I’m delinquent on my promise of a couple weeks ago to reveal the solution to my PV array’s under-production problem. I hadn’t wanted to chatter about the issue until we’d actually solved the problem. Our installer, SPG Solar, and the panel manufacturer Kyocera have since come through with the ultimate solution, so I’m ready to tell the tale.
Back in August, we thought we’d found the problem right away when we flipped open the DC disconnect switchbox and a gallon of brown water jumped out.
Yes, jumped. It was alive. And seriously nasty… we’re having to repaint the splash zone.
Long-term readers may recall that the DC switchbox flooded once before. Statistically speaking, this is nearly impossible. As the field supervisor and chief system debugger from SPG commented, “I’ve only ever seen this problem two times in six years… and for both of them I was standing right here on your porch.”
The problem, this time, was different — not so much that water was getting in, but that it couldn’t get back out. The counterintuitive reality is that outdoor switchboxes like these are built on the assumption that moisture will get in, so they come equipped with punch-outs to allow for draining and ventilation. In January 2004, when our first DC switchbox was replaced, the installer (who is no longer with SPG) failed to pop open the drain holes. Ironic, no?
So, the first step was to replace, and actually upgrade the DC disconnect switchbox. We were confident that this would solve the problem. But it didn’t.
SPG’s crew then tested each module on the roof. One was found to create a voltage drop, and was removed, but this didn’t affect total system output.
Next they tried a spare inverter… no change.
Next they bypassed the original wiring harness… still no change.
Yet under full sun, the array was producing significantly less power than it should have, even when correcting for the angle of the sun (using an irradiance meter). To confuse matters, the array worked fine up to a certain level of generation, and then fell off as the sun climbed in the sky. All very mysterious, and frankly, frustrating.
SPG finally called Kyocera’s regional office for assistance. We’d replaced or bypassed every component in the system except for the modules on the roof. There wasn’t anything else to do but pull down the 23 remaining modules and put up new ones.
The modules are warrantied to produce 90% of their rated power for 10 years, and at least 80% for the next 10 years. We’re well within the first decade, yet clearly below 90% production, so Kyocera stepped up or a full replacement, including the labor costs. I’m impressed with this level of service — warranty or not, to see any company do the right thing without a fuss seems remarkable, and worth noting.
SPG deserves a commendation too. They have north of 70 man-hours invested in the diagnosis and repair of my PV system, yet I haven’t paid them a cent since October 2003. But I’ll say this: PV is a long-term investment, so if you’re considering it, choose a company that’s going to be around in 20 years. Sure, you might save a few dollars now buying through a co-op, but before you do please find out who you’re supposed to call five years down the road when production is off.
That’s not to say that most PV systems fail. You sort of have to figure that the reason manufacturers offer 20-year warranties is that most systems work fine for that long. Still, I recommend using an installer with a track record, and more importantly, a future.
Here’s the money shot… 15 minutes after turning up the new system, production hit 2531 watts, the highest I’ve ever seen.