vfd booster pump control 3

[evan63]

WE have a problem with water pressure at our plant.The pressure ranges from 70psi to 40psi and often there is a significant enough change in pressure to shake the water lines.The booster pump we have is driven by constant speed motor at 1750rpm which we would like to be able to control the speed and try to maintain a smoother water pressure.We
have installed a vfd(Yaskawa G515) and inverter duty motor for the booster pump.The way we are controlling the vfd is by a pressure transmitter(4-20ma)signal insalled on water line, to a honeywell controller(udc2500) and have the honeywell control the vfd and tune the PID on the controller so there is the correct response.Do you guy's think this will work and if you have any suggestions I would appreciate the advise.

 

[jraef]

That's the way to do it!

Some VFDs have the PID control function built-in, I would have thought the Yaskawa was among them. But no matter, I actually favor doing it the way you did because if anything goes wrong with the VFD, the loop controller stays behind and is already set up to control another VFD, of any make, model etc, that you could get in a hurry.

 

[evan63]

Thank's jraef,We haven't got it online yet but I will let you know how it goes.

 

[evan63]

The reason we were using the honeywell controller is that we are familar with the PID loop tuning on it.We don't have any experiance with the Yaskawa drive PID tuning at this point.If you have any input you could give on tuning the PID on the Yaskawa drive I would like to learn.I'm familar with the basic set up of vfd's,like choosing the correct control method which in this booster pump we selected open loop vector because we don't have a closed loop(no encoder)because it doesn't need to be that pricise.

 

[itsmoked]

Your description of the magnitude of the problem would probably have me NOT trying to cure it with a VFD. You don't get pressure drops radical enough to "rattle the pipes" from just common flow variations.

What you need is what we electrical engineers use a to eliminate or control sudden voltage fluctuations - a capacitor! In your water realm this would be a bladder tank. Every domestic well typically has one. It is a tank with a rubber bladder stretched across the middle. The bladder has about 8psi on one side. The other side is directly hooked to your water system. The water will fill the tank almost full which compresses the air on the other side to the water pressure.

In a well system this will greatly reduce the pump from thrashing as it allows the pump to fill the bladder tank while storing energy. Once the system pressure reaches cut-off the pump is stopped and the user can continue to use water for quite a while until the energy stored in the compressed air diminishes to a fraction of the charged state.

In your case this would represent extra water capacity that could be distributed throughout your facility. I'd put them near any large use point that has a cyclic nature. You would also want to put them at the end of long runs away from your source. This would allow small runs of pipes or runs that have become too small over time, to continue to service large sudden demands. This prevents the water velocities in the long pipe runs from having to skyrocket to meet a sudden demand, not being able to accelerate the water, resulting in banging and pressure drops.

Most large plant air distribution systems use this same method with Accumulators spread around to greatly reduce air pressure spikes and variations.

Since you've already installed the VFD, fear not, as your VFD would work exceptionally well with this distributed storage. It would also have something to work with rather like the aforementioned domestic pump, working with the storage to keep the pressure up, and steady. I don't believe you will have any real success without distributed storage.

Bladder tanks come in all sizes from about 1/2 a gallon to hundreds of gallons.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[evan63]

Thank's itsmoked,your suggestion makes sense.I think my boss is trying to correct the problem without looking at the whole picture.Correct me if I'm wrong but I think what your saying is the water pressure will still bang and drop even with the vfd trying to level out the pressure because of the large volume of water distributed throughout the plant.

 

[itsmoked]

The VFD by itself will not solve your problem because of the inertial content of your medium, water. If the pressure drops it has to be detected by the sensor which is mounted where(?) in your facility's matrix of plumbing. The sensor must relate this to the VFD, the VFD must then command the motor to speed up. This speeds up the pump's impeller which has to then accelerate the water in the pipes to get the pressure back up. You can easily see where a sudden demand way down the pipe occurs, then the demand ends just about the time the pump has responded... Too late for the pressure drop that occurred but now you're headed for a pump created overpressure. If you had bladder tanks the immediate demand is entirely covered by the tanks. The VFD/pump then just has the job of recharging the tank in a nice linear fashion as rapidly as possible, or desired.

