8 replies to this topic

[merac]

Hi

I'm a chemical engineering who has finished the degree this year and I am reading some things of PSV locations and I have a lot of doubts.

I have attached an excel sheet with the possible locations of a PSV and my doubts are the following:

A.Related to Location A (see figure attached):
In the book I am reading (Distillation operation, from Kister) it is said : "the above strategy is usually preferred because the trays and the liquid on them may severely restrict vapor downflow toward a low placed relief device".
1.My question is how vapor can go downfload if the pressure is higher in the bottom of the column than upstairs?? The fluid is supposed to go from higher pressure to lower pressure, so how it is going to go to a PSV located in a low place if the pressure there is high than above?


B.Related to Location B I have the same question: 2.How can the PSV be located just below the bottom tray or paking supports? How is the vapor going to go downfload if the pressure driving force is to go from higher pressure to low pressure?

C.Related to Location C(PSV on the exit of the incondensables of the reflux drum).
In the book it is said that "this location must be avoided when the dicharging vapor is hot enough to boil the coolant during a cooler water failure". My question is 3. What coolant is it referred to if the top of the column uses an air cooler (not a heat exchanger with cooling water)?? I don't really understand why this location must be avoided.
In the book it is said that "5.LOcation C offers a shorter blowdown line", WHY?

Attached Files

•  PSV_LOCATION.xls   113KB   374 downloads

[Art Montemayor]

Sara:

This is a very important, practical subject and merits a detailed and correct explanation(s).

You are absolutely in your right to challenge the author(s) of text book when you find descriptions or explanations that don’t make sense. Many times, people like Kister (Richardson & Coulson, etc. & many others) don’t know what they are talking about simply because they have rarely – if ever, operated actual, industrial Unit Operations equipment. You simply don’t have time in your career to teach, study, write, and lead design teams within Engineering & Construction (E&C) companies. What winds up happening is that the majority of engineering authors (like the majority of engineering professors) never actually get their hands “dirty” by actually implementing, installing, starting up, shutting down, operating, and maintaining the equipment they write about and teach how to design. This is not meant to downplay the role that they play in your preparation as an engineer. This is merely fact and logical results when you simply can’t do or experience two, three, or four things at the same time during your career. As a Chemical Engineer out in the production plants you will probably accumulate much more real, factual, and practical experience than Kister has up to now. That is reality. One can’t expect an engineer who has spent the greatest majority of his life qualifying for prestige and recognition (which means purely academic and engineering society membership) to be that familiar with hands-on engineering. There isn’t enough time and one has to decide early-on what career path to take.

I am not surprised if what you quote Kister as writing is correct. What you have quoted doesn’t make for good communication or writing. But then, Kister is writing to sell books – not to design and build successful operating plants. You are interested the latter; he is interested in the former (primarily).

Now to explain to you what is REALLY happening inside a distillation column:

The PSV is located directly on the overhead vapor product line. I always consider this to be an amateurish decision and a bad place to place PSVs on columns. I prefer to design early and plan to protect the column directly – by locating the PSV on the top head of the column. Nevertheless, the location on the vapor line will work – but with the caveat that it must account for the 3% maximum inlet pressure loss. First – and foremost – understand this: you are trying to relieve vapor flow, NOT LIQUID flow. Therefore, you place the PSV where the vapor is being generated or where the vessel is more readily protected. You are protecting the top of the column primarily because the cooling medium on the overhead condenser can fail and the pressure will rise quicker at that vicinity. We experienced engineers prefer the top of vessels (or overhead lines) simply because it allows for pure vapor relief and does it quickly and in an accessible location. We certainly do not want 2-phase flow (or liquid flow)! That is the main reason – not what you have quoted.

In your second question, you are now making the mistake of mis-interpreting the application. When the PSV at the top can’t handle any excess reboiler vapor load, one usually locates a PSV at that location to handle the extra load. That’s the reason it is located where it is – directly in the VAPOR space of the bottom of the column, connected to the reboiler outlet. It is relieving vapor from the reboiler – and it is doing it in a normal, conventional manner. It is taking an overpressure and relieving it into a relief manifold (probably) at a lower pressure. It has to. Otherwise, it wouldn’t work. In this case, Kister is correct.

I agree with you. You must certainly have a PSV at the reflux condenser drum. It is a pressure vessel and, as such, has to be protected – according to almost all local codes (including ASME Section VIII). Therefore, the statement that you quote: “this location must be avoided when the discharging vapor is hot enough to boil the coolant during a cooler water failure” is not correct as written. You MUST install a PSV at this location – it is mandatory to protect the vessel. This is simple enough to understand and challenge.

