What size electric boiler to put in a 2950sf ICF attached garage? It is in Minneapolis MN area The PEX is in a 5" slab on 2" of XPS. There is a 12' ceiling with R-65 in a vented attic. There are three glass OH doors facing South, although they are insulated glass there is no insulation in the aluminum frames. One insulated steel service door and 53sf of double thick glass block windows.

It will make a big difference if the heat is intermittent and you want it to heat up in a reasonable time frame.

The space will be maintained in the 40s and it is not expected to quickly bring it up to the 60s. A secondary heat source will be added if a quick recovery is desired, but will be expected to maintain the 60s.

You are probably going to need a second source for more immediate comfort as that slab will probably take quite a while to warm up to make that much space comfortable. Since you didn't give the dimensions, a similar space of 35' X 80' = 2800 SQ/FT. 230' of linear wall at 12' = 2760 SQ/FT of wall area. lets take out the 3 overhead doors at say 144SF EA, the 21SF for the walkin and the 53SF you mentioned for the glass block leaves 2254SF of ICF at say R25/.04U. 2800 SF @ R10/.5U floor and 2800 SF @ R65/U.015 ceiling. The 99% design temp in Minneapolis is -26F
For a building the size described above@ 40F inside and -26F outside:
The ICF wall will loose 5950 BTU/HR(the concrete wall mass will have an effect on this overall performance how thick is the core?).
Don't have an estimate on your earth temp. If the floor/footing perimeter is insulated, the floor shouldn't loose all that much untill you start to heat for that 60F internal temp.
The ceiling will shed 2800 BTU/HR
@ an R2 estimate for the overhead doors(probably being very kind here) they will loose 14,256 BTU/HR
the other openings say 1000+ BTU/HR
With all the WAG's in here That is 24000 BTU/HR but makes no account for all the other variables(air infiltration, solar heat gain ect) not to mention the actual numbers for the overhead doors. To hit 60F inside under these conditions you would need to add 25%+ as the floor will loose more to the earth as the floor temp rises...

Your Mileage will really vary

Ronmar provided a good example of what needs to be accomplished, namely a good heat loss analysis for this space. You could use our heat loss analysis software to accomplish this:

Borst Heat Loss Analysis Software

If you want an even more refined answer, you could then use our hydronic radiant floor heating design software for your exact hydronic radiant floor heating design and which will also account for the additional heat loss resulting from the slab actually being at a higher temp than what was used for the initial heat loss analysis (i.e., the indoor space design temp):

Borst Hydronic Radiant Floor Heating Design Software

The space is 46.5'x63.5' with one of the 6" BuildBlock (2.5" eps each side of 6" concrete) ICF walls against the attached house. The three 10'x10' glass overhead doors have no R or U rating, neither do the 53sf of glass block windows. The glass blocks are doubled, flushed with the inside and more flush on the outside of the ICF wall. We finished blowing in Cellulose in the attic Friday and had temp heat. Early this morning the temp in the garage was 39 while it was -7 outside. With the sun it reached 42 as of noon. The glass block window surface temps were 34 at the same time.

OK so 220LF of wall. X 12' +2640SF - 374SF for openings = 2266SF of R25/U0.04 ICF wall Multiply the U factor X SF X temp difference to get BTU per hour 0.04 X 2266 X 66F +5982 BTU/HR at -26 outside
2950 SF floor @ R10/U0.5 again not a lot untill you start to push the internal temp above that 40F as the slab temp to maintain 40F in the room is probably not much above your deep earth temp. You didn't mention if you used perimeter insulation around your footings to help isolate the earth under the slab from the earth around the build that is exposed to that -26F air in the winter.
2950 SF @R65/U0.015 (2950X.015X66) for 2920 BTU/HR.
A single 3" glass block may be in the area of R1.75/U.57 so 53 X .57 X 66 = 1994 BTU If they are doubled(6") it could be half that or 997 BTU/HR
The rollup door manufacturer should have something for R value just for doing this sort of calculation. 300SF @ R2/U.5 9900 BTU/HR Again severe wild a$$ guess here it could be better it could be worse.
The 21SF walkout could be as high as R10/U0.1 or 140 BTU/HR when it is -26 outside.

At your dimensions, that is around 20,935 BTU/HR The biggest difference between this and my earlier example being mainly for the rollup door size being smaller than in the example. Makes a big difference, as will the R value of the door if it is significantly higher.

Again no accounting in these numbers for air leakage/infiltration. With the nature of rollup doors, you could have significant air exchange which could easilly double your heat loss.

+1 for using Sailawayrb's calculators. They simplify this somewhat and account for things in better detail(plus explaining what the info is for/how it is used in the instructions). But like any calculator, garbage in = garbage out, so you need good R values for the components.

The fact you can put heat into the space now is fantastic, as you can monitor the energy you are inputting to maintain a temp vis the outside temp and confirm the calcs.

