Matching pump and controller box

[Mozzie]

Greetings from a new member! I write concerning a solar pumping system I bought from a supplier in Johannesburg, South Africa, and installed in Mozambique. It is not performing as per specs and the supplier has no answers for me. I get contradictory advice from other suppliers. The factory (Wenling Jintai) tells me the pump doesn't have the capacity (pressure) required, but according to the test curves it does. The details are:

Pump model: JS3 2.1-100
Control box: 48V
Solar panels: 4 x 36V/165W, arranged to give 72V output
Distance from solar panels to controller: 4m
Cable thickness used for solar panels: 4mm
Distance from controller to pump (down borehole): 80m
Cable thickness used: 6mm
Voltage from the solar panels into controller with pump switched off: 80V
Voltage from the solar panels into controller with pump running: 71-75V
Pipe thickness: 32mm HDPE (also from borehole to tank)
Total head between pump and tank: 85m (but dynamic water level is at 70m, so net total head = 75m)
Distance between borehole and tank: 33m

The pump was delivering a trickle of water, at best. Both pump and control box were taken back to the supplier for testing on their test bench (with a 48V battery array) and produced 650 l/h at 80m, but on site only about 150 l/h. Now no water at all is reaching the tank although the pump is running. One supplier says the problem is the control box which should be a 72V one, not 48V. Please could you advise what the problem could be.

[inetdog]

If the pump motor runs slowly compared to the design value it will reduce the head pressure against which the pump can supply water. What is the rated voltage of the pump and what voltage are you supplying to it when you test on the bench?

Never mind. I see that this is a nominal 48V pump. Driving the controller directly from 72V worth of panels is going to waste panel power, and depending on how the control box works may result in the controller delivering less than 48V to the pump.
You need to get some more measurements, such as voltage at the output of the control box and current into and out of the box.

[Mozzie]

Thanks for the response. Measurements have varied a lot over time. When it was installed (Aug 2014) I was getting 5A at around 40V from controller to pump. Then the flow dropped, I took both pump and controller for testing, the motor was found to be faulty and the supplier gave me a new pump unit. Tested on their test bench, with 48V batteries, it gave about 700 l/hr at 75m (much less than spec) drawing 8A. But back on site the flow was much less. Measurements I made in Jan 2016, when it was pumping well, gave 5,6A - 6,6A at 71-75V going into the box but only 4,7A going to the pump. I couldn't get a good DC voltage reading and on the AC setting the meter gave only around 13V. I assume this can't be correct. The factory tells me the panel and controller combination is fine and suggests trying a new screw. However, I have in the past tried several new screws and it makes no difference. It does seem logical, though, that if a positive displacement pump is turning (quite fast) but can't raise water to the design head, there must be leakage past the screw.

[Mozzie]

Just to clarify: From August 2014 to about September 2015 the pump ran daily but only delivering about 200 l/h. Then, without anyone touching anything and for no apparent reason, it started pumping about three times this amount or more (I was not there and the reports I got related to how fast the tank was being filled). This continued until January 2016 when, due to a severe drought in the region for the past two years, which caused the water table to drop, the water level sensor started operating and switching the pump off. So I dropped the pump down from 70m to 80m and now no water reaches the top, although the pump runs. The voltage and current readings are the ones I mentioned above.

[inetdog]

Any chance that as you dropped the pump down 10 meters you introduced a leak into the pipe somewhere between the pump and the surface outlet?

[Wy_White_Wolf]

I would have to guess that you either have a leak between the pump and surface or a bad wiring connection that is limiting amperage.

WWW

[Mozzie]

Thanks for all the suggestions. I've decided to buy a new controller (the new ones are apparently more efficient), change the screw, check all connections and clean the panels. Hopefully this will solve the problem.

[Bala]

In my experience for almost any pump, or just about anything for that matter you can generally take at least 10% off the manufacturers specs straight up.

So you are now down to a 90m max head pump.

You have at least 80m of delivery pipe plus the distance to the tank, so you have losses from friction which can be substantial.

The 6mm cable has been there for over 12 months so depending on quality may already have lost efficiency. like a lot of things customers are demanding low prices so suppliers are delivering cheap cable. I work in automotive and use a lot of small cable and have found brand new cable where the copper is already black.

So combine the facts you are on the very upper limit of pump specs, factor in friction losses, and voltage drop I would not expect the pup to be delivering.

When bench testing you normally just test pressure which gives you a theoretical head, but does not allow for friction losses and they would not likely have tested through 80m of cable so have full 48v power.

The factory told you that and dropping the pump another 10m proves that you are expecting to much.

You need to spec a pump for real world, so probably around 50 to 75% rating, if you spec close to max as soon as they wear a bit you will lose flow.