My company has been installing rooftop solar equipment since the 1970s, when the words “solar panel” meant solar hot water, and photovoltaic modules — now commonplace in both residential and commercial settings — were found mostly on satellites, being far too expensive for most earthly projects.
But PV costs have come down to earth. The grid-connected systems we install today typically cost around $4.30 a watt, or about what the PV modules alone cost six years ago. For a 9-kw system like the one shown in this article, that brings the gross cost to $39,000 or so, and municipal, state, and federal incentives may cut the out-of-pocket cost by half or even more (see “SRECS and Tax Credits,” sidebar).
Estimating system output. We use a free solar calculator called PVWatts to develop an accurate preconstruction estimate of a proposed system’s output. It’s designed for grid-connected systems and is available from the National Renewable Energy Laboratory at nrel.gov/rredc/pvwatts/grid.html. Using it is a matter of plugging in the system parameters and the location and letting the program run. But because this will give you the expected annual power production on a hypothetical site with full sun throughout the day — and many installations experience partial shading at some point during the solar day — that initial figure has to be de-rated to account for site-specific conditions.
To perform that fine-tuning, we use a solar site-assessment tool called the Solar Pathfinder (solarpathfinder.com). It’s a reliable and relatively simple-to-use analog device that lets us factor in local shading from trees and other obstructions. Although more sophisticated — and costly — electronic site-assessment calculators are now available, we’ve had such good results with our current approach we see no need to change. By the time we’ve crunched all the numbers, our estimate usually comes in within 5% of the actual production of the installed system — typically on the low side, since we want our customers to be pleased at getting a little more power than expected (rather than the other way around).
Cost and payback. With a kitchen or bath remodel, value is subjective. That’s not the case with a PV installation. Our clients tend to be environmentally aware and concerned about issues like global climate change, but at the end of the day their decisions are based on payback expectations measurable in dollars and cents.
To help with those decisions, we provide homeowners with a document that lays out the upfront cost of the system, factors in available incentives and the value of the power produced, and calculates its expected payback time. Most of the systems we install will pay for themselves in five years or less.
Roofing matters. It’s important to make sure existing roof shingles have lots of life remaining before covering them with a solar array. We might consider mounting panels on a 10-year-old roof, but if the material is older than that it’s a good idea to reroof first. We’ve removed a lot of solar-thermal and PV systems to permit reroofing before reinstalling them afterward, and this isn’t something you want to do more often than you have to.
Interestingly, the roofing beneath the solar array is almost always in vastly better shape than areas exposed to sun, wind, and weather. Unfortunately, unless a house has been designed with available module dimensions in mind, it’s seldom possible to extend the life of an entire roof plane by covering it with panels from eaves to ridge. More often than not, obstructions like dormers and partially shaded areas limit the available roof area to some fraction of its overall area.
The project pictured on these pages is typical in that respect. The following photos offer a step-by-step look at how the key details in this installation went together.
Conrad Geyser is the principal of Cotuit Solar in Cotuit, Mass.