Powering Grundfos A/C Pump in off-grid application

[cjb80]

I currently have a SunPumps T-135 well pump in an well that is positioned 50 feet down. The water level in the well is essentially at grade and the well is a part-time artesian well. The pump fills a 2600 gallon tank that I use as a buffer for a much bigger pump that pressurizes a drip irrigation system. The T-135 pump is a brush-type DC pump that (IIRC) consumes something on the order of 150W or so; however, the pump is apparently dying according to the readings that I have taken with my volt-meter and the advice from SunPumps. I have 4 T-105 batteries wired in series (24V), ~5.4 Kwh of storage. I have 3 285W (36V) panels charging the batteries - they are wired in parallel, and I have a Xantrex C40 charge controller.

To make a long story, less long, I need to increase my flow rate from the well to fill the tank so that we can do two waterings per day. The well will support the increased flow rates, but the current pump is no capable of doing that and it's dying anyways. The increased fill rate will enable us to do two waterings per day and also the filling of the tank will occur during daylight hours. We are targeting a flow rate of 6-7 GPM, an increase from ~1.5-2 GPM using the T-135.

I have been looking at the Grundfos SP pumps because they are much less expensive than DC brushless pumps. This will require an inverter, obviously, and I have initially chosen the Samlex SA-2000K-124, 24V Pure Sine Inverter as it is 24V and relatively cheap and I have another Samlex inverter that has been working really well for the last year or so.

I am looking to use the 7S05-11 pump, 115V 2 wire. The Pump's documentation can be found here: http://us.grundfos.com/content/gpu/e...-SP-PG-001.pdf on page 22 and 23. This should meet my flow rate requirements and has a reasonable cost.

OK, so that's the background. Here's the questions:

1) In the pump's documentation it shows the power consumption "per stage", but the documents never define what a stage is. I am assuming that the number of stages = the number of impellers. Thus, the peak steady-state power consumption of the 7S05-11 should be ~ 0.05 * 11 = 0.55 Kw = 550W. This sounds about right, assuming that the pump is extremely efficient (i.e., that is only about 100W more than a DC brushless pump with equivalent capabilities). Does this sound right? Is my assumption about stages = # of impellers correct?

2) Would a 2Kw inverter (4Kw surge) have any issues with this pump? Can I go with a smaller inverter (e.g., 15000W).

3) Are there other A/C submersible well pumps with similar efficiency?

4) Assuming that the panels produce a "real world" 600W (855W nameplate) then it seems like the panels should be sufficient to run the pump and keep the batteries topped off. Does this seam reasonable?

Thanks!

PS - I realize that this pump will be a huge load on the batteries. The irrigation pump is panel-direct, so in low light conditions it will not pump out of the tank. Thus, the only time the tank should drain is when the sun is shining. In theory, the power should "pass through" the batteries and they should be fully charged much of the time. In any case, if this doesn't turn out to be true then I realize that I will need to replace the battery bank. The last watering of the day starts at 1 PM, so the tank should be filled again by approximately 4:30 PM. Finally this is only running during the summer, and there is still light out at 9:30 PM here in the summers.

[Texas Wellman]

Do you use a controller for your sun pump and if so is the controller compatible with their brushless line of DC pumps? I would recommend you look into one of sun pumps brushless pumps if they will work with your controller.

Here is the breakdown for you on what a submersible motor consumes by Horsepower:
1/2 HP = 600-950 watts
3/4 HP = 940-1300 watts
1 HP = 1200-1600 watts.
Source: http://www.franklin-electric.com/aim...l/page-13.aspx

These numbers will vary depending on manufacturer and pumping conditions. The amps are dependent on the amount of water you are moving. With a high water level you're going to have max. flow and thus max amps.

If you're dead-set on using an AC pump look at the Grundfos SQ series that use a 3" pump/motor and have built-in soft start. They have one available for 115V and they are all 2-wire. The 115V Grundfos SQ is the only pump that I am aware of that will start on a Honda EU2000i. The inverter you mentioned will probably not start any other pump due to the surge amps.

