Adding Solar, Part 3: Installation

I've documented the research and proposal phases of our residential solar power project here and here. After a long spell of bad weather, and my being out of town for two weeks, EcoHouse Solar began installing our photovoltaic system on April 15, 2015.

Their first step, after unloading gear, was to place the mounting brackets on the roof.

This is what the brackets look like.

By the end of the day, all the mounting brackets were up and the rails attached. Two brackets were placed too close to the shading tree, and were removed later.

The twelve solar panels and Enphase microinverters were stored in our garage for the next day.

This is one side of a microinverter.

Thursday the 16th had one-tenth of an inch of rain, not a day to be up on the roof. EcoHouse Solar came out on the 17th to mount and wire the panels and microinverters.

Joan and I left about noon for a weekend at the Arc of Appalachia annual Wildflower Pilgrimage. We chatted with the EcoSolar guys as we closed up the house, and learned that they were also familiar with the Arc. This turned out to be meaningful ...

On Saturday we participated in the Ohio River Bluffs hike, with the blessings of fair weather and peak blooms, especially of Virginia Bluebells, Dwarf Larkspur, and Blue-eyed Mary. When we returned to the Forest Museum for dinner and the evening program, there was a message for us at the front desk. From Kevin of EcoHouse Solar. Please call him back. Uh-oh.

I called him back. A spool of heavy grounding wire had gotten away from them, bounced off the trellis for the trumpet vines, and smacked into the outside faucet on that side of the house. Something you could not do if you tried. They thought they had heard water inside the house, and called the Reynoldsburg Water Department to turn off the water to the house. Kevin had arranged for a plumber to show up at 3pm Sunday. There was nothing to be done right away, but it was a restless sleep that night.

Joan and I were able to attend the breakfast, ceremonies, and part of a hike on Sunday before heading back, and it had started raining, so we were thankful that it was Sunday and not Saturday that had been shortened. We arrived home and saw that Kevin had left two jugs of distilled water for us on the front porch.

Dashing down to the (unfinished) basement, we encountered a scene that wasn't nearly as bad as it might have been. First off, the pipe hadn't snapped and produced a gusher. The seal of the fitting had, however, been broken, and sprayed water until the house supply was shut off.

Nearby boxes and insulation had been sprayed, but nothing important was damaged. The water had accumulated and flowed out across the concrete floor, absorbed along the way by cardboard boxes, old rug fragments, two pieces of foam, and the like. The tide reached only a fraction of the basement, and our losses were confined to several old cardboard boxes and some insulation. The cardboard boxes were pitched, and the rug fragments hung out to dry.

The plumber didn't show up at 3pm. At 4pm I called Kevin, who called the plumbing company, Water Works. To make a long story short, it turns out that the plumber had an arrival window, starting at 3pm, and arrived just after 8pm, the we'll-go-to-a-motel cutoff time Joan and I had agreed upon. A very friendly and capable young man replaced the faucet with a Woodford freezeless model including anti-siphon; neither feature belonged to the old faucet.

He also put in a shutoff valve and was able to turn our water back on, after struggling with the street valve, hidden underneath a bush. All this on a windy, rainy night. Thank you! No motel for us.

I will also give credit to the Reynoldsburg Water Department, which Kevin said came out quickly when he reported the leak. They were also willing to come out and turn on the water Sunday night if the plumber couldn't manage it, and that's service!

By Tuesday it was time for connecting the panels, on the south side of the house, to the electrical breaker, on the north side of the house. The attic only runs partway. In this photo, the flexible conduit containing the wires comes in from the roof.

It snakes along the attic to the furthest spot where it can take a straight shot down the outside wall.

Down through the soffit,

and the back of the house.

Then into the basement for a run along the joists to the north wall.

Once the wires had been run, the electrician spent half of Wednesday the 22nd installing the new electrical complex around the breaker box.

The solar power comes in from the upper left of the photo, into a manual disconnect box. From there it travels into a "revenue-grade" meter,

required for reporting production when applying for an SREC (Solar Renewable Energy Credit). Then it goes to the outside of the house, where there is another manual disconnect box, on the right with the red sticker,

and finally into the breaker box, from whence it can flow into the house or out onto the grid.

