20 replies to this topic

[Nate River]

Hello Everyone,

I'm a university student working as a summer intern at a refinery. I was just wondering if there's a creditable source (API Standard, manufacturer's handbook, etc.) that mentions the three percent rule. Is it just a rule of thumb or is it actually published somewhere? I've tried doing a google search and a search in Crosby's Pressure Relief Valve Engineering Handbook but have come up empty.

As I understand it, the three percent rule is a guideline that states that the pressure drop in the inlet piping to the PSV must be less than 3% of the vessel's relieving pressure. It is meant to prevent chattering and subsequent damage to the relief valve.

Thanks in advance,

Nate

[pleckner]

The 3% Rule falls more towards good engineering practice than a rule-of-thumb but is not as strict as some people would like to believe. It is referenced to the PSV set pressure NOT the vessel's relieving pressure. Also the pressure drop calculated only considers non-recoverable losses and it still doesn't consider losses from static height.

References:

1. ASME Section VIII, Appendix M

2. API RP520 Part II - Installation, 5th Ed., August 2003

[jprocess]

The 3% Rule falls more towards good engineering practice than a rule-of-thumb but is not as strict as some people would like to believe. It is referenced to the PSV set pressure NOT the vessel's relieving pressure. Also the pressure drop calculated only considers non-recoverable losses and it still doesn't consider losses from static height.

References:

1. ASME Section VIII, Appendix M

2. API RP520 Part II - Installation, 5th Ed., August 2003

Dear Phil,
1.Why the value is fixed on "3%"?Why not a lower or higher value?In the other hand I am interested of knowing about the origin of "3%" value.
2.Should we calculate the non-recoverable losses based on "rated capacity" of PSV?If yes as you know this value should be specified by VENDOR,so what approach the designer should follow?Should He/She hold the sizing of inlet line to psv up to the supply of data from vendor or it is possible to apply a rough margin to calculated relieving load and obtain the rated capacity?
3.Is the 3% rule applicable for inlet line to blowdown valves?
Thanks in advance.

[pleckner]

What goes up must come down. In the case of PSVs, what opens must close. A PSV closes when the system reaches the PSV blowdown pressure. This pressure is mainlly a function of the mechanical PSV and varies between vendors and models. However you will find that most (not all) have blowdown pressures in the range of 7 to 10% of the PSV set pressure. Some point had to be set to limit the inlet line pressure loss to prevent the valves from chattering during a relief; chatter being the violent and quick opening and closing of the valve. So, the ASME committee decided on 3% of set pressrue as being that number, which should provide an ample safety margin.

The flow rate to use in calculating the inlet line pressure drop is the PSV rated (stamped) capacity. As the Process Engineer, you are responsible for sizing the PSV. Your company probably has one or two vendors they deal with so you can choose the vendor and model of PSV your company tells you to use. With the vendor, model and size now known, you then get the stamped capacity of the PSV directly from the vendor cataloge.

The 3% Rule applies to spring operated, pop-open PSVs.

[pleckner]

Correction: The 3% Rule also applies to pilot operated PSVs.

[jprocess]

Dear Phil,
In calculation of inlet line pressure drop,should we consider the density value at winter conditions?
Cheers.

[pleckner]

You use the conditions at relief, whatever that may be.

[pleckner]

You use the conditions at relief, whatever that may be.

CORRECTION!!!

This is what happens when your mouth goes before the brain. You use the conditions at SET PRESSURE, whatever that may be. NOT RELIEF.

[jprocess]

You use the conditions at relief, whatever that may be.

CORRECTION!!!

This is what happens when your mouth goes before the brain. You use the conditions at SET PRESSURE, whatever that may be. NOT RELIEF.

But I think that using the conditions at relief seems to be more logical.

[pleckner]

What logic? You're dealing with a committee.

That's the way it is.

[JoeWong]

Phil,
Again, sorry for cut-in...

jprocess,
PSV will open at setpressure, pressure possible accumulate upto 110% of setpressure (non-fire case).

