The Green Bank’s sunny report, part I: Solar power on the rise

Print More

The Connecticut Green Bank’s vaults are full of data, and the doors are open. Taking as an example our program that helps deploy solar residential power, we collect measurements on a host of useful characteristics for every installation that receives a state incentive. These include project cost, the size of the solar setup, the electricity output, the maker and model of hardware and customer information.

We use this to verify that systems receiving state incentives are producing their intended amount of power, achieving our mandated policy goals, and ensuring that public dollars are being invested wisely and productively.

Policymakers need data to gauge whether program investments are cost-effectively advancing Connecticut toward its goals, since state incentives for solar are investments drawn from electricity ratepayers. Having a robust data set on the full fleet of rooftop solar systems in Connecticut gives insights into things like hardware capability, weather pattern impact on fleet performance, and the amount of electricity generation no longer needed from the power grid.

We maintain a public spreadsheet online with all installations to date. The spreadsheet includes which company did the work, the system’s cost to install and the value of subsidy it received, and how much power it is designed to produce. One reason we make this data public is to provide transparency around solar pricing so consumers can benefit from a competitive market.

Residential solar deployment over time

Residential solar photovoltaic (PV) systems are growing in popularity in Connecticut — exponentially. More than 10,000 households have gone solar to date, with just under half of them doing so in 2014. Data from the Connecticut Green Bank’s Residential Solar Investment Program shows us how this plays out on geographically over time.

TrendCT readers probably know that every statistic has a story behind it. In this case, several measures appear to correlate — towns with lots of money spent to install systems also have the highest average annual electricity usage among solar adopters and also the highest expected output on their solar systems.

This suggests these towns not only had a large volume of installations but in many cases may also have large houses where a greater amount of electricity consumption needed to be offset. Also, towns may have varying proportions of suitable roofs for solar: tree canopies can cast shade, north-south streets can orient roof slopes away from southern exposures, and cities in particular have more non-owner occupied residential units where the benefits of solar don’t accrue cleanly to either the owner or renter. It is estimated only around 25 percent of U.S. residential roofs are suitable for solar, and 33.7% of total occupied housing units are occupied by renters.


Our data begins in 2004 when the Connecticut Clean Energy Fund was established and started providing incentives for Connecticut residents to install rooftop solar power. Money from ratepayer charges was redistributed to subsidize residential and commercial systems, as well as other forms of renewable power. However under this model, when the money ran out so did the mission.

From 2004 to 2008 the cost to install solar PV hovered at more than $8 per watt. From 2005 through 2008 there were 558 solar PV installations in Connecticut; with costs still being high, the customer base tended to consist of either environmentalists or residents seeking a degree of independence from the grid. Incentives from the Clean Energy Fund were also higher then. A system’s purchase price was split nearly evenly — consumers paid for half and were subsidized for the other half with electric ratepayer-funded state incentives through the Connecticut Clean Energy Fund. During this period, federal tax incentives were also lower. Now the federal incentives are 30 percent of whatever portion the system owner pays – until after 2016 when federal incentives are slated to step down to 10 percent for leased systems and zero for purchased systems.

Real-time statistics on solar generation across Connecticut

The Connecticut Green Bank recently created a widget that displayed real-time statistics on solar generation across the state.

After 2008 the cost per watt began creeping downward as demand rose modestly. During this period, international markets for solar PV were changing dynamically in Europe and Asia resulting in worldwide reductions in hardware costs (e.g., solar photovoltaic modules). In 2008 there were as many installations as in all previous years combined. However this growth leveled out during the next few years and showed no momentum.

The Green Bank’s creation in 2011 included a newly imagined, data-rich Residential Solar Investment Program. It includes both a solar purchase program and a solar lease program, in which an outside company owns the solar panels on a home and the homeowner pays to use the power. Both programs provide state incentives in declining amounts over time. Strategies at various levels attacked the overall cost drivers of solar to push down prices, while the Green Bank attracted private capital sources to create long-term, low-interest financing.

