Pressure drop across a control valve

[deccansher]

Hi,

For a flow control valve, having a fixed upstream pressure and downstream pressure, how does the pressure drop across the control valve vary with change in flow...or does the velocity change and pressure drop is constant?

 

[zdas04]

As MortonA said in your other thread, a control valve will control a single parmaeter. Either upstream pressure or downstream pressure, not both.

David

 

[ifreeman]

As the pressure drop between two points are fixed, the total pressure drop is fixed.
And in this case, the total pressure drop constists of line pressure drop and control valve dP.

For example, assume that you have total pressure drop of 5 bar and control valve dP=2bar.
Available pressure drop for piping is 5-2 = 3bar
Now, you throttle your control valve so that control valve pressure drop becomes 4 bar.
You have only 1 bar pressure drop available for the piping therefore your flow through piping will be reduced.
Generally the pressure drop across piping is proportional to square of volumetric flow rate.

The resultant flow will be

Qthrottle = Q1*sqrt(1/3)
where Q1 = original flow rate
and Qthr = throttled flow rate

If Q1 was 100m3/hr then throttled flow rate will be 57.7 m3/hr.

Regards,
ifreeman

p.s. [ "Realistic Control-Valve Pressure Drops", Chemical Engineering 1987 Sep., J.R. Connel ] <-- This will be a good starting point.

 

[BigInch]

So far you've got a couple of answers, I haven't looked at the other thread, and ifreeman discusses (quite accruately) the total available pressure drop in the system, but he also assumes that the outlet pressure of the downstream piping is NOT reduced to attempt to compensate and maintain flow in that piping to the same level before throttling the valve. I think that if you throttle the valve, you may also reduce system pressure at all points downstream, so that scenario is not necessarily true.

You ask about the pressure drop across the valve only, so I don't see how that addresses the question you asked, so I'll try...

You ask, "For a flow control valve, having a fixed upstream pressure and downstream pressure, how does the pressure drop across the control valve vary with change in flow...or does the velocity change and pressure drop is constant?"

Pressure drop across a valve is described by the valve's Cv Flow coefficient. Cv has units of gpm/(psi)^0.5, where gpm is flowrate in gallons/minute and psi is the differential pressure across the valve in psi.

So a Cv of 100 would mean (Pi-Po)^2 * 100 = Flow_gpm
where Pi = valve inlet pressure
Po = valve outlet pressure

This yields, (Pi-Po)^2 = Flow_gpm/100

So, discussing only the pressure drop at the valve and disregarding what might happen with total pressure drops in the upstream or downstream piping system, you can see that if Pi and Po remain fixed, and the Cv of 100 remains constant, there is NO change in flow across the valve, nor consequently, velocity either.





BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[BigInch]

I just saw the other thread. Since this is the latest one, let's stay with this one.

My comments to the other post are, that I agree totally only with Quark, for the following reasons,

MortenA says,

Your valve can't both control upstream and downstream pressure. Its either or.

My response there is,

Actually its neither, your valve (position and corresponding Cv) controls the differential pressure across the valve alone, nothing else.

Quark says,

no change.

My response is, Correct I agree! I'll go back and put a star there.

dcasto says,

We may want to quiz this question more. The upstream pressure may be fixed by another system (Valve, gas supply system with multiple control devices), hence fixed upstream, then the valve he is describing may be the one controlling the downstream pressure. Or the opposite where another system (Valve, speed controlled pump) is controlling downstream pressure and this new valve the upstream?

As for change in flowrate with a, lets used a different word, controlled up and down stream pressures, then yes the valve will be changing positions as the flowrate changes given the assumptions in the first paragraph.

deccansher, any more information on how you can get both up/down pressured fixed?

He sees that the new valve might not have any effect on upstream and downstream pressures if some other equipment is controlling those, and seems to imply that the control valve will change position to match whatever flowrate is going through it, since the flowrate through the new control valve is actually being controlled by the upstream and downstream equipment setpoints.

However my response there is,

It could work something like that, but why is the control valve needed, if upstream and downstream devices are controlling pressure and flow through that pipe segment now?

CJKruger says,

No, the pressure drop stays the same and the valve opening changes with flow rate.

My response is,

If the pressure drop stays the same, any change in valve position would probably change the pressure drop, if it was not stagnated, hence the valve position change would change the Cv and the flow would not stay the same.

I leave the floor open for return fire.






BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[CJKruger]

BigInch,

The situation deccansher describes occurs when you have a high pressure and low pressure separator with a control valve in the liquid line between them. The separator pressures are controlled by other control systems.

Thus, the dP across the liquid control valve remains almost constant, and the valve opening changes to change the flow rate.

