Pump + Pressure Xducer + VFD = FUN ?

[itsmoked]

I want to run a 5hp 3ph pump (230V 60Hz) at about 4hp using 1ph 230V. I also want to set it up so it maintains the water pressure at various flows at a fairly steady 70psi.

I plan to use a 0~10V pressure transducer to feed the VFD analog in. I expect to set the pressure thru button pushing up or down to reach the desired pressure. Once that's set it shouldn't be messed with.

A problem is, once all consumption stops the VFD knows nothing of this and will happily keep the water boiling at 70psi. To detect this I plan to have a check valve follow the pump. I will periodically reduce the pump speed a minuet amount and check the pressure on the down stream side of the check-valve. If the pressure doesn't drop with pump speed there cannot be any consumption occurring and I can stop the pump. If there is a pressure drop when I've slowed the pump then consumption is still occurring and I run the pump speed back up. I can do this perhaps once every 5 seconds.

I will also use a 7.5hp VFD to cover for the single phase supply.

See any problems with this grand plan?





Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

Forbidden frequencies?? The first or possibly the second elbow downstream of the pump may reflect back the pressure pulses caused by each impeller blade passing the pump outlet. The period depends on the length of the pipe section. If this period matches a multiple of the period of the blades passing the outlet the pressure pulses may reinforce.
Back in the day, I saw a 60 HP pump blow a piece of cast the size of a mans hand out of the pump casing with a VFD inadvertently running at a forbidden frequency. TWICE!!!
The mechanical guys can give you more information on mitigation strategies.
Yours
Bill

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[DRWeig]

I've installed and programmed many VFD / pump applications with a sequence of operation that is very similar to yours.

1. There are many VFDs in that size range that take 230V single-phase input and give 480V three-phase output if you haven't already selected the pump motor. It sounds like you may be working with an existing installation though.

2. Bill has a very important point. Your VFD should have the ability to avoid speeds that create resonance, all you need to do is find those frequencies and program them in. We typically run the pump (or fan) speed up slowly and find them, most often with a vibration monitor. An alternative is to put a vibration transmitter on the pump and program the controller to get the speed up or down a notch when high levels are sensed. If the speeds at which this occurs are consistent over time, the controller can memorize them as forbidden.

3. There is some lower flow that is above zero, at which you must stop the pump to prevent overheating and improper lubrication (pump manufacturer will usually provide this flow rate). We typically use either a flow switch or a flow transmitter downstream and make a stop decision based on flow. With a flow switch, you can hard-wire it to the stop input on the VFD and you don't need a smart controller -- but it sounds like you already have something programmable. Your check valve may not close until the flow reaches very close to zero and becomes dangerous. If it can be set to slam at a particular flow rate, it may work. I can't remember a check valve that could function as a flow switch, though. I usually don't get into valves that aren't automated.

Good luck with it!





Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[jonr12]

Just get a drive with “sleep mode”, and use a set point of 70 PSI. Then adjust the minimum frequency, according to the pump curve, that will produce the minimum flow required for the pump at 70 PSI. This way you don’t need flow switches or sensors and can spend your time mitigating the other problems you will have like critical frequencies and demands below the minimum flow rate.

 

[ScottyUK]

Keith,

Maybe a link in the forum403 would attract some folks with direct experience of this kind of application?

 

[LionelHutz]

Yes, what jonr12 posted. Use a drive with a PID controller that has sleep mode as part of the setup and it all works internally. It's also safer to use a 10hp drive.

 

[itsmoked]

Bill; Thanks for the heads-up on the forbidden freaks. Never would've thought of that until something broke. In this case I'll probably add a 3 gallon accumulator so when demand is asked for there is already a pressure reserve available. This will give the pump time to build head without a pressure hiccup. I'm betting this same accumulator would nullify the resonances.


Dave; !! I don't think I've ever seen a 230 ~ 460 boost VFD. Very good point about really LOW flow support. I really wanted to avoid flow measurement.. Hmmm.

jonr12; I'm not following your logic. I don't see how setting things for the minimum flow at 70psi is going to supply 70psi at full flow?

jonr12 & LionelHutz; What is "sleep mode"?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[ozmosis]

Keith
An example of sleep mode:

http://www.danfoss.com/NR/rdonlyres...E18577DFCA5/0/FC202Closed_loop_sleep_mode.pdf

It is also called hibernation mode in some drives.
It basically operates in closed loop control with a sensor providing the feedback and if the VFD runs at minimum frequency for a preset period of time, it stops but is still 'alive'. It will awaken once the fb signal passes a preset point that causes the VFD to run again.

I'm not sure about the 230v 1 ph to 480v 3ph, it is probably 230v 3 ph and change the motor connection to suit the lower voltage.

 

[DRWeig]

Oops, my bad on the 1-phase 230 to 3-phase 480. It's 1-phase 230 to 3-phase 208/230. Honeywell SMARTVFD is one brand.

Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[Compositepro]

I suspect that a VFD is not going to be doing much for you except convert single phase to 3-phase. If you need 70 psi, you will probably need to run your pump close to full speed just to reach this pressure before any flow will start.

 

[waross]

Can you let the pressure drift a little? I would consider pumping through a check valve to the accumulator and using a pressure droop. At no flow the pressure will build to possibly 75 PSI and the pump will stop. When the pressure drops below a small dead band the pump will start and run in proportional control down to 70 PSI or 65 PSI at full flow. When the flow stops, the pressure will build up to 75 PSI and the VFD will go into sleep mode. Be aware that when at shutoff pressure, it will take only a few RPM drop to get to no flow and no further increase in pressure. Keep the connection to the accumulator short and large. I have seen issues with a long connection to the accumulator that was a couple of pipe sizes too small. The accumulator should limit short cycling for dripping taps or similar consumption. Good ides; I like it.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[jonr12]

Compositepro is right. Most pumps can only be slowed down by about 10% and still produce the static head, which in your case is 70 PSI. With a 60 Hz motor that would be a minimum of 54 Hz. This greatly limits any benefit from a VFD. You could save yourself a lot of trouble and just use a standard single phase motor, control it with a Constant Pressure Valve (CPV) set at 70 PSI, and use a small accumulator tank with a 65/75 pressure switch. This way you don’t have to worry about de-rating, converting single phase to three phase, critical frequencies, resonance, pressure hiccups, short cycling, harmonics, and other problems that need mitigating when using a VFD.