Farm Off Grid Poultry Lighting System
Here is a project I recently undertook. An off grid renewable energy system for poultry buildings. The lighting will turn on and off at predetermined times during the night and early morning to increase production of the egg layers.
There are also street lighting for the pathways to get to/from the buildings during the dark. These lamps will need to be kept on the whole night.
To minimize losses, it will have to use a higher voltage during distribution and then step down on site for the low voltage LED lamps.
To accomplish this, I designed a 24V power distribution system and the lighting will be 12V as it is most commonly available. To prevent wiring another run for the street lights, I devised a way to transmit power through the existing wiring and turn off the building lights but keep the street lights lit.
Rather than remote control the buildings or individually turn them off manually, the converters I designed to step down from 24V to 12V will switch the output off when the input is below 19V. So during the sleeping period of the lchicken, the bus voltage is dropped to 18V keeping the pathway lighting on but the building lamps will turn off.
Bunch of parts to build 15 units of the custom step down converters that I designed.
Boards populated. Ready to wind the toroids.
Power mosfets and Schottky diodes
One converter under burn in and overload test.
One step down converter completed. 14 more to go.
These will supply 20 LED lamps per building for a total measured draw of 3A.
The inductor was too big (no local sources for smaller T106 toroids) so I mounted it under the board.
For the automatic mains charger, I used two 12V 15A open frame PSUs in series. Each one is set to 14.4V output and one is modified so that it goes to constant current mode at 12.5A.
I decided to check the PSUs before modification
PCB Version PCB250-V2.2
This capacitor wants to stand taller than the others.
Shocked to see a solder splatter shorting between the mains to earth capacitor! Quality check? What quality check?!?
Removed one of the current shunts so the powersupply goes into constant current mode at 12.5A
Powersupply under burn in test.
Powersupply thermal image with no load.
Powersupply thermal image with 12.5A load
Now for the fun part. This is how it all goes together.
Click me for full size.
(The 18V control system is not yet installed. The three switches below are for manually turning on each bus when loading chickens into the cages at night)
The circuit is a slightly modified version of Mike's 555 solar/wind controller to operate at 24V and switch the relay on when the voltage is below instead of above the threshold setting.
The charger turns on at 24.0V and turns off at 28.6V then the solar panels will take over in the morning to top off the batteries.
Automatic charger controller in standby mode.
Controller in Charging mode.
8pcs of 12V 100Ah batteries for a total of 24V 400Ah bank capacity.
Four 150W panels on the office roof.
Converter, fuse and switch mounted on a plank of wood.
Cover just snaps on and off.
LED lamps hanging on the trusses.
LED lamp detail.
Here is the LED lighting in one of the buildings at night.
05 Jan 2015
Had some more time to work on the project. Initially, the plan was to drop the 24V line to 18V to prevent the buildings lighting up but allow the streetlights to get power without having to run separate lines. After connecting and testing the system, I found 18V was a little too close and some buildings were lighting up so I dropped it lower to 17V and it worked great!
I made a bunch of step down buck converters to drop the 17V-24V input to 12V used by the LED lamps.
The fully populated panel board.
I also added a digital voltmeter as the LCD display is not a permanent part of this system, it is only used for setting up and adjustments.
This portion controls the street lighting.
The relays are wired as a simple control system that switches power to the DC-DC converter when there is no more voltage coming from the PV array.
The DC-DC converter (right) is the same as used in the buildings but is tweaked so that the output is 17V instead of 12V. Output is distributed to the three busses via the diodes on the terminal block (bottom).
Part of the street lighting at night. The lines are running at 17V this time and there are clearly no building lighting lit up.
There are about 14 installed street lights total and there is plan to add some more soon.
26 May 2015
Due to the additional loads and to have some margin against less sunlight due to dust or rainy season, we opted to add another pair of 150W panels for a total of 900W PV array.
Not sure if it was age or the dirt but the old panels were outputting about 4.5A at the time while the new ones were pushing 8A! Time will tell if the new ones are really better as the supplier says so.
I decided to take the FLIR for a spin and checked the roof temperature.
The max reading on the roof sheeting is about 55degC.
Under the PV modules, the temp drops to about 39degC.
With the FLIR pointed on the PV modules, I see interesting hot spots...
I can also see some hot spots where the junction boxes are.
14 Sep 2015
I replaced the 20A LS2024B with a new Tracer4215BN 40A controller for upgrade headroom and additional harvest improvement
There had also been a lightning strike to a nearby tree and the resulting EMP killed about 25 LED bulbs in two buildings and streetlights plus a few step down converters for the streetlights. Besides those, there were no damaged electronics in the main office and the DC-DC converters in the buildings.
Updated system diagram.
Click me for full size.
25 Feb 2016
Can you guess what has been added?
I made an arduino based battery metering system.
This is a much better way of monitoring the battery state at any time.
I did have to add a buck converter since my battery metering circuit can only take up to 16V max.
A 50mV 50A shunt has also been installed on the battery negative to measure incoming and outgoing energy.
This is an animated GIF of what are being displayed in actual speed. Available measurements are: realtime volts, amps, watts and amp-hour DoD, min volts since last full charge, peak amps, peak watts, highest amp-hour DoD since last full charge, and finally, time in hours or days since last full charge.
Some interesting data: The battery bank is 400Ah 24V so a max DoD of 87Ah is 22% daily. With all the lights on at night, it reads 400W and about 80W when the building lights turn off while the street lighting remain on. Peak power was 750W for a 900W PV array. The pic was also taken at almost noon and it was showing 18Hrs since last full charge meaning the batts were 100% at around 6pm. It has been cloudy these past few days too.
17 Oct 2016
Array has been increased to 1.75kW. One 150W panel has been removed and 10x 100W panels added in 5S3P wiring.
It was interesting that while it has been in use with 6x 150W panels, we only saw up to 850W peak on a 900W array. During rewiring, We removed one 150W panel and left 5x 150W for a total of 750W but that array peaked to 900W! More than the six panels.
The final config has been optimised for low light output so the peak is intentionally clipped and only peaks to 1.3kW max as the controller self limits to 41amps. This was to ensure batteries are always fully charged even in poor weather and handle the additional loads as some buildings will be expanded for more capacity.
Page created and copyright R.Quan © 13 Oct 2014.