flare the vented vapors?   I've stored a lot of amine tanks in my time and never had to flare the tank vapors.  What are your reasons?   Local regulations?

#6 breizh


    Gold Member

  • ChE Plus Subscriber
  • 3,789 posts

Posted 06 February 2018 - 09:38 PM

Hi ,

Great comments from Art .

A paper to support .




Attached Files

Similar Topics