The Benefits of Three Phase AC Pumps For Solar Water Pumping

Steve777

New Member
While I would agree with the OP, that 3 phase AC pumps are a better solution than 120 or 240VAC single phase, I think that DC submersibles are an even better solution. There are a few out there but the worm pump mechanism ones with controller at the surface seem to offer much more efficiency, ability to change setup params, and serviceability.

I have been using such a DC pump in my well for a bit over 3 years. I got two used 160W 24v (72 cell) panels to power it, and it has worked well for us so far. With some AGM batts, it can pump overnight and refill the cistern after a day of heavy water use. I also especially like the ability to dial the speed of pumping up/down at the controller. This allowed me to match the pump speed to my well's recovery rate, so that it rarely hits a well low condition; more efficient use of power and actually higher overall pump volume per time.

I may change my mind if something bad happens in the future with this setup, but as of now I would never go back to an AC pump.
 

Johncfii

Solar Addict
Steve777 ….

Good information. DC pumps are a worthwhile alternative to consider for shallow wells, or low water volume pumping,

But for deep wells, or higher water volumes, it seems to me that they are not so suitable. I’m no expert on DC pumps, but in general, deeper than 150 feet and more than 6 or 7 gallons per minute, looks like about the practical limit to me for a 48 volt system. Of course depth and volume can be traded off to some degree, but the actual pump curve tells the tale. Jump up to a 200 or 300 volt DC system, and a lot more should be practical.
 
Last edited:

Hedges

I See Electromagnetic Fields!
So Grudnfos, which can be PV direct? That's DC (but high voltage) in the sense it takes DC in and uses an inverter to drive a 3-phase motor (which is what "brushless DC" is.)
 

Guido1

New Member
Hello

After the tutorials gleaned from this group, and fingers crossed, I took the dive into some components. Prompted by the above discussion on low frequency transformer type inverters leaving less noise signature, I scored a Sigineer MK12W split-phase inverter. It appears to have massive transformers at its heart (285 lb. shipping weight) with huge surge capability. From what I could find, the brand enjoys better than average reliability.
I'm only feeding it 4Kw from the PV's at this point, with (16) 250 watt panels. I'm waiting on the last three 206AH SOK batteries from Current Connected, I'll have (4) x 412AH (4 x 2 per leg) to channel the 48V required for the Sigineer when they arrive and I get them hooked up.
I've been trolling the sub. pump pages a bit, I would like to score a Franklin "Sub 75 Solar" 1-1/2hp setup, but Franklin is extremely proud of their VFD boxes. Would I run into any issues running a Franklin 3 ph 1-1/2hp motor and pump with one of the Hitachi VFD components?
The Franklin box allows a direct PV DC connection, would I be foolish to ignore that? From what I learned in the pump class (above) the 3ph motor doesn't really require a surge to start, anyway.
So, the biggest power consumer would be my heat pump, I haven't been able to test the amperage of it yet, but I'm thinking even in AC mode, it shouldn't be too big of an amp suck.
I know I'll need more panels. As I plot the logistics for mounting even 16 panels, I can appreciate the higher-rated 350-400 watt panels for reducing the sum of panels to mount.
Should I be able to route the charging current through the 100 amp Sigineer MPPT even though the panels will be separated into four groups to maintain the (4 x 12v) 48 volt current requirement, or can I just charge them all as a single group?

Thanks again for all of ya'll's help here. If I have really screwed up on any of the above, tighten me up, please. I'm very durable.