FYI a 3-wire pump normally pulls about 1/2 the starting power that a 2-wire pump will so I suggest you stay away from any 2-wire models except either the Grundfos SQ or CentriPro (Goulds) 2-wire which uses a PSC (capacitor start). Franklin motors (most common) use a biac switch and pull 2x the starting load.

The best solution is probably the Grundfos SQ Flex. Wire the panels in Series to get above 90V and you will really move some water. The pump will cost you about $2K but has built in MPPT (no controller needed). You can add controllers that only act as an interface with level etc. but they have multiple models.

I currently have a SunPumps T-135 well pump in an well that is positioned 50 feet down. The water level in the well is essentially at grade and the well is a part-time artesian well. The pump fills a 2600 gallon tank that I use as a buffer for a much bigger pump that pressurizes a drip irrigation system. The T-135 pump is a brush-type DC pump that (IIRC) consumes something on the order of 150W or so; however, the pump is apparently dying according to the readings that I have taken with my volt-meter and the advice from SunPumps. I have 4 T-105 batteries wired in series (24V), ~5.4 Kwh of storage. I have 3 285W (36V) panels charging the batteries - they are wired in parallel, and I have a Xantrex C40 charge controller.

To make a long story, less long, I need to increase my flow rate from the well to fill the tank so that we can do two waterings per day. The well will support the increased flow rates, but the current pump is no capable of doing that and it's dying anyways. The increased fill rate will enable us to do two waterings per day and also the filling of the tank will occur during daylight hours. We are targeting a flow rate of 6-7 GPM, an increase from ~1.5-2 GPM using the T-135.

I have been looking at the Grundfos SP pumps because they are much less expensive than DC brushless pumps. This will require an inverter, obviously, and I have initially chosen the Samlex SA-2000K-124, 24V Pure Sine Inverter as it is 24V and relatively cheap and I have another Samlex inverter that has been working really well for the last year or so.

I am looking to use the 7S05-11 pump, 115V 2 wire. The Pump's documentation can be found here: http://us.grundfos.com/content/gpu/e...-SP-PG-001.pdf on page 22 and 23. This should meet my flow rate requirements and has a reasonable cost.

OK, so that's the background. Here's the questions:

1) In the pump's documentation it shows the power consumption "per stage", but the documents never define what a stage is. I am assuming that the number of stages = the number of impellers. Thus, the peak steady-state power consumption of the 7S05-11 should be ~ 0.05 * 11 = 0.55 Kw = 550W. This sounds about right, assuming that the pump is extremely efficient (i.e., that is only about 100W more than a DC brushless pump with equivalent capabilities). Does this sound right? Is my assumption about stages = # of impellers correct?

2) Would a 2Kw inverter (4Kw surge) have any issues with this pump? Can I go with a smaller inverter (e.g., 15000W).

3) Are there other A/C submersible well pumps with similar efficiency?

4) Assuming that the panels produce a "real world" 600W (855W nameplate) then it seems like the panels should be sufficient to run the pump and keep the batteries topped off. Does this seam reasonable?

Thanks!

PS - I realize that this pump will be a huge load on the batteries. The irrigation pump is panel-direct, so in low light conditions it will not pump out of the tank. Thus, the only time the tank should drain is when the sun is shining. In theory, the power should "pass through" the batteries and they should be fully charged much of the time. In any case, if this doesn't turn out to be true then I realize that I will need to replace the battery bank. The last watering of the day starts at 1 PM, so the tank should be filled again by approximately 4:30 PM. Finally this is only running during the summer, and there is still light out at 9:30 PM here in the summers.

[cjb80]

The SunPumps controller that I have is a voltage booster that goes up to 30V (I think) and it is not really useful for anything besides controlling the SunPumps brush-type DC pumps.

The reason why I was interested in the SP series of pumps is because they are much cheaper than the SQ series. Specifically, I found one online for $6xx (it was one of the first results in a Google search). Thus, the inverter cost (just under $900) plus the pump is less than the cost of the SQ series pump. The other issue with the SQ series pumps is that I would not be able to run it off of my batteries due to the increased voltage required.

