We had a requirement at The Hostel in the Forest http://foresthostel.com
to pump water from the lake up to the new lake garden and orchard area that we were beginning to be developed. It was troublesome to carry watering cans back and forth, the banks of the lake were steep enough to make it a struggle and various experiments with a small hand-operated lift-pump had not been successful . After one particularly good bottle of wine I even dreamed up the idea of creating an Archimedes screw that could be rotated by hand to lift the water simply by wrapping hose pipe around a 4" piece of PVC tubing... but the maintenance issues looked daunting!
Ideally, we wanted a system that was expandable and could scale as the garden developed and as more trees were added to the orchard. The idea was to place two or three 50 gallon barrels up high enough to develop some water pressure and then pump the water up into the barrels as they emptied. The lift was about 20 foot from the water level in the lake to the top of the barrels, depending on the time of year and the lake water level. The system was going to have to be as simple as possible with the minimal of maintenance.
I truly wanted a stand-alone solar system as it was several hundred meters from the nearest power source but the concept of having a car or boat battery was not appealing. Traditionally, in a solar-powered system, the solar panels charge a battery, the battery is then available to run the device(s) as needed. At some point, the battery would fail and have to be replaced with the cost of a new one and the even more arduous task of sensibly disposing of the old one with all the environmental issues of a lead/acid battery. So I was determined NOT to have a battery in the system. I wanted to use the smallest solar panel to keep it cost-effective but if you simply connect the pump to a solar panel. If this was connected directly to the pump, the chances were that it would only have enough power to turn the pump in direct sunlight and even then maybe for only an hour at mid-day.
I calculated the 'lift' required to get the water to the pump and 'push' required to get the water up and into the barrels. I would need about 7.5 psi to lift the water including the expected losses. Selecting a high-grade 12 volt water pump from the extensive SHUREflo series was quite simple. I found one that met the required specification and was a diaphragm pump that would tolerate a little sand and dirt should some make it up the pipe. This pump would require a 45 watt solar panel to make it work in full sunlight.
In a motor, the current represents the torque, and the voltage represents the the speed. I was prepared to reduce the speed of the pump, if I could drop the voltage and gain the benefit of increasing the current (torque). By introducing a device called a Linear Current Booster you can do just that. The Linear Current Booster boosts the current to produce the torque to turn the pump motor. It sacrifices voltage to the motor, hence the motor runs slower. This results in a motor that pumps slower in low sunlight rather than having a stalled pump.
Lets assume it takes 3 amps of current to produce enough torque to turn the pump. So there must be enough sunshine to produce the 3 amps of current. With the panel we are using, this will happen at around 11.00 a.m. When the panel can produce 3 amps of current the motor will start to turn, and pumping will start at whatever voltage (speed) the panel can support the 3 amps of current. With a Linear Current Booster, as the current in the motor rises to 3 amps, pumping starts. The question is how much current is required of the solar panel to support this current in the motor? If the output voltage is 1 V and the input is at 15 V, the output is 1/15th the input in voltage. For Power in = Power out (neglecting the approx. 4 % loss), the output current must be 15X the input, or conversely the panel current must be 1/15th the output. If it takes 3 amps to run the motor at 1 V it takes only 3/15 = 0.2 amps of panel current at the panel 15 V. We get about 0.2 amps from the panel at about 8:00 a.m. in the morning. So the motor starts pumping at 8:00 am instead of 11:00 am. Conversely, the solar direct will stop pumping when the panel can't produce 3 amps, say about 2:00 pm whereas the boosted motor will stop when the panel cannot produce the 0.2 amps, likely say 5:00 pm. So the motor is pumping a lot longer each day and moving a lot more water! A direct solar system would only pump from maybe 11:30 am to 2:00 pm, while the boosted motor runs from 8:00 am to 5:00 pm, just at a lower speed.
Another issue was that because it take a 15 feet of pipe to reach out to some clean water in the lake, every time a cloud would cover the sun, the pump would stop and all of the water would drain out of the pipe. While the pump was self-priming, it did require a couple of minutes to suck the water back up the pipe and start to push it up to the barrels. By installing a small foot-valve and filter at the inlet to the pipe, whenever the pump stopped sucking water, the water would stay in the pipe and not drain back into the lake. So every bit of sunshine was put to use pumping and not just priming. The last addition to the design was a float valve in one of the tanks so that when the tank was full it would shut the pump off and as the water level got down a 10% of full, it would switch itself back on again.
The system was installed at the end of 2006 and has given us nearly four years of trouble-free service. A small in-line filter makes sure not too much algae gets into the pump and this is simply cleaned about once a month. The energy is free free we do not have an ugly battery issue to deal with.