Meanwhile solar has crept toward grid-parity — a price where electricity from solar power is the same as the cost of grid power — and homeowners’ installation costs are getting financed over longer periods of time. The Connecticut Green Bank’s model also attracted the largest installer of leased residential solar power in the United States — Solar City — to plant its flag with a warehouse in Rocky Hill where we are also headquartered.

All of this together is spurring brisk consumer interest. Nowadays buyers and lessors install solar so they can save money — and that represents a larger consumer motivation than environmentalism or independent-mindedness. It’s also worth noting that households have lately been picking up much greater shares of unsubsidized costs in their pursuit of a future cash flow.

In 2012-2013 there were more solar installations than during the previous eight years combined. In 2014 alone, there were more installations than that. And with the exploding popularity of solar leasing, we will probably see installations more than double that amount in 2015.

From our first layers of data we can generate additional ones representing public benefits such as avoided emissions, trees planted, or jobs created.

Data help inform the level of public incentive that the Green Bank provides to Connecticut homeowners, and whether future incentive levels should be adjusted gently or aggressively. The Residential Solar Investment Program scales down incentives over time in order to spur demand for solar with a minimal sacrifice of public funds. Here we see data not just on already completed projects, but also those that have applied and been approved for incentives, and those whose installation is in progress.

The Green Bank regularly updates the growing community of solar installation companies to give advance notice on when public subsidy levels (cost per installed watt) are stepping down, so installers can better plan their business. The result is that installers race to acquire customers before the subsidy reduces again.

Through marketing and outreach efforts, the Green Bank began experimenting with targeted deployment strategies. Programs such as Solarize CT focus time-limited discounts on one town at a time, or a cluster of small towns. The short-lived discount opportunity helps consumers overcome inertia and close deals. Programs like this inject certain jurisdictions with bursts of deployment; even afterward these places experience a sustained level of improved customer interest over their pre-Solarize benchmarks.

Up next: Part II of our series will go deeper into our experiments with residential solar deployment, how different towns adapt to the growing market, and some of the barriers and costs solar installers encounter.

The Connecticut Green Bank (formerly the Clean Energy Finance and Investment Authority) was established by the Connecticut General Assembly on July 1, 2011, as a part of Public Act 11-80. As the nation’s first full-scale green bank, it is leading the clean energy finance movement by leveraging public and private funds to scale-up renewable energy deployment and energy efficiency projects across Connecticut. The Green Bank’s success in accelerating private investment in clean energy is helping Connecticut create jobs, increase economic prosperity, promote energy security and address climate change. For more information about the Connecticut Green Bank, please visit

Trend CT welcomes contributors to share data, analysis and perspectives that follow our guidelines.

What do you think?

  • Tom Weston

    I think that Mansfield’s Projects per 1000 Residents value below 5 is way off – probably skewed by counting UConn students that cannot possibly install solar on their dorm room.

  • Matt Macunas

    Great catch Tom. 2010 census data was the basis for population, and college students living either on-campus or off-campus are counted in that. Same for residents in correctional facilities and military bases. Unless it’s controlled for then that issue can skew per capita measures. We’ll work around that for our next data sets, which include towns where our Solarize campaigns were actively driving volume – such as Mansfield. Happy to keep an eye on the post and field other questions.

  • This article seems sunny but actually has a darker side: the vast majority of the solar PV equipment being installed is imported, so state and federal incentives end up subsidizing foreign industries even as American manufacturing is washed away by these imports. In the last 4 years I’ve seen at least 5 brands of formerly US-made solar PV equipment disappear either to bankruptcy or offshoring, and every few days I see offers for cheap imported solar panels at costs of 50-70% of US-made. The Free Trade policy coming out of Washington is insane, and solar is following the loss of US manufacturing in electronics, apparel, furniture, auto parts and a wide range of other goods.

  • Matt Macunas

    Will – thanks for the comment. This is an interesting issue that certainly deserves some time. We’ll provide data in our next article that goes deeper into panel manufacture and origin.