Regarding the power consumption: 550W seems like it could be within the realm of possibility (granted, it's definitely on the low end!); so it seems like my assumption about the number of impellers = the number of "stages" is correct.

Thanks for the response.

Chris

[Texas Wellman]

You have SQ confused with SQ Flex. There is a difference.

The SunPumps controller that I have is a voltage booster that goes up to 30V (I think) and it is not really useful for anything besides controlling the SunPumps brush-type DC pumps.

The reason why I was interested in the SP series of pumps is because they are much cheaper than the SQ series. Specifically, I found one online for $6xx (it was one of the first results in a Google search). Thus, the inverter cost (just under $900) plus the pump is less than the cost of the SQ series pump. The other issue with the SQ series pumps is that I would not be able to run it off of my batteries due to the increased voltage required.

Regarding the power consumption: 550W seems like it could be within the realm of possibility (granted, it's definitely on the low end!); so it seems like my assumption about the number of impellers = the number of "stages" is correct.

Thanks for the response.

Chris

[Mike90250]

Here is the breakdown for you on what a submersible motor consumes by Horsepower:
1/2 HP = 600-950 watts
3/4 HP = 940-1300 watts
1 HP = 1200-1600 watts.
Source: http://www.franklin-electric.com/aim...l/page-13.aspx....

All I can add is that my 1/2 hp Franklin pump, logs 1Kw on my inverters' control panel. It's a 3 wire, 240v unit, and my 6KW inverter does just fine with it and other loads. I'm pumping about 160' lift.

[cjb80]

You have SQ confused with SQ Flex. There is a difference.

I looked up the SQ pumps and the 5SQ05-90 looks pretty good. It looks like the normal running power consumption is around 500 watts. What should I expect the power consumption to be during start-up with this pump? Does the soft-start make the starting power consumption similar to the running power consumption?

Thanks,

Chris

[Texas Wellman]

Running watts is starting watts. There is no surge. Please post back up how you like the pump and set-up so others will know.

I looked up the SQ pumps and the 5SQ05-90 looks pretty good. It looks like the normal running power consumption is around 500 watts. What should I expect the power consumption to be during start-up with this pump? Does the soft-start make the starting power consumption similar to the running power consumption?

Thanks,

Chris

[russ]

Running watts is starting watts. There is no surge. Please post back up how you like the pump and set-up so others will know.

Yes! Please do - there are always questions about the solar pumps.

[cjb80]

OK, so to follow up on this thread, I have finally installed an inverter and Grundfos pump.

The inverter is a Samlex 2Kw pure sine wave inverter and I chose the 5SQ05-90 pump. Everything is working but I have noticed a few 'issues' along the way.

1) The soft start does not trigger the "sleep" mode to turn off in the inverter, so I have to run the inverter with the sleep mode disabled. This causes a higher power draw when the pump is not running. I need to double check this, but I believe I tested it with 40W as the threshold to turn off the sleep mode. So apparently the pump is not even drawing 40W when it performs the soft start.

2) The water level in the well is currently about 30 feet below grade, and it is pumping into a tank. The pump is ~120 feet down from grade. So there should be something on the order of 45 feet of head; which should yield about 6 gallons/minute. I am seeing about half of that flow rate right now; it started at about 3.9 GPM, and then dropped to 3.33 GPM after about 30 minutes of running. If I look up the flow rate on the Grundfos charts to work backwards and determine the "observed head", it appears that I am starting at about 115 ft of head and then progressing towards about 125 feet of head (I assume this is due to draw-down of the well). I can't explain why I am observing ~120 ft of head (in regard to flow rate) but the with the water level I should only have about ~45 of head..? I have double checked the system and I don't believe there are any restrictions. The drop pipe is 1", so there shouldn't be any restrictions in the line...

3) My PWM charge controller is not keeping up with the additional draw on the line, so I am swapping out to an MPPT controller in the hopes that the panels are always producing more power than the pump is drawing... I estimate that the power consumption is ~600W including the inverter losses.

