Cascading thermal relief system in oil terminal

[yahu55]

Hello,

I am designing a "cascading" thermal relief system for an oil storage terminal.
The cascade will terminate at an atmospheric storage tank.
Product level in the tank will be variable and as such the "static" back pressure will vary.

See an example sketch below:

My questions are:
a) Does TRV-2 need to be of balanced type due to the variable back pressure (only source of back pressure would be the varying tank head).
b) Is the following approach correct: Assuming TRV-2 is of balanced type and its set pressure is 10 barg. Then the maximum backpressure seen by TRV-1 will be 10barg. If TRV-1 is then specified as balanced valve too, then its minimum set pressure shall be >20 barg (so that the backpressure is <50% of the set pressure as stipulated in API521)?

I'm looking forward to your advice

 

[georgeverghese]

a)TRV-2 should be a balanced PSV if max backpressure from the tank is greater than 10% of set pressure
b)If TRV-1 is a balanced type PSV, max permissible backpressure is 80% of set pressure. This is permitted only for thermal RVs' (liquid expansion due to solar radiation heat input), since relief load, though it is affected, is very small. Balanced PSV set pressure is not affected for backpressure up to 80% of set pressure. See API521 for this detail.

 

[yahu55]

Georgeverghese, many thanks for your help. Where did you take the 80% value from? I can't seem to find it in API521...
And why would 80% apply to TRV-1 and not TRV-2?

 

[Latexman]

The capacity of these balanced PRVs begins to decrease when the backpressure exceeds 30 % to 50 % of the set pressure due to subsonic flow and/or physical responses to the high backpressure.

Without knowing the performance of a specific PSV, use 30% value.

georgeverghese, can you point to that reference in API 521? I could not find it.

Good Luck,
Latexman

 

[LittleInch]

I think you might really need pilot operated valves to do this as you have no real idea what the back pressure is.

Section 2 might already be relieving before section 1 gets to the design pressure.

You might really need to set cascading design pressures to allow the furthest most section the chance to discharge into the next section down.

Much better just to install a 1" closed drain header....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[RVAmeche]

Look up UFC 3-460-01 for standard details on US Government/military installations of this type.

The industry norm, which I do not agree with, is for all upstream devices (TRV-1 and others further to the left in your sketch) are not true ASME relief devices. They're unstamped "relief regulators" which function the same way but are not stamped and tested. Only the last device at the tank, TRV-2 in your sketch, is a true ASME relief device. I strongly disagree with this approach but the response I got was "it's working, it's fine".

 

[georgeverghese]

Sorry, cant find this in API521 either. It may be in API520 Part I, which I dont have at the moment. I've used this rule in setting up these cascading TRVs' for long high temp piping segments.
As I indicated, TRV-2 should be a balanced type PSV if backpressure exceeds 10% of set. This is because the set pressure of a conventional PSV shifts higher if backpressure exceeds 10% of set. A typical 3/4 x 1 inch TRV would have relief capacity far in excess of the required capacity for a thermal liquid relief scenario where heat input is from heat tracing or solar radiation.

 

[Latexman]

I also looked in the current API 520 Part I and II and didn’t see it there. Maybe it was in an earlier Edition.

Good Luck,
Latexman

 

[georgeverghese]

Found a pdf copy of Part 1 on the internet. In the Jan 2000 edition, Section 3.3.4.2 on page 37 says set press is not affected by backpressure. This probably where I found additional guidance, perhaps in some previous edition. Figure 31 on page 38 stops at 50% backpressure on the x axis - you could alternatively ask the TRV- PSV supplier what the Kw factor is at backpressure of up to 80% of set pressure for a balanced bellows PSV in liquid service; I'd say even a Kw of 0.1 would be acceptable since actual relief load is a tiny fraction of rated capacity in this case.