Thermal Images of Modules with failed MLPE RSDs

Status
Not open for further replies.

Greg H

Member
Location
Durham, NC, USA
Occupation
Electrician
A thermographer colleague of mine sent me some thermal images they had taken of systems that were experiencing frequent failures of their SMA JMS-FSun Spec rapid shutdown devices. He said that many of the failed RSDs were overheated and the heat transfer to the module manifested as hot spots that were visible on an IR scan. So far, so sensible.

What was interesting though is that for the most severely damaged RSDs the IR scan showed a checkerboard pattern on the module (see pictures). He asked me what I thought might be causing this. I told him my best guess is that the failed RSD had somehow cascaded to the PV modules and this was manifesting as some kind of cell level thermal anomaly. Wasn't really sure though about exactly how that might work though, so, thought I would share some of the images here and see if y'all had any thoughts about it.

Thanks
 

Attachments

  • image13.jpg
    image13.jpg
    1.6 MB · Views: 17
  • image10.jpg
    image10.jpg
    317.3 KB · Views: 14
  • image9.jpg
    image9.jpg
    1.4 MB · Views: 13
  • image8.jpg
    image8.jpg
    1.2 MB · Views: 12
  • image7.jpg
    image7.jpg
    1.2 MB · Views: 16
  • image6.png
    image6.png
    896.9 KB · Views: 15
  • image5.jpg
    image5.jpg
    1.1 MB · Views: 15
  • image4.jpg
    image4.jpg
    296.8 KB · Views: 14
  • image3.jpg
    image3.jpg
    341 KB · Views: 14
  • image2.jpg
    image2.jpg
    1.4 MB · Views: 14
All I know is the new residential SMA inverters will no longer work with the JMS-F. Just APSmart.
Perhaps this is why?
 
A thermographer colleague of mine sent me some thermal images they had taken of systems that were experiencing frequent failures of their SMA JMS-FSun Spec rapid shutdown devices. He said that many of the failed RSDs were overheated and the heat transfer to the module manifested as hot spots that were visible on an IR scan. So far, so sensible.

What was interesting though is that for the most severely damaged RSDs the IR scan showed a checkerboard pattern on the module (see pictures). He asked me what I thought might be causing this. I told him my best guess is that the failed RSD had somehow cascaded to the PV modules and this was manifesting as some kind of cell level thermal anomaly. Wasn't really sure though about exactly how that might work though, so, thought I would share some of the images here and see if y'all had any thoughts about it.

Thanks
Those are TIGOs, not the JMS version from SMA. TIGOs are trash and a known fire hazard at this point.

I've seen that thermal signature on modules without MLSD devices so I'm not convinced the two are related.

Also, if it matters, the meta data is still tagged to the thermal images showing the lat/long of the site that was photographed.
 
My guess:
The signatures might simply be of modules that aren't producing. After all, the sun's energy is still hitting the modules and that's got to go somewhere. Got any unplugged ones to compare to?
 
Have Tigo RSDs caused any specific fires that you know of?
Yes. Some are localized events, some have spread to the membrane, some are just a pffffffft when the magic smoke gets out. During a site walk with an industry expert, he mentioned the typical failure method allows the units to fail a second time. It fails once, then can arc over again due to the incident energy still present.
 

Attachments

  • TIGO2.jpg
    TIGO2.jpg
    120 KB · Views: 4
  • TIGO1.jpg
    TIGO1.jpg
    138.5 KB · Views: 4
My guess:
The signatures might simply be of modules that aren't producing. After all, the sun's energy is still hitting the modules and that's got to go somewhere. Got any unplugged ones to compare to?
Yes, To some extent loaded cells will be hotter cooler than open circuited cells with the same insolation since no power is being exported from the cells.
(Unless a loaded cell absorbs rather than reflects more sunlight than an unloaded one?)
 
Last edited:
Yes, To some extent loaded cells will be hotter than open circuited cells with the same insolation since no power is being exported from the cells.
If "loaded" means "producing electrical power through an external circuit", then I think you mean that the loaded cells will be cooler? This is easily observed with IR cameras.

Cheers, Wayne
 
image9, and maybe image2, seem to show unloaded modules that are hotter throughout. Images13, 8, 7, and 5 show something more complicated, but I still wonder if they could be unloaded or partially loaded modules, perhaps taken earlier in the day before all the cells heated up, or with the uneven heating due to peculiarities in the manufacturing.
 
There has been concern in Oregon about RSD & fires for a while now. There was a large fire in Oregon caused by one and there have been proposals to take it out of the code.
 
There has been concern in Oregon about RSD & fires for a while now. There was a large fire in Oregon caused by one and there have been proposals to take it out of the code.
RSD, another genius move by our esteemed code making panels. Let's put a zillion pieces of power electronics scattered all over, yeah there we go , that will make things safer 🙄
 
That sounds like a conspiracy "theory" to me.
I've been involved in a couple of forensic engineering reviews of C&I PV fires. Generally, several stakeholders are looking for who to sue and usually an insurance company looking for ways to limit the pay out for damages. So several groups are looking over the event together, and everyone in a group is under NDA to keep information from getting to the other groups. The only time I have seen engineering reviews of a C&I fire go public is when an AHJ was involved and public building departments don't sign NDAs so they can talk. That is the reason we got the rundown on the Dietz & Watson PV file in NJ.
 
Last edited:
RSD, another genius move by our esteemed code making panels. Let's put a zillion pieces of power electronics scattered all over, yeah there we go , that will make things safer 🙄
I know people love to hate on RSD. Keep in mind that RSD was always intended to be a quick and dirty fix to a problem and it was only a stop-gap until it was to be replaced by listed PV hazard control equipment to make PV arrays electrically safe for firefighters to work around. Eventually, no one will need RSD, just listed PV array equipment. The vestiges of RSD will only live on as array isolators needed for discontinuous arrays or to move inverters away from arrays.
 
I've been involved in a couple of forensic engineering reviews of C&I PV fires.
I know people love to hate on RSD. Keep in mind ...
It sounds like you saying there is problem with Rapid shutdown and you have investigated fires caused by them?
Were all professionals here its not that we pick some piece of a PV system to 'hate' on,
rather code requirements should make things more safe not make things less safe, fail to work and endanger people.
These devices sound like the Federal Pacific breakers of tomorrow, people will be paying to remove these for decades to come.
Besides just causing fires they are not supported by the manufacturers very long and when parts break in older models they tend to fail 'on'. I have seen several older RSD systems that do nothing when you hit the button.
Firefighters that I have asked have never herd of them and dont train on them.
 
Status
Not open for further replies.
Top