Further Wrestling with Recessed-Can Lights
How do you air seal IC-rated light fixtures that aren't accessible from above? Do the air sealing from below.
Since my first article on recessed-can lighting (“A Recessed Can of Worms,” HE Jan/Feb ’01, p. 42), there have been great strides in the development of new and improved insulated ceiling airtight (ICAT) recessed-can light fixtures for new and remodel construction. And new compact fluorescent models have become more available for reducing electricity use. The driving force for these developments may be new building codes mandating the use of the safer and more efficient lighting. What concerns me as a private consultant is how to deal with standard IC-rated recessedcan lights that were installed in cathedral ceilings or attic spaces that are not accessible and therefore cannot be air sealed from above. Standard IC-rated fixtures are still being installed in new construction, as well as in remodel projects, by some contractors and by homeowners themselves. (Non IC-rated lighting fixtures are another story all together.)
Yes, I still get calls from homeowners every winter when they notice holes in the snow on their roofs above the locations of all their recessed-can light fixtures. Even better yet, the calls I truly love: “Why is my house the only house in the neighborhood that never has any snow on the roof? Do you think it has anything to do with the thirty-six recessed-can lights I recently installed?” Well, it just may have something to do with it, or maybe the aliens have grown tired of crop circles! During the winter of 2004/2005, that is exactly what happened with me and my business (not the aliens-on-theroof thing).To try to address the problem of inaccessible can lights, I’ve gone back to the lab and into the field to test some ideas and different commercially available retrofit kits (see “First a Little History”).
In the Lab
How do you make a standard ICrated recessed-can light fixture more airtight when you can’t build an air-sealed box from above? What air sealing material can you use to withstand the temperatures expected inside the fixture? What materials will withstand summer temperatures of 140°F? What lamp is the best choice to keep the thermal overload from shutting the fixture off? How much air reduction can we expect after air sealing these fixtures? How costeffective are manufactured retrofit kits?
To answer these questions I started back in the lab with my original test mockup (see photo above). I removed the 13 inch x 16 inch x 10 1/4 inch airsealed drywall box from over the top of the fixture, attached a thermal sensor to the side of the fixture, and placed R-19 fiberglass batt insulation over and around the fixture. I placed a second sensor above the insulation to record the ambient temperature and then placed the original 28 inch x 53 inch x 24 1/2 inch box made with insulation board over the entire mockup.The inside of the insulated box was heated with a 300 watt lamp controlled by a Chromolox thermostat, in order to simulate a hot ambient attic air temperature of 140°F.All the tests were performed at this summer ambient air temperature. I disabled the thermal cutout switch on the fixtures before testing them.
The lamps I chose for the test are a 100W A-base lamp, a 90W Capsylite flood, a 20W compact fluorescent flood, and a 20W triple biax compact fluorescent. I chose the 100W A-base lamp to start the testing, since I expected this to be the worst lamp of the group. The temperature of the fixture, after three hours of operation, went to 310°F— well over the 194° limit of the thermal cutout switch that I had disengaged. The wire nut that I used to jumper out the thermal cutout switch was on the floor; it had melted off the connection (see photo on p. 33).The good news was that the metal foil tape held in place.
After returning from the store with a new fire extinguisher, I started to test the next lamp, the 90W Capsylite flood. After three hours of operation the fixture reached a temperature of 217°F, also above the thermal cutout switch rating. The good news is that the new wire nut didn’t burn off this time and the metal foil tape was still intact. I did think of calling my insurance agent, though, to check my coverage.The 20W compact fluorescent flood lamp raised the fixture temperature to 144°F after five hours of operation. This is a good thing. The 20W triple biax compact fluorescent produced a fixture temperature of 162°F after six hours of operation (see Table 1).
Duct Tape and Air Leakage At Home
Using my home as the next test site, and after trying to convince my wife that I was only cleaning the insides of our light fixtures, I conducted a blower door/pressure pan test on the four IC-rated recessed can lights above the kitchen countertop. These fixtures are in an inaccessible attic space that is insulated with cellulose insulation. The trims are the open black baffle type. The fifth recessed can I tested at home is an uninsulated eyeball that is located
in the center of the kitchen, between the first and second floors. I tested three additional recessed-can fixtures in the first-floor master bathroom. The gasketed glass-trim shower light and two open trims in the bathroom are insulated with cellulose insulation.