Example: A large demand is made near a distant storage tank. The demand is a step function like a valve opened to fill a product mixing tank. The otherwise sudden drop that would occur is covered by the storage tank but the flow continues and the pressure would drop as related to how big the demand is verses the tank's sizing. Meanwhile the VFD/pump has time to bring the pressure back to the desired level and the new open-valve flow, coming in smoothly to the target.

Here's the first link I hit on Bladder Tanks, also called diaphragm tanks.

http://www.watertanks.com/category/48/

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Valvecrazy]

I agree with itsmoked. The problem is probably the rapid changes in demand, not the pump. A few strategically placed bladder tanks might solve the problem. You can also incorporate a pump control valve that is made to work with a bladder tank. This will have much better effect than a VFD. A VFD will most likely amplify the transients instead of eliminating them. You already have at least one problem. Adding a VFD will also amplify your number of problems. See the following thread where this was discussed earlier.

http://www.eng-tips.com/viewthread.cfm?qid=191266&page=2

 

[DickDV]

I note that the original poster stated that his pressure ranged from 40-70psi. This problem, if it is a problem, is exactly what a VFD/motor/pump system can correct.

The sudden transients are a separate issue and the accumulator tanks sound like a good idea to correct that.

Actually, as already mentioned, the combination of accumulators and VFD pressure control make for premium performance and would be highly recommended for the situation described.

 

[Valvecrazy]

Accumulators simply give the VFD a little extra time to respond. The delay in response of a drive can still amplify the transients. A drive can maintain a fairly constant pressure by increasing flow when the pressure drops and decreasing flow when the pressure rises but, so can a simple pump control valve without the complications and side effects of a drive. Some pump control valves can react almost instantly to changes in pressure. This fast reaction is what helps keep from amplifying the transients.

 

[LionelHutz]

What pressure do you want? I noticed you said the pump is operating at 1750rpm and the pressure varies between 40psi and 70psi. The VFD won't make the pump capable of pumping more water so the pressure will still drop to 40psi at high flow demands. Are you using some type of pressure relief valve to limit the pressure at 70psi?

 

[evan63]

We finally fired up the vfd today to see if we could maintain pressure.First we opened up the inlet valve going to the pump,then we opened the outlet valve from the pump,then we started the pump(vfd) and closed the main supply.The water supply was then routed through pump and into the plant.Looked good at first as the vfd was controling and maintaining about 70psi,but as soon as we had a strong demand for water the drive was unable to do what we hoped.The pressure still dropped to around 40psi and the pipes banged when the demand was gone.The only thing the drive(vfd) improved was that it shortened the time it took to get back up to pressure(70psi).We watched it cycle a few times and it was the same.We even set the honeywell controller up as on/off instead of using the PID loop control so when there was demand the controller would go to 100% output instantly but the pump just couldn't make up that much of a pressure drop.So, with that being said I Understad what you guy's mean about the need for a bladder tank.I'm not sure what my boss plans on doing from here but I'll do a follow up post when I find out.I did mention what you guy's had posted and appreciate all of the expert advise as I am a maintinance electrician not and engineer but I'm very intersted in learning about things just like this problem.

 

[itsmoked]

We would love to see a sketch of your main plumbing situation. We have a truck load of pump exspurts here that could provide some great suggestions with that info.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Valvecrazy]

You must set it at 40 PSI or get a bigger pump. A VFD can only reduce the flow from a pump to maintain a set pressure, not increase it. This is why a simple valve can do the same thing with a lot less problems.

 

[DickDV]

Actually, valvecrazy, a VFD can increase flow and pressure in most cases by using overspeed to bring the pump hp up to match the motor hp. The pump has to be able to cope with the small amount of overspeed but that is usually not a problem.