The garage has a 4' deep frost wall of the same ICF. I think we have a deep earth temp of about 58 or 59 here. The garage door did not, would not provide a U rating, I asked Haas. http://www.haasdoor.com/commercial/ca/index.html

Deep subsoil temps in Minneapolis are nowhere near 58-59F. Try 45-ish.

http://www.builditsolar.com/Projects/Cooling/US-ground-temps.gif

Deep subsoil temps tend to track within a few degrees of the 100 year average outdoor temp in an area, unless there are geological anomlies such as volcanic calderas, etc. Deep subsoil temps of 58-59F are found in the warm edge of US climate zone 4, or the cold edge of zone 3. Minneapolis is in the cool half of zone 6.

Right, sorry, I meant 48-49, per the water from well for my geo.

So what size boiler? Lets say maintain 65 degrees, but can take a long time to get there.

Describe the windows used in the doors? How thick, 1 3/4"? 2 panes? Glass or poly? how large are the rails and stiles(2"X3"?) Are there any markings on the window assembly? Not uncommon for a door or window manufacturer to use a glazing package from another manufacturer. That might get you to a little closer loss number for the doors. OR at the very least a little more info might allow us to help you with a little more accurate SWAG. Since these doors are potentially such a large percentage of the heat loss, you really need to nail their loss numbers down to make an educated choice.

You said you used temp heat to get it to 39F with -7 outside? Can you describe that temp heater? Did it cycle on-off or did it run full time to get you to 39F. If we know how much heat you put in, and since we can take a good guess at what most of the structure will shed at a 46F temp difference, the rest you could attribute to the overhead doors and infiltration...

There is a link in my post to the Haas door but it is all 7/16" dual pane clear glass with 2-4" wide aluminum. Except the bottom section is 7/16" insulated panels set in the same non-insulated aluminum rails. No more info on doors

Friday it was warm here 35 and we, I am guessing had it to 50 without attic insulation, with one of my guys homemade propane heater of unknown BTU, finished insulation late Friday. Saturday cold, Sunday colder 0 for a high about. Monday morning -7 outside, garage 39 at sunrise, with the morning sun it got 42 without any heat other than passive solar through the OH doors. Afternoon was cloudy and end of the day temp fell to 39 inside and 6 out.

As Ronmar indicated, you can validate and improve the accuracy of your heat loss analysis:

Borst Existing Building Energy Usage Analysis Software

Thanks for the use of the software. I will try again when I have some time.

I punched your numbers into Sailawarb's heat loss calc and came up with 45,097 BTU/HR @ -26F outside and 65F inside. It only takes a minute to do. The only remotely tricky part is the floor type. When using type 3(slab on grade) you only need to put in the length of floor perimeter exposed to the outside temp. I am guessing the slab is inside the ICF, so I only put in 33' for where the slab meets the doors.

Of course the unknowns here are the real drivers of the numbers.

With the uninsulated metal frames on those OH doors and 7/16 pane spacing, I don't think they will be much better than R2(if they even reach that). That is the number I used and that accounted for 13KBTU/HR(for 353SF of window area). The double stacked glass block is probably a little better but it is also a pretty small percentage of the total envelope. I used R8 for the walkout door, R10 for the floor, R25 for the walls(@2266SF) and R65 for the ceiling @ 2950SF

As is probably pretty common, the biggest variable is air infiltration. Did you do a blower door test? In the calc, I used .33/one third air change per hour(probably pretty tight with rollup doors in the mix) and that accounted for nearly 20,000 BTU/HR of that 45KBTU/HR. So you will have some decisions to make or perform some tests to define the actual numbers. The cool thing with the calc is you can make changes in the calc and see the results easy enough.

One way you could confirm these numbers is to put a known quantity of heat into the space. That was why I asked about the heater you were using, to see if it was something you could get an estimate on how much heat it was putting out... Homemade heater unfortunately dosn't help, but the ammount of fuel used over the time heated might help a little, but the unknown of combustion efficiency then becomes a factor.

If you could round up a known quantity of electric heaters you could test that way, as heat output per KW is pretty well defined(~3400 BTU/KW). If the numbers I used in the calc are accurate, then for -26 outside you should need 13KW of electric heat to maintain 65F inside.

Good luck, sounds like a nice garage...

Can the loss in temp be used to calculate this? Overnight, 16 hours, the space lost 11 degrees from 56 to 45 degrees. While the temp outside was relatively constant at 20 degrees.

Posted By ICFconstruction on 14 Jan 2016 08:41 AM
Can the loss in temp be used to calculate this? Overnight, 16 hours, the space lost 11 degrees from 56 to 45 degrees. While the temp outside was relatively constant at 20 degrees.

The unknown there is how much energy was stored in the structure to be released over those 16 hours?  IE: You have a bucket with a hole in it.  It took it 16 hours to empty, but it is hard to say how much water was in the bucket to begin with...    It is way easier if you can add a known quantity of energy and maintain a relatively stable internal temp.

If it is 20F outside, and the estimates used in the calc are accurate, it should only take you around 4500W to maintain 50F.  Three 1500W space heaters spread around the space? If you can get the temperature to stabelize at some number with a steady energy input, it is easier to gauge how large the hole in the bucket really is.  If you did this and record what the internal temp stabelized at, and what the exterior temp was over that time period, then you could more confidently confirm or refute the calculator results...