N2 Control Valve Sizing to Prevent Vacuum for Steam Stripper Shutdown

[ETBass]

I am working on sizing a N2 control valve for a steam injection stripper at my plant. When we stop steam to this vessel, we obviously pull a vacuum. We are currently connecting a N2 hose to prevent this from happening. Is there a calculation somewhere that can tell me how much N2 is needed to prevent pulling a vacuum on the vessel? Thanks. ETBass

 

[bchoate]

bchoate
One can make a rough estimate of the nitrogen requirement.
The free volume in the stripper is known or can be estimated. At shutdown stripper conditions (T & P) are known. Assume the free space is filled with air. Calculate (from PV = nRT) the number of moles of air in the stripper at shutdown. Assume the stripper cools to ambient temperature. Using V, ambient T, and n calculated above calculate the new P at ambient. This estimates how low the pressure may drop due to cooling (no leakage is assumed).
Using V, ambient T, and atmospheric pressure calculate the total moles of gas needed to preserve atmospheric pressure.
Subtract the moles of air at shutdown from the moles needed at ambient T & P. This is an estimate of how much N2 is needed to hold ambient P under cooling in the vessel.

The vessel can be protected by installing a pressure switch to sense the drop in pressure. A small (1/4" Baumann) control valve can be used to feed nitrogen to the column when the switch shows a deviation alarm.
Bill C

 

[sshep]

The condenser is collapsing some amount (volume/time) of vapor, and will start to pull a vacuum when the steam is cut if an equivalent volume is not replaced. I believe that a make-up N2 volume flow (at tower pressure) equal to the max steam rate that could be handled by the condenser is sufficient to size the control valve. I would envision a quick opening valve and a pressure controller set reasonably below the operating pressure.

This simple basis assumes that no credit need be given to vaporization from sensible liquid heat as the pressure falls, differences in Hvap between the steam and vapor going to the condenser, etc. If the only heat loss is to ambient (i.e. you have no condenser and there is a check valve in the overhead line to prevent back flow), then this basis would not apply.

 

[Montemayor]

ETBass:

The following is the manner I have designed storage tanks and process vessels for a potential "suck-in" (sudden vacuum) condition. I would install a buckling pin vacuum relief on the stripper in order to protect the vessel in any case. You cannot ensure that the Nitrogen source will always be available in an emergency. I have fought this out time and time again in various HaZops and the only credible way is to have the buckling pin installed and ready to save the vessel. I am assuming you are using Nitrogen to break the vacuum due to a process or safety need to avoid using atmospheric air to normally break the vacuum.

First, you must use a credible rate equation that represents the rate formation of the potential vacuum. This is available in the form of Nusselt's classic derivation of the condensation steam theory. I have used this relationship with the following logic to back it up:

1) Your stripper will be operating normally with steam sparging directly into the vessel; the stripper will have a liquid phase and a steam-rich vapor phase within it.

2) There is no air or non-condensables inside the stripper - this is to be expected as part of the normal operation; all vapor inside the stripper is condensable and poses a vacuum hazard if allowed to condense since it will pull an almost perfect vacuum.

3) At the moment of steam failure, I allow for the worst possible credible scenario to occur: a rain shower is drenching the stripper at that same moment or directly after the steam inlet fails.

The rate at which the vacuum is formed is the rate at which the condensables are condensed. In order to avoid a vacuum being formed, you must (in the worst case scenario) inject Nitrogen at the rate of condensation, plus a contingency. The Nitrogen control valve and all associated piping must be sized for that condition in order to ensure that the stripper is protected at the worst condition and the N2 flowrate should remain sub-sonic. If the Nitrogen supply should fail, you have the buckling pin as a fall-back position.

You can calculate the Nusselt condensing coefficient (found in Kern's "Process Heat Transfer"; McGraw-Hill; 1950 - another classic) or use what I have often applied: an overall "U" of 400-500 Btu/hr-ft2-oF. You apply the heat transfer equation:

Q = UA (T2-T1), Btu/hr

The heat transfer area is the exposed stripper's system surface and the temperatures are those across the stripper shell. I normally assume a rain temperature of 50 oF.

You assume the worst case is pure steam with its latent heat condenses at the rate of heat transfer and you obtain the amount of vapor that "disappears" in the stripper due to condensation. This is the rate of vapor you must replace to avoid the stripper going to vacuum.

This is the logic and reasoning I have used in the past - with some variations and modifications to allow for local conditions and process characteristics. You may have some modifications you need to apply yourself - but the basic need to safely ensure that no vacuum is suffered by the stripper follows the above logic. I have done this on various applications with plant management approval of the logic and calculations, which by the way must be done and documented in accordance with OSHA's Management of Change policy and the customary HaZop. I assume you are in the USA and will be following the same rules and guidelines as well.

I hope this helps you resolve the problem.

Art Montemayor
Spring, TX

 

[EGT01]

Perry's Chemical Engineers' Handbook, 7th edition, now has a Process Safety section (26). Among other topics, there is one covering Hazards of Vacuum. They cite the following study which was made to allow prediction of the rate at which air must enter a tank with and without internal condensation to prevent a pressure difference from arising:

Influence of Product Vapour Condensation on Venting of Storage Tanks,
Chemical Engineering and Processing, Volume 22, Issue 3, November 1987, Pages 137-144
D. Fullarton, J. Evripidis and E. U. Schlünder
published by Elsevier-Sequoia

The study focuses on vessel inbreathing due to weather effects and gives the background on how the method was developed as well as an example problem to follow.

I don't have a lot of experience with "vacuum relief" due to condensation but the article seems reasonable and would offer a more rigorous approach to determining your nitrogen requirements.

If interested and your local library doesn't have the reference, you can obtain it from Science Direct for a small fee

http://www.sciencedirect.com/science/journal/02552701

The following is a link that will take you directly to the article abstract but I'm afraid you will have to copy the full address (between the quotes) and paste it in your browser address box
"

http://dx.doi.org/10.1016/0255-2701(87)80040-5"

You will see some text in German but the article is in English.

As an alternative you may be able to obtain it from the following for a lesser fee though you may have to wait a day or two before delivery.

http://www.lindahall.org/services/document_delivery/index.shtml