How not to insulate a cold floor
Does anyone know how to *not* insulate a semi-cold floor safely? Is this allowed by code?
here is my scenario: detached garage with half garage, half carport. Above those is a conditioned office that I plan on installing a diy minisplit. The office I will keep at 60-70F. The garage however I want to just keep above 50F and plan on installing a cheap resistance heater (likely radiant). It seems then that by insulating the floor above the garage I’m just losing a giant, free, and far more efficient, radiant ceiling heater. Is the 20F temperature difference enough to reach dew point in the winter?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
It would work but you would end up with bitterly cold floors. Air film above a floor is around R1, and uninsulated cavity is R3. The floor temperature ratio is the ratio of the air film R value to the whole assembly R value. So in case of 70F room 50F garage, that would be 65F floor temp. Definitely not bare foot friendly.
About the only way to do this and still have a comfortable office is with floor heat.
Your best bet is to insulate the garage well and get decent garage doors. This way you need much less heat to keep the garage warm.
Also insulate and air seal the floor of the office. You don't want any fumes from the garage making its way into the house.
Happy to wear slippers in the winter, my main motivator here is saving on heating costs. Though I don't know, perhaps it's only pennies to heat the 400sqft garage with resistance heat in the winter of zone 5b. Air sealing - was planning on double layer of air sealing - taped subflooring on office floor and taped drywall on garage ceiling. I could spray foam the edges of the joists from below for a triple layer.
Just insulate the ceiling with regular batts then just get yourself a small propane set up. Lots of cheap propane wall wall heaters with or without a built in thermostat that add heat and ambiance.
My understanding of code is that you have to seal and insulate the building envelope. The code doesn't say you can't also insulate between rooms that are within the envelope, or insulate outside of the envelope.
So I can see two strategies: put the garage outside of the envelope, and then insulate the walls of the garage anyway. OR: put the garage inside the envelope, and insulate between the garage and the office as well.
They're not that different, the difference is in the details of how the insulation is finished. In either case you have a stack: the ground; an unheated, insulated garage; and a heated, insulated office. I think that will get you what you want.
DC, the goal here is not to insulate between the garage and office. The reason is to be able to use heat loss from the higher temperature office to the lower temp garage and not require installing 2 heat pumps.
Two points: First, what you are proposing is not a code-compliant way of building.
More important, I get what you want, and I don't think you need to have no insulation to achieve that. In building science no insulation equals on the same side of the envelope and at the same temperature, which I don't think is what you want. The question is what the right amount of insulation is.
I'm assuming your garage is going to have an uninsulated concrete slab. The problem is that the heat loss models have a hard time dealing with earth contact. Earth can sink heat, and it can source it, depending on the conditions. My experience is that a well-insulated, unheated building that is connected to the ground below will stay above freezing even when it is very cold out. But there's no place in Manual J to plug in the heat contribution from the ground.
My gut feeling is that if there is good insulation, the garage is going to stay at an acceptable temperature even with insulation between floors, between the heat it gets from the ground and from above. But we don't like going on gut here. Instead, assume the ground is heat-neutral. Then you could model the heat loss of the walls using normal techniques. There would be a heat gain from above -- the room above is at a higher temperature, insulation isn't perfect and it's not absolute. From the model you can calculate the temperature in the garage where heat gain from above balances heat loss out the walls. Typically Manual J is based on keeping the house above 68 at the 99th percentile low temperature, but for this application I think it would be more reasonable to calculate keeping the garage at 50F on an average January/February day and above freezing on the coldest (99th percentile) days.
I suspect what you'll find is you want lots of insulation in the garage walls, and more than you'd think between the floor and ceiling. But don't go on what I think, run the numbers.
The detached garage will be insulated, so if it will stay above freezing without any heat at all just on its own this is indeed a pointless discussion as I don't need to run any resistance heaters. I'll try a manualj using coolcalc, I just don't know how accurate that will be.
You haven't explained why you want it kept above 50F. Is that a hard and fast requirement, or is there some flexibility in that number? DC's advice above could work well if you are willing to let the temperature drop to near freezing during cold snaps. If you set up the Manual J model, you would have the numbers to look at some "what-if" scenarios. By playing with the relative amount of insulation in the garage walls and ceiling, you will see the heat required to maintain 50F in the garage for various scenarios. Then you can balance the cost of that heat with the cost of additional insulation and/or adding a bit of resistance heat to the garage when necessary.
It’s not a hard number. Mostly to keep paints and finishing supplies, as well as plumbing, from freezing in the winter. 48F is also is the min temp on the thermostat of the resistance heater I was thinking of getting. I can run a manualj on coolcalc, just not sure how accurate it was. If DC is correct and the ground should keep the garage above freezing anyway then its a non-issue, I could only run the heaters when I’m there and the cost should be minimal.
Code and enforcement is very different from town to town. Sometimes one inspector in the same town with the same set of rules will see things very differently than the next.
Anything anyone says here is a meaningless opinion to your local inspector.
I say ask your inspector as only his opinion matters.
Most of the inspectors are big on fire safety and will require a egress size window, fire walls/ceiling and fire rated door between your office and the garage and do not care about insulation.
I say there is no “free” heat to send to the garage. If the office equipment is big enough to heat the garage it is oversized and is likely to do a poor job air condensing the office.
I think you could buy a lot of paint with the money you will spend to heat a garage, now if you are running plumbing in the garage someone did some poor planning.
I'm looking more for advise in terms of best practice than what is allowed by code, especially if anyone has done anything similar but perhaps my situation is too unique. As for equipment oversizing, yes the minisplit will be a bit oversized for the office. This is because the smallest one I can buy is too large for it so there's really no way around that. The heat from the office is not free, but it will be over 3x more efficient than the heat from a resistance heater. Regarding plumbing, the office above has a bathroom and it has to go down. The office and garage are detached from the main house.