VFD controlled water well pump thoughts

viper · May 25, 2011

One of the common controls today is an "on demand" system that simply throttle are vary the pump speed depending on demand. They are simply controlled with a pressure transducer. As the pressure drops, the pump controller commands a higher speed to the pump to maintain pressure.

Just curious how a guy might set this up with a VFD instead of a standard pump controller which I find to be nothing more than a stripped down VFD. However, there seems to be a slight amount of PLC type calculations here regarding the pressure transducer. It would easier and cheaper to just run a more standard VFD if possible.

I am also kind of wondering about a slight over speed of the pump system. Seeing this load as a variable torque application, it might not be a good idea to try and overspeed the pump for fear of needing more torque than the motor can safely provide.

tmt · May 25, 2011

I think you nailed when stating that these controllers power end is nothing more than a VFD.

Most, not all vfd's allow a 0-10v analog input to control the speed, where 10v equals 60hz, that helps keep you from over speeding the pump.

The most work involved would be the selection of your pressure transducer to closely match the pressures you are trying to monitor. Also you would want to use the curve that reprsents the flow rate of the pump that you are using. You might get lucky and find a nolinear transducer that closely matches the pump curve, but I wouldn't hold my breath.

The most complex part is to convert the 0-10v analog input voltage to the VFD to closely approximate the curve for the pump from the signal of the pressure transducer.

The simplest method would be to have the tranducer that will output 0-10v analog. To match the curve more closely, well that's were all the hard work begins and there many ways to do that.

Miguels244 · May 25, 2011

Hitachi L100 VFD's have PID functionality built in.

DaveKamp · May 25, 2011

Actually, it doesn't really matter...

the 'demand flow' pump controls are built for-purpose, but what they're doing, is basically limiting motor speed once you've reached the pressure setpoint.

Let's say you want to run a system as 45psi. Check valve holds system pressure (a very small air-chamber type tank gives you a 'little' buffer volume)... so there's pressure immediately when you crack the valve.

Once the transducer sees pressure fall below 45psi, it starts the motor, and runs it up until you hit 45psi... and if it goes to say 46psi, it backs down on motor speed, and if it falls to 42, it cranks the speed up.

Using the VFD's input, basically what you'd do is loop the pressure transducer back into the feedback input, and establish the running ramp features (if the VFD HAS them) accordingly.

Personally, I wouldn't do it... I wouldn't bother trying to use a VFD to control the well pump, and frankly, I wouldn't use the demand-based system. Reason being, lightning is hard enough on well pumps as it is. I've got four neighbors here (this is rural, mind you), and in a storm last summer, one of 'em took a hit to the pump... I had to change one of the caps, but the pump still runs, but pulls pretty hard and fails the megger at about 350v. We're all still fine for water, though... and of course, when his pump fails, he'll hafta submit it for insurance.

Guy about a quarter mile northeast of me had a demand system... he's lost his pump twice, but gone through five controllers... every close strike toasted it.

Lightning strikes cause ground-potential rises... strike current flowing through the ground induces a sympathetic current through the pump wiring, and kills that controller.

The other reason... I wouldn't nix the pressure tank for anything. When we have long-term power outages, I can run one of my generators, run up the pressure tank... shut the gens down, and have more'n enough water to wash hands and flush toilets all night... not start it up 'till morning. Can't beat that kind of flexibility.

Jraef · May 25, 2011

viper said:
One of the common controls today is an "on demand" system that simply throttle are vary the pump speed depending on demand. They are simply controlled with a pressure transducer. As the pressure drops, the pump controller commands a higher speed to the pump to maintain pressure.

Just curious how a guy might set this up with a VFD instead of a standard pump controller which I find to be nothing more than a stripped down VFD. However, there seems to be a slight amount of PLC type calculations here regarding the pressure transducer. It would easier and cheaper to just run a more standard VFD if possible.

I am also kind of wondering about a slight over speed of the pump system. Seeing this load as a variable torque application, it might not be a good idea to try and overspeed the pump for fear of needing more torque than the motor can safely provide.