Chris

[inetdog]

Depending on how many hours per day your pump runs, and the difference in inverter power consumption (tare) between sleep and simple idle, you may just want to put a 40W or smaller light bulb in parallel with the pump on the output side of the pressure switch or other control.
Even a 10W light bulb will have a low enough cold resistance to reliably bring the inverter out of sleep.

If the switch is used only as a set of dry contacts for an input to the controller, you could look for a switch with two sets of contacts (as is required in some cases for 240V pumps) and use the second set to control the 120V light bulb.
If you are using a float switch and cannot find one with dual contacts, you may have to use a relay to drive both the controller and the light bulb.

[Mike90250]

A MPPT controller will only "create" power from extra voltage, if you have 19V panels, MPPT might get you another 20 watts. If you have 35V panels, MPPT will more than double your harvest from the extra voltage over 12V. (I'm assuming a 12V system, every 12V needs about 19V to charge with, 24V batteries need 38V ....)

2) The water level in the well is currently about 30 feet below grade, and it is pumping into a tank. The pump is ~120 feet down from grade. So there should be something on the order of 45 feet of head; which should yield about 6 gallons/minute. I am seeing about half of that flow rate right now; it started at about 3.9 GPM, and then dropped to 3.33 GPM after about 30 minutes of running.

I'd consider the drop from 3.9 to 3.33 ignorable. But the discrepancy from expected 6gpm to 3.9gpm a issue. Call the pump shop.

[Texas Wellman]

How do you know the starting level of the water? It would be difficult to measure with the pump in the well. Are you using an air-line?

[cjb80]

A MPPT controller will only "create" power from extra voltage, if you have 19V panels, MPPT might get you another 20 watts. If you have 35V panels, MPPT will more than double your harvest from the extra voltage over 12V. (I'm assuming a 12V system, every 12V needs about 19V to charge with, 24V batteries need 38V ....)



I'd consider the drop from 3.9 to 3.33 ignorable. But the discrepancy from expected 6gpm to 3.9gpm a issue. Call the pump shop.

I called the pump shop and the feedback that I got was that I did not account for the pressure in the line or any sort of elbows, couplers, etc in the line when calculating the head. This does not make much sense to me as the pump is pumping water into an open tank; hence the pressure should be negligible. I would also be astonished if the PVC connections were that much of a restriction. In any case, to achieve the amount of head that I am observing, I would need to have something on the order of 20 PSI of pressure in the line. When the water comes out of the pipe it is similar to water flowing out of a gallon junk, not like it is spraying out of a garden hose. So it is my judgement that the pressure is not anywhere close to 20 psi....

In response to Texas Wellman, I know the water level from looking down in the well. Also, when I took the old well pump out there was dry pipe and wet pipe, and I measure the length to the wet pipe.

Anyways, the question that I have is: Does the SQ series of pump always run at full speed (besides start up?). The pump curve says that it is at something like 10,000 RPM, but how do I know that it's actually trying to work at full speed?

Chris

[paulcheung]

That inverter say 50/60 HZ. is it selectable? if it run on 50 cycle the pump performance will be cut significantly.

[Texas Wellman]

I guarantee you cannot tell the water level in the well at 30' just by looking. There is no point of reference. I've been doing this for 20+ years and I can't do it accurately. I'll bet your level is deeper than you think.

I called the pump shop and the feedback that I got was that I did not account for the pressure in the line or any sort of elbows, couplers, etc in the line when calculating the head. This does not make much sense to me as the pump is pumping water into an open tank; hence the pressure should be negligible. I would also be astonished if the PVC connections were that much of a restriction. In any case, to achieve the amount of head that I am observing, I would need to have something on the order of 20 PSI of pressure in the line. When the water comes out of the pipe it is similar to water flowing out of a gallon junk, not like it is spraying out of a garden hose. So it is my judgement that the pressure is not anywhere close to 20 psi....

In response to Texas Wellman, I know the water level from looking down in the well. Also, when I took the old well pump out there was dry pipe and wet pipe, and I measure the length to the wet pipe.

Anyways, the question that I have is: Does the SQ series of pump always run at full speed (besides start up?). The pump curve says that it is at something like 10,000 RPM, but how do I know that it's actually trying to work at full speed?

Chris