latent heat of vapourisation near critical point 3

[ChemEngSquirrel]

Fire relief calc, not dense phase.

My hydrocarbon liquid is very close to critical point. I calc latent heat using hysys at 40 kj/kg. Can i justify ignoring my number and using the API 521 figure of 115 kj/kg. Much smaller valve wi api no.

 

[Latexman]

Did you use Search (between Forum and FAQs) to find similar threads? There must be dozens. Here's one I recall: thread1203-316374


Good luck,
Latexman

 

[ChemEngSquirrel]

Yes i did. The linked thread doesn't answer my specific question.

 

[TD2K]

That's really your call as the process engineer. The advantage of doing what you want is that you can specify a much smaller valve. However if the latent heat really is that low, you won't have enough relief capacity. Depending if you are working for the owner/operator or as a contractor, there may be specifications that you have to meet that dictate the minimum latent heat.

Can you change the set point to get you a bit away from the critical point where the latent heat is higher? Have you tried doing an isentropic sizing approach as per Appendix B or C, API 520 part 1? That will give you the maximum mass flux and I think is one of the better approaches to sizing difficult relief valves.

 

[don1980]

If you're that close to critical, it's likely you relieving pressure is above critical. In cases like this it's unlikely the safety valve is going to be of any real value at all in protecting from fire exposure.

My advice is to acknowledge that fact and focus your attention (and the client's too) on protective measures that are effective (water spray, fire resistant insulation, auto-depressurization).

It's a waste of time and money to proceed with a sizing calculation for a scenario in which the PSV isn't effective. And worse, this misleads the client into a false sense of security. The client needs to be clearly aware whenever the PSV isn't effective at protecting against fire exposure. If there's nothing boiling inside the vessel, then the PSV isn't providing any protection from fire exposure.

You have to have a PSV for code compliance, but it's entirely up to the users to choose a sizing basis. There are no pressure vessel codes that require you to size for fire exposure. That's a user decision. Unfortunately, a lot of engineers don't understand this and they proceed with designing relief devices that won't work, rather than considering other layers of protection that are effective. Those measures don't completely solve the problem. That is, they don't render the vessel completely safe, but they do buy some valuable time, and reduce the severity of a failure if it does eventually occur.

BTW, we often overlook this fact, but PSVs don't provide complete protection either, even when there's a boiling liquid.

 

[PaoloPemi]

if you mean to size the valve (given the mass flux and discharging conditions) see this thread

http://eng-tips.com/viewthread.cfm?qid=329093

if you mean to simulate the fire event and calculate the different conditions at a predefined discharging pressure (PSV set) then the evaluation of latent heat at the different conditions is done automatically by simulator, you can control compositions, set pressure, heat contribute by fire and some other parameters, this is how the utility in Prode Properties works and I suppose your simulator does the same

finally, if you are calculating the discharging flow (based on some internal conditions and external heat), then possibly you may adopt the predefined API values for latent heat (as mentioned by TD2K) but I prefer to utilize the value of latent heat calculated by Prode Properties (which should be the same as that calculated with your simulator)

 

[ChemEngSquirrel]

To Don1980,

My relieving pressure is below critical. I'm not in the dense phase and have liquid at relieving condition (121% overpressure).

 

[chemlite]

if you have the possibility to calculate the value for latent heat then use that as suggested by PaoloPemi and others
make sure you are far from critical point at discharging conditions,
for mixtures you should simulate the process since the evaporation of some components change all the properties

 

[Latexman]

There is excellent input above from everyone, but notice (I'm sure you did), no one really gave a yes or no answer to your specific question, "Can I justify ignoring my number and using the API 521 figure of 115 kj/kg?". Why? Many, many reasons. To name a few in no particular order, it depends on owner policies. We don't know the scale. How big is the risk? We don't know the dynamics. Does the scenario take minutes, hours, or days to reach the moment of consequence? Is the layers of protection and emergency response big enough? Is the emergency response quick enough? Is the LOPA done correctly and is it closed?

Several folks above correctly said, it's up to you/your company/your client. IMO, if a higher [Δ]H_vap than reality is used to reduce the size of the relief below that required to stop the scenario, no LOPA credit can be taken for the PSV. A token PSV to meet Code doesn't cut it on a LOPA.

So, it depends on a lot factors we have not been told about, and I'm with the others, it's up to you/your company/your client.

Good luck,
Latexman

 

[MortenA]

I think that there is reason to be cautious NOT using the API value close to the critical point. At the critical point you go into the dense phase. This transition happens with no apperant energy needed (and theory tells you it happens all at once). So you would expect the latent heat to go down as you approached the critical point (or more specifically the boundary to dense phase).

Are you sure that the change in latens heat would increase valve size (ok its a factor 3 so chances are high). Maybe you caould go look for litterature data on similar systems? I assume that you are dealing with rater light components C3-C5?

Best regards

Morten