What are some good techniques for dealing with recessed light cans that protrude into an unfinished attic?
I would like to add to the blown insulation in my Florida attic and tighten up (reduce) air intrusion.  I know
that light cans get extremely hot and you are not supposed to cover them.  If you are blowing in more
cellulose insulation, what is a good technique for keeping the required clearance around light cans?  How
can I close up the gap between the cans and the ceiling drywall?

Replacing the cans with air-tight insulation-contact versions with gaskted to meet WA state standards, then using only high efficiency CFL or LED lamps in the suckers to keep the heat down would be best.

Short of that, airtight steel enclosures that fit over the fixture with at least 3" of clearance to the fixture, caulked with silicone for air-tightness to the top side of the gypsum works. The tough part is getting it air-tight around the electrical penetrations, etc. There are commercially available retrofit cardboard versions treated with fire retardents for this purpose, but I wouldn't trust them. These solutions still don't meet spec for fixures not rated for insulation contact, and you may get blow-back from inspectors etc. And if you are using 75W+ in 'em you likely WILL see an early failure of the fixture (but hopefully not a fire.) If lamped with sub-20W high efficiency goods it'll last pretty much forever though.

It's a big enough PITA that for my money, if there aren't 1001 of 'em re-fixturing with air-tight insulation contact versions is more reliable. If there's a gap between the gypsum & fixture that the gasket won't handle, silicone caulk should do it. It has to be something that can handle moderately high temps without deterioration- spray foam sealants may not be up to the task.

If they are insulation contact rated already, then you could insulate right up to the cans. The question or challenge then becomes how to "air seal" them. I suppose you could foam around the base of the can light on the attic side to eliminate air penetration from the can-drywall junction. The question I have is if the can is not "air seal/tight" rated, does that mean the housing will also have penetrations for air to escape?

The construction is from 1980, so the cans are not contact rated.  I believe that there are adjustment slots, and other holes and slots in the housing
that allow air movement between attic and conditioned space.  So sealing the gap between housing and dry wall will obviously not change much.
Replacing the cans seems like the best solution, but is somewhat expensive though I can do the wiring myself.

I am not familiar with the metal covers mentioned earlier.   I might assume though that using them could still result in heat build up in the
fixture.  I would use CFLs but who knows about the next owner.   I have experienced from an accidental covering of one can (which had
an incandescent bulb installed) that the wiring to the socket will suffer damage to its insulation.

Does anyone have a technique for computing savings by eliminating heat gain from a recessed light exposed to an unconditioned attic (in Florida)?

In my experience the standard cans listed as airtight are far from it. I build boxes out of foamboard and use expanding foam to attach and fill in around the elec. wire and do so with standard ic rated cans.

Foamboard seems like a easy product to use for the enclosures.  How much larger than the cans do you make your boxes? 
Since you can bury your IC rated cans in insulation have you found that the insulation doesn't stop all air intrusion in these
cans?  With my non IC rated cans I think they call for 6 in of clearance to any insulation, I can build boxes like that, but I
don't know if its legit to close the top and bury the box, with the non IC fixture inside, in cellulose insulation.

Short answer, no, but it's big.

It's a function of several factors including the wattage of the bulb and how many hours it stays on.  A 75W bulb while on increases the flow through the fixture by more than 10x what it is when off.

http://www.pct.edu/wtc/docs/articles/BB0502-Air-Leakage-in-Recessed-Lights.pdf

http://arkansasenergy.org/business_development/energy/files/Clearinghouse/Air%20&%20Moisture%20thru%20recessed%20light%20fixtures.pdf

In FL recessed lights also increase cooling loads as well as heating loads, since it provides a large infiltration path that the air-handlers WILL drive.  All forced air systems (heating or cooling) create pressure differentials between rooms, and if a return path exists through the attic, it'll use that path as easily as any other.  The difference during the heating system is that it allows a "stack effect" even when the air handlers aren't running, putting the house under a slight negative pressure, sucking air into conditioned space from every other air leak in the place.

