Control valve - flow and pressure

[jack880]

Hi. I have a flow from a pump, and soon after the pump discharge want to be able to take off half the flow at times, and regulate the pressure. Is this possible with an isolation valve and single control valve as shown in the attached diagram? Thanks in advance

 

[gerhardl]

Which pressure do you want to regulate, and how?

The flow through a valve (roughly speaking) is Q = Cv (Cv of valve or total constriction) multiplied by the root av delta p (pressure difference before and after the valve) / divided by specific gravity of fluid.

After the control valve you can either regulate the flow, or you can regulate the pressure, not both at the same time.

If you have a direct pilot operated control valve (operated by spring and liquid pressure before and after the valve) it is normal to install and set the valve to keep a constant pressure before or after the valve.

If 'everything' is constant (pump variatons, pipelines, pressures etc) this will also give a constant flow through the valve (Note!: only if variations are damped enough to be considered constant within tolerance limits for he valve)

If you want to limit the flow, you will either have to select components that at max possible pressure difference do not have a througlet (a CV) that gives a flow larger than your limitation, or you have to measure the throughlet and poweroperate or power-close your valve.

 

[jack880]

Hi, thanks a lot for the reply.

I think it's my fundamental understanding of flow and pressure that's probably lacking.

Please see the attached sketch. When the valve is closed, all the flow from the pump goes round a recirculation system. I want to be able to open a valve (close to the pump discharge) and divert up to half the flow down another branch. I think that for the flow to split in half the pressure drop across the valve needs to be the same as the pressure drop through the rest of the system (i.e. the pressure drop in each branch needs to be equal)? But will this then leave no pressure downstream of the valve, rather than the required residual 3-5 m?

The end of the side branch is a free discharge. So does that mean that the pressure will always be 0, so i need to open the valve such that the flow is at the upper limit, and when an additional restriction equating to 3-5 m headloss is applied to the pipe outlet the flow will reduce to some flow in the required range? Hope that makes sense...?

Thanks again

 

[jack880]

Just noticed a mistake on the sketch - in the scenario with the take-off valve open, the flow going straight up will be 0.4 l/s minus the flow down the take-off branch, i.e. the total flow from the pump will remain at 0.4 l/s

 

[jack880]

I've been thinking a bit more and think that the client's requirement of 3-5m pressure at the free discharge tap is ambiguous.

If they want the ability to stop the flow with 3-5m back pressure, I'd need a pressure control valve to stop the flow when a downstream pressure of 3-5m is reached, to protect anything connected to the tap from the possibiliy of seeing 30m pressure.

Alternatlvely they may want it such that they can connect a device to the tap which will add 3-5m headloss to the system, and need to still be able to achieve the required flow.

I will ask them what they mean....

 

[tr1ntx]

talk to OCV, website: controlvalves.com

Yes, you're correct, for the flow to be split equally, the head in each branch would need to be equal. Then if they're talking about periodically adding a device which increases head on the diversion line, that will change the flow. Perhaps they can live with the momentary /short-term change and the equal flow requirement is not set in stone for all conditions. I've seem that, where you want to split the flow evenly so that your sample line will have the same mixture/concentration as the main flow, but once you get it flowing down the line, you can grab your sample quickly and not worry about the change you're inducing back at the split. If they need a long duration sample though, you're looking at a more complex system with pressure control valves on both lines and a high pressure relief valve.

OCV (and others) can supply you with a pilot-operated pressure equalization valve which will give you equal pressure in both branches, providing there is a certain minimum level in the main line to begin with.

 

[jack880]

Hi, thanks for the reply. I was thinking of putting an orifice plate in the short diversion line to equal the headloss in the much longer other line, so then the tap on the diversion line was opened the flow would be equal on these 2 lines. As you say, if they then increased the headloss slightly on the diversion line, the flow will drop a bit, which should be fine. Thanks for the link - I'll check them out once the client has confirmed what they actually want.

 

[tr1ntx]

Orifice plate works too, the old-fashioned simple solution (provided your flow conditions allow it).

One more thing, use a "Y" to split the flow. Don't simply "T" into the side of the main pipe at 90 deg. (I mean you can come off the main straight run at 45 deg or probably even 60 deg, I don't mean to say that you have to install a true "Y" splitter.)

 

[jack880]

Hi, thanks for the reply. Are you saying that to minimise headloss? I don't care about headloss, I have way too much head anyway. If I used a T instead of a Y wouldn't that just mean my orifice plate would have a slightly bigger hole?

Thanks again

 

[tr1ntx]

The only reason I said Y in to the pipe is that I'm assuming that they want to get the same mixture/concentration down the sampling system as they have in the main flow. Coming off at 45 does a better job of that than a T. Well, in some situations it makes a difference. If it's a homogeneous fluid mixture, it doesn't matter, or fine suspended solids.

I saw a situation once where they were pumping sand plant wash tailings and had a 2" sample line T'd into the top of the pipe. The pump vendor had calculated the speed for a new pump and it was delivered skid-mounted with drive system installed. It started up fine pumping water, but very quickly after the material started flowing, they plugged the line. Their samples had not been representative of the true % solids, in fact were WAY off.

 

[jack880]

Ah ok, i see your point. It's clean water from a service reservoir but not sure what they're testing the sample for - should have only very fine solids in if any I guess. Still, I'll raise the point. Thanks again.