12 replies to this topic

[alexzo1990]

Hi all,

 

new member here, although I've been following the forum for some time now finding answers to common engineering problems!

 

I am currently sizing a pressure relief valve for a fire case in a vertical  carbon steel vessel. I have followed API521 to calculate the wetted area and the relief rate and I'm now onto sizing the device. My problem is around specifying the relieving temperature in a fire case scenario. 

 

The vessel had a design pressure of FV/3.5 barg and a design temperature of 100degC. The device is set at the design pressure. To calculate the temperature at the relieving pressure (including appropriate overpressure) I am using a physical properties software for my single component liquid (acetone). My problem is that for a saturated vapour pressure of 3.85barg the temperature is 110degC so over the design temperature of the vessel. I understand the difference between the two temperatures may be small, however I have similar issues with vessels holding other liquids were I calculate the temperate at 200 or 250degC for a saturated vapour pressure of 3.85barg.

 

Any thoughts on how to approach this? What I was thinking is along the lines of the recommendations from API521 and provide measures such vapour depressuring and/or liquid dumping.

 

Best Regards,

Alexzo1990

Edited by alexzo1990, 06 November 2024 - 11:58 AM.

[latexman]

The allowable stress of carbon steels does not start reducing until about 400 F (204 C). There is no real problem here, just a technicality. Whoever specified this vessel was short sighted. They could have made the MAWT 400 F and the cost would not have increased.

Rerate the vessel to 400 F, so those that follow you do not have this issue.

[alexzo1990]

Hi Latexman,

 

thanks for your quick reply. This actually would save from a lot of issues and technical discussions on how to approach the solution correctly if the design temp. remains 100degC. Some times it's the simple things eh?

 

Thanks for your input.

 

Kind regards,

Alexzo1990

[latexman]

Hi Alexzo1990,

You are welcome. I saw this identical issue for the entire 45 years of my career. I’m retired now. I think I personally had over a dozen vessels rerated during that time. I saw a lot of vessels rated for 100, 200, and 212 F. The original designers could have done a lot better!

[latexman]

Hi Alexzo1990,

 

The above discussion was about increasing MAWT of a vessel such that the relieving temperature was less than the new MAWT.  There were cases in my experience when the relieving temperature turned out to be greater than the new MAWT.  In that case, we looked at reducing the MAWP slightly using the reduced allowable stress at the relieving temperature such that Code is met.  IIRC, this is a trial and error calculation with the relieving temperature of the component at the coincident relieving pressure/accumulation (1.21 x MAWP), MAWP, and MAWT.  My experience is that allowable stresses don't change quickly, so MAWPs don't change that much in this case.

[shvet1]

Wait a minute, gentlemen. Let's look onto what you talking about

 

The PV has MAWT 100°C and MAWT can be increased to 200°C with no consequences to the PV.

in opposite to this

The relief temp is 110°C. This still poses no consequences to the PV because of the metal stress and bla-bla.

but

One shall increase MAWT higher than relief temp.

 

What for? 110°C is more than 100°C, and what? PV wall temp will be 110°C instead of 100°C, and what? Will be relief temp 110°C instead of 100°C and what goes wrong?

What problem the topicstarter is intending to solve by increasing MAWT? Or decreasing MAWP, does not matter.

 

Hope the idea above is clear.

Edited by shvet1, 07 November 2024 - 11:46 PM.

[latexman]

Rerate the vessel to 400 F, so Engineers that follow OP do not have to repeat and repeat and repeat this silly issue.

[alexzo1990]

Hi all,

 

again thanks for all the replies. 

 

What you are suggesting there latexman also crossed my mind and it (sort of) was the angle I was coming from when I discussed this with my supervisor. I argued how we could lower the set pressure of the RV so that in the case of a fire the valve lifts prior to exceeding the vessel's MAWT. That almost by default would mean re-rating the vessel for a lower pressure. That led to a wider discussion around DP and DT in the project.

 

Shvet1, if I am following you correctly, I think I understand what you're saying that a difference of 10degC is probably not a massive issue. However, as I stated in my original post I have similar issues with other vessels containing different liquids where I calculate the relieving temperature between 200-250degC at 3.85barg depending on the case. In those cases, and in absence of any other measures such as what I mentioned in my original post, my understanding is that there needs to be a balance between relieving temp, relieving pressure, MAWP and MAWT as latexman suggests. However, I am quite new to pressure relief so I wanted to gauge opinions from more experienced members.

 

Cheers,

Alexzo1990

Edited by alexzo1990, 12 November 2024 - 05:23 AM.

[latexman]

What you are suggesting there latexman also crossed my mind and it (sort of) was the angle I was coming from when I discussed this with my supervisor. I argued how we could lower the set pressure of the RV so that in the case of a fire the valve lifts prior to exceeding the vessel's MAWT. That almost by default would mean re-rating the vessel for a lower pressure. That led to a wider discussion around DP and DT in the project.

For the acetone case in the OP, it seems to me, and, of course, I don’t have all the data you have, the re-rate would be to just raise the MAWT. Now the other cases you mentioned sounds like the re-rate would be to raise MAWT and slightly reduce the PSV set pressure such that the MAWP + allowable overpressure is not exceeded. IIRC we didn’t reduce MAWP often; we slightly lowered set pressure of the PSV so the allowable accumulation (MAWP + 10% or 21%) AT the new MAWT is not exceeded, but there were a few cases where we re-rated to a slightly lower MAWP also to make it crystal clear to everyone. It’s a case by case investigation/decision.