Note: if the utility power goes out, the microinverters automatically disconnect, for the safety of repair personnel.  Also, the microinverters require the external power to synchronize the phases of their AC current with the grid's phases. To have power during an outage, batteries or a generator are required.

At lower left in the earlier photo is an Enphase Envoy, which collects data from the twelve microinverters and sends it, via the Internet, to Enphase. At the Enphase website the installer and the owner can look at power production, per-panel statistics, and other geek-out goodies. The face of the Envoy also reports some statistics:

The WiFi signal at the Envoy location, I had discovered earlier, was poor. I created an Internet connection to the Envoy from the router at the far side of the house using MoCA, the Multimedia over Coaxial Alliance standard. MoCA adapters communicate over the house coaxial using frequencies that the cable company doesn't. Here, one adapter is plugged into the coaxial and into an Ethernet port on the router.

One floor below on the far side of the house, another adapter sits, connected to the coaxial and translating that back into Ethernet. The Ethernet cable sneaks down into the basement through a hole in the bottom of a cabinet!

On Thursday, April 23rd, the Reynoldsburg inspector came out to review the electrical and structural work. Vivian was thorough and went up on the roof with the EcoHouse guys, which she called her second worst (steepest) roof, the first-worst being a silo. There were two minor points to fix and have re-inspected, but it was OK to turn on the juice. The EcoHouse team put the final panel on the roof -- it had been left off to give Vivian the ability to see the mounting sytem -- and, it being a sunny day, by 12:30 our meter was spinning backwards.

On April 30th AEP-Ohio came out to replace the old analog meter with a new digital meter that records power-from-the-grid and power-to-the-grid separately.

On May 6th I received the signed AEP-Ohio "OAD - Net Energy Metering Service Application" in the mail, and Kevin came by to deliver more materials and information, and to settle up the remaining installation bill.

Now, it's time to see how power production works out through the various changes of the year: longer days vs. shorter days, leaves vs. no leaves, high sun vs. low sun, clear weather vs. cloudy. My hope is that at least 3600 kilowatt-hours will be generated each year. We'll see!

Adding Solar, Part 2: Proposals and Economics

In the previous post, I described various factors affecting the suitability of our house for a solar installation. Now I'll review the talks we had with solar installation companies, including configuration and economics.

We ended up with proposals from three firms: EcoHouse Solar, Paradise Energy Solutions, and Third Sun Solar. Joan and I had met Kevin Eigel of EcoHouse Solar eariler, during the 2014 Green Ohio Tour, and we also met with Sheldon Stutzman from Paradise and Kyle Kidd from Third Sun, companies that had advertised in the Green Ohio Tour brochure.

The first thing that all the companies made clear was that we had shade issues, particularly due to the locust east of the roof section. The economics were iffy. This picture shows the situation, albeit with the leaves off. Also note the woods to the west, which will truncate the late afternoon sun.

The companies were willing to discuss specifics once Joan and I made it clear that our primary motivation was not to make money, that is, to reach a payback on the installation as rapidly as possible, but rather we were intrigued by the technology, and wanted to do our part for clean electricity generation in Ohio. This led us to our catch phrase, You do what you can (ydwyc). Our house is not a fresh construction located in a corn field with a sapling or two in the front yard, nor are many other houses, but in such cases it is still appropriate to investigate what we can do, given where we are. You do what you can.

Costs
All of the final bids ended up being for twelve panels, two strings of six, with varying wattages per panel. At first Paradise Energy proposed sixteen panels, but when I commented on this, Sheldon admitted that twelve, leaving empty space next to the locust, would make more sense. Looking at raw costs, but including the 30% Federal tax credit, the proposal numbers looked like this:

There were many points to consider besides our shade issues and raw cost.

Other Incentives
There are minimal state-level incentives for renewable energy sources in Ohio. Indeed, the legislature has frozen the state's renewable-energy goals for the next two years, which has caused the value of SRECs, or State Renewable Energy Credits, to drop substantially. At this writing an SREC, representing 1 MW-hour or 1,000 kW-hours of production, fetches $30, so our system might produce $100/year through SRECs.