Non-IC fixtures are intentionally designed for enhanced air flow while on as a means of cooling the fixture.  Blocking the flow with a box-over approach defeats that design goal, and not a good idea if there's a possibility if it'll be lamped with incandescents.   If you have enough fixtures that a complete swap-out to WA state compliant air-tight gasketed versions is prohibitively expensive, you may want to look into using spray foam on the underside of the roof deck and an unventilated attic configuration instead.  It's a trade off between cheap insulation + fixture cost vs. expensive insulation & no fixture cost.  Either solution can work.  Boxing over un-rated fixtures (even when done well) is at best a half-measure.

Thanks for the explanation and references.  Up to a 30 cfm loss of conditioned air is certainly an incentive to swap out my old
cans and install air-tight units.   Pay back should be less than 5 years if I do the work myself.   I have discovered that a
kitchen remodel resulted in 9 IC rated cans, but they aren't air-tight.  Still have 10 old non-IC rated cans in other areas.
None (even the IC rated ones in the kitchen) are the "remodel" type that I might be able to remove from below.  I better
get to this attic work before the weather heats up.  Thanks to all for the information to help make this decision.

Foam box ic cans instead of replacing them. Replace your non ic cans with ic cans and foam box them. That is the most cost effective method and delivers the best seal of not only the can, but the penetration of the ceiling. It will be way cheaper and give you better results.

A new construction 6" ic can is $5, how much is an airtight WA state compliant can cost?

At my nearest big box store an IC fixture is about $6 and an air-tight IC is about $9.
I don't know if the later meets WA state requirements.  But if I'm in the attic anyway
seems reasonable to make a foam box and adhear it with a little spray foam.  Should
save a few bucks as well.  Thanks.

Foam applied directly to an IC-rated can isn't a good idea. Polystyrene & polyurethane break down at temps below that of an IC can with a 75W bulb in it & a bunch of cellulose heaped on top.  Polyisocyanurate gives you another ~100F of margin- if you're going to go that route use foil-faced iso and seal the edge/corners with FSK tape & silicone caulk, not 1-part polyurethane foam.

By the time you've bought the foam board & additional sealants, I'm not sure you're saving anything relative to the $3 delta in air-tight vs. econo-grade IC fixtures.

Dana, I dont think your on the same page here, this is done all the time with ic cans. I've been in attics after the fact and the boxes are fine. Unless the WA cans are not much more and do an excellent job at sealing around the drywall, this works much better. The airtight cans that are sold in my area are the same standard can with a flimsy gasket, maybe a good idea on paper but in field conditions it's junk. See pic.

I s'pose as long as the foam isn't in direct contact and has sufficient clearance you'll be good to go then.

In the configuration depicted the heat is conducted out via the gypsum, so surface area of gypsum per watt will determine the peak temp inside the foam box.

The photo helped.   Saves me time figuring out the best configuration.  I probably would have used 3/4 in poly-iso foil faced on both sides
rather than the 2 in shown in the photo.   Most likely would have taped the joints as well.  So the photos saved me those errors.  
Thanks.

Posted By 1FloridaNative on 09 Feb 2010 02:03 PM
The photo helped.   Saves me time figuring out the best configuration.  I probably would have used 3/4 in poly-iso foil faced on both sides
rather than the 2 in shown in the photo.   Most likely would have taped the joints as well.  So the photos saved me those errors.  
Thanks.

 

3/4 foil faced iso would be fine,FSK- taped joints also fine.  My issue was keeping the foam sealants under their operating temp, which is a clearance-to-the-fixture at max wattage issue.  Commercial retrofit boxes are made out of fire-retardent cardboard.  It's only functions are to be air-tight, and provide sufficient clearance to the insulation.  The insulative value of the box itself is irrelevant, but it's flammability and operating temp is.  Any air-tight box is good, but low flammability is better.