The only change in valve dP is due to the change in friction loss in the piping to/from the separators. This effect is normally small.

 

[25362]

Consider n pipes of different diameters D_n, carrying a turbulent flow of an incompressible fluid with a constant [&Delta;]P_f. The velocity on each pipe would be proportional to D_n^0.5, that is to say the flow rates would be proportional to D_n^2.5.

It seems to me that, as with any other pipe flow restriction, when keeping the [&Delta;]P_f constant by external means, the flow rate through the valve would be directly affected by its % opening (stem position) in some, probably non-linear, relation.

 

[BigInch]

CJ,

Yes I agree with your answer and that your scenario is possible. I just wanted to emphisize those cases where its difficult to maintain constant pressure in both vessels. When I said "probably", I said that to discount the case where the upstream and downstream pressures can actually be held constant at whatever flow is going across the valve. Sometimes constant pressure is the desired operating method, however at large flows across the valve, the upstream vessel pressure tends to be reduced and the downstream vessel pressure tends to increase and the practicality of keeping them both constant is likely to decrease. If it is really possible to hold upstream and downstream constant... at any flowrate, I should/could have kept my big mouth shut.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[ifreeman]

BigInch,

You are right if you are talking about "dynamic behavior"
of piping system with relatively small system volume.

But, as you know, most of our work is done on "steady
state" basis. And the question given by deccansher was not
about maintaining the constant pressure at upstream/downstream. So I can say this is a steady state problem.

To make it more clear, I would like to suggest more defined problem here.
Imagine that there is a water tap which supply water to grass.
The source(water supply) pressure will be 'almost' constant
as the water system is very big and the destination
pressure(atmospheric pressure) is also constant.

Now I think we can concentrate on the control valve
drop vs. flow rate(therefore velocity etc.).

Regards,
ifreeeman

 

[BigInch]

Yes, I apparently have a hard time taking stead state as anything more than an ideal design concept.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[deccansher]

Hey Guys,

Thanks for the replies. Biginch and ifreeman in particular thanks a lot.

Let me make things little simpler......as i understood in school that we can have two types of control valves....pressure control valves and flow control valves.
My question is with regard to the 2nd option of a flow control valve.....

To elucidate with an example.....take a storage tank that has a valve at its outlet which discharges to ambient pressure. Now assuming negligible pressure drop for the line and the overhead tank always maintains the same liquid head.....the dP across the valve is fixed???

By changing the control valve opening....i control the flow! Does the dP ever change???

Hope this will help to resolve....or am i missing something?

Thanks again,
deccansher

 

[pleckner]

Why does a fluid move from point A to point B? It does so becuase it has been motivated to do so. This motivation in fluid flow is differential pressure, nothing else. It is not alive and does not decide its own fate. Heck, even you have a motivation to move from one state to another don't you? If there were no motivation for you to move, you wouldn't. If there is no change in the differential pressure, whatever the state at the time will remain so.

The control valve has a pressure drop associated with it as does the piping. The control valve is nothing more than a variable orifice. If I do not change the size of the orifice and piping, keeping everything the same, flow will not change, it can't. There is no motivatio to do so. As soon as I change the orifice size, I've changed the differential pressure and this will affect flow. The overall system pressure will not change (point A and point B) but the inbetween pressure drop(combination pipe and control valve) have changed. This is what happens at home in your sink. Opening the valve reduces the pressure drop of the control valve but this must be compensated by an increase in piping pressure drop. This is accomplished by a flow increase, and vise-versa.

 

[BigInch]

I know exactly where you're coming from. I think I spent quite a while figuring out that the only difference between a pressure control valve and a flow control valve is from where they get their input variables. Opening tends to increases flow and tends to reduce differential pressure drop no matter what kind of valve you call it.

Opening the valve will tend to reduce the system hydraulic gradient, however if the system pressures upstream and downstream somehow compensate for the reduced hydraulic gradient and maintain their same potential at a new flowrate, the new flowrate will be achieved at steady state. Likewise the reverse is also true.

Taking a "free body" diagram of the valve alone, disregarding what equipment is upstream or downstream, a flow change TENDS to change dP and a dP change TENDS to change flow. If the change, whatever it was, is not compensated for by a set point at some other piece of equipment, the change propagates through the entire system until it finds a new equilibrium point.

In your constant liquid level tank question, the flow will increase when you open the valve and pressure drop will remain the same. The REASON for that is that YOU are specifying a boundry condition, "The system somehow can compensate for the reduced hydraulic gradient caused by the reduced dP when the valve opens. YOU say, "THE LIQUID LEVEL REMAINS CONSTANT. A "natural" system boundary condition (one requiring no intervention) at the tank would be one where the water level would normally decrease, if you didn't keep on refilling it "somehow". The other boundry condition at the valve discharge is that, pressure remains constant at atmospheric pressure, which is another "natural" boundry condition that requires no intervention to maintain.



BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[25362]

A pilot-plant chap could tell us that a constant level upstream water tank can easily be achieved by providing a convenient overflow to the tank which is being continuously fed with an adequate excess.

Which brings us back to the ubiquitous Darcy-Weisbach equation and its applications to pipe fittings.

 

[BigInch]

The significance to the first system is still the same; defining a tank with infinite volume at a given level. That was tried 10,000 years ago at the Bosphorous Straits, but even the Black Sea eventually filled up.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[MortenA]

Biginch as i have posted in the other post: I may have been too strict in my answer. He writes: "For a flow control valve, having a fixed upstream pressure and downstream pressure, how does the pressure drop across the control valve vary" and how can the dP change if the upstream and downstream pressure is fixed?

Of course there is an infinite combinations of Pup, Pdown, Cv and flow rates that will produce the same Pup and Pdown but the dP remains the same. But strictly thats reading something into what deccansher writes that isn't there?

 

[quark]

I consider this as a good question after the brief about gravity flow from a tank. BigInch and Pleckner both supplemented the discussion with clear answers. Initially I was puzzled by the fixed upstream and downstream pressures but varying flowrates (incase of a gravity flow). Before visiting BigInch and Pleckner's posts I thought about it quite a while and did some calcs to reconfirm and I now clearly understood what they meant.

My conclusion (and also the suggestion from both the members) is that the upstream and downstream pressures are to be considered in the immediate vicinity of the valve. The gross upstream and downstream pressures give us wrong direction and thus violate natural laws. The pressure drop is very significant even for 1 m length pipe when acted upon 1m static head in a pure gravity flow.

The pressure, just upstream the valve, is lower incase of a open valve than a partially closed valve due to higher flowrate. But the extra pressure drop required during the reduced flow is contributed by the valve. Even the downstream pressure is held constant, say by not providing any pipe section downstream the valve, the pressure drop varies for every valve position and thus flowrate.

I did the same thought experiment of various orifice sizes as pleckner did and that solves the issue. Still, I am not sure about a case where a sandwich type buttervalve is directly bolted to a tank without any downstream piping. Now if we operate the valve at various positions, the flowrate should change yet maintaining constant upstream and downstream pressures, if we top up the tank at a rate equal to that of outflow. We shouldn't bother whether this is practically possible or not. We should have the theoretical answer.

There are two quack ideas as I never observed the actual case. Either the flow shouldn't change or the entire tank should offer resistance to the flow. That is the pressure at the bottom of tank may vary at various flowrates

I am interested to continue the discussion, in any case.

 

[BigInch]

I'll try.

I think this latest discussion is all due to the fact that, to find total pressure drop, either the inlet or the outlet pressure must be known and its sometimes difficult to separate the theoretical pressure drop of an individual element (or sum of several) from that practical reality. P1 - Cv dP^2 = P2
Now let P1 and P2 float. Note that it can be Cv for a valve, k for a pipe, maybe L*(r1^2-r2^2)*Q for a reducer, or whatever for any individual flow element, but the sum of all of them only ever equals dP for the system of elements, until you specify either P1 or P2. Even if flow is known, either P1 or P2 must also be known to solve the hydraulic system equations.

Even with a butterfly or sandwich check valve bolted to the tank, theoretically there is still an infitesimal length of pipe both upstream and downstream of the valve's "orifice".

OK, let's take it another step farther and consider just a hole in the side of a tank.

Now, theoretically and practically, inside the tank itself, as a fluid particle goes from rest away from the tank outlet to some infitesimal velocity, there is the transformation from static to dynamic head, and there is also the exit flow coefficient, Cd, to deal with that drops the hydraulic gradient as a fluid particle drifts from an area under the influence of only the liquid surface static pressure, accelerates and moves into the tank outlet and out into free air or into the inlet just before arriving at a valve's orifice. That loss can be significant, depending on shape and form of the outlet at the tank wall, a nozzle protrusion, [] or O x-section, curved or square, sharp or blunt edge, etc.

But, as soon as you say, "constant level"... you've fixed the upstream pressure and can then calculate downstream pressure, or v/v.

Was that it?


BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[reena1957]

Dear Big inch, Can't we say the valve exgorges pressure drop(energy) by opening its mouth wide or consumes it by closing shut, to maintain the flow? Why bother for the small losses at tank inlet and outlet? In other words, the CV(or any throttling valve) is capable of releasing or absorbing the available head to maintain the desired flow by varying its opening.

 

[BigInch]

Yes. I certainly think so.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com