[shvet1]

Shvet1, if I am following you correctly, I think I understand what you're saying that a difference of 10degC is probably not a massive issue. However, as I stated in my original post I have similar issues with other vessels containing different liquids where I calculate the relieving temperature between 200-250degC at 3.85barg depending on the case. In those cases, and in absence of any other measures such as what I mentioned in my original post, my understanding is that there needs to be a balance between relieving temp, relieving pressure, MAWP and MAWT as latexman suggests. However, I am quite new to pressure relief so I wanted to gauge opinions from more experienced members.

 

Sorry for late response, but I have not so much free time as other active forummembers do. Sorry for english, I have a rare practice and lack of time.

To reply your comment in a constructive manner I have spend a couple dozens hours screening my archive for applicable corporate engineering practices. I have looked through 10+ of those in total just to find references that I would be able to cite as my personal view point/opinion (or someone else) is not relevant for cases like this is.

 

Find below my advise for a newcomer to process industry

 

The idea is the abnormal/emergency conditions like overpressure cases are, especially a firecase, do pose a serious threat and shall be treated with care. But what is not obvious is (1) probability of such events; (2) the overprice resulted from rash actions taken; (3) the incredible amount of assumptions behind abnormal/emergency calculations. You would be amazed if you assess all those assumptions that industry has historically incrementally accepted to implement ovepressure practices. From an inside look ovepressure phenomenon more looks like a public agreemnt than an a sort of engineering.

 

You may open a process safety handbook and do find the probability of the death as a result of a direct strike of lightning or a meteorite. Although such event seems exaggerated and more like a humor those do have a rate that the industry can't deny. Does this mean an engineer, particularly alexzo, should implement protection measures from a meteorite? The answer th industry has been practicing is "no" as the value resulted for industry/society in this case would be less then the value for a victim of a such direct strike.

 

Most Majors' corporate engineering practices omit the instructions to ignore emergency relief temperature during specifying a PV design temperature. I guess this is because of the point is quite obvious for their developers&recipients so there is no need to codify it. I have not provided a really deep search as I have time only for screening, but the only statement that I have come across is from FosterWheeler's std.:

 

 

1.3.3 Overheating and Corrosion

Pressure relieving devices protect a vessel or item of equipment against overpressure and not against failure due to high temperature when exposed to fire, nor failure due to corrosion. Safety in  connection with such failures must be considered independently. For instance, a vessel exposed to fire is protected for pressure by the relief valve ; however, the vessel can fail because of excessive metal temperature, especially in those cases where the vessel contains mainly gases. Vessel safety can be improved by a depressuring system to reduce the pressure to a safe low level (refer to API  521, § 3.19) or by cooling the vessel with water spray, or with fire-proof insulation.

 

The Majors having the most conservative approach to the safety permit specifying relieving temperature as PV design temperature but only in particular cases, e.g. ExxonMobil's std.

 

Remote Contingency

An abnormal condition which could result in exceeding design pressure at the coincident temperature, but whose probability of occurrence is so low it is not considered as a design contingency. Note  that temperatures above the design temperature may also be permitted under remote contingency conditions.

 

I can't cite here all practices related to peculiarities of relieving and design temperatures as those are extensive and complex and do not comply with the format of this forum. But I hope the idea above is clear.

 

Relating to your questions posted - sure you can rise this PV's DT up to 110 or 150 or 200°C and this will be believed as a good move. The idea I tried to express in the previous post is:

(1) this is not a widelly spread or generally accepted industry practice, many users do not follow this logic

(2) you or your client will get no value from this move other than a sort of gambling with some figures not knowing to other except you and maybe 1-2 persons that will have been invoilved in this design

Edited by shvet1, 15 November 2024 - 05:31 AM.

[breizh]

Hi,

Consider reading this document (link):

https://www.aiche.or.../20131068_r.pdf
Breizh

[alexzo1990]

Hi Shvet,

 

thank you for spending, what by the sounds of it is, a considerable amount of time going through established engineering practises. I get the point around assessing the value of an engineering solution in the wider context of societal benefit. It is always good to be reminded of that as it gives wider perspective when we get tunnel vision such as maybe in this case.

 

In the example above, I am not an advocate of any solution in particular. I am just trying to form an opinion based on logical conclusions. I understand why when setting DP and DT in the Basis of Design for example, one would choose to disregard the conditions developed during a fire or any other extreme event for the reasons that you presented. As you (and API521) rightly suggest "Vessel safety can be improved by a depressuring system to reduce the pressure to a safe low level (refer to API  521, § 3.19) or by cooling the vessel with water spray, or with fire-proof" insulation." . If this is not preferred,the other option would be the "slightly more unorthodox" decision of raising the DT and/or finding a balance between the DT and DP (as discussed in the replies with latexman).

 

From my point of view though you couldn't argue that you will choose to neither provide additional safety systems nor change the DT/DP, especially in a case like this where the fire case is the governing case for sizing. Because if you do, you would end up sizing a valve for a purely hypothetical scenario where the vessel will fail prior to the valve lifting. Would that not be considered designing from a meteorite as well? Fundamentally, this is the question I am trying to answer. So opinions like yours and latexman's are very valuable as the answer seems to require a mixture of (operational) experience and company practise/policy. 

 

Latexman, thank you for your reply. Yes you are right, setting the DP of a vessel relies on other things as well, maybe it's best to just lower the set pressure first.

 

Also, big shout-out to Breizh! Thanks for the AIChE article. Very concise description of the procedures for sizing!

 

Kind regards,

Alex

[breizh]

Hi,

I encourage you and others to get a copy of:

Chemical process safety

fundamentals with applications 3rd edition

by Daniel A Crowl and Joseph F Louvar 

publisher Prentice Hall

 

a lot of examples and calculations.

 

Breizh