design temperature of exchanger shell and tube side 1

[ENG-RAY]

hello everybody,
we have some shell and tube heat exchangers for cooling of LPG with different inlet temperature ( min 4deg C, Max 150 deg C).As cooling medium we use Ammonia. Ammonia flow to shell side and LPG in tube side. Design temperature of shell side is -28/50 deg C. The question is that what happens to tube side if control valve located on Ammonia outlet piping fails open. If it is fails open we can reach up to -29 to -33 deg C in shell side. But it will be atmospheric condition. What happens to tube which are not designed for minus temperature? it is realistic to assume the tube will break because of brittle fracture ( temperature of LPG inside tube is different in different heat exchangers as said 4 to 150 deg C), LPG will release to shell side and if we pressurise the shell side again, shell side will also crack and LPG releases to environment.
or only tube will crack?
on the other hand design temperature of tube side is -28 deg C but we can reach to -33 deg C in case of control valve failure? is the 5 deg difference really important?
Hope I explain my question well and looking forward to your reply!

 

[don1980]

Equipment should not be allowed to operate outside of it's design conditions. So, the options are: (1) design the equipment for the worst case temperature/pressure conditions, or (2) protect the equipment from exceeding its temp/pressure limits using relief valve and/or high-integrity instrumentation. With shell and tube exchangers, it's sometime necessary to include a bellows expansion joint in the shell. Present the full range of worst case conditions (temp/pressure combinations) to a mechanical designer for an analysis.

 

[ENG-RAY]

yes you are right. But I need to know what exactly can happen and need to add safegaurd with certain probability depends to the risk. Exchanger has been designed for the operation range but we still need to know what happens in case of failure and define safegaurd for it. Even if the design temperature of shell side was -33 deg C, cracking of piping in case of control valve failure should be considered.

 

[EdStainless]

Then there should be redundant valve mechanisms.
The situation should either be designed for or made practically impossible.

In your case you need to go back to the design documents. Will the materials really fail at -33? You will need to go back to the material specs. Will the larger delta T hurt? You will need to run the numbers.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[don1980]

Based on your follow-up message, it sounds like you're running LOPA or some similar risk management tool, and you're trying to assess the gap between an acceptable degree of risk and the degree of risk posed by the existing installation. If that's not correct, then please clarify the question again.

The answer to your question, "what can happen?", is that a mechanical failure can occur, and in this case that will result in high pressure fluid (LPG) entering the low pressure side of the exchanger. What happens next (will LPG escape to the atm, and if so, where will that occur, and in what quantity?) is completely dependent on the details of this system/process design.

 

[ENG-RAY]

yes exactly I am doing Lopa but I doubt if the consequence analysis of HAZOP was correct in first place.The assumption was that both tube and shell side crack and LPG will escape to Atm.
As you said it depends to system. Therefore I described design temperature for both side of exchanger. Since the tube side has only design temperature of +90 deg C, I think in case of -28 to -33 deg C in shell side of exchanger we will have crack in tube side and release of LPG in shell side but not in Atm.

 

[horacio Torres]

check the risk assessment.  PSM sets the risk in 10-6 probability in a single event. You can calculate the probability of this event.

I think that you should consider installing and adicional safety device for the process,,, I have not enough data from the process but a fail close valve seems to be a better solution, ask your instrument men.

horacio