I;m considering adding solar panels to my garage, or carport.
Has anybody used a few Lux sensors, scattered in different locations around their house/garden, to asses potential locations and angles?
I thought this would be a good idea to asses best location for some panels (already have some on the house roof).
Its not possible to see the garage or carport roof, without ladders, to see how often the sun is on them.
You can use https://www.sunearthtools.com/ to generate a Sun path which should help you figure out the best placement without the need for Lux meters.
You mean you have shadows from trees buildings etc.?
Best angles can be calculated, no need for sensors.
Yes, very rural house, have some trees around the perimeter. It could make a huge difference if panels were one side/end, or another.
Garage roof is almost 60m2, car-port around 20m2 and there are also other locations around the garden. (and potential new garden room).
I answered your question in my Lux sensor topic.
You can estimate the “base-value” for the production just measuring how many degrees it’s off from south. But the shading is trickier and you have to consider also the difference over the year. Some orientation could be perfect during the summer, but heavily shaded during winter when the sun is lower.
Anyway, lux sensor could be used, but they are usually not designed to measure direct sun light, so chances are high that they get saturated. Photodiode would probably work better. Both need diffuser.
Some small cheap solar cells with shunt resistor connected to Esp ADC could be best option. Or some ready made logger.
I disagree. The BH1750 is chap and perfectly suited to this as it can measure up to 100,000 lx.
That’s pretty good, definitely usable here. I didn’t refer any specific sensor, some other sensor tuned for indoor/indirect measurements might saturate on the half way.
One inconvenience with lux-meter approach is that it senses 400-700nm and peaks at about 550nm while low angle sun irradiance peaks >700nm. That would cause an error if measuring east/west orientations, while for south orientation it would match well.
Interestingly, just installed a Ecowitt 7 in 1 weather station and that records Lux levels. Its on the corner of the garage and is showing highs of just over 70,000 at one point today.
Its particular location will likely get the most sunlight.
I want to avoid having to spend money on optimisers, a solar solution for the garage roof will be done on a budget, so I would be trying to get the most from some ÂŁ60 panels (400-500w panels) and will be installing them myself.
Some of the trees may be trimmed back, to assist with generation, they can be orientated due South, the trees are mainly to the east.
It could be argued that the actual real Lux level is irrelevant, its more about the difference/range between potential locations on the roof. And whether its getting direct sun, and at what times in the day. The sensors could even be mounted in a clear enclosure etc, as long as they are consistent.
Any technical issue with one ESP32 having 6 sensors attached, to different IO ports?
That likely has sensor optimized for direct sun light.
Long wiring.
Only if using amorphous solar panels (unlikely). Silicone cells under glass match that response well enough.
And only if in the northern hemisphere.
Down here we point ours north.
This is a simple linear equation I use to compare the PV power output to my BH1750 lux reading:
No significant roll-off in the mornings or evenings is present. Though as stated, they both point north.
- name: "PV A Efficency"
unique_id: b450d48f-1018-4cf2-8d1d-45a2a337a259
unit_of_measurement: "%"
state_class: 'measurement'
icon: 'mdi:solar-power-variant-outline'
state: "{{ ( 100 * states('sensor.pv_power_a')|float(0) / (states('sensor.outside_light_level')|float(1) * 0.05) )|round(0,default=none) }}"
availability: "{{ states('sensor.outside_light_level')|float(0) > 0 and has_value('sensor.pv_power_a') }}"
Commonly used solar panels absorb about 300nm-1um spectrum. My point was not the angle of the sun related to the panel, but the wavelength of irradiance when sun is low (warm white) That’s dominant at the morning and evening.
That means they don’t receive direct light at all, different from east/west orientation.
Anyway, I’m not saying BH1750 isn’t good option for solar irradiance sensing, actually specs look very good. I’m saying that some random smarthome lux sensor might be bad choice. And small PV cell as a sensor likely gives best results. Setup might be trickier though.
ps. if you mount the panels on a flat roof facing up, the whole discussion about wavelengths becomes irrelevant.