Max. backpressure conventional PSV for fire case 6

[Guidoo]

API RP520 part I, section 3.3.3.1.3 (page 36) reads:
"In a conventional pressure relief valve application, built-up backpressure should not exceed 10% of the set pressure at 10% allowable overpressure. A higher maximum allowable built-up backpressure may be used for allowable overpressures greater than 10% provided the built-up backpressure does not exceed the allowable overpressure"

It is my interpretation that when we have a PSV for fire case only, I can use a conventional relief valve provided that the built-up backpressure stays below 21% of the set pressure. However, the wording "may be used" in the API 520 sounds quite weak to me.

In this particular case we have a relief valve that reliefs directly to atmosphere.

What is your opinion here?

 

[22082002]

Though the discharge is to atmosphere, there will be little increase in back pressure on PSV because of release, you can neglect it, if it's very short distance. Or take 10% & be safe.
This is my personal opinion.

 

[Guidoo]

In this case they calculated a built-up backpressure of 17%. Probably the tail-end line turned out te be longer or have more bends than expected...

 

[SooCS]

I think there is a little confusion here. May be this will clear up a little.

Built-up back pressure refers to the pressure downstream of the PSV.

Allowable over pressure refers to the pressure upstream of the PSV.

In your case, provided that piping system downstream of the PSV to atm is the same size of PSV flange connection, you should not have any problem. For other applicatiion where the relief discharges tied-into a common header to the flare system, the back pressure should not exceed 10% of the PSV setting otherwise, the PSV relief capacity will required to be derated accordingly.

As for the allowable overpressure in the equipment/vessel the PSV is protecting, should not exceed 10% of the equipment/vessel design pressure on block outlet case or other process upset case. However on fire case, the max allowable overpressure may exceed the design pressure by 21%.

Hope the above clear thing up.

 

[aarti]

i think SooCS has hit the nail on the head. overpressure has to do with the psv relief pressure i.e relief pressure = set pressure+overpressure. however the built up back pressure is due to the psv popping i.e it exsist at the psv downstream after the psv has lifted. API says that the overpressure can be 10% or 21% of the set pressure in the blocked outlet (or any other case except fire case) & fire case resp. & for the conventional psv the built up back pressure < 10% of the set pressure.
besides i dont think u will want to increase your back pressure on the psv... than needed since higher back pressures mean a correction in the psv relief rate.

 

[Guidoo]

I do not think I am confused anywhere.
Point is with a conventional relief valve, the backpressure reduces the lift force that is holding the valve open. When there is a fire, pressure in the vessel increases until relief valve setpressure. At this pressure the relief valve starts to open. When the fire is not extinguished, pressure may increase inside the vessel to 1.21 times the relief valve set pressure (assuming the relief valve is correctly sized).
If the relief valve pressure increases more than the increase in vessel pressure, this may result in (partly) closing of the valve which reduces the valve capacity and this is of course not exeptable.

For example: if the relief valve set pressure is 10 barg, the allowable overpressure is 2.1 bar. So, I can use a conventional relief valve, provided that pressure drop over the tail-end pipe to atmosphere at max. flowrate is less than 2.1 bar.
In that case the delta P over the relief valve remains at least 10 bar in all circumstances.

 

[Guidoo]

To prevent any confusions:

If the relief valve backpressure increases more than the increase in vessel pressure, this may result in (partly) closing of the valve which reduces the valve capacity and this is of course not acceptable.

 

[pleckner]

To Guidoo:

Your interpretation of API is 100% correct and I can't see where the confusion is. Personally, I don't even see the need for interpretation as I think API is quite clear. A conventional valve is dependent on the forces acting on the spring. If you have a greater upstream pressure, then the backpressure at the relief valve can also be greater before the valve tries to close and you loose capacity, period.

A possible reason why the built-up backpressue is high maybe due to a choked flow condition. Said another way, with a given flow and pipe configuration, you may very well have reached critical pressure at a point just near the pipe exit. Therefore, the calculation to determine backpressure at the relief valve outlet flange (you start at the pipe exit and work back) must start with this critical pressure, NOT atmospheric pressure.

 

[MortenA]

Guido:

If this is for a vessel in europe be aware that the new PED does ONLY allow 10% overpressure no matter what. The API 520 sizing rule for fire is no longer valid in europe. Only 10% overpressure.

Best Regards

Morten

 

[Guidoo]

MortenA,

Yes, this is for a vessel in europe. But...

Official Guideline 5/2 (see

http://ped.eurodyn.com)

reads:

Guideline related to: Annex I Section 2.11.2, Annex I Section 2.12.

Question:In respect of pressure limiting devices, does the PED require that the permitted short duration pressure surge of 1,1 PS be maintained when the equipment is exposed to external fire conditions ?

Answer: The 1,1 PS restriction does not apply to fire.

