Relief Rate Through a Full Open Globe Bypass Valve.

[Pavan Kumar]

Hi All,

I want to calculate the flow through a full open 1 1/2" globe valve for pressure relief valve sizing . This globe valve is the bypass valve to a control valve and I want to calculate the relief rate for in-advertant opening of this bypass valve. I have evaluated this scenario and found that in-advertant opening of this bypass valve can lead to over-pressure of the vessel. The process conditions upstream and downstream of the globe valve are as below:

Inputs:
Valve Size = 1 1/2"
Line size upstream and downstream of the valve = 1 1/2" Sch 40
Pipe ID = 1.610 inches = 0.13416 feet.
PSV Set Pressure = 72.5 psig
Relief Stream = Water
Relief Temperature = 248 Deg F
Upstream pressure, P1 = 116 psig = 116+14.7 = 130.7 psia
Downstream Pressure,P2 = 1.1* PSV Set pressure + 14.7 = 1.1*72.5 +14.7 = 79.75+14.7 = 94.45 psia

Calculations:

As an intial guess assuming the flow to be turbulent, I got the Darcy friction factor, fD from Pipe friction data table Crane Paper from Appendix A-26.

for 1 1/2" pipe, fT = 0.021

The equivalent Length, L/D for full open globe valve = 340 ( from A-27)

so , K factor is K = fT(L/D)= 0.021*340= 7.14

From the formula relating K to Cv, I calculated CV for the full open Globe valve as below.

CV = 29.9 *(d^2)/ SQRT(K) - Equation 3-16 from Page 3-4 of Crane TP410M

CV= 29.9*(1.610^2)/SQRT(7.14) = 29.9*2.5921/2.6720 = 29.005 US gpm/psi^0.5

Density of water ( relief stream ) at conditions of globe valve

is Rho = 58.8982 lbm/ft3
Specific gravity , G = 58.8982/62.4 = 0.9438

I calculated the flow through the full open globe valve using the formula below:

Q = CV * SQRT(DP/G)

DP = P1-P2 = 130.7-94.45 = 36.35 psi
CV = 29 ( calculated above)
G = 0.9438

Q = 29*SQRT(36.25/0.9438) = 179.7 US gpm

Relief Rate = 179.7*0.133681*60*58.8982= 84905.3 lbm/hr

Now that we know the flow, I calculated the friction factor through 1 1/2" Sch 40 line

Pipe ID,D = 1.610" = 0.13416 feet
Flow Area,A = (PI()/4)*(0.13416^2) = 0.01414 feet^2
Velocity, V = Q/ A = (179.7*0.133681/60)/0.01414 = 28.3 feet/sec
Density, Rho = 58.8982 lbm/ft3
Viscosity, Mu = 0.232 cP = 1.558X10^-4 lbm/ft-sec

Pipe roughness = 0.0018 inches

NRe = 0.13416*28.3*58.8982/1.558X10^-4 = 1434432.45

Using Chruchill Eqn

1/SQRT(f) = -4*Log(0.27e/D + (7/Nre)^0.9), f calculated using this eqn is fanning friction factor.

ffanning = 0.00511
fdarcy = 4* 0.00511 = 0.02044

This is close to the 0.021 I assumed initially, so the K factor and CV calculated is going to stay.

My question is, if the flow capacity of full open globe valve calculated above is correct. If not what corrections need to be made?. Any reference material that you can provide will be great help to me.

Thanks and Regards,
Pavan Kumar

 

[LittleInch]

Remove the valve?

I always argue that adding a manual valve of any sort, globe or otherwise, around a control valve removes your control function and creates hazards like this, where the risk IMHO, outweighs the benefits, which are pretty dubious. Once you put a valve there some fool will open it when they shouldn't and then leave it open.

If you really really want to do this then simply limit the valve opening by a mechanical stop to say 50% open or order a valve with a higher max open CV.

Or fit an RO in the bypass leg.

Far easier and cheaper just to not have it in the first place....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Pavan Kumar]

Hi LittleInch,

I will take your suggestion and suggest as one option. I will still have to size the relief valve for in-advertant opening of by pass valve scenario as the valve is already in the field right now. I wanted to know if the flow rate calculation I did using the formula below for the full open by pass valve is correct or not?

Q = CV SQRT(DP/SG)

where Q is in US gpm
CV is the valve CV for full open case
DP is pressure difference across the valve in psi
SG is the specific gravity of the fluid at upstream conditions.

Thanks and Regards,
Pavan Kumar

 

[LittleInch]

I believe that is correct

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

Its common on refinery high maintenance control valves. Don't forget the PI on the downstream side so the operator can easily check that outlet pressure.

Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[georgeverghese]

Opening the bypass valve will be supervised if a work permit is required, and this would be the case if the block valve next to the globe is normally locked close. Under these conditions, (and assuming there is a healthy work permit controls/management culture at this plant), at least one Operating Company allows the relieving pressure to be hydrotest pressure for RV sizing purposes.

 

[jari001]

Is there a reason why you can't use the actual Cv of the valve?

