A couple months ago, I wrote in a post that we were preparing to install more solar panels. As of a couple weeks ago, this was completed and we’ve been producing a lot of power lately as a result (as it’s been unseasonably sunny).
We were able to complete the install over about 5½ days, most of which was consecutive. Here’s a summary of what we did.
Pulling New Wire
The first step was to replace some of the wires between our house and the original array. The wires were big enough for the amount of power that array could produce, but were not appropriate for twice as much power. In addition, I had discovered that the total distance between the house breaker panel and the arrays, about 300 ft, was a potential problem due to the resistance of that much wire. So we replaced a lot of wire, some of which is in a conduit in the house, some overhead outside, and the rest in a conduit out to the array site. I did this with my dad, and it took the two of us about a day.
The nature of a ground-mounted solar array is that it has to be held to the ground somehow! In our case, this meant eight concrete footings 12 inches in diameter; half of them 3 ft deep and the others 5 ft deep. I rented an auger from a local hardware store and dug these on Friday, March 21st. The same day, Lowe’s delivered an entire pallet of Quikrete that we would be using to fill the holes.
On Saturday, we had several friends over to help. We started by putting the vertical pipes in their respective holes. We carefully lined them up with the old array and held them in place with 2x4s. Then we hauled several dozen 60-lb bags of Quikrete over to the site and dumped them in the holes. It was great to have a few friends help — it would have been extremely hard for Trish and I to do this task ourselves! We were done with this step by about 2pm.
Setup Ground Mount
Next, on Sunday, Trish and I built the ground mount. I started by removing the 2x4s we had supporting the poles; by now the concrete was cured enough to hold them up just fine. We used a transit level to mark where each pole would need to be cut off, to match the old array, and cut them off with a band saw. I was surprised, and happy, at how easy it was to cut the pipe this way!
Using various couplings, we attached all the rest of the pipes, which I’d ordered pre-cut to the appropriate lengths. This went very well, and pretty soon we had the ground mount completed.
Also on Sunday, we were able to install the rails. The rails run vertically on ground-mounted systems, and the solar panels attach to them with clamps, this is how they do it in San Diego based solar companies. The aluminum rails, too, were pre-cut to the appropriate length. We carefully lined them up with the existing array and bolted them on. But after looking at it, I realized it was not lined up well enough, so we ended up moving all the rails a bit to get them square, before we decided to call it a day.
When I went out Monday morning, looking at the setup from a distance, I noticed that we had obviously not got the new one lined up exactly with the old one. I started taking measurements, and sure enough, it was not what we thought it was the day before! I again adjusted things and squared it up. Finally, I was very confident that we had it right.
Microinverters & Cables
The next step was to lay out the “Engage” cable that goes up and down the rails to connect the Enphase microinverters together. We used two cables, one for each half of the array. The cable gets attached to the rails with special clips. Then we started attaching the microinverters themselves. A special bolt is made for attaching Enphase inverters directly to the rails. We positioned the inverters so that they would be as close as we could reasonably get them to the vertical center of each panel that would eventually go on top of them.
Time for Solar Panels
Now we were ready to start install the solar panels. A very helpful friend, Keith, drove up from Bellevue, about 100 miles, in his Nissan LEAF to help us out. He showed up just in time to help with this part (and some wiring). At first we had a little trouble getting the panels lining up exactly the way we wanted, but soon we got the hang of it, and it went somewhat smoothly. However, each panel had to be carried by hand some 300 ft from my shop where they’d been stored. This turned out to be one of the biggest chores of the whole project!
We tried each carrying a panel at a time by ourselves, but that was too much, so Keith and I ended up carrying one panel at a time together. We were working as fast as we could. But eventually Earth’s rotation caught up to us, and we had to bring out some lights so we could see what we were doing. Finally, by about 9:30pm, we had all the panels in place (but not connected or anything). We were all exhausted. Keith went home and Trish and I went to bed.
Connecting it all
The next morning, Tuesday, I was still exhausted, so I didn’t immediately go out and start working on the array. After a few hours, though, I’d built up enough energy to get out and keep working. I connected all of the panels to their microinverters and hid the cables in the rails, connected the Engage cables to regular wiring that we’d run through a conduit the day before, connected a grounding wire to all the rails, and connected the wires in the breaker panel to their breakers. There were still a lot of cleanup tasks to do, but everything was hooked up and ready to test out. So I turned on the breaker, and watched (on my phone) as each of the panels started producing, one by one!
It took me a couple more evenings to finish the cleanup tasks, including trimming cable ties, putting end-caps on the rails and pipes, and picking up a bunch of garbage.
Shortly after the array was turned on, it became apparent that we were going to have an issue due to the length of wire, despite our efforts of increasing the wire size. In order for the inverters to push power out into the grid, they have to push with a higher voltage than the grid is pushing in with. But the resistance of wires decreases voltage, and more so when more amps are being pushed. As a result, when the panels are in full sun, there’s a voltage drop of close to 20 volts along the wires. If the panels were pushing 240 volts, only about 220 volts would be left by the time it got to the house.
This is a problem because the voltage given to us by the utility is about 245 volts, meaning the voltage required by the inverters is 260 to 265 volts at peak time. By default, the inverters are configured to turn off if they exceed 262 volts. This started happening almost immediately.
Luckily, Enphase, the company that makes the inverters, was able to deploy an update to them over-the-wire that increased their maximum voltage slightly. Since then we haven’t had any problems with the inverters turning off for this reason.
It’s worth noting that this voltage drop equates to a loss of about 700 watts of power over the length of the wire when we’re at peak production (11.4 kW AC). This will have a slight effect on the benefits of our solar array, but I don’t think it’s significant in the long run.
In the first couple weeks since the full array has been live, our production has far exceeded our expectations. That’s probably mostly because it’s been sunnier than normal, however. I expect it will even out over time. I’ll probably start writing up reports quarterly, or at least annually, covering our production, the amount of money we’ve made, and so on. Until then, enjoy the sun!