I don’t have the slightest idea of why Location C offers a shorter blowdown line. It all depends on the equipment layout – pure and simple. Additionally, I object to calling the discharge of a PSV a “blowdown” line. That is simply (and semantically) not what it does or is supposed to do. That line RELIEVES the pressure inside the system it protects. We design and construction engineers have other means to blowdown vessels and other equipment, such as compressors. It is an entirely different procedure and done for different purposes and reasons.

Locating PSVs is not nuclear science. It is merely using your common sense. Always direct your design towards complying with the basic goal: relieve the fluid causing the overpressure; do it safely, quickly, and in a device that is located as close as physically possible to the vessel (or pressure source). Additionally, always allow for installation and maintenance space and room around the PSV.

My colleague Phil Leckner probably has a lot of professional and personal opinions on this subject, and I will eagerly await his comments - which I have no doubt will appear as soon as he is aware of this thread.

[merac]

HI
Thanks a lot for your reply. I totally agree with you about the fact that some university teachers are very theoretical in their explanations, and it would be more practical for the students if they teach us in a more real and industrial point of view.

1.Related to place the PSV directly on the top of the column or in the overhead vapor product line
You said that the location on the vapor line will workbut with the caveat that it must account for the 3% maximum inlet pressure loss. Doesn't it the same if you put the PSV on the top of the column?. How is the PSV connected to top of the column, with extra piping? because if there is extrapiping between the top of the column and the PSV the line will also have to be designed with a maximum pressure loss.

2. Related to size the hole area of the PSV
There are some formulas to size the hole are of the PSV and they depend on the fluid you are relieving: vapor, liquid or water vapor.
When is it used the formula to relieve liquids? Because if I have understood well, the objective to reduce the overpressure is to relieve vapor flow not liquid, so?

3.Related to the PSV located on the noncondensables that exit the reflux drum.
I thought the column only need one PSV not more. And that if you choose to locate the PSV on the reflux drum (only if you have a partial condenser because if not you cannot do it this way), that PSV will relieve vapor flow if the column pressure rises above design pressure.
But you said that it is neccessary to put 2 PSV, one on the top of the column and one on the reflux drum, is that true? And in the case that the condenser is total (not partial), would it be neccessary to put the PSV on the top of the reflux drum?

4.When calculating the discharge flows of the PSV in different situations, if the biggest flow (i.e. 22000 kg/h) is the case of blocked outlet, is the PSV designed with this case? or should you take into consideration fire case (ie.3000 kg/h) also although is smaller and put 2 PSV, one for blocked outlet and one for fire case?
I have this doubt because I have seen one basic engineering specification sheet in which two relief valve are considered for two different cases (fire and blocked outlet) and it shocked me because I thought that with only one PSV you can handle both cases, doesn't it right?

Thanks for your help.

[pleckner]

In response to your last post:

1. It is ideal to put a PSV right on top of a column by adding a top mounted nozzle; you would not need additional piping. This way the only pressure losses you would account for are the inlet to the nozzle and the short (6" to 8") straight pipe run, better known as the nozzle projection. However in design we usually can't accomplish this. Especially for columns, vessels can be long lead items and must be specified with the purchase order sent to the manufacturer as early as possible. Relief calculations are often not completed until much later in the project. The good news is that the column overhead line is usually several sizes larger than what is ever needed for pressure relief. It is usually much less expensive to add a tee into the overhead line for the PSV later in the project than it would to try to add a new nozzle to a completed vessel or to issue a change order to the manufacturer for the additional nozzle. Also the PSV discharge piping orientation can be critical and thus it may be not practical to add the PSV directly to the column later in the project. You may have much more flexibility to add it on the overhead piping.

2. A PSV has no "hole" or orifice. Calling a PSV an orifice was be best thing ASME could ever have done to create massive confusion! It is a converging nozzle with a short but definite straight flow path exiting into an open cavity if you will (the pressure relief valve's body).

You use the equation that relates to the relieving scenario. There may be instances where a liquid relief is credible in your system so you would use the liquid relieving equation to find the "orifice" size. For example, if I have a small surge drum feeding a small column, can I over pressure this surge drum by filling it with liquid from a pump? For that matter, can't I liquid fill a small column?

Many of the most commonly used pressure relief valves can handle both liquid and vapor flows but you must use the appropriate equation to size the PSV.

You are required to calculate the PSV size for each and every credible relieving scenario you come up with and install the largest size PSV from all of these (a heads-up to my answer for your last question).