Respectfully posted,
-Paul
 

Johncfii

Solar Addict
Hello

After the tutorials gleaned from this group, and fingers crossed, I took the dive into some components. Prompted by the above discussion on low frequency transformer type inverters leaving less noise signature, I scored a Sigineer MK12W split-phase inverter. It appears to have massive transformers at its heart (285 lb. shipping weight) with huge surge capability. From what I could find, the brand enjoys better than average reliability.
I'm only feeding it 4Kw from the PV's at this point, with (16) 250 watt panels. I'm waiting on the last three 206AH SOK batteries from Current Connected, I'll have (4) x 412AH (4 x 2 per leg) to channel the 48V required for the Sigineer when they arrive and I get them hooked up.
I've been trolling the sub. pump pages a bit, I would like to score a Franklin "Sub 75 Solar" 1-1/2hp setup, but Franklin is extremely proud of their VFD boxes. Would I run into any issues running a Franklin 3 ph 1-1/2hp motor and pump with one of the Hitachi VFD components?
The Franklin box allows a direct PV DC connection, would I be foolish to ignore that? From what I learned in the pump class (above) the 3ph motor doesn't really require a surge to start, anyway.
So, the biggest power consumer would be my heat pump, I haven't been able to test the amperage of it yet, but I'm thinking even in AC mode, it shouldn't be too big of an amp suck.
I know I'll need more panels. As I plot the logistics for mounting even 16 panels, I can appreciate the higher-rated 350-400 watt panels for reducing the sum of panels to mount.
Should I be able to route the charging current through the 100 amp Sigineer MPPT even though the panels will be separated into four groups to maintain the (4 x 12v) 48 volt current requirement, or can I just charge them all as a single group?

Thanks again for all of ya'll's help here. If I have really screwed up on any of the above, tighten me up, please. I'm very durable.

Respectfully posted,
-Paul
Lots of potentially complex considerations. Hard to answer without looking at exact equipment specs. Can you post links to data sheets for the Franklin items? Save us the trouble of looking them up.

My first initial thought is that you will probably pay significantly more for unique items, such as a dedicated solar VFD, or unique “solar “ pump. It is probably significantly more cost effective to use a common 3-phase pump motor, driven by a common VFD. But efficiency will probably be a bit better with a DC driven VFD, because of one less conversion loss in the circuit.

At first consideration, I don’t think it would be at all foolish to pass on the special high-cost Franklin VFD box …. Calculate what lower conversion losses you would realize, and compare that saving to the added capital cost

All is complicated by the unique considerations of your install, such as the length and convenience of wire runs, and proximity of panels and inverter to where you plan to install the VFD.

What is this about a 3 HP motor not “requiring a surge” for starting? Can you post a link? Is that a feature of the motor, or the drive? As noted above, no 3-phase motor driven by a VFD needs a starting surge if the motor speed is ramped up over some seconds - easily set within the drive parameters.

4kW of panels can all potentially feed the single 100 amp charge controllers. Depends on VOC of the panels, how they are wired in series/parallel, and the charge controller specs. Again, can you post a link to the charge controller data sheet?
 
Last edited:

Porch

Solar Enthusiast
Ditto on the VFD idea. I converted a 5HP air compressor to 3 phase with a cheap VFD a few years ago to test the idea. The dual Outback VFX3648 had issues starting it if the tank was near full. With the VFD, the slow ramp up is wonderful and no issues. In fact, I run it at 35Hz to keep the noise down in the shop.

I just did a 3HP well 6 months ago. Pulled the normal split phase 240V motor and dropped in an off the shelf 3ph 3HP motor. I used a 10HP Chinese VFD and a used inductor on the output. System is a dual Outback VFX3648 again. All the startup surge issues went away. I set the ramp up time to be about 6 seconds. It gives the slow reacting Outback inverters and charge controllers a chance to keep up. This well is also used for pressure (no booster), so I found 50Hz is the sweet spot where I still get good pressure and flow, but the watts drawn is down quite a bit. Under 50Hz and it has a hard time building pressure.

I had a Grundfos well (400 feet, but not used for pressure) about 10 years ago. Run on high voltage solar panels. It did fine, but the pump only lasted about 3 years. I pulled it and replaced it, but the new one only lasted a few years more. They replaced it, but I got tired of pulling 400 feet of pipe so I didn't install it. If the well was out in the middle of nowhere and didn't have power or wires going to it already, I can see the benefits of a direct solar DC pump. But if there is power there, an off the shelf 3phase pump and VFD is the better option.
 

Pyrofx

Flux capacitor builder
So what vfd did you use? 220 split to 3 ph?
I'm about to pick up a 3hp pump in the next couple weeks
 

Hedges

I See Electromagnetic Fields!
Ditto on the VFD idea. I converted a 5HP air compressor to 3 phase with a cheap VFD a few years ago to test the idea. The dual Outback VFX3648 had issues starting it if the tank was near full. With the VFD, the slow ramp up is wonderful and no issues. In fact, I run it at 35Hz to keep the noise down in the shop.

