Valve Flow coefficient Cv

[al1979]

Hi All,

Go through this thread

http://www.eng-tips.com/viewthread.cfm?qid=209037

We are having exactly same problem as per above thread. The Cv values we are getting through software are very less but they are matching when valve is treated as CONTROL VALVE and are 40/50% less when valve is considered as ON/OFF valve for given size/class. Most of the Cv values of ON/OFF valves are derived from formulas in Crane Technical paper 410.

Is there any seperate procedure / standard practice to measure Cv of ON/OFF valves?

 

[MortenA]

No, for an open ball valve the pressure drop is very low (even if its not full bore) - the loss would be around 1-2 velocity head

 

[al1979]

I am concerned about Cv values I am getting from flow simulation software.
If I use procedure as per ISA/IEC for Cv computation of CONTROL VALVE, the values are 40/50% less than ON/OFF valves.

So I am confused which procedure to be simulated on software to get correct Cv values for ON/OFF Ball valve.

 

[MortenA]

I will maintain, for an open ball valve it dosnt make sence to assign a Cv value but you could do it this way (but you have to choose/guess a velocity head loss say in the range 1-2 - but that why you are an engineer ;-) ).

Say 1 velocity head, water and a velocity in the pipe of 1 m/s in a 4" opening => dP= 1000*1^2 = 1060 Pa= .153 psi

and the flow rate is .0083 m3/sec or 110 gal/min

Cv= 110*sqr(1/0.153)=281

Best regards, Morten

 

[katmar]

It is not clear to me what your problem is. Are you saying that for a given flow rate and pressure drop, changing the specification in your simulator from "control valve" to "on/off valve" changes the required Cv? This could be because an on/off valve is assumed to operate fully open, while a control valve would be assumed to operate at around 70% open. If the calculated Cv for the control valve has to be provided at 70% open the full open Cv could be double the calculated required Cv. Does your simulator allow you to specify the percentage opening of the control valve at the rated flow?

Perhaps include some actual numbers as an example to make clear what your problem is.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[al1979]

MortenA
Theoretically and by formula I can get the desired Cv but when I give same inputs to CFD Software it simply doesn’t give me desired flow (Cv) for ON/OFF valve.
Katmar
For 12” ON/OFF ball valve Cv is around 22,000 whereas for same size CONTROL valve it is 12,000.
When we physically measure Cv of full open ball valve, as per ISA/IEC procedure we get Cv around 12,000.
When I simulate this condition in CFD software I get Cv around 12,000.
Now my question is what condition shall I simulate in software to get Cv = 22,000?
I’m asking this cause I have to freeze the PROCEDURE / METHOD in CFD software. The same procedure can be used for partial open valve. This is because now days some crazy people are asking for % opening vs Cv curve for ON/OFF valves also. God knows why they want this.

Thanks & Regards,

 

[MortenA]

Double checked with engineering toolbox and they give a head loss of 0.05 - much below my assumed value of 12.

Replacing 1 with 0.05 gives 1257 in the 4" case and 11310 in the 12" case (close to what you mention). If you had a value why do you ask?

The enggineering tool box (

http://www.engineeringtoolbox.com/minor-loss-coefficients-pipes-d_626.html)

dont mention the assumption - but its probably "old school rule of thumb values". A full bore modern large valve could have much migher Cv/lower loss. Solving my little excel sheet that i made just for you gives me a head loss coefficient of 0.01 (assumed ID=12" exactly)

If you know its 22,000 then use 22,000???

I have recently approved that a characteristic similar to "quick opening" was used in a case similar to what you describe. I think this is optimistic but acceptable.

 

[al1979]

Thanks MortenA,

I want procedure to generate 22,000 value not 12,000 value. This procedure will be used for computation (in CFD software) of Cv at different openings to plot % opening vs Cv curve.

Cheers !!!

 

[katmar]

Although 12,000 seems very different from 22,000, when you express the differences in terms of pressure drop rather than Cv you are talking about very small differences. If you want this level of accuracy then you need to have all the dimensions exactly right. What inside diameters are you using for the pipe and valve bore in your simulation? What flow rate have you assumed? Have you allowed for pressure recovery downstream of the valve?

Modeling the ball valve as a thick orifice gives me a required valve bore of 10.9" if the pipe ID is 11.94" and I want a Cv of 12,000. I would be surprised if your CFD numbers are so different from this. There seems to be good agreement amongst the ball valve manufacturers that the Cv for a full bore 12" ball valve is in the region of 22,000 to 24,000.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[MortenA]

katmar has an important point! The actual pressuredrop varies little. In my excell sheet for a 12" valve with a 5 m3/hr flow rate the dP drops from 65 pa to 13 pa when ching the cv from 11300 to 25300 - both very hard to measure.