Erica's Solar Panel

Sometime between June and August 2023, Erica's solar panel (originally fitted in August or September 2020) ceased to function. Had it been a standard panel, screwed into brackets, we would undoubtedly have whipped it off and replaced it in a matter of weeks, but with it being a flat panel that was well adhered to the roof, I got as far as buying a windscreen removal saw (cheese wire type), then we procrastinated on the job until 3 years had passed. 

We managed just fine for those three years, with the leisure battery only being charged via the split charger, but when the engine battery gave up due to old age in June, we decided that a new panel would be useful (the split charger works both ways, so with a solar panel, the engine battery would have been kept topped up too). This time I didn't opt for the expensive aluminium-backed semi-rigid job, but a cheap and cheerful alternative with a longer warranty. 

Removing the old panel seemed, as we started the job, like it was going to be an absolute nightmare of a task, but once we worked out the technique with the cheese wire saw, it went reasonably swiftly (the first 1/10th of the removal probably took as long as the final 9/10ths). 

With the old panel removed, that still left me with quite a clean-up job:

The old, dead panel, half way through sawing it off. The new panel doesn't look anywhere near as nice as this one, but it was a third of the price (and being on the roof, it's not like anyone looks at it very often!)
 

Starting the clean-up process, seems I didn't take a single photo once I had the roof in a nice clean and shiny state. I also didn't photograph the new panel in situ.

So, I spent a chunk of hours, spread over a week or so, sitting on Erica's roof, cleaning it up. Being a very hot and sunny summer was restrictive; every time the sun came out, I had to stop, as it was just too hot up there. The glorious weather was also restrictive for getting the new panel glued on as our chosen Sikaflex adhesive relies on humidity to go off, and after weeks of sunshine and heat, the daytime humidity levels were lower than was ideal. So, we waited and we waited for a cloudy afternoon, then we waited until evening to minimise the length of time until humidity would rise as night fell. 

Finally, three weeks after starting the task, the new panel was installed.

As for the failure of the old panel, I suspect it was (kinda) our fault, and thus we did things a little differently this time.  We used fewer beads of adhesive and used nylon spacers to make sure the adhesive beads stayed at least 5mm deep, thus maintaining an adequate air gap underneath. Perhaps most importantly, we didn't apply silicon along the sides of the panel, thus maximising under-panel airflow, and also allowing water to escape. I think the silicon along the sides was probably one of the biggest culprits for the failure of the old panel, as water had clearly pooled on one side, and if it did that in winter and then froze, it would have put undue pressure on the panel, eventually causing broken connections between cells - however, I'm not fully taking responsibility for this poor choice, as it was exactly what the original panel's fitting instructions told us to do. My theory, back in 2020, that water that got under the panel would have a clear route out via the troughs in the roof didn't take into account that water that was under there when Erica was parked up wouldn't be able to escape from the lower side of the roof until she next moved. The question now (and only time will tell) is whether flexing of the panel due to it being a less rigid design combined with having less adhesive coverage, will cancel out having solved the issue of water getting trapped underneath it.