Reasons: The requirement in Annex I section 2.12 for external fire refers to damage limitation, and does not serve the purpose of pressure limiting device in normal operation.

So we can still use API 520 sizing rule for fire. (The reasons given are a bit strange though...)

In fact, we have used the API sizing rule for fire on several projects that complied to PED. This was accepted by the notified bodies.

 

[MortenA]

Im actually only fully aware of the danish implimentation. This is a quite important issue for me.

I will look further into it.

Best regards

Morten Andersen

 

[CHD01]

1. Ref: Original Question: In reponse to the initial question, the selection of a conventional versus a balanced bellows valve should be made INDEPENDENT of whether you are sizing the relief valve to protect a vessel from exceeding a 10% or 21% allowable over-pressure! The answer to the original question is that you will limit the valve to a 10% TOTAL VARIABLE backpressure; which means (for the 17%VARIABLE backpressure you report) you will select a BALANCED BELLOWS valve or install larger discharge piping. This assumes a new installation; if existing, I recommend upgrading the valve to a balanced bellows or installing a larger line without question . If for some strange reason (does happen) a balanced bellows design is not available and you cannot increase line size, you must derate the valve per mfg. instructions. Derating of the valve is based on the sum of the built up AND superimposed backpressure as a percent of the relief valve set pressure expressed in terms of ABSOLUTE pressure(PSIA). Note this is different than the original comparison to select a conventional or balanced-bellows valve which is based on a percent of set pressure expressed in GAGE pressure (PSIG).

2. Ref Sonic Velocity at discharge to atmosphere: If sonic velocity is reached at the discharge, this is INCLUDED as part of the TOTAL VARIABLE backpressure on the valve. For example, if the relief valve is set at 100 psig and the vessel design pressure is 100 psig and the built-up back pressure due to flow (friction) is 5 psig and the pressure at the end to the discharge pipe is 10 psig with 10+5 = 15 psig pressure immediately after relief valve as it enters discharge line (due to sonic conditions being reached) with an atmospheric pressure of 14.7 psia, then the TOTAL VARIABLE backpressure is equal to 5 psi +(24.7-14.7) = 5+10 = 15 psi which is equivalent to 15% backpressure. So a balanced bellows valve is required.

3. Ref Europe vs API: To the best of my knowledge you still consider both 10% and 21 % overpressure based on whether a fire-case is applicable. Unfortunately, I do not have the needed regulations on file to verify this. Allowable accumulation for fire will vary between API, ASME and PIPING regulations, however. The more you learn, the less you are certain of.

 

[pleckner]

First, the cost difference between a typical (FARRIS) balanced bellows valve and a conventional valve is small enough to justify almost exclusive purchase of balanced bellows valves. This way one never really has to be concerned about the built-up backpressure affects.

Second, I guess we'll just have to agree to disagree as far as CHD01's point number one goes. Yes, if at the outset it is determined that the variable backpressure exceeds 10% of the set pressure during a non-fire scenario, then the balanced bellows valve must be chosen. However, just for a fire scenario if the variable backpressure exceeds 10% of the set pressure but still remains less than 21%, this should not eliminate the conventional valve from contention (see quote from API in original post).

The allowable variable backpressure is directly related to the allowable overpressure and should not always be assumed to be based on a 10% allowable overpressure (again, see quote from API in original post). Look at the case when you have two or more valves in the same service. You are allowed a 16% overpressure for the same non-fire scenario! In this case, the allowable overpressure in the system is NEVER calculated at 10%. We are dealing with opposing forces on a spring.

Most people just take the API statement of &quot;DO NOT LET THE VARIABLE BACKPRESSURE EXCEED 10% OF THE SET PRESSURE FOR CONVENTIONAL VALVES&quot; without reading further. The use of one or two statements in the Codes and practices do not preclude the need to read and understand the whole thing.

One last thing I have to say on this. No matter what I think or anyone else thinks, each company or institution has their own safety standards (or should). These cannot be any less stringent that given by the applicable Codes but can be more. So, you must always follow what your particular standard is no matter what. If your standard for conventional relief valves is never to allow the variable backpressure to exceed 10% of the set pressure, then so be it and enough said.

Just for the record, I agree with CHD01's second and third points.

 

[EGT01]

I think a lot of the confusion about allowable backpressure is related to the language in ASME Sec VIII, Appendix M, which in brief states &quot;will not exceed 10% of the set pressure when flowing at stamp capacity.&quot;

I believe earlier versions of API recommended practices used to cite this the same way. However, later versions have changed the language to reflect the original post.

In most cases, vendors stamp relief valve capacities at 10% overpressure which makes sense for the ASME language. But Appendix M is a non-mandatory section and I think would accept further engineering assessment of the guidelines. I think that the analysis of balanced forces on the relief valve disc presented by pleckner is a reasonable engineering assessment to say that allowable backpressure is related to overpressure. So of course this would mean checking the outlet line backpressure for the valve capacity at overpressure.

