4
\$\begingroup\$

I have a project to power a 12V well pump from a solar panel, no battery. My plan is to use a 270W panel and a 12V DC-to-DC converter.

Under reasonable illumination the solar panel is overkill providing much more than the 100W required for the pump.

So my question is: There are 2 possible solutions (operating points) on the power for the solar panel. Which one will it choose?

\$\endgroup\$
4
  • 4
    \$\begingroup\$ This is a repeat question. You have an impedance mismatch and you will never get 270W starting up. The motor load draws 10x current on startup so the converter may collapse. It will be terribly unstable. [Edited by a moderator.] \$\endgroup\$ Jul 28, 2021 at 6:51
  • 2
    \$\begingroup\$ If you connect the panel to the converter before you connect the motor, then it may choose the higher voltage operating point. If you connect the motor to the converter before connecting the panel to the converter, then it may choose the lower voltage operating point. Or, the whole thing may not work at all for a lot of different reasons, including motor startup current demand, like Tony says. \$\endgroup\$
    – user57037
    Jul 28, 2021 at 7:26
  • 2
    \$\begingroup\$ The operation point for solar panel is actually a more complex issue. Search for Maximum Power Point Tracking \$\endgroup\$ Jul 28, 2021 at 7:35
  • \$\begingroup\$ Thanks to mkeith. That is the answer I was looking for. \$\endgroup\$
    – AI6MK
    Jul 28, 2021 at 16:32

3 Answers 3

18
\$\begingroup\$

I have tested similar arrangement with 350W solar panel and 120W 24V centrifugal pump via DC-DC buck converter set at 24V and no current limit. Idea was to have solar panel produce voltage above 24V and directly power the pump.

It does not work well.

Pump was in a 12m deep shaft with a float switch set to go ON if the water level rises above it. The bilge well had sufficient volume to allow for overnight standstill of pump in worst case scenario so the pump as the sun comes up would pump the water during the day, and stop if level is lower than the switch.

First problem was obviously power availability. I found that this arrangement worked only during peak sun input. Secondly, as this was centrifugal pump if it was started without enughuah power to overcome the head of the water column it would just be running and heating up the water inside the housing, as there was no flow. This caused premature seal failure and water ingress problems.

Much better was Solar > Charger > Battery > Switch > Pump! This system even though more expensive for the price of the charger and battery is much more reliable, the pump can start and keep the well dry even during the night. No rain but cloudy days would charge the battery and keep it charged as needed. I found that even 18Ah battery was sufficient. System is in operation since April 2014 still using the same equipment.

\$\endgroup\$
1
  • \$\begingroup\$ I did purchase an MPPT CC and batteries and was about to implement it, but then had the thought about just using a DC-DC converter. What inspired me was that many commercial solar pumps appear to not use a battery, but I guess they would need to provide sufficient energy to start the motor. Thanks for sharing your experience. I will return to the conventional solution (panel > cc> battery > pump). \$\endgroup\$
    – AI6MK
    Jul 29, 2021 at 17:39
3
\$\begingroup\$

This is the reality of solar power.

Power sequencing with no load until DC OK is essential with some storage capacitance. Batteries , even tiny ones have huge C values like 10k Farads.

The rating for a PV is max. Solarity into a perfect impedance matched converter called MPPT since **V/I=R(mpt) = Voc/Isc at Pmax ** is exactly the resistance required , R(mpt) needed to get maximum Vmpt*I=Pmax.

This means R(mpt) has a minimum and rises towards high R from reverse diode protection in the dark.

Without a MPT controller if the battery voltage is between 72% ~ 82% of open circuit Vpv or Voc then you at least have a good chance to get > 80% of PV available power from 10% to 100% available solar power. The 82% of Voc is at optimum conditions +/- x depending on PV type.

The input impedance of a DC motor , you might guess is V/I but is actual the coil DC resistance (DCR) to startup, which for an efficient more is 8 to 12% of this ratio which means the start surge from a Voltage source is V/DCR=I or the same as locked rotor current.

Boost converters are like step up transformers where the load is transformed by Rin ~= Rout * (Vi/Vo)^2 * efficiency for steady state. But on startup it’s like a motor with a coil DCR driven by a FET low Ron into an uncharged Cap low ESR, so the startup power demand can be as much as 20x the max rated load power unless it has a soft start. This is to store the energy in the reactive LC components.

The PV is a current source with a Voc limit, but make no mistake, MPT needs a matched impedance to satisfy the Maximum Power Transfer theorem. It is also estimated by Rmpt = Voc/ Isc for the short circuit current. Where in either case, Voc, Isc there is zero power supplied.

So the reality is you need to match the voltage ratio Vmpt/Voc (82% down to less sun @ 72% to convert the charge from a current source into a voltage source ( battery ) with a battery charge regulator ( according the chemistry type) then if necessary, DCDC convert to load with the understanding of the load impedance ratio for input and output gives a product VI = P for perfect efficiency and a motor load and converter load that yield less than 1% of Pmax rating of,PV on startup, so until it overcomes this, it won’t work.

\$\endgroup\$
0
2
\$\begingroup\$

In regard to your main question:

Then you need less than the maximum power from the solar panel, both points (towards higher voltage/lower current and lower voltage/higher current) will do.

Choosing one of them can be related to:

  1. The panel aging - generally, the less voltage, the less panel aging. The effect is pretty minor and also see next.
  2. The panel developing hotspots - lower quality (or already aged) panels can have a considerable disbalance between cells. The weakest cell in such a panel gets heated by the current and because of the heat it ages even faster. Using the higher voltage / lower current point may mitigate the undesired process.
  3. Wire loss (and heating, and the risk of a contact deterioration) will be lower if you use the higher voltage point as well.
  4. The power loss (and heating, and aging) in most DC-DC circuits is current-dependent and not voltage dependent.

In addition to these two options you have at least three more:

  1. Use PWM regulation. Cheaper (in regard to the regulator), simpler and you don't care about the panel working point as long as long as the panel gets you enough voltage and current. The panel efficiency is lowered and you need to explicitly match the panel to the load.
  2. Use a dump load. Operate the panel always at its MPPT, the power you don't need can be dumped to a separate (resistive) load.
  3. (esp. easy for vibrational pumps) make the pump vary its power in order to match the available power from the panel.

Considering all these factors, most solar systems use the higher-voltage point when under partial load.


As others already mentioned, water pumps are not the best load for a batery-less solar power source.

Of them, the best ones are vibrational pumps.

Be aware that vibrational pumps are not the best in regard to the well maintenance and longevity.

\$\endgroup\$
1
  • \$\begingroup\$ Thanks to all those who took the trouble to reply. \$\endgroup\$
    – AI6MK
    Jul 29, 2021 at 17:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.