Cost/benefit analysis for cool roofs

We're doing a remodeling project which includes installing a new roof. Here in California, we get a lot of sun, so the impact of solar irradiance on solar heat gain is a major concern — either for A/C costs (and thus peak summer energy loads) or on comfort (for those of us who don’t have A/C).

Thus, I’ve been looking into solar reflectivity and what has been called the “cool roofs“ movement. There is the Cool Roof Rating Council, “created in 1998 to develop accurate and credible methods for evaluating and labeling the solar reflectance and thermal emittance (radiative properties) of roofing products and to disseminate the information to all interested parties.”

There are also cool roof pages from the California Energy Commission (both consumer and business oriented pages) and at the EPA. (Unlike my smug friends praising German solar initiatives, cool roofs seems to be one place where the Europeans are copying the US).

A March 2009 presentation by Sheila Blake of the City of Houston summarizes the issues and their public policy implications — demonstrating how important a factor this is for energy efficiency, especially in the Southwest and Southeast. There’s also a March 2009 article at the McGraw-Hill continuing education site for the construction industry.

Normally one would assume that it‘s just a matter of lighter colors, but it’s not that simple.

Standard roofing materials have been developed to last a decade or more, and many of these materials (e.g. slate) are natural materials that come in specific colors. Lighter colored dyes, coatings and other treatments will fade or weather over time, reducing or eliminating the benefits of such treatment. (Apparently 19th century tin roofs provided durability and superior reflectivity.)

Another issue is that reflection isn’t enough. A roof also needs to emit heat (via infrared radiation) or it will raise the temperature of the roof and thus the house. The Metal Building Manufacturers Association has explains why this is important.

The solution to combined solar reflectance and emissivity is the SRI. Here is a succinct explanation from Astec Paints of Australia:

Total Solar Reflectance (T.S.R.) figures are expressed as a percentage falling between 0% and 100% dependant on a product’s Total Solar Reflectance as tested to ASTM C-1549 or ASTM E-903.

Emissivity or (Infrared emittance), is a measure of the ability of a surface to shed some of it’s heat in the form of infrared radiation away from the surface. The results from tests conducted to ASTM C-1371, express the emittance value as a percentage falling between 0% and 100% depending on the product’s performance.

Solar reflectance index (SRI), combines both the T.S.R.% reflectivity value and emittance value as a measure of a coating’s overall ability to reject solar heat.

The SRI is normalized to be in the range of [0,100] based on nominal “black” and “white” values. Lawrence Berkeley National Laboratory has an Excel spreadsheet that turns TSR and IE values into an SRI value, while Oakridge National Laboratory has a calculator to convert the ratings into energy savings. (These are among several calculators listed on the CCRC website).

Not everyone wants a white roof, particularly for a peaked roof (as is common here) seen by all your neighbors. The Cool Colors Project of LBNL and ORNL is trying to address this question. (It is also raising questions about the revising measurement standards for reflectivity, but let’s set that aside for now).

Starting in 2005, California’s Title 24 began requiring Cool Roofs for business buildings — perhaps because of their impact on daytime A/C and peak energy consumption — but I’d bet money that residential buildings will be covered in the next go-round. The California standards are being copied by other states.

What strikes me is how little of this is filtering down to the homeowner, perhaps because it’s not (yet) mandated. I’m painfully familiar with the state’s energy conservation regulations — Title 24 — that are either praised as the ultimate either in environmental leadership or bureaucratic micromanagement

However, I’ve heard nothing about roofing colors from my builder or roofing contractor. The only reason I investigated this is due to a stupid decision 20 years ago on re-roofing my home — replacing a light gray asphalt shingle with a medium gray — that raised the summer temperature 20°F until I added roof vents.

All the roofing bidders recommended a specific brand of 4.5mm thick APP modified bitumen membrane as the roofing material. After reading about it, it appears this material (guaranteed for 20 years) is now the preferred solution for low-slope roofs, replacing asphalt, hot mop and other solutions.

I couldn’t find the suggested roofing material in the CCRC database or the LBNL database. The manufacturer (Johns Manville) doesn’t provide data online, but a phone call provided numbers for the four lightest colors. In addition, a competitor (CertainTeed) has created a “CoolStar” variant of its competing APP product which is designed for Cool Roof initiatives. Clearly the latter has much better performance

Mfr.	Product	Solar Reflectivity	Thermal Emissivity	SRI
JM	White	.33	.84	35
JM	Tan	.26	.84	26
JM	Desert Blend	.21	.85	19
JM	Grey	.16	.84	13
CT	GTA	.74	.88	91†
CT	GTA-FR	.75	.92	93†

† SRI estimated using LBNL calculator

I ran the numbers using the ORNL calculator, with several California locations. Given some reasonable assumptions, the lightest color roof provided huge savings for hot inland locations (Bakersfield, Fresno, Sacramento), while coastal locations in Central and Northern California (Santa Maria, San Francisco, Arcata) with low cooling needs actually increased costs due to reduced winter sun. Southern California locations (San Diego, Long Beach, Los Angeles) showed more modest gains. (The results were magnified or attenuated based on the thermal resistance of the roofing assembly).

Even in the Southernmost clime, the difference between the white roof and the medium colored roof was about $40/year in air conditioning. Since we won’t have air conditioning, I need to figure out what this means for the interior temperature on hot days, but the data suggests that rooftop temperature will be at least 20°F cooler with the white roof.

Pickens Plan peters out

The New York Times, Wall Street Journal and others have reported that T. Boone Pickens has given up (at least for now) on his plans to to build a four-gigawatt wind farm in the Texas panhandle.

The stated (and undoubtedly important) reason was that his remote wind farm needed transmission lines. At one point he hoped to borrow $2 billion to build his own transmission line, but financing in today’s credit crisis made that impossible. A line is expected to be completed in 2013.

Of course, another reason is that wind power is less competitive due to declining fossil fuel prices — in this case natural gas, which produces about 21% of US electricity.

The 81-year-old oilman was so confident (or aggressive) of his plan that he ordered 687 (some say 667) wind turbines from GE for the first phase, and now has to find something to do with them — either place them in other wind farms or “put ’em in the garage.” With $2 billion tied up in these turbines, it’s quite possible that he’ll never see the turbines used.

Of course, this is a stark reminder of the dependence of wind and solar on transmission capacity, and also (as if we needed it) their vulnerability to shifts in the prices of substitute fuels. But more generally, this is an example of the Achilles heel of both technologies — their low operating costs come due to massive up front capital costs, magnifying the risk to private sector investors who might otherwise eagerly embrace these technologies.

California’s renewable power mandate is one way to reduce that risk, by providing a relatively predictable demand for those building plants and thus committing utility cash flow to keeping such plants open. However, as with all government interventions in the marketplace, there is the risk (in this case, with ratepayer dollars) that such mandates prove to be foolishly bone-headed distortions of the market (something we can’t know until we try).

California’s 33% mandate by 2020 seems particularly risky; a figure of 20-25% after the first decade would be more realistic, giving economists and policymakers a chance to access program success. Instead, the Sacramento politicians (including our governator) want to brag about their legacy to the voters, who will forget (a decade later) about who was responsible if it all turns out badly.

Still, this demonstrates the benefit of the US system of federalism and local policy initiatives: California can try its experiments while others watch. If it’s a great idea, California ratepayers benefit and all the other states will copy it. If it’s a terrible idea — or needs fine-tuning — other states can try something different and only Californians pay the price.