I would tend to disagree with the arbitrary use of bellows valves though. Bellows valves are inherently less reliable since bellows can fail over time for various reasons. Bellows valves require proper venting of the bonnet to function properly and this can become very difficult when dealing with highly toxic chemicals that would not be suitable for venting to atmosphere. If at all possible, try to design you discharge piping system to preclude the need for bellows type valves as a first choice.

 

[CHD01]

Yes, there are requirements and best practices (or plant standards). I can agree with you all (and do) and can also still support my position. While getting to less than 10% should be our first objective, I thought I made it clear I did allow for some flexibility. Also, I agree that a Bellows trim is indeed not that expensive; and while you do have to vent the bonnet and perhaps treat the discharge that should not preclude its use. Sometimes easier than larger pipe.

One comment, if I am over 10% I might sometimes stay with a conventional valve for an existing installation. But before doing so, I would insist on recalculating the required orifice size if the Built Up Backpressure exceeded 10% enough that it resulted in a derating of the conventional valve capacity. This is just to be very sure that it did not push us to the next larger orifice size; plus a recalculation of pressure drops for an increased capacity valve. One cannot just assume derating a valve results in an acceptable installation.


The more you learn, the less you are certain of.

 

[CHD01]

One other point, which is what is REQUIRED versus RECOMMENDED for the ASME Code and for API. Our bottom line is first complying with what is REQUIRED. The more you learn, the less you are certain of.

 

[MarkkraM]

I may be a little confused here so correct me if I'm wrong.

* To provide enough force to open a conventional PSV fully we require an inlet pressure 10% above the PSV set pressure. So at the starting point we are already 10% above the vessel design pressure.

* When the conventional PSV opens and thus fluid is flowing an additional back pressure generated in the discharge piping, which needs to be compensated for by a further increase in inlet pressure to keep the valve fully open

* If discharging a conventional PSV into a closed system the back pressure present needs to be equally adjusted for

* The piping run from the protected piece of equipment to the PSV inlet can have a pressure loss of up to 3%. This also may need to be considered as it could tip you over a 10% or 21% overpressurisation of the protected piece of equipment.

 

[CHD01]

Markham:

Seems to me you are basically correct in everything you say, if you agree I can re-state your note and include your observations as follows in calculating your required area initially:

1) Assuming your required relief flow is exactly equal to the capacity of the valve, and 2) the orifice is sized to relieve that flow at a 10% overpressure, and 3) you reduced your inlet pressure to allow for a 3% drop in calculating the orifice area, and 4) you further adjust your dp across the orifice for the outlet built up plus superimposed backpressure in calculating the orifice area required, then all of what you say is exactly the case. If flow is less then you will not get full lift and will get less than 10% overpressure. 5) There is one more wrinkle to this - it assumes that sonic flow occurs at the orifice, this is where the flow correction factor comes into play if sonic flow is the case depending on the type of valve used and the total superimposed and variable backpressure. While superimposed backpressure affects the set pressure as you say, builtup and superimposed pressure must exceed a limit determined by the manufacturere before the factor is less than 1. If sonic flow is not the case we use another equation to determine the relief area.

Hope I did not mistate something (its easy to do when you try not to write a book), but if I did someone will probably pick it up. The more you learn, the less you are certain of.

 

[EGT01]

Let's take a another look at MarkkraM's second point and make sure there is no misunderstanding.

When a PSV opens, the additional back pressure created is builtup back pressure. I would not say that builtup back pressure needs to be compensated by a &quot;further increase in inlet pressure to keep the valve fully open&quot;.

If we are talking about a case where we are using 10% overpressure, then the builtup back pressure could be anything up to 10% of the PSV set pressure without any concern at all. Ignoring superimposed back pressure for the moment, the balance of forces on the valve disc should be such that

set pressure + allowable overpressure > spring force + builtup back pressure

and the disc remains in position (fully open). If builtup back pressure is greater than the overpressure (builtup>10%), the forces become

set pressure + allowable overpressure < spring force + builtup back pressure

and the disc will start to close. Then you must start looking at ways to reduce the builtup backpressure or its effects by increasing the outlet line size or using other than a conventional type valve. I don't think increasing the valve overpressure is a viable solution.

Now MarkkraM's last point about the 3% inlet loss is one I've always had some questions about but that could be a whole thread to itself.

 

[MarkkraM]

I don't understand the logic that the PSV will stay fully open. Could someone have another go at explaining it.

To fully open a conventional PSV we need the force created from moving 10% above the set pressure. If the built-up back pressure is 10% does that not effectively cancel out the driving force applied against the spring to keep the valve fully open and thus it would start closing? Have I missed something fundamental regarding reseating of PSVs here?