3. A problem within our industry. Code allows you to protect multiple vessels with a single PSV if they are connected without any chance of being blocked. If there are no valves between the column, condenser and reflux drum (or there are valves but will be locked opened), then you are permitted to protect all these vessels with a single PSV. Many of my clients do it this way. I'm not a big fan of this for many reasons. API will tell you that if you do this you must take into account the pressure drop from the furthest vessel at the time of relief. If my PSV is on my reflux drum my vapor needs to travel from inside the column, into the outlet nozzle, through the pipe, into and out of the condenser and then into and out of the reflux drum. That could be a pretty nice pressure drop to contend with. I'm not only concerned about the 3% Rule but the built-up pressure in the column during relief. Depending on the PSV set pressure, this could actually violate Code. For this arangement to work, the PSV would have to be set at a lower pressure than you might otherwise set it for. So one could put a PSV on the column and on the relflux drum. Calculations can become more comples because now you have to decide how to treat these two. You can treat them as separate entities or you can size them as two PSVs relieving the same scenario with staggered relief settings. Read up on API RP520 and API Standard 521.

If there are isoation valves between these items that can indeed be closed at any time, then each require their own PSVs. As a general note, with some very special exceptions, all ASME coded vessels must be protected by a relief device.

4. This can become complicated. Code allows you to treat fire cases separately from other relieving scenarios. It allows you to install a secondary pressure relief device with a different set pressure if it is to handle a fire scenario exclusively.

In reference to your first Post that Art has already addressed:

My only comment above what Art has already said is that it is not typical to put a relief device at the bottom of the column. I have seen it a few times but again, it is not common. I'm in favor of it for packed columns especially because of the potential plugging that can happen during a relief.

Whether you can get a large vapor load from the boiler depends on the type of reboiler. The only ones I'm concerned about are fired heaters. Those on steam will typically "stall out" if you will. What happens is that as the pressure in the column rises due to the excessive boiling the boiling point of the liquid in the column bottoms goes up. The LMTD in the reboiler drops and the amount of boiling reduces. Many people don't take credit for the pinch point in the reboiler, I do.

By the way, in general, for a column system, the controlling relieving scenario is more typically loss of cooling in the condenser than anything else, including fire.

It would do you some good to get API RP520 and API Standard 521 and go throught these documents.

That was a lot of writing. I'm sure I've opened pandora's box of more questions.

[merac]

Hi
Thank you for your reply
I have thought about what you have said and I have still some doubts:
1.Related to fire scenario.
You said that "Code allows you to treat fire cases separately from other relieving scenarios. It allows you to install a secondary pressure relief device with a different set pressure if it is to handle a fire scenario exclusively."
But what I thought was that the set pressure of the secondary pressure relief would be the same as the other PSV (because the set pressure is usually the design pressure of the vessel) and what the thing that would be different would be the overpressure because in case of fire it is allowed a 21% (instead of a 10% used in other cases).
Is that all right? Or am I wrong?

2.Related to cases in which the PSV must relief LIQUID
"For example, if I have a small surge drum feeding a small column, can I over pressure this surge drum by filling it with liquid from a pump? "
Of course, the pressure vessel will increase but it is supposed that the design pressure of the vessel is the shut off pressure of the pump, so when pressure increases it will only go up until shut off pressure because then the pump will stop, doesn't it? And so there is no necessity to put a PSV in the vessel.
Are there any case in which the design pressure of a vessel that is after a pump is not the shut off pressure of the before pump?
And how do you decide when to design the vessel to the shut off pressure and when not?
If the vessel is designed whith the shut off pressure of the pump, would it be necessary to include a PSV in the vessel?


3.Related to the arguement you give to explain that it is better to put 2 PSV, one the overhead vapor line and the other on the reflux drum, I have read your explanation a few times but I don't understand what you mean when you said:
"I'm not only concerned about the 3% Rule (that part I have understood, I think so) but the built-up pressure in the column during relief (Why? What happens to the built-up pressure if you put one PSV or two?What is the difference?). Depending on the PSV set pressure, this could actually violate Code (Why?). For this arangement to work, the PSV would have to be set at a lower pressure than you might otherwise set it for (Why?). So one could put a PSV on the column and on the relflux drum."
Can you explain me again please?


Thanks a lot for your help. I would like to understand as easy as you, but for me is not so easy.I hope you can help me.

[pleckner]

In response to your last post:

1. My original statement is correct. The set pressure for this secondary PSV protecting against a fire can be as high as 10% over MAWP. The allowable overpressure/accumulation will still be 21% over MAWP.

In addition, if I were to be protecting a vessel from any scenario, other than fire, with more than one PSV, then those PSVs also do not have to have equal set pressures. The allowable overpressure/accumulation in this case is 16% over MAWP.

2. I also said that the liquid relieving scenario would have to be credible. If it is not a credible relieving scenario then you would not need to calculate the PSV size based on a liquid relief.

What makes the scenario credible in many instances is you, the Process Engineer. Have you designed the vessel for pump shut-off? Have you taken into account that the pump is not centrifugal but a positive displacement pump? How about if I am tranfering liquid using pressurized gas?

I would indeed try to design out the scenario. Sometimes it is only a matter of cost. Making equipment inherently safe is always the way to go. Relying on a relief device should be your last resort.

