3-way ball valve Cv

[jpolise]

I have an application where a 4" 3-way ball valve has a Cv of 840 in the 'through' position. Knowing that 3-way plug valves are notoriously poor when it comes to Cv, I found that the 4" plug valve has a value of 160.

Since we're having issues with the ball valve I wanted to compare what the effect would be with a plug valve. The difference in Cv's represents ~80% reduction.

Suppose I were to replace the ball valve with the plug valve. The existing flow rate is 350 gpm of a starch slurry (1.15 SG).

My first inclination is to evaluate the individual pressure drop expected as follows:

ball valve: 350gpm*1psi/840gpm ~ 0.4 psi
plug valve: 350gpm*1psi/160gpm ~ 2.2 psi

The additional pressure drop of about 1.8 psi would yield a drop in flow that would hardly be seen when applied to the pump performance curve. Yet when you consider the 80% difference in Cv, you'd think that the overall flow would be affected, i.e., 350*0.20 ~ 70 gpm?

My gut tells me to consider the individual pressure drop and that there would be little impact on capacity. Am I going wrong anywhere?

 

[MikeHalloran]

ball plug
350 350 gpm
840 160 Cv := gpm/sqrt(psid)
0.417 2.188 sqrt(psid)
0.645 1.479 psid



Mike Halloran
Pembroke Pines, FL, USA

 

[jpolise]

Mike,

The specific gravity is 1.15. The formula I used was Q = Cv*sqrt((psid/SG)).

I got the following:

ball valve = 0.209 psi

plug valve = 5.5 psi

That represents about 11 ft head to be added to the pump discharge. From the performance curve the resultant flow would be about 300 gpm. This would be about a 15% difference in capacity.

Do you agree? If so, this helps in the evaluation.

Thanks,
Jack

 

[gerhardl]

... totally apart from the difficulties in evaluating and balancing the Cv's, let me add to the confusion by mentioning that in my experience it is in most cases better to use two separate regulating valves (for instance v-ports), and if necessary two additional separate on/off valves.

This both to obtain better and more accurate regulation and better thightness (when closed) and longer standtime/less maintenance with cheaper, and more easily found on the market standard components.

... just a thought!

Good luck!

 

[MikeHalloran]

I was ignoring the SG contribution.
Adding that,

ball plug
350 350 Q, gpm
840 160 Cv := gpm/sqrt(psi)
0.417 2.188 sqrt(psid)
0.645 1.479 psid (water)

0.417 2.188 sqrt(psid/SG=1)
1.15 1.15 SG slurry
0.479 2.516 sqrt(psid/Sgslurry)
0.692 1.586 psid (slurry)

One of us is doing this somewhat less than perfectly right.
It could well be me; I will admit to considerable rust.



Mike Halloran
Pembroke Pines, FL, USA

 

[katmar]

You both seem to be a bit rusty - not the first time I have seen rust cause a valve problem!

The definition of Cv is
Cv = Q x sqrt(SG/[Δ]P) Q in USgpm and [Δ]P in psi

Rearranging for [Δ]P gives
[Δ]P = (Q/Cv)² x SG

This gives me 0.20 psi for the ball valve and 5.50 psi for the plug valve at 350 USgpm.

It is quite possible and reasonable for this to give a change of only 15% in the flow rate because there can be considerable pressure drop in the rest of the piping, and this will remain constant - or even decrease slightly if the flow rate decreases.

Katmar Software - Engineering & Risk Analysis Software

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[MikeHalloran]

Thanks, katmar. Rust removed. Okay, burnished a little.



Mike Halloran
Pembroke Pines, FL, USA

 

[cvg]

Losses occur in pipes as major loss and in system components as minor loss. The minor loss can be significant compared to the major loss - in fact when a valve is closed or nearly closed the minor loss is infinite. For an open valve the minor loss may often be neglected (a full bore ball valve).

http://www.engineeringtoolbox.com/total-pressure-loss-ducts-pipes-d_625.html

 

[jpolise]

Thanks to all for your help and responses.

 

[ports394]

What about installing a Tee and 2 full port ball valves? Is that an option?

 

[jpolise]

It's in a location where space is very limited as there is other piping and equipment nearby. I'm going to go with another 3-way ball valve with a different seat material that will hopefully address the seat wear issue we're having.

The TFE seats on the existing 3-way ball valve are not lasting more than 3 months before we start to get leakage by. This creates a plugging problem downstream as the solids in the slurry will settle out.

My plan is to go with a similar ball valve but with UHMW polyethylene seats that is supposed to withstand abrasive wear.

The application is a suspended starch slurry (~35% solids) that in our view is not particulary abrasive. If anyone has any other thoughts I'd be glad to listen.

Thanks for asking.

jp

 

[ports394]

it can be abrasive on a molecular level though.. we have the same issue with whole corn being very abrasive.

What other seat materials are available?

 

[cvg]

abrasion is also related to the flow velocity. reducing the velocity from 9 fps to something lower might increase the life of your valves. A larger valve would also reduce your pressure drop.

 

[gerhardl]

Now that the particular problem have beenn pinpointed:

1. Two valves for this purpose is almost always far better than one.
2. If ball valve: is inside V-port (not ball - shell form with V) possible/available? (Always better than normal ball for regulation)
3. I suspect starch solutions could gather and have peculiar properties (behave like solids and be highly abrasive) under certain circumstances (?)
4. In addition to flow velocity reduction, if possible: ceramics, or possibly other materials with long standtime for solid suspensions, should be preferred.

 

[jpolise]

I estimate velocity at about 8 ft/sec. I don't think that is excessive.

You mention a V-port above (Note 2). Remember this is not a throttling application. It merely diverts flow 90 deg from the straight through flow pattern.

The idea of two 2-way valves is good but is somewhat problematic in this case due to space and control system available points, etc.