8 replies to this topic

[pleckner]

We typically don't design for failure cases, we design to protect against failure. That is the purpose of the independent backup vacuum breaker valve, designing to protect against failure of the primary vacuum relief. Saying this, it is not uncommon to do as you are suggesting. Just make sure that the backup system is highly reliable. I would set up a frequent preventive maintenance program to ensure the integrity of the backup system as this would be the last line of defense against a catastrophic tank failure.

[proinwv]

Cesar,

I agree that we need to design to avoid the problem.

Let me ask you a question. If you blanket with N2 rather than fuel gas, and there is a vacuum valve actuation, would it be less likely to reach a dangerous atmosphere, being that you are starting both O2 and fuel free?

Also, a properly designed and maintained blanketing system should provide proper pressure control to eliminate the opening of a vacuum vent valve.

One method that might interest you would be to use a cascade blanketing system in which you had two blanketing valves with set points somewhat displaced from one another. For instance, one set to 1 in.wc and the second set to -(minus) 1/2 in.wc. The vacuum vent would be set to say -2 in.wc.The valves would open whenever the tank pressure was below their setpoint, so under normal operating the higher set point valve would operate, and should it fail the negative pressure valve would come on to keep the vacuum at or above a certain level. Should it also fail, then the atmospheric vacuum vent opens and an alarm system can be employed to warn of this and shut down any pump off operation.

The pressures are for example only. Actual values depend on the tank MAWP and MAWV.

Please exercise design care should you use such a system and be in contact with suppliers to understand the operanting curves of the hardware.

Best wishes
PAUL

[César]

Thank you all for your coments!

I agree that the most important thing in this case is ensure a reliable system to provide the blanket gas.

Paul, you are right, it would be better to blanket with N2, taking in account that it´s a non explosive gas, however, this decision implies the instalation of a new system of blanketing.

I like the alarm system idea.

My decision it was the following:

I think that for a vacuum pressure in the tank, it should matter two things at the same time, 1) Fail in the blanketing system 2) Decrease in the level of the tank. The probability of this two events at the same time is very small. What i´m going to do is to put an alarm system when the pressure in the tank is 0 psig.

Thanks you all for your contribution!
Best regards!!!

César

[proinwv]

Cesar,

There is another situation where the failure can occur; sudden cooling of the tank, caused by a rainstorm.

The probability of a multiple failure is finite.

The penalty for a failure could be loss of the vessel due to vacuum collapse, damage to other property, loss or contamination of product, injury or death. It is difficult to justify such a decision after an accident.

Further, should the tank pressure be dropping and your alarm sounds, then what do you do? Run?

I wish you luck.

PAUL

[César]

Good day Paul.

>

>Cesar,

>There is another situation where the failure can occur; sudden cooling of the tank, caused by a rainstorm.

In this case the blanket system will provide more gas to increse the pressure.

>The probability of a multiple failure is finite.

Yes, you´re right

>The penalty for a failure could be loss of the vessel due to vacuum collapse, damage to other property, loss >or contamination of product, injury or death. It is difficult to justify such a decision after an accident.

Exactly, and the vacuum breaker valve is going to be put to avoid a vacuum collapse. Remember that what i don´t want is to let the atmospheric air comes into the tank through the Vacuum Breaker Valve, because this situation it would generate an explosive atmosphere.

>Further, should the tank pressure be dropping and your alarm sounds, then what do you do? Run?

The alarm will alert the operator that the pressure is coming down and he will take actions to avoid the pressure fall below the set of the vacuum breaker valve. One action he can take is, for example, decrease the outlet flow of the tank. If the operator can´t avoid the pressure fall below the set of the vacuum breaker valve this valve will let air enter in the tank, after that the operator will stop the plant until the blanket system can be repaired.

>I wish you luck.

Thank!!

>PAUL
>

Best Regards.
César

[proinwv]

Cesar,

I understand what you are saying, and the sequence.

However, my concerns are that you relying on a human response, and even a judgement, and relying on this response being timely.

I simply do not believe that this is a safe way to proceed. Possibly some others who are monitoring this thread will jump in and give us their advice.

PAUL

[César]

"Unregistered" i agree with you about the way how will take place the mixtura, so the fact is, i would have a flammable mixture. I agree with you that i have to see how much the tank breathes, and that´s why i decided not to connect the combustible gas to the vacuum breaker valve.

Proinwv, i understand your concern about the human response, and his intervention timely but this is not a normal condition, because this case will take place when the combustible gas supply fail (i should give reliability to this system) and the level of the tank (...of sour water) decrease fastly.

Thank you all for your comments, and your concern about this situation.

Greetings!

César