Let review the effect of pressure on flow (hence inlet line loss) with selected PSV orifice area


Effect 1 : Lower pressure ==> Lower density ==> larger volumetric flow (with same mass flow) ==> Higher inlet line loss
Effect 2 : Lower pressure ==> Lower DP (less significant on critical flow) across PSV ==> lower driving force ==> Lower flow ==> Lower inlet line loss

Maximum inlet line loss may/may not at setpressure nor Relieving conditions.

If you use rated flow (based on higher DP) will properties obtain from setpressure, then it will be conservative.
I would recommend you to conduct a little more studies to convince yourself (prove my comments), the best would publish you results here for benefits of all.

Just a little extra on Phil's statement "The 3% Rule also applies to pilot operated PSVs.", Phil statement is correct. However, you may have exception if the pilot line tapping is located elsewhere i.e. vessel. Refer to API RP 520 Pt II, section 4.2.


I would encourage you to read "Using the ideal gas specific heat ratio for Relief-valve Sizing", CE, Nov 2003, by Aubry SHACKELFORD. This effect will be more significant in sizing PSV and as well as inlet line loss.



JoeWong

[pleckner]

@Joe:

You're reading too much into why the API Committee says to use set pressure for the 3% Rule. As a note, there is discussion about changing it to relief conditions instead. We are only talking about a 3% (+ or -) pressure loss, not a very significant change to vapor/gas desnisy in the scheme of things. As a matter of fact, the 3% Rule also applies to liquid flow where there is not only no density change but you still aren't supposed to take into account static head; doesn't sound very conservative to me. And your statment, "If you use rated flow (based on higher DP) will properties obtain from setpressure, then it will be conservative." confuses me. you are supposed to use the PSV rated (stamped) capacity in the calculation wheather you use the set pressure or not so there is no additional conservatism introduced.

In your brief review of the effect of pressure on flow with a selected orifice size, your Effect 1 is basically negligible and your Effect 2 to me is non-existent. It isn't less significant on critical flow; it is no significance on critical flow. And what driving force across the PSV?

So I stand by my statement, "What logic?"

[JoeWong]

Phil,
Thanks for knocking my head to think more...

You're reading too much into why the API Committee says to use set pressure for the 3% Rule. As a note, there is discussion about changing it to relief conditions instead.

You may be right that i read too much into it. However question raised by my young engineers really push me to read and think more into it so that my statement / judgment is convincing.

Moreover, your second statement already shows that API committee still in discussion on which conditions to be used.

We are only talking about a 3% (+ or -) pressure loss, not a very significant change to vapor/gas desnisy in the scheme of things.

Most peoples aware that 3% rule is just a guideline to comply but it still can be deviated. I personally has deviated at least once (upto 4.5%) but selected a PSV with high blowdown in order to maintain the margin between opening & closing of PSV. So the 3% rule really is not strictly needs to be comply and one may exceed if one understand the background on the margin.

As a matter of fact, the 3% Rule also applies to liquid flow where there is not only no density change but you still aren't supposed to take into account static head; doesn't sound very conservative to me.

I remember in one of your post, you mentioned that 3% rule does NOT include the static head. I have not put much thought on this issue. I do agree (prelim) as margin loss on the line pressure drop during PSV opening, you will gain it back when the PSV is closing / reseating. I think i should put extra thought on this subject.

And your statment, "If you use rated flow (based on higher DP) will properties obtain from setpressure, then it will be conservative." confuses me. You are supposed to use the PSV rated (stamped) capacity in the calculation wheather you use the set pressure or not so there is no additional conservatism introduced.

In your brief review of the effect of pressure on flow with a selected orifice size, your Effect 1 is basically negligible and your Effect 2 to me is non-existent. It isn't less significant on critical flow; it is no significance on critical flow. And what driving force across the PSV?

Yes. This may create some confusion to certain level.

My starting point is ....a PSV will open at setpressure, pressure possible accumulate upto 110% of setpressure (non-fire case). Those the properties at 100% setpressure to 110% setpressure will be different (i agree it may be insignificant).