And, of course, if the pump hp and motor hp match at motor base speed, overspeed isn't possible but that is a rare case. Almost always, the motor is oversized for the pump so you can take advantage of that.

 

[Valvecrazy]

The motor is usually oversized for a VFD because it needs to be derated. Because of the harmonics, loss of efficiency with pulsing DC voltage, parasitic losses, and extra internal heat from running on harmonic frequencies, the motor must be oversized. Now because the HP increases by the cube of the speed, a slight increase in speed and the HP goes up dramatically. Any increase in speed and HP and the motor is no longer derated. Not to mention that an increase in speed above nominal pump speed also gets the pump and motor into possible resonance which causes vibrations the same as a decrease in speed.

So, unless you oversized the motor even more than it needed to be derated to run on the drive, then you cannot increase the speed above nominal.

If he needs the same flow rate at 70 PSI that he is now getting at 40 PSI, he will need to almost double the size of the pump and motor. The pressure will increase by the square of the speed but, the flow will only increase in exact proportion to the speed. So I don't think he would be able to speed up the pump enough to get this done.

Therefore, a bigger pump is needed to maintain 70 PSI. Then you need to vary the flow rate with a valve or a drive to lower the flow when needed and still maintain 70 PSI.

 

[LionelHutz]

I figured that would happen. If the pump can't produce 70psi under all flow conditions then the VFD won't make it.

You probably need the pressure tank for 2 reasons. #1 is to absorb the surges in the water system. #2 is to provide extra water capacity during the high flow demands. So, you need pressure tank(s) large enough to source the water and maintain the pressure during the high flow demand periods.

Of course, if the pressure of the water is not that important then you could attempt to maintain something like 40psi or 70psi (or some value in between) and just let the pressure fall when the demand is high.

Is it possible to slow down the operation of the water valves in the system? The water hammer sounds like you have electric solenoid valves that are slamming shut very fast. I doubt you can slow them down but it's worth at least a quick look. The valves switching the water on and off are the source of the problem so look there first.

 

[DickDV]

valvecrazy, you are simply incorrect to state that motors need to be derated on drives. Anyone who routinely does this simply doesn't understand or is soaking the customer for capacity they don't need.

I don't recall the original poster saying he needs 70psi. If less is acceptable, the drive/motor/pump might just barely be big enough.

You are correct about the pump hp increasing rapidly with overspeed. My post says "the small amount of overspeed" just for that reason.

My whole point is that the OP has the existing equipment but isn't using all of it. Why not use it to the max and see if that's enough. If not, its time to buy something, maybe even the valves you are so in favor of!

 

[waross]

The bangs are water hammer. It is caused by rapidly closing a valve on a large column of moving water.
Forget the pump for now, it won't affect water hammer that much except to possibly make it worse.
At each point that the flow is suddenly reduced (usually a valve closing) install a surge suppressor such as itsmoked recommended. It is plumbed in ahead of the valve so as to provide a cushion for the moving column of water (the moving water in the pipe supplying the valve).
If you have pipes banging at 40 PSI, you may well have pipes breaking if you are able to maintain 70 PSI.
Sorry but pipes banging when the flow is stopped is the wrong problem for the solution of installing a VFD.
If the varying pressure is causing other problems in your plant then the VFD will probably correct those problems but it will do nothing for water hammer and will very possibly make the water hammer worse.
respectfully

 

[evan63]

I'll try a make up a rough sketch of the water distribution system and post it.Lionelhutz I believe your on the right track with the electric solenoid valves.WE have 2 water tanks that are used for supply for production.Each tank is supplied by a 2" pipe that branches off the main water supply line.The tanks are being filled with electric motorized valve witch opens slow but springs shut fast.As far as we can tell this is probably the biggest of the bangers.We had the production line shut down and simulated a tank fill and sure enough the banging was coming from the fast action of the water valve as it was closing.We figured the pressure drop was the result of the tanks being filled but didn't realize that the fast action closing of the valves were the cause of the banging pipes.So at least we learned something with the help of you all.