Sounds good, what I imagine is a PID circuit (or algorithm) monitoring line frequency from Sunny Island and varying VFD analog input. It would seek to keep frequency at 61 Hz, so Sunny Boys (which curtail output linearly from 61 Hz to 62 Hz) deliver full power and pump uses all of it if possible.
 

Pyrofx

Flux capacitor builder
Do you think the inductor is necessary? 3 ph is a new set of learning for me. The machines in my shop run off a roto phase
 

Johncfii

Solar Addict
Do you think the inductor is necessary? 3 ph is a new set of learning for me. The machines in my shop run off a roto phase
Not absolutely necessary, but I think a good practice.

The inductor (which the industry calls a “line reactor”) is primarily intended to suppress RFI (radio frequency interference) from affecting nearby devices, especially for long runs of wire, as to a well head. The RFI is an artifact of the audio frequency switching (typically something like 10kHz) used by the VFD to generate the three-phase power. While 10kHz isn’t radio frequency, many harmonics of that frequency can also be generated.

The reactors also can benefit the motor, and the inverter by suppressing that 10kHz ringing that is otherwise superimposed on the leads to the motor, and is reflected back to the inverters through the supply leads. It is non-productive energy, that gets dissipated as heat somewhere, and produces higher voltage spikes that can potentially break down components. I have line reactors on both input and output leads. They aren’t that expensive, my own practice is to always install them.
 
Last edited:

Pyrofx

Flux capacitor builder
I'll start looking.
I know Hedge had a little trouble with his pool pump and the Sma inverters
 

Porch

Solar Enthusiast
Do you think the inductor is necessary? 3 ph is a new set of learning for me. The machines in my shop run off a roto phase

The idea of an line reactor on the output is that the VFD makes a choppy and spiky waveform and the long wire run amplifies the spikes, creating higher then expected voltage and can cause the insulation on the motor to break down. I figured there is something to that as they make VFD rated motors with more insulation (or marketing hype).
The VFD manufacture recommends a line reactor at the end of the long wire run, at the motor side. I can't do that in a well, so I did the second best thing and put it right after the VFD.
 

Hedges

I See Electromagnetic Fields!
I'll start looking.
I know Hedge had a little trouble with his pool pump and the Sma inverters

Yes, 10000TLUS said "IGBT error", not sure what state it had been in.
5000US, when it checked line voltage before reconnecting, said something like relay error. Turning off VFD didn't fix, but power cycling PV input did. I think it fed OK with VFD on, but had the problem connecting.

Maybe yet another inductor on input of VFD would help? Slight improvement in PF and waveform but not ideal.

What we want is a front-end that follows AC voltage sine wave and synthesizes sine wave current to charge the capacitors. That does exist, but don't know of an add-on unit.
 

Pyrofx

Flux capacitor builder
So how close in amperage does the reactor need to be? Pump is 17 amp. Most reactors I'm finding are 25 to 32 amps
 

Hedges

I See Electromagnetic Fields!
Reactor has to be able to handle the amps. Damaged if too many amps put through it, not if fewer amps. It is just a coil of wire on a core.
When an inductor is sold as a "reactor", it is spec'd according to percentage voltage drop, inductor in series with resistor (impedances add up with Pythagorean theorem, which takes care of imaginary numbers.)
You can select something that gives the percentage voltage drop you want.
 

Johncfii

Solar Addict
So how close in amperage does the reactor need to be? Pump is 17 amp. Most reactors I'm finding are 25 to 32 amps
As Hedges says, a reactor rated for current higher than your maximum pump draw is perfectly fine.
it is just like using a 6 or 8 or 10 gauge wire to carry current that needs at least a 12 gauge wire.
 

Pyrofx

Flux capacitor builder
Ya. Here's my issue.
Pump 3 hp. Can go split or 3ph.
Bore hole is 1030 feet down.
Have 2 sunny island 6048 , 2 sunny island pv 7.7 , and 1 sunny island tu6000us.
18.5 kw of panels , and 78kwh of lifepo4.

Right now it's all set up in my shop to work out bugs before hauling 30 miles out in the desert.

I have the split to 3ph vfd arriving tomorrow. I'll test it out on the mills and lathes.

Hedge got some errors on his setup on the pool pump.

I have a backup plan to use my shop rotary phase converter if need be , or just use a straight 220 pump. It's to expensive to have to do over.
G
 
Top