At 100% setpressure, density lower, higher pressure drop is expected with same mass flow.
At 110% setpressure, density increases, lower pressure drop is expected with same mass flow.
Thus lower density, higher volumetric flow, which probably result higher pressure drop.


At 100% setpressure, the driving force across a PSV (selected) orifice will be 100% minus backpressure (subcirtical flow) or Critical pressure (generally 50-55% of PSV inlet pressure)

At 110% setpressure, the driving force across a PSV (selected) orifice will be 110%* minus backpressure (subcirtical flow) or Critical pressure (generally 50-55% of PSV inlet pressure)

* to be exact we may need to minus inlet line loss

If we go in detail, we may find that driving force for 110% case is slightly higher which potential gives higher flow across PSV. High flow will cause higher inlet line loss.


Increasing pressure ==> increase density ==> increases volumetric flow ==> less pressure drop
Increasing pressure ==> increase driving force across PSV ==> increase flow passing PSV ==> higher pressure drop

Thus, maximum inlet line loss may/may not at setpressure nor Relieving conditions.

If one use properties at setpressure (lower density), calculate rated flow considering PSV inlet is at 110% set pressure (higher driving force), then probably the inlet line loss will be conservative. This scenario is not a feasible scenario but just built in the conservatism.



Write upto here...looks like a bit confuse...i would like some of you to point out mis-concept if any...

Anyway, i have done some studies years ago, the pressure drop is not so significant whether you use set pressure condition or relieving condition.

The consideration of "using ideal K for sizing PSV" is really give significant flow different as compare to using set pressure condition or relieving condition to determine the PSV rated flow.

Any of you who is reading this post, i would encourage you to read the findings in "Using the ideal gas specific heat ratio for Relief-valve Sizing", CE, Nov 2003, by Aubry SHACKELFORD.


JoeWong

[latexman]

We calculate inlet (and outlet) dP at the rated flow of the selected PSV at 10% overpressure.

The 3% rule is met on essentially all new installations.

On existing installations, we try real hard to meet the 3% rule. If there are good reasons and it is exceeded, there's extra documentation to inform and archive the reasons and the blowdown must be adjusted. No matter what, inlet dP must not exceed 5%.

[d2205]

On existing installations, we try real hard to meet the 3% rule. If there are good reasons and it is exceeded, there's extra documentation to inform and archive the reasons and the blowdown must be adjusted. No matter what, inlet dP must not exceed 5%.

i'm a student, currently interning at a refinery. I have been reading about this topic and would like to know the consequences if this 3% rule is not obeyed.

I have read Merkblatt AD-A2 and ISO 4126 which requires the the implementation of this 3 % rule. but I have also been told, that these guidelines are optional, BUT in a way a company is taking a risk by not adhering to these guidelines.

what if implementing the 3 % rule is not possible in real situation? (as far as i have been told, costs to enlarge/shorten/adjust pipings and also pilot operated valves are very expensive) Other than chattering, what else can possibly occur if the 3% rule is not obeyed?

I have also read from a source(which I have forgotten to write down), that with more than 3% pressure drop in the inlet pipe an overpressure could occur in instruments. Is it true? Is it caused by the pressure accumulating in the instruments, not being able to overcome the large pressure drop in the pipeline?(self-derived solution.if this is not the case, appreciate a further explanation.)

Lats but not leas, what are the 'good reasons' and where can i view the documentations mentioned above. And why 5%?

Appreciate any help I can get!

