Infrared Camera Inspection of Photo-voltaic Solar Farm

[ecoems]

Hi all,

This may require someone with direct experience with Solar IR inspection but you never know.

I am currently looking into using an IR Compact camera and Drone to visual inspection a Solar Farm which we have build. The solar farm is not reaching its expected performance ratio targets... I understand this could be down to a large number of things, but as i understand it an IR camera could quickly and easily identify issues with the modules them else and there strings.

What i am unsure of is what issues the IR camera will pick up on the modules and how they will show up on the IR camera?
Could the IR camera also be used to evaluate terminations and switchgear issues in both the string boxes and inverters?

Thanks in advance

ECOEMS

My first post on Eng-Tips and i can guarantee its not the last!!

 

[MacGyverS2000]

What are you looking for with the cameras? Are you looking for hotspots in the switching gear, or is there some odd benefit I'm not aware of at imaging the panel itself?

Dan - Owner

http://www.Hi-TecDesigns.com

 

[ecoems]

From the limited information i can find online in english... I understand that an IR camera can be used to see 'hot spots' on the solar module its self but unsure that the 'hot spots' are showing... would it show reverse polarity, installation faults/ damage, manufacturing faults and what would they like look through the camera?

What would the 'hot spots' indicate in the switch gear? Loose terminations, faulty fuses....

 

[DRWeig]

ecoems,

Hot spots are relative things in an infrared picture. If you have a field of panels and a few of them are glowing brighter than the others, the bright ones need to be investigated. You don't know why they are hotter yet, you only know that they are hotter than their identical counterparts under identical conditions. A close-up of one of those panels might show a bright area near the lead connections -- a possible corroded or loose connection. A bright area mid-panel indicates a defect or damage perhaps. You'll have to lay eyes on it close up. Note that in a solar panel field, dark spots are areas to investigate too. Ideally, you should see a homogenous picture, all alike. Investigate the ones that are different.

In electrical equipment, you are again looking for things that are not like the others, and also for things that are very bright. Each bright spot needs investigating. If the incoming terminations to a 3-phase panel show one phase brighter than the other two, I'd start with a clamp-on ammeter. Perhaps the hot phase is carrying more load. If the load is the same, though, the connection may be loose or corroded inside. Check then with an infrared temperature sensor, see if the actual hot spot temperature is just 50C versus the other two phases 30C, or if the hot one is getting near the equipment rating (very bad). In either case, if the load is the same but the temperature is different, get a good electrician over there.

Looking at a row of 20A breakers, they all might be slightly different colors. The more load passing thru them, the warmer they are. The brightest ones need a quick amp check, just to make sure they're carrying loads below their ratings. However, if the brightest one is not carrying the highest load, it may be loose. Electrician time.

If you remember that all is relative, that you're looking for differences, and that your imager only shows you temperatures (not problems), you'll do well. Hot spots are things to check further. You can't know what they are until you get eyes and meters on them close up.

Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[IRstuff]

As a general rule, exposed surfaces are at a premium in a solar installation. Therefore, it may be extremely unlikely that an aerial drone would be able to see much, since the electronics for each panel will be mostly be behind the panel and possibly inaccessible from the air.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[jraef]

If anything, I would think that the only relevant information you could get from a drone flying over the field with an IR camera would be either a bad solder joint in a panel showing up as a hot spot, or a dead panel or section of a panel showing up as being cooler than the others because no current is flowing (open circuit somewhere). As mentioned, you would not be able to see anything inside of the combiner boxes or gear through the doors unless you had clear GLASS (not plexiglass) on the doors, which is impractical.

"Will work for (the memory of) salami"

 

[IRstuff]

" unless you had clear GLASS (not plexiglass) on the doors, which is impractical."

Actually, most IR imagers operate in the LWIR, which is not passed by normal glass, but is passed by certain plastics (in the all-encompassing sense) like teflon and some acrylics

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[racookpe1978]

Instinctively and intuitively, your idea of being able to investigate the entire solar collectin area at one time looking for hots spots and cold spots sounds good.

I cannot quantify exactly what you will find though - and what you may find is a "negative" - but ANY information will help you troubleshoot further.

You will have eliminated "dead areas" where connections or entire panel sections may have failed. You will have eliminated bad wires, failed solder connections, and areas where excessive dirt and grit and contaminates are not getting washed off as expected. Repeat the IR inspection every month and you'll start building up trend data that will also be useful.

But whether you will actually find the problem is unknown. Still, NEVER discard the positive effect on your client of simply trying different solutions (especially innovative "new" solutions") to show that you are working on his array.

 

[Sparweb]

It might be more cost-effective and quicker to start the inspection from the ground. There are solar PV installers that use hand-held infrared cameras already to inspect their work this way.
Using hand-held equipment would also allow you to inspect the BACK of the panel, which would be impossible from the air. All of the wiring is on the back, so I don't see what looking at the top will do for you.

STF

 

[GregLocock]

My guess is that you are on the track of a good idea, but I bet the drone idea doesn't work out.

A PV array with 22% efficiency is typically running at about 60-70 deg C in direct Australian sun, which is around 1 solar, if it has only passive ccoling. Ambient air temp is about 35. An O/C cell is only going to be 5 degrees hotter than a working cell, not too sure if you could spot that from the air. Obviously a short circuit could be anything up to the melting point of your solder, so you'd spot that.

I'd be more inclined to wander round the farm with a wattmeter and identify which panels are bad rather than remote sensing. Once you have a bad panel then debugging individual strings and cells is tedious but not difficult.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[LionelHutz]

It would be 1 tool to use. I'd tend to agree that a walk under the panels looking for anything abnormal would be more useful then taking images from above. Using a simple clamp-on DC ammeter to measure the current output of each string could help find any possible issues too.

Depending on your inverter arrangement, you might be able to compare output metering to determine groups of panels you need to investigate further.

 

[MacGyverS2000]

A random idea... put some dirt-cheap sensing circuitry into the panel's control box, wiring a single LED (IR?) along the edge of the panel facing towards the sky. If it senses a problem, the LED goes on... a drone with a CCD camera should be able to spot that. It's hands-off and remotely checkable.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[IRstuff]

Drones don't necessarily have to airborne. The same infrared cameras could be mounted on a robot which could be programmed to look at the backsides of the panels. With DGPS, the robot's position can be known to better than 1 ft.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[GregLocock]

On one solar car we had a led wired into each string, which was back driven so if the string voltage was less than the bus voltage it switched on.

Yes there was probably a cost in efficiency, but boy did it make debugging easy.

We could also see each string dropping out as people walked in front of the array, which was fun.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?