Currently the Ohio SREC market will cease in 2024. 

Roof replacement
If the roof is more than 10 or so years old, the homeowner must realize that taking the solar panels off to re-roof a few years down the road will be a significant additional expense. Our roof is twelve years old, in the gray zone, so none of the companies would offer an opinion whether to replace he surface on which the panels would be placed. I had two roofing companies come out to evaluate the roof, and their estimates for the remaining lifetime of the roof were between fifteen and twenty-five (!) years. (The roof is composed of "40-year" dimensional shingles.) Joan and I decided not to replace.

Optimization Tech
In the past, a drop in the power output of one panel, say due to shade, could affect output of an entire string of panels. There are now two technologies for reducing these losses, microinverters and DC optimizers. There's a lot of information on the web comparing the two technologies, much of it suspect due to conflict of interest. Even the proponents of power optimizers will admit that the microinverter technology is likely better for installations with large shade issues, and we went with that.

Monitoring
Modern systems include a monitoring package based on the solar installation feeding performance data back to the manufacturer, over the Internet. Both the owner and installer can log in and review the information, including per-panel output numbers. The per-panel information incurs an extra charge, but that is usually included in the quote from the installer.

Net Metering vs. Net Metering
The term "net metering" has been applied to several different arrangements, leading to much confusion. With the old spinning-dial electric meters, when a photovoltaic system put a surplus of energy back onto the grid, the dial would spin backwards, erasing the corresponding amount of electrical consumption from the meter. That doesn't apply any more; the first thing that our utility (AEP-Ohio) did was swap our old meter for a new, digital meter that records energy-from-the-grid and energy-to-the-grid separately. This provides flexibility in billing.

For a long time I, and the solar companies, had the impression that the only way to receive any credit for putting energy back into the grid was to use AEP-Ohio as the electricity provider, and that the alternative energy suppliers who have sprung up with deregulation in Ohio would not provide any credit at all. But Reynoldsburg, our municipality, has an aggregation agreement with AEP-Energy (now a separate entity from AEP-Ohio) that gives us a less expensive rate than AEP-Ohio. The question became, should we switch to AEP-Ohio, and receive net metering, or stick with AEP-Energy, without net metering but with a cheaper rate. It would depend on the charged rates and how much electricity our system would put back on the grid. That's a head scratcher, especially in a rapidly changing electricity market, and when AEP-Ohio and the Ohio Public Utilities Commission are frequently at loggerheads.

Then, after receiving my Interconnection Agreement from AEP-Ohio, I had a conversation with the Distributed Generation Coordinator. He explained that most if not all CRES (Competitive Retail Electric Service) companies would recognize the from-grid and to-grid information that AEP-Ohio (the distributor and billing agent) sent to them, and charge only for the net. What's the difference with AEP-Ohio, then? AEP-Ohio will carry a negative balance (so much electricity put back on the grid that your monthly bill was a credit) forward to the next month, but a CRES would not. In fact, his office maintains staff to work through billing issues with the CRES companies. He advised sticking with AEP-Energy CRES, because our system would rarely produce a negative monthly balance. And if my first bill isn't right, call him. He was a pleasure to work with.

Does your head hurt yet? Mine did! And we'll see how the distribution charges are affected when the first monthly bill arrives.

Payback Times
The payback time, how long it takes to recoup the installation cost, depends on too many factors to have much confidence in the final result. What will the real-life electricity production be, after shade and other "derating" influences? What is the rate of inflation of electricity cost? What will SRECs be worth? Will the utility regulations change? Will the state pass more anti-green legislation?

Also, in my experience, when a solar company gives you a payback estimate, that doesn't include making up deferred income. That is, you could have invested the money you paid for the solar installation. Or, if you took out a loan, you're paying interest.

The estimates for our system were for a payback period of 14 to 18 years. (In states with expensive electricity, such as New Jersey, it might be half that.) The net cost of our system will diminish over time, but there has to be some intrinsic satisfaction with the project, some non-financial value to justify the outlay.

You do what you can.

