Electronic Design

# A Solar Story

“The meter’s running backwards!” It’s 1 p.m. on July 27, which means the sun is at its zenith, and the contractor has finally connected our solar system to the power grid for the first time. My wife Vicky is hypnotized, watching the black mark on the edge of the disc in our electricity meter move from right to left—moving pretty fast, truth be told.

This meter, the one that came with the house, is temporary. We’re signed up with the utility for “Schedule 7” rates. During the “middle” of the day, the cost of electricity will be around 28 cents per kWh. (The definition of “middle” depends on the season.) At other times, the rate will be around 4 cents per kWh.

We hope that when we’re putting power onto the grid, our electric bill will decrease by the larger figure. When we’re consuming grid power, much of it will be billed at the lesser rate. We still will be sending money to the utility every month, but we’re hoping it will be substantially less than it was before we got our solar system.

But capital and labor costs were pretty high. How long is the payback? I don’t know. Fifteen years is what’s bandied about. I figure there are too many unknown variables to say with any certainty. But if you’re interested, I can give you some basic numbers.

On the roof, there are 16 Kyocera KC190GT modules on a Pro Solar rail-mount system (Fig. 1). Solar panels are rated according to “PV USA Test Conditions” (PTC), which were developed at the Photovoltaic (PV) USA test site at the University of California, Davis.

The PTC rating refers to the output of a panel under conditions of 1000 W/m2 solar irradiance, 1.5 air mass, and 20ºC ambient temperature at 10 m above ground level and a wind speed of 1 m/s. Each module on my roof has a PTC rating of 167.7 W dc. The dc output of the total panel array is nominally 2683 W (under PTC conditions). Our array appears to be better than that, based on voltage and current measurements made by the inverter (Fig. 2).

That dc-ac inverter with the built-in measuring capability is on a wall on a second-floor deck. It’s a Fronius IG-series unit rated for 3 kW. The inverter’s nominal efficiency is 94%. The ac output from the inverter goes through an emergency cutoff switch located next to the house’s service entrance and goes on to a pair of breakers in the electrical panel.

The system’s estimated peak ac output is 2522 W. With good weather, it should produce 4500 kWh/year. If all of that output decreased our electric bill to the tune of 28 cents per kWh, that would amount to \$1260/year.

So what did it cost? That’s where it starts to get tricky.

To mount the panels, we needed some plywood under the shingles. The brackets for the mounting rails attach to that. But the old roof needed replacing, not just under the relatively small area where the panels would be mounted, but everywhere. All told, the roofing work came to around \$10,000. I might arbitrarily allocate \$2000 of that to the solar system.

The contractor charged \$24,600 for materials and labor, but we’re supposed to get a \$7000 rebate from the utility. (I’m doing some rounding-off here.) So non-roof costs were \$14,600. A county permit (we’re outside the city) cost around \$750, and the contractor charged \$250 for steering that through. (Permit charges for solar installations have been eliminated or reduced to trivial amounts all around the area, but not in San Mateo County.)

So our costs sum out around \$17,600. If we actually reduced our utility bill by \$1250 a year, payoff would be a little more than 14 years. Ask me in 2030.

You may be wondering how long it takes to spend \$17,600. Vicky got the ball rolling back in February. She came up with two contractors with experience. One didn’t return our calls, but the other one produced really good vibes. I wrote about him in the Megatrends issue of Electronic Design in June. He’s the guy with the all-electric RAV4 (see “Alternative Fuels Look To Solve Petrol’s Plunder,” ED Online 12857).

The roof job happened in March. It was fast and painless, except for writing the check. What made the rest of the system installation last until the end of July was panel availability. As I remarked in an e-mail to an ED reader, I don’t recommend solar panels as a recipe for instant gratification.

The problem seems to be that while European demand for panels (lots of subsidies) is still very strong, demand in the U.S. is ratcheting higher, and supply isn’t keeping up. When we started this exercise last winter, the contractor bid certain Kyocera panels. Then it appeared that they couldn’t be obtained with love or money, and he found some Sanyos he though he could get his hands on.

The Sanyos would have boosted our cost by a couple thousand dollars. But they used three-layer cells, with different bandgap energies at each layer, so the output would have been higher over a wider range of sun angles. The source for those dried up, though. The contractor then found some other Kyoceras that will cost less than the original bid, but they’re a little less efficient—at least nominally. As I said, they seem to be quite a bit better than they’re rated.

Fairly soon after we got the new roof, the solar contractor installed the inverter and ran conduit between it and the roof and between the inverter and an emergency cutoff down by the service entrance. And then we waited, while the contractor gave us periodic updates.

Finally, the contractor scored the panels—at exactly the time that local midday temperatures soared to new records well beyond 100°F. But once the panels were in hand, even just working early mornings, everything was in and wired up. The county inspector signed off on them by the early afternoon of the third day.

Now we’ll just have to wait and see. If I have any interesting tidbits to add, I’ll post them here.

UPDATE: In the system’s first week of long, sunny summer days, it put 124 kWh on the grid. In my initial writeup, I also left out an expense: \$277 to the utility for a new electric meter that can handle different rates at different times of day. It hasn’t been installed yet, and we’re still on the standard flat-rate billing plan.

SECOND UPDATE: It's a week later. The guy from the utility is coming later today to install the smart electric meter.

THIRD UPDATE: About 20 minutes after this story was posted on-line, I received an email from a reader at Sandia. He directed me to a Web page (www.fsec.ucf.edu/EPact-05.htm) that describes The Energy Policy Act of 2005 (EPAct 2005). In brief, the act says we're eligible for a \$2000 tax credit. As Mr Campbell notes in his email, that shortens our payback time by about a year and a half!

And the smart meter is now installed. It looks like a conventional meter, except that a multi-function LCD display replaces the old clockwise/counterclockwise dials. Amusingly, the display also includes an LCD analog of the old rotating silver disc—a series of marching dots and a pair of arrowheads that indicate which way an actual disc would be spinning. The man who installed the meter also told Vicky that we would now be billed for electricity only once every six months.

You can e-mail Don Tuite at [email protected].