 

[pierreick]

Hi,
You may want to compare your Cv's value with the data in the chart attached.
Note: as other stated I don't like the idea of bypassing a safety valve. You need to have a work permit procedure in place to allow this kind of operation validated by management.
Note: Review Car seal valve equipment.

https://www.wermac.org/valves/valves_car_seal.html

Good luck.
Pierre

 

[Pavan Kumar]

Hi jari001,

Yes I could have used had I known that there are standard CV values listed. Not sure why my CV is twice as that listed in the chart ( 29 vs. 14 ).

Thanks and Regards,
Pavan Kumar

 

[Pavan Kumar]

Hi georgeverghese,

This 1 1/2" globe valve is the bypass valve to the Level Control Valve which has isolation valves which are normally open. The bypass valve is not car sealed but can be made to be car seal closed to be able to add safety using the work permit system.

Under these conditions, (and assuming there is a healthy work permit controls/management culture at this plant), at least one Operating Company allows the relieving pressure to be hydrotest pressure for RV sizing purposes.

Does this mean that if there are administrative controls to prevent in-advertant opening of block valves, (that can over pressurize the equipment protected by the PSV under question when opened)then the scenario can be written off if the pressure upstream to this block valve does not exceed the protected equipment's hydrotest pressure?.

Thanks and Regards,
Pavan Kumar

 

[Pavan Kumar]

Hi pierreick,

Thanks for the CV chart. I will use this hence forth. Can you please share the source of this chart?. Yes we can consider car-seal closing the bypass valve as everybody has suggested.

Thanks and Regards,
Pavan Kumar.

 

[georgeverghese]

Yes, that is correct. But check the process safety authority in your Operating Co buys into this approach. And the pressure upstream of the globe would be the PAHH pressure, at the least.

 

[shvet]

@Pavan Kumar
for info
1/ Flow area in calculation refers to inlet/outlet piping while actual flow area inside of globe valve is less as this type is not full-bore. Check velocity in seat to confirm that relief is not reduced by cavitation or even choked.
2/ This may help for admin control of inadvertent opening.

https://www.youtube.com/watch?v=YbxD-wowEjM

https://www.youtube.com/watch?v=qINy-gBdErM

 

[pierreick]

Hi Pavan,
Reference as requested:

http://www.cvltech.com/tech_iframe_Cvs_of_Valves_Chart.php

Note: We have left school for a very long time, don't ask us to check your work, you should have peers within your organization to support you.

Pierre

 

[Pavan Kumar]

Hi shvet,

So from what you are saying then it is not possible to calculate the CV for a globe valve as the inner area is not known and have to rely on the standard CV values per the chart ( or similar) goven by Mr. Pierre.

Thanks for the clarification.

Thanks and Regards,
Pavan Kumar

 

[Pavan Kumar]

Hi pierreick,

Thanks for sharing the document. I only wanted to make sure that the formula that I am using to calculate the liquid flow through the full open globe valve is correct. To make sure that I do not miss out on anything I decided to show my calculations. The inputs and the suggestions from this forum are very helpful and give the correct direction to me. Thanks to all for sharing your expertise with me.

Thanks and Regards,
Pavan Kumar

 

[jari001]

If it's an existing valve, you can ask the vendor for the valve details.

 

[georgeverghese]

In the GPSA chapter on Fluid Flow, L=55ft for 1.5inch globe valve.

 

[shvet]

So from what you are saying then it is not possible to calculate the CV for a globe valve as the inner area is not known and have to rely on the standard CV values per the chart ( or similar) goven by Mr. Pierre.

Misunderstood. Idea is:
Your calculation is ok provided that water flowrate is 180 gpm. This 180 gpm refers to valve inlet/outlet point. Actual flowrate in seat can become say 360 gpm or 1800 gpm as a consequence of cavitation. Or even fluid can be choked in case of sonic velocity.
That is the point to be checked as velocity in seat is obviously higher that in inlet/outlet nozzle.

Note Crane measured that K factor for a particular globe valve while friction losses for valve this kind of depends on design and conditions (e.g. seat erosion). This fact means your calculation refers to a hypothetical globe valve having nothing to do with a real safety valve during a real overpressure. Actual flowrate will be higher/lower depending on procurement history and similar.
I am trying to caution that your calculation should be reflected in procurement&construction&operating somehow otherwise this is just a funny engineering exercise.

This is the point why some engineers prefer controlled choking conditions for cases your kind of.

 

[GBTorpenhow]

for 1 1/2" pipe, fT = 0.021

This is close to the 0.021 I assumed initially, so the K factor and CV calculated is going to stay.

Just a note that the Crane fT referenced in the K equations is not the friction factor at the actual flow conditions. They somewhat confusingly use fT, the friction factor at fully turbulent flow for clean SCH 40 commercial steel pipe, to modify K per size for a given component. Crane's K value for a given component does not change with Reynolds #, connecting pipe wall thickness, pipe surface roughness, etc.

To echo sentiments already offered:
- a locked/car seal closed block valve upstream of the globe is usually the solution I lean to when a manual bypass is unavoidable
- advisable to use a 'real' vendor supplied Cv for the actual valve installed for something as sensitive as PSV sizing