Autoclean a soiled solar panel

I am working on how to autoclean a solar panel soiled with dust.

The project is divided into two parts:
The first part is on how to detect when the solar panel is dirty. In this part I need ideas on which light sensor to use and how to go about designing the circuit. Someone suggested this idea of placing an LDR in a glass near the solar panel, so when the glass is dirty the solar panel should also be dirty. But is clear to me on how to achieve this. I need suggestions based on this or any other method which would yield the same result. Thanks for your response in advance.

The second part is on how to automate the cleaning process

An LDR would think it is dirty every night. How about shining a IR LED at the solar panel and measuring the reflection with a photo transistor or photo diode. Maybe just do that at a certain time every night so the sun doesn't affect the measurement.

victorgozan said:

Sir, please I would like you to give a more explanation on how I could use phototransistor to measure the reflected rays from IR LED and use it to determine when there is dust on the panel. thanks in advance for your concern.

Click to expand...

This is what I had in mind.
(Quote from web page)
Light sensor using photodiode

**broken link removed**

The project about light sensor circuit shows the application of photodiode to detect the presence of light. This light sensor can be used as an intermediate circuit in various applications to detect the presence or absence of light. The sensitivity of the sensor can be adjusted using the variable resistor VR2

What lies Inside the Circuit?
This circuit is based around a LM339 comparator along with variable resistors, photodiode, LED and a solo resistor.

A photodiode, used as a photo detector, generates current in the circuit when light incidents on it. The circuit of this project uses the photodiode in reverse bias mode with resistor R1(10k ohm). This resistor does not allow too much current to flow through the photodiode in case a large amount of light falls on the detector. Initially when no light falls on the photodiode, it results in high potential at the inverting input of a comparator (pin6) of LM339. When light falls on the photodiode, it allows current to flow through the diode, and thus drops the voltage across it. The non-inverting input (pin7) is connected to a variable resistor VR2 (preset) which is used to set the reference voltage of the comparator. A comparator works on the principle that its output remains high as long as the non-inverting input is at higher level than that at its inverting input. Here, the output (pin1) is connected to an LED. The reference voltage is set to correspond to a threshold illumination through a preset VR1 (10K ohm).

If the solar panel is dirty, it wouldn't reflect as much light. The voltage will be lower. You can adjust the variable resistor VR2 to tell the comparator when to turn on. (Off with high voltage, On with low voltage) That can start your wash cycle. Maybe a car wash type system?

Maybe use the alarm signal from an electric alarm clock to power this at the same time every night.

Rather than reflection loss, why not use transmission loss with a stable light source at night?

The output current will be proportional to efficacy of panel with dirt loss, so use something like a 6W LED bulb at a suitable distance.
The output can be calibrated with any photo diode, PD, which is very accurate compared to any LDR. THus you measure the ratio of current from PV to PD or adjust the light source to give a fixed PD reference level. THere are other factors that will affect accuracy, but this is the easiest way I know to give accurate readings.

The PD may be a wide angle detector for low geometric sensitivity from a suitable dust-free distance.

To detect if the panel is dirty, you need a reference of cleanliness. That can be a solar cell at a location/reach that can be manually wiped clean and with the same sun exposure.
A comparator circuit checking the panels output and the clean cell output may do it to trigger the washing part.
Now, the washing part, not tackled by other responses...
Plain sprinklers, fixed or *. Manual or activated by the reference sense/comparator enabled only during daylight. That can even get rid of some snow if your location deserves it considered.
If you want to fine tune, the sprinklers can be fed by some detergent dosifier as garden hoses with fertilizer dispenser do. Any pipes rising to a roof (if that is were the panels are) must be drained empty by a bottom level purging solenoid after the wash, or weeping tiny hole so there will be no frozen/burst prone pipes.

*----> **broken link removed**
*----> https://i.stack.imgur.com/9vB38.jpg

You may find with time that the use of a dirt sensor/comparator circuit can be replaced by a no-matter-what-once-a-month timed wash.

victorgozan said:

The second part is on how to automate the cleaning process

Click to expand...

Two options I can think of:
1) High pressure: similar to a pressure washer
2) A windshield wiper

Mechanics could be different for a small panel or a larger array.

Snow clearing is also cleaning.

ClydeCrashKop said:

An LDR would think it is dirty every night. How about shining a IR LED at the solar panel and measuring the reflection with a photo transistor or photo diode. Maybe just do that at a certain time every night so the sun doesn't affect the measurement.

Click to expand...

Kinda like the reflective object sensor and measure at night. Those you can buy. Difference sensing is possible. keep one aimed at a enclosed clean piece of glass and the other at the dirty panel. As the sensors age, they may track.

Then there is just periodic cleaning.

Say, use an irrigation controller that gets weather data from the Internet. They don;t water when it rains.

**broken link removed**

Thanks guys for your replies

KeepItSimpleStupid said:

Kinda like the reflective object sensor and measure at night. Those you can buy. Difference sensing is possible. keep one aimed at a enclosed clean piece of glass and the other at the dirty panel. As the sensors age, they may track.

Then there is just periodic cleaning.

Say, use an irrigation controller that gets weather data from the Internet. They don;t water when it rains.

Click to expand...

Or just use a beam splitter. Emitters also age. That would probably eliminate the need for a reference reflector too. Just sample a small percentage of the beam.
My bet is on periodic cleaning or monitoring solar brightness and compare with PV output.

John