Liquid thermal expansion overpressure protection 4

[Sawsan311]

Dear All,

for liquid thermal expansion of liquid full piping which can be blocked in with more than 95% and where presence of noncondensable vapors is not envisaged to preclude liquid expansion as explained in section 4.4.12 in API 521 7th edition. Section 4.4.12.4.2 is also discussing the role of the trapped volume in determining the requirement for the TERV, appreciate your views on the following:

- typically any liquid system which can be blocked in with liquid full section at 95% and above required TERV.
- what is the volume criteria dictating the thermal expansion, I think this can't be generalized, some guidelines calling for 500 L of LPG and volatile/toxic liquids where the leak is quite of a concern.
- Do you agree that the lower the initial operating temperature in comparison with the maximum expected solar Black body temperature, the greater is the concern of thermal expansion associated pressure rise?
- do you agree that for water full system, the concern is higher than other high API crude oil since the cubical thermal expansion is relatively higher with associated higher density contributing to faster pressurization rate for the same incremental increase in the temperature.

I am explicitly addressing liquid thermal expansion and NOT thermal vapor relief considering that the liquid will not reach its boiling point at the relieving pressure nor have a vapor pressure exceeding the design pressure.

regards,

 

[danschwind]

My personal rule of thumb is: if liquid may be trapped, I install a thermal PSV.

I don't care what the liquid, the initial temperature or the volume is.

The only situation where I would envision not doing it is if there are an explicit operational procedure to ensure there would be no trapped liquid-full pipes.

Daniel
Rio de Janeiro - Brazil

 

[Sawsan311]

Thanks Mr.danschwind for your reply, I agree with you that the for liquids specifically the ones which have high cubical thermal expansion with greater extent to expand (below the boiling point) require mitigation against overpressurization through TERV. However, in specific cases as explained in API 521 section 4.4.12.4 some conditions can exempt such requirements : temperature being higher than maximum expected ambient temperature from solar radiation, batch operation, presence of non condensable vapors.

I agree with you that it is always a safe design to consider the trapped liquid inventories subject to thermal expansion, however, the associated isolation (to enable removing these TERV online and prevent hazardous personnel exposure to the process), maintenance and inspection requirements with these TERV can impose more demand on operations personnel along with higher cost.

Regards,

 

[don1980]

My view is very similar to that explained above by danschwind. Engineers commonly over-think and over-analyze this question of whether liquid thermal expansion is a credible risk. They often waste time trying to calculating the pressure rise based on the liquid expansion coefficient, or calculating the time required to reach the limiting pressure. Such calculations have value in the academic world but not the real/practical world. To determine whether excessive pressure is a credible risk, one only needs to answer whether the following three things are true: (1) is the system truly liquid-full, (2) can the system be isolated, (3) while isolate, can the temperature rise. If all of those things are true, then one needs to either install a thermal relief valve, or implement measures that prevent the system from being isolated.

 

[crshears]

Anecdote:

A one-million-Imperial-gallon fuel tank left over after the decommissioning of the gas turbines it supplied was being emptied of the last of its #2 furnace oil. Emptying was into the ignition oil tanks for the closest pair of coal-fired generating units via a rubber hose running across the service yard. Oil had to be supplied intermittently, as there was too much to just transfer it all at once. There were valves at both the tank outlet into the hose and at the outlet end of the hose into the ignition oil tanks.

The auxiliary plant operators were told to continuously monitor oil admission into the ignition oil tanks to preclude tank overfilling. Upon completion of each transfer, it was stipulated that once the valve at the outlet end of the hose was secured, the supply valve from the tank into the hose was to be closed as well.

I objected: "Oil trapped in hose between valves will expand and burst hose!" [Small spill.] "Please allow us to just secure hose outlet vale, leaving tank outlet valve open so oil can expand back into tank."

This was rejected, based on concern hose could rupture and tank would start to empty all over service yard; really big spill.

Result? oil expanded and burst hose, causing small spill.


I guess I should have asked if we could close outlet valve from hose, close outlet valve from tank, then re-open hose outlet valve to relieve all pressure, but I didn't.

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[LittleInch]

What you should have done was ask for the hose inlet valve to be "cracked" open (~~5%) so it allowed pressure relief. But only a small spill if the hose ruptured otherwise.

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

 

[pierreick]

Hi,
Consider the document attached, very detailed.
Good luck
Pierre

 

[Sawsan311]

I would like to thank you all for sharing your experience and lessons learnt.