And as I said before, with some very special exceptions, EVERY ASME coded vessel must be provided with some type of pressure relief device. If you can't find a credible scenario, we usually just use thermal expansion as our scenario and install a minimal 1/2" x 1" or 3/4" x 1" PSV (depending on the manufacturer and model) and move on to other things.

3. If I have one PSV protecting the column, condenser and reflux drum and the PSV is on the reflux drum (many feet away from the column) what will the pressure be in the column during relief? The pressure at the PSV is essentially set pressure + allowable overpressure. What does that make the pressure at the column?

API wants you to put the PSV as close to the source of overpressure as possible, not many feet away. If you have to put the PSV many feet away, drop the set pressure so the pressure in the column won't exceed its allowable maximum during relief. Another solution would be putting a PSV on the column, as we've been mentioning, so it is closer to the source of overpressure. And again, if we have some isoation between at least one of the equipment pieces we've been discussing here a second PSV is going to be needed.

If you are going to get involved in relief systems then you need the proper reference library:

1. API RP520, 7th Edition, January 2000
2. API Standard 521
3. ASME Section VIII, Div 1 (or your location's pertinent Codes)
4. "Guidelines for Pressure Relief and Effluent Handling Systems", Center for Chemical Process Safety (CCPS) for the American Institute of Chemical Engineers (AIChE), (1998) New York
5. Fisher, H. G., Forrest, H. S., Grossel, S. S., Huff, J. E., Muller, A. R., Noronha, J. A., Shaw, D. A., and Tilley, B. J. (1992). "Emergency Relief System Design Using DIERS Technology: The DIERS Project Manual. AIChE, New York.
6. Fauske, H. K. "Revisitng DIERS Two-Phase Methodology for Reactive Systems Twenty Years later", Process Safety Progress, (AIChE; Vol. 25, No. 3), September 2006
7. "Easily Size Relief Devices and Piping for Two-Phase Flow", J. Leung, CEP, December 1996
8. Darby, R., Self, F. E., and Edwards, V. H. "Properly Size Pressure-Relief Valves for Two-Phase flow", Chemical Engineering Magazine, June 2002
9. Darby, R. "Size Safety-Relief Valves for Any Conditions", Chemical Engineering Magazine, September 2005
10. Simpson, L. L. "Estimate Two-Phase Flow in Safety Devices", Chemical Engineering Magazine, August 1991
11. Ouderkirk, R. "Rigorously Size Relief Valves for Supercritical Fluids", CEP, August 2002
12. Leckner, P. Six Part Series on Rupture Disks, Chemical Engineers' Resource Page

Various Websites including:

(1) The Chemical Engineers' Resource Page Forum on Pressure Relief Devices
(2) www.fauske.com

Is this extensive? You bet it is! Will you have time to read and understand all of this? NO WAY!!!! I've been doing this stuff for 20 years and still haven't gotten it all down. But I consider these to be MUST HAVES in any library if you are indeed serious about pressure relief system design. If you only are doing this occasionally, then References 1 thru 3 will be good enough as long as there is someone around that has the rest and is over-seeing your work.

[Art Montemayor]

Sara & all others reading this thread:

I was not kidding when I mentioned Phil’s vast experience and knowledge in the area of safety pressure relief – this thread and his posts will probably be frequent source of reference and learning for young graduate engineers and, hopefully, for ambitious and resourceful students looking for self-improvement.

I want to extend my personal thanks to Phil for this excellent series of posts on a subject that is very important and of practical worth and enrichment to many young engineers starved of mentorship in this very important area. I wish everyone reading and gaining knowledge and information from this thread would also extend their sincere thanks to Phil, who has obviously gone to time-consuming lengths in organizing, preparing, and publishing his comments.

Phil, once again, thank you for a thorough, experienced, and valuable lesson in not only designing PSVs, but also on installing them. I believe this thread will become one of the most popular and visited ones on this Forum. Your detailed reference list was a task to list out and I thank you for that as well.

[pleckner]

Art, you got me blushing!

Thank you very much for those kind words. I'm glad to help wherever I can and share any information that will help us all become better engineers.

[JoeWong]

Art, you got me blushing!

Thank you very much for those kind words. I'm glad to help wherever I can and share any information that will help us all become better engineers.

Phil,
Thumb-up for you !!!

As for maximum allowable overpressure / accumulation, ASME Section VIII, Div 1 allow upto 21% for fire case. But if your vessel is designed to other location's pertinent codes e.g. BS5500, GB150, etc, they may not allow upto 21%. (This has been highlighted in Phil's post "..........3. ASME Section VIII, Div 1 (or your location's pertinent Codes)......"

I have seen many engineers who work on international project make mistake on this. Just be careful whenever you are dealing with vessel not designed to ASME Section VIII, Div 1.

JoeWong