The Monkey House

< Writing : Other : PV Solar >


Installing Solar Panels

Based on my PV system modeling work, and with a roof that should last comfortably longer than 10 years, I called a local installer to come out and have a look. I have a southwest-facing roof that is only slightly shaded and is big enough for almost 6KW of panels, which should cover roughly my annual usage. I have another roof plane facing southeast which can accommodate another 2.5-3KW for future expansion. Because there is some shading, I decided on a SolarEdge inverter with DC optimizers (offering per-module maximum power point tracking rather than per-string tracking) and had the electrical work done to support the additional panels while the installers were already crawling about in the attic. Also, because we have a healthy population of squirrels in the yard, I had the space below the panels sealed off with a coated steel mesh barrier.

My various system design changes made the base system fairly expensive, though future expansion should be much cheaper as a result. Time will tell whether that was a wise investment or not. Even if I leave the system alone the payback period is expected to be under 11 years. But if it does get expanded in the future, that payback period should drop nicely. Here are the system costs at installation:

$17,550Base price for 5700Wdc system$3.08/W
$ 2,180Upgrade to SolarEdge power optimizers and inverter$0.38/W
$ 770Steel mesh and installation$0.14/W
($2400)Focus on Energy rebate$0.42/W
($5431)Federal tax incentive (ITC)$0.95/W
$12,673Actual system cost$2.22/W

Predicting actual value of the solar system is difficult. Common modeling tools (eg SAM) allow you to enter an electric rate structure to have your pre-solar and post-solar costs calculated and compared. However, a more realistic comparison may be pre-solar costs calculated with flat-rate electricity pricing, and post-solar costs calculated with Time of Use rates, since the addition of the solar system makes the ToU rate structure economically viable. Furthermore, considerable guesswork goes into initial modeling of ToU rates, because most homeowners - myself included - will not have fine-grained usage data before signing up for net metering. I know what my month-by-month power usage looks like, but not day-by-day or hour-by-hour. For my modeling, I found example hourly residental usage data, scaled it to my expected annual usage excluding the Leaf, and added 1.8MWh of off-peak energy usage for the car. Two different versions of SAM calculated expected payback periods of approximately one year less and one year more than the estimate provided by my PV solar contractor. With three different models predicting payback periods of 10-12 years, I decided to go ahead with the install.

The installation process went smoothly and took about four days in all, plus an hour for system commissioning. First, a pair of electricians came in to run rigid conduit from my basement where the inverter will live, up through the garage to the lower sub-array of panels, and then up to the attic above the second story to the higher sub-array and the potential third sub-array. They installed the inverter and wired it through an external mechanical disconnect switch mounted on the side of the house and into my electrical panel. I have 200A service running through a 200A rated panel; apparently in such cases an additional 20% power input is allowed at the opposite end of the panel from the main breaker to accommodate PV solar, wind, or other sources of power other than the main utility feed. This electrical work took almost two days, and then there was a brief hiatus before the panel installation crew was available. Panel installation took about a day and a half, with installation of the racking being the major use of time.

Continued in part three, analyzing year one results.

Version 0.2     |     Originally written: July, 2017     |     Latest revision: August 18, 2018     |     Page last generated: 2018-08-18 18:04 CDT