[skearse]

The 3% rule seems to be one of those ASME "recommendations" that API does not necessarily agree with. I say recommendation because the actual wording of the ASME Code in UG-135(B)(1) doesn't actually state 3%; the only place that value actually appears is in Appendix M, as Phil mentioned. Appendix M, though, is actaully a non-mandatory Appendix, and even then, applies to compressible flow. Now, having said that, since it has been "adopted" as an industry standard, if you go higher than that, you do open up yourself to additional risk, and API 520 Part II even states that if you go higher than 3% that must be supported with an "engineering analysis" (section 4.2.2). The 5% value that latex mentioned seems to be sort of an adopted industry absolute maximum, but I don't know that there is any factual basis for going to 5%. In some cases, 3% is physically impossible to obtain-larger valves with low set pressures, combination installations with rupture disks (in these cases you must include the pressure drop through a burst disk), etc. can easily chew up the 3% even if installed directly on a nozzle simply by virtue of the entrance losses.

Last I heard, API commissioned a study to look into the 3% rule and revise it if possible.

We take the same appraoch as latex with our installtations-all new installations are 3% or less, and retrofits try to get below 3% if possible-and never above 5%. Anything above 3% gets extra attention, particularly to the manufacturers stated blowdown values.

The other thing I would mention is that for pilot valves, the 3% rule applies to the sensing lines for flowing-type pilots; for non-flowing types API 520 Part II paragraph 4.2.3.2 recommends lines with a flow area of 0.07 sq. in. (45 sq. mm).

[fallah]

Lats but not leas, what are the 'good reasons' and where can i view the documentations mentioned above. And why 5%?

Appreciate any help I can get!

Normally,all types of PSVs are fitted with a blowdown adjustment ring set at 5%.

If pressure drop of inlet line exceeded 3%,let say 5%,when relieving to be started the pressure sensed by PSV is to be equal to blowdown pressure and would be closed without complete relieving and following that chattering would be happened.

If we have to have inlet pressure drop higher than 3%,we shall adjust blowdown with higher value,let say 7%,and at this time the pressure drop could be up to 5%.

[demank]

Lats but not leas, what are the 'good reasons' and where can i view the documentations mentioned above. And why 5%?

Appreciate any help I can get!

Normally,all types of PSVs are fitted with a blowdown adjustment ring set at 5%.

If pressure drop of inlet line exceeded 3%,let say 5%,when relieving to be started the pressure sensed by PSV is to be equal to blowdown pressure and would be closed without complete relieving and following that chattering would be happened.

If we have to have inlet pressure drop higher than 3%,we shall adjust blowdown with higher value,let say 7%,and at this time the pressure drop could be up to 5%.

I think 3 percent is necessary for PSV body lifetime. because is more higher pressure drop of inline, the PSV will fast to open and to stop which will damage the body.

[JoeWong]

I think 3 percent is necessary for PSV body lifetime. because is more higher pressure drop of inline, the PSV will fast to open and to stop which will damage the body.

I guess what you inferred was the CHATTERING.

[phosty]

I have been reviewing the article "Easily Size Relief Devices and Piping for Two-Phase Flow" by J. Leung, Chemical Engineering Progress, Dec 1996.

I realise it is an old article, but in it Leung presents an equation (Eqn 55) that is used later in the article to calculate the max inlet line length for a given ID assuming the maximum 3% inlet line losses:

dP_inlet = 0.03(P_set - P_back)

P in absolute pressure

i.e. the allowable inlet line pressure drop is limited to 3% of the absolute set pressure less the psv absolute back pressure.

I realise the wording in API 521 Part II, section 4.2.2 is quite specific "...the pressure-relief valve should not exceed 3 percent of the set pressure of the valve..." but I can see where he is coming from. Is there any merit in this approach for conventional PSVs (i.e. not balanced bellows or pilot operated PSVs)?

If so, would P_back not have to be the superimposed back pressure at the rated flow rather than the static back pressure used in the PSV sizing equations?

[pandora]

Just a few thoughts to consider:

1 - The 3% rule applies to the stagnation pressure NOT just pressure, which includes kinetic energy
2 - The rule exists to prevetn "chatter" and potential "self destruct" of the relief valve - pressure rises to set pressure, pressure drop from flow cause pressure to drop at valve throat, relief valve reseats, pressure rises, relief valve lifts - rinse and repeat

As a side effect the valves capacity will be severely reduced and a much higher overpressure will be required to keep the valve open