I depressurized the house and did pressure pan tests over each fixture.After documenting all the pressure pan readings, I removed all the open black baffle trims from the recessed-can fixtures in the kitchen and sealed all of the holes with the same aluminum duct tape that I had used in the lab. In all but one of the fixtures, I installed the tape on the inside of the fixture and on the drywall just under the reveal of the trim so that it would not be seen. I used tape to seal the holes inside one of the fixtures above the kitchen countertop, but I didn’t seal it to the drywall, as I did with all the other fixtures. All the fixtures were equipped with 75W reflector flood lamps. (See Table 2 for the air flow around the fixtures before and after air sealing.)
The results of the pressure pan tests indicate that a substantial reduction of leakage is possible by simply taping all the holes inside the fixture as well as taping the fixture to the drywall. Also, there was no reduction of leakage through the fixture above the kitchen countertop that I sealed inside the fixture but not to the drywall.
This air-sealing method was very time consuming. The biggest problem with this method is trying to reattach the small springs that hold the trim to the ceiling. I found that I had to stretch the spring to attach it to the lamp bracket that holds the lamp socket. But the work paid off. The tested kitchen lights, which are equipped with thermal overloads, have been in normal operation for more than five months and have not tripped off since I sealed them with the tape.
Since taping the fixtures was so time consuming—it took roughly 30 to 45 minutes per fixture—not to mention the frustration of trying to reattach the trim with the small springs, I decided to try manufactured retrofit kits.To analyze the air sealing effectiveness of the manufactured trims, I conducted a blower door/pressure pan test on a recessed-can frame that was insulated with fiberglass batt insulation.The area where the fixture was installed was depressurized to -50 Pa. I did a smoke test at each fixture to identify air leakage locations. A standard, open black baffle trim indicated a pressure pan reading of -48 Pa, which I used as a baseline.The pressure pan test was done first with the standard trim and then again after the retrofit trim was installed (see Table 3).The installation time for all the retrofit kits was approximately ten minutes each. I liked the large springs on the retrofit trims that mounted the trim to the ceiling. Less frustration is also a good thing.
Can Man’s Latest Wisdom
Air leakage through recessed-can lights is still a problem and I expect this problem to continue in the future. Although the manufactured trims or retrofit kits I tested indicated some reduction in air flow during the blower door and pressure pan test, they shouldn’t be relied on solely to stop this flow of air. Most of the kits were successful in reducing the amount of air leakage inside the fixture. However, most did not address the leakage associated with the trim and the drywall connection, where 50% or more of the leakage occurs.The cost of the retrofit kits made no difference in how well they reduced air flow. The Lithonia Air-Tite foam tape kit, at a cost of $2.35, did as well as the Builders Best retrofit kit, at a cost of $34.All the manufactured kits tested were easy to install.
The duct tape manufactured by Nashua Tyco Adhesives was best for both air sealing recessed-can lights inside the fixture and sealing the trim to the drywall. Nashua manufactures a butyl-backed tape with a 5-mil adhesive that may be another solution.However, I did not test this product. Caulking may be another way to air seal the fixtures, although that may get a little messy and caulking material may not bridge larger gaps between the drywall and the fixture.The drawback with taping was the time needed to complete the air-sealing and to reinstall the trim. The problem that added the most time to the air sealing was the small springs that attach the trim to the ceiling—did I mention how annoying they were? Using the large wing type of springs that are on the trims in the manufactured retrofit kits could reduce the installation time. The cost of a roll of tape, which could be used for dozens of fixtures,was less than $20.The Halo Air Tite Super trim and the duct tape may be another option.
Although the duct tape method of air sealing IC-rated recessed-can lights is not in the International Energy Conservation Code or the National Electrical Code, it appears to be the most successful method of reducing air flow through ICrated recessed-can lighting located in inaccessible attic spaces—not to mention keeping the aliens off your roof. If you decide to use this method of air sealing, don’t ever use it on non IC-rated lights, make sure to use lower-wattage reflector spotlights or fluorescent spotlights (see “New Reflector CFLs That Can Take the Heat,”HE March/April ’05, p. 9), and make sure that the fixtures have thermal overload protection (IC rated).
Enter your comments in the box below:
(Please note that all comments are subject to review prior to posting.)