As alluded to by Miguels, you need a controller in the system, camonly called a PID (Proportional, Integral, Derivative) control loop. Without it, your VFD will "hunt" up and down because as the speed increases, the pressure increases and the demand decreases, so it just repeats itself ad nauseum. A PID control loop dampens the response and smooths it out over time. The Cruise Control in your car is a PI loop controller: you set the speed you want to maintain, if you start to go up hill, the controller responds by goosing the gas, but it doesn't just floor it until you over speed and then apply the brakes, it smoothly and gently increases the speed until it detects that you are close, then eases off on the rate of acceleration.

Most decent VFDs now have PI or PID control built-in, but if you've never done one, they can be daunting. In my opinion, the ABB ACS (and ACH) drives have the simplest and easiest PID setup macro there is. once you tell it you want to do that, it steps you through the entire process by asking you simple questions on the keypad. For most others, you pretty much already have to know how to program a PID loop before you even start. and if you were unaware of what one was, that's going to be difficult.

You can also buy a separate stand-alone PID controller. I actually recommend this in many cases for this simple reason: if anything were to happen to the VFD 3 years from now, chances are whatever that mfr has on the market is now obsolete and you will have to RE_LEAN how to program the replacement drive. If you use a stand alone PID controller, the VFD becomes essentially generic, one will work the same as another so if you ever have to replace it, your PID controller doesn't change.

viper · May 25, 2011

DaveKamp said:
the 'demand flow' pump controls are built for-purpose, but what they're doing, is basically limiting motor speed once you've reached the pressure setpoint.

Let's say you want to run a system as 45psi. Check valve holds system pressure (a very small air-chamber type tank gives you a 'little' buffer volume)... so there's pressure immediately when you crack the valve.

Once the transducer sees pressure fall below 45psi, it starts the motor, and runs it up until you hit 45psi... and if it goes to say 46psi, it backs down on motor speed, and if it falls to 42, it cranks the speed up.

Using the VFD's input, basically what you'd do is loop the pressure transducer back into the feedback input, and establish the running ramp features (if the VFD HAS them) accordingly.

Personally, I wouldn't do it... I wouldn't bother trying to use a VFD to control the well pump, and frankly, I wouldn't use the demand-based system. Reason being, lightning is hard enough on well pumps as it is. I've got four neighbors here (this is rural, mind you), and in a storm last summer, one of 'em took a hit to the pump... I had to change one of the caps, but the pump still runs, but pulls pretty hard and fails the megger at about 350v. We're all still fine for water, though... and of course, when his pump fails, he'll hafta submit it for insurance.

Guy about a quarter mile northeast of me had a demand system... he's lost his pump twice, but gone through five controllers... every close strike toasted it.

Lightning strikes cause ground-potential rises... strike current flowing through the ground induces a sympathetic current through the pump wiring, and kills that controller.

The other reason... I wouldn't nix the pressure tank for anything. When we have long-term power outages, I can run one of my generators, run up the pressure tank... shut the gens down, and have more'n enough water to wash hands and flush toilets all night... not start it up 'till morning. Can't beat that kind of flexibility.

That is actually a pretty good way of looking at it. We have been running the old pressure switch and pressure tank setup for many years. The only concern I had was that I might install a 3-5HP 4" cased well pump and that is a way overkill for any home and was worried about the big surge of water but I guess maybe a larger volume tank will fix that.

I am also hunting for a little water well 101. Dad used to drill oil wells but I am a little green on advanced techniques in water. I will post in the general section to see if there is any wisdom out there.

adama · May 30, 2011

Pump and Well Forum

Have a search around there. Controling a well pump that way just over complicates things. GO with a CSV. it just restricts the pumps output as demand drops, achieves exactly the same thing, both electrical consumption wise and water wise.

Basically as you load a pump, the pressure increases. Limit the output flow and the pressure goes up, but water moved drops significantly, hence reducing electrical load.

Running a normal three phase pump at reduced speeds a bad idea do to the design of the bearings in them, they rely on the flow to both cool and lubricate the bearings. Most of the inverter type pumps are rated at 10Krpm, not the usual 3.6Krpm.

Search

Largest Manufacturing Technology Community on the Web

Search

VFD controlled water well pump thoughts

viper

Titanium

tmt

Hot Rolled

Miguels244

Diamond

DaveKamp

Titanium

Jraef

Titanium

viper

Titanium

adama

Diamond

Similar threads