The companies all seemed competent; some had been in business longer than others; the bid costs were all different. Some had small extras (free annual inspection for the first two years). In the end, given the high expense, the low bidder won our job.

At one point after signing the contract, EcoHouse Solar proposed using larger (315 watt vs. 305 watt) panels with power optimizers at no change in cost. (The microinverters aren't appropriate over 305 watts.) This sounded good. However, after measuring our roof to double check, EcoHouse reported that the larger panels would completely cover the height of the roof, from gutter to ridge vent. We decided to stick with the original proposal.

Adding Solar, Part 1: Location, Orientation, Expectations

Joan and I have always had a general interest in alternative, or "green," energy. When a neighbor lost several immense ash trees to the emerald ash borer our view to the south improved, and I began to speculate about putting photovoltaic panels on our roof.

In October 2014 we had a chance to investigate further through the Green Ohio Tour. In earlier years we had been out of town or had other conflicts preventing checking out such events. One of the houses on this tour was only a few miles from our house, with a recently installed a photovoltaic system, and at 2pm the installer, Kevin Eigel of EcoHouse Solar, would be there to answer questions. This was a prime opportunity to find out more.

The homeowners were thrilled with their system, and said the installation process had been painless. Several other people were there to pepper Kevin with questions, and our interest was piqued. A few days later we invited EcoHouse Solar to come out and evaluate the suitability of our house for solar. That visit, and others, and follow-up research on the web, revealed a lot. This blog post will focus on the elements related our house's location, orientation, and shade, and thus our prospects for power generation.

Roof Pitch
The south-facing portion of our roof is at a 45º pitch, which is steeper than ideal.

This chart, originally from the tilt page of pveducation.org, shows the insolation (solar radiation) through the year with a 30º roof. Note that the green line, the module power, closely follows the ideal ("incident") blue line during the middle of the year, but doesn't capture all the possible energy during the winter, because the panel is aiming "above" the sun's lower track.

Here are the curves with our 45º roof. The output is "capped" at the middle of the year, due to the panel aiming "below" the sun's higher track, but there's almost no loss due to the angle during the winter months.

The two curves overlaid, with the 45º having an advantage shown in orange and the 30º advantage in yellow.

Azimuth
Azimuth refers to the compass orientation of the panel. The rooftop holding the solar panels faces 188º, or 8º west of due south. This is not a significant deviation from ideal. Using the Solmetric Insolation Tool, it would appear that the overall loss from both tilt and azimuth isn't so bad, with 97.8% of maximum insolation -- a figure almost too good to believe. 

This, however, doesn't take our location-specific shade into account. There is a large locust that's going to block a significant amount of morning sunlight, and trees to the west that will block the late afternoon sun. This picture was taken when there weren't any leaves on the trees. As you can see, the south-facing portion of our roof isn't huge, either.

The online PVWatts calcuator from the National Renewable Energy Laboratory helped me estimate the impact of shade. With this calculator, after specifying our location (I used the Port Columbus airport), I supplied several specifics about the proposed installation (all three companies that made proposals suggested a twelve-panel system).

The 'loss calculator' in the above image allows you to specify percentages of loss due to such factors as dust, soiling, snow cover, wiring mismatches, and so forth.  With the default value of 3% shade, the PVWatts estimate is 4,410 kilowatt-hours; back-calculating, that's 1,205 hours of full-sun equivalent per year hitting the 3.66 kwh array. For our case, I bumped the shade losses up to 20%. The results?

PVWatts predicts a typical year's output would be 3,694 kilowatt-hours, or only 1,009 hours of full-sun equivalent on the 3.66 kwh solar array. Interestingly, the proposals from three companies, including EcoHouse, all predicted an annual full-sun equivalent of only 900 hours; to match this the shade losses would need to be 29%. With judicious tree trimming I hope we can harvest 1,000 hours, but that's too complicated to predict. The shading effect will be worst when the leaves have popped out on the large, tall trees, and less when the leaves are gone -- five to sometimes six months of the year around here. The angle of the sun, and hence the spots where the shade falls, will vary throughout the year.

Clearly, comparing results to predictions will be a long-term effort.

In the next post I'll discuss the proposals and their cost and economics.