I need help to how calculate vent size for vacuum relief by CIP process

[Franc_DL]

Hi all,

I need help to how calculate vent size\ air flow rate to venting of insulated tank for vacuum relief by a CIP process.

Process:
spray ball water injects inside tank with hot water to 95C degr - flow rate 18 m3/h pressure 2 bar

Then, in second order, spray ball with cold water to 9C degr flow rate 18 m3/h pressure 2 bar

The problem is the colapse of the tank produced by the vacuum for this process.

The max. allowable vacuum pressure is 6,3 mbar and the design pressure 5 mbar.

Thanks in advance,

 

[RVAmeche]

I think the absolute worst case would be to assume you're injecting steam that 100% condenses and determine the required vacuum relief flow from that.

 

[Latexman]

If the relief area gets too large, consider injecting some non-condensable gas, like air, after the hot water and before the cold water.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[don1980]

Sudden vapor collapse is a serious risk whenever a tank is completely filled with a condensable vapor (no noncondensable air or N2 present). Tank failure can happen extremely fast when/if this vapor is suddenly cooled. For uninsulated outdoor tank, we're usually concerned about a sudden cool rain shower. In your case, the cooling is from the direct contact with cold water that is sprayed into the tank, which is probably worse than rainwater on the outside walls.

Because sudden vapor collapse occurs so fast, it's a scenario that must be prevented rather than controlled by use of a relief device. The required opening size is impractically large, and the response time of a vacuum relief device too slow.

During the initial step (spraying the inside of the tank with 95C water), air is being purged from the tank by the water vapor. Depending on the duration of this step, the air concentration may approach zero by the time the second step (spraying with cold water) starts. Notice that this is very similar to a designed experiment in which the objective is to create a sudden vapor collapse. So, implement layers of protection aimed at prevention. For example, leave a large manway open, or inject a high amount of air into the vessel for the duration of this cleaning operation.

 

[Franc_DL]

RVAmeche:
I agree with you, this is the worts case, but not a real, my tank capacity is 65 m3 and have more components in the roof, very large vent size is a problem, I need be more exact in my calculus.

Latexman and don1980:

This is the most real case, I thinked calculate the heat transfer coefficient with Nusselt theory, reduced by the presence of noncondesable gas (air) then the condesate rate so the air flow rate equal of volume desplaced. What do you think? I auggested to my client leave open the upper lid manhole during the cleaning. But is not possible for your process.

 

[georgeverghese]

You need 2 levels of control here; one for operational vacuum break, for every time this CIP sequence happens; and another for safety protection with a VRV or RD or similar. In either case, we are looking at large instantaneous inflows here to keep operating pressure within mechanical limits. There may be thermal fatigue concerns too with this rapid rate of cooling if this happens frequently.

Do you have a liquid level in this tank prior to injecting this cold water? Can you inject cold water into the liquid phase instead of directly into the vapor phase ? That will reduce the rate of cooling and the required inflow rates considerably. Then, once the tank has cooled down, finally inject into the vapor phase if you must.

Otherwise, even though this "tank " is 65m3, if you cannot leave the tank open to atm, the only solution that comes to mind would be to ask for the lower design pressure for this vessel to be full vacuum.

 

[Franc_DL]

Georgeverghese:

The tank is fully emptyied before to Hot water inject, I cannot inject cold water in liquid phase because this cleaning process is defined by my client and is not possible modified,so for my inderstanding, I will calculate the vent size for full vacuum with a pressure drop in vent pipe lower than max. allowable tank pressure for vacuum. Do you know any procedure to calculate this situation? Is correct the calculation described in my comment above?

 

[georgeverghese]

Think you will need to alert your lead process engineer to the practical limitations of the current setup, and get him to talk to the plant owner's process engineer to modify operating / control philosophy, or else enable full vacuum mechanical design pressure. All who contributed responses have alerted you to the difficulties with enabling a suitable vacuum break here.