Boiler question

[Mothra]

Hi. I'm looking to replace my home's current heating system and have a few (many) questions about what boiler is right for me. I purchased this home three years ago, however, due to work movement, I haven't lived in it and am currently in a massive renovation before I do. To get down to it, I'm looking to have a propane boiler heat my house through buderus radiant panels. I'm also going to have pex installed in my basement slab, which will be roughly 2400sqft. The home itself will be roughly 3500 sqft. I've included sketches of the layout of the house, however, this is was done prior to adding a ~1000 sq ft addition. I live in the 44802 area code, where the average lowest temperature is 18 degrees F, however, my heat loss calculations are from delta 60-100. The house will be mostly spray foam insulated and fairly tight. I guess i'm looking on suggestions on sizing and/or any suggestions on brands that would be reliable and capable of running this system. Thank you very much!

 

[Dana]

What method was used for heat load calculations? And what outside design temperature was used? (Your 99% outside design temp in ZIP 44802 is probably more like +2F. Fremont's is +1F, Lima's is +4F.) A heat load of 86,683 BTU/hr @ 70F indoors, 0F outdoors (a 70F delta) is a ratio of 25 BTU/hr per square foot which is pretty high for a tight 3500' 2x4 framed house, and insanely high for a 2400' house. The house doesn't have an unusually large number of corners or a gazillion square feet of window. Most pretty-tight 2x4 framed houses will come in ~15 BTU/ft, maybe 20 if the foundation isn't insulated. A new tight IRC 2015 code-minimum house (you're in zone 5A) would usually come in around 12 BTU/ft, sometimes less.

The delta-T that should be used for establishing the 99% heat load in your neighborhood is ~65-67. Yes, it gets colder than that, but not long enough to matter, and there will always be some oversizing factor, but you don't want the oversizing factor to get out of hand. A 1.4x sizing factor on the 99% design heat load is almost always enough to cover Polar Vortex lows.

If you have access to full winter's heating fuel use history on the place, first run a fuel-use based load calculation on it, which will establish a fairly firm upper bound on the boiler size needed. Your 99% outside design temp in ZIP 44802 is probably more like +2F, (Fremont's is +1F, Lima's is +4F.) Do the calculations with an assumed balance point of 60F and another for 65F.

For modulating condensing boilers figure out how much radiation their is per zone, and find a boiler with a minimum modulated output low enough that it won't short-cycle on zone calls when operated at condensing temperatures.

With the spray foam insulation, be aware that the thermal bridging of the framing cuts the legs off any potential advantage from closed cell foam. In a 2x6 assembly 5.25" of closed cell foam adds only about R1.7 to the "whole wall-R" compared to 5.5" of open cell foam, making it about the most expensive R1.7 one can buy. There is much cheaper/better ways of buying better performance.

Also, spray foam only air-seals the wall cavities, not the seams between doubled up top plates, or the seam between the bottom plate & subfloor, etc. Any framing seams outside the cavity itself needs a bead of polyurethane caulk.

 

[Sponsor]

 

[Mothra]

I apologize about the late reply, i just now saw your response. I had used http://www.usboiler.net/heat-loss-calculator.html for my calculations. I'm attaching my up to date calculations based upon that website that include the addition. It does not take into account the basement area, roughly 2400-2500 that will have pex ran in the slab.

A little more information on the house, it was originally built in 1895. However, most of the first floor was reinsulated prior to my purchase and i'm having every outside wall of the second floor as well as the first floor addition sprayed with an inch or so of closed cell insulation then backed with R-19 insulation. Unfortunately, i don't have access to any prior heating bills.

I recently had a contractor come out to give me a quote on a boiler installation, as well as a sizing estimate. He came back with my heating need as being 158 btu/hr. i was skeptical of this number from my research and own calculations, as i have read not to terribly oversize my potential boiler. I really appreciate the help!

 

[Dana]

The heat losses of basements are pretty low per square foot. If you put at least R10 under the radiant slab the losses of the slab itself will be tiny.

The crud heat load calculator is going to overshoot reality by at least 25% even with conventional framing and new construction but would fine as long as the boiler selected has output no larger than those numbers. But your house is a bit different, and you're adding R6 to the exterior, which is going to lower the wall loss numbers by quite a bit. I'll be more accurate if you run your own room-by-room I=B=R type heat load calculation based on U-factors calculated on the actual construction (including the inch of closed cell foam.) The only surfaces/walls that matter are what's on the exterior (roof, exterior walls & windows, foundation, etc.) Doing this with a standard spreadsheet tool is easier than doing it with pencil & paper, and changes/updates can be more easily made.

To figure out U-factors for the older construction, are the old studs are full-dimension 2x4 (not 1.5" x 3.5" milled lumber), and you're saying they compressed some R19s into the 4" space? Is there any plank sheathing & siding that will be foamed-over (if yes, please describe type and approximate thicknesses.) Is the interior finish wall now half-inch wallboard, or is it something else?

The type & age of the windows also matter, including any storm windows.

A 158K heat load for a house that size would only be true with some windows wide open. HVAC contractors tend to use lousy rules of thumb rather than I=B=R or ACCA Manual-J type methods. Sadly, the rules of thumb used typically oversizes by 2-3x (as if you really need to cover the heat load at -150F outdoors!)

BTW: In US climate zone 5A (that's you) an inch of closed cell foam on the exterior is not sufficient for dew point control at the sheathing on 2x6 framing, per the IRC, but it'll be fine for 2x4 (even full dimension 2x4s.)

 

[Mothra]

Thank you for your continued replies and awesome information!

I used the slant-fin app and it came back at a design temp of 70 degrees to need roughly 93kBTU/hr, which i had read is a more accurate portrayal of actual heat loss... which is way under what my contractor said. However, I didn't take into account the hallway and stairwell space, as well as the basement. Yes, i am putting R-10 under the slab. I will look into the link you have provided.

The old construction is full dimension 2x4's. It is plank sheathing behind wood siding covered in aluminum siding. The planks are approximately 3/4" thick. Yes, the interior finish will be 1/2 wallboard. The windows have been mostly replaced with double paned. The ones that haven't been replaced will be within the next couple of years.

Thank you for the info on the closed cell. I will look into that, also!

 

[Dana]

The Slantfin tool is another I=B=R tool, definitely better than the other one, but it typically overshoots reality by 15-25%, sometimes more, depending on the underlying air infiltration rate assumptions compared to the actual air leakage.

But it also probably doesn't have your wall stackup in their standard wall assemblies, and that too makes a difference.

Is the older building ballo0n-framed, with studs that extend from the foundation to the attic, or is it more conventional, with band joists between floor? Is it it full-dimension 2x4s 16" on center?

A balloon framed wall 16" o.c. with full-dimension 2x4s will have a framing fraction of about 20-22%, whereas conventionally framed with band joists the framing fraction will be about 27%, all of which affects the U-factor of the wall. If you're leaving the older wood siding on and spray-foaming over it, the type and thickness of the old siding also makes a difference.

Also, is it compressed R19s in 4" deep cavities, or is it something else? An R19 compressed to 5.5" runs about R18, but in a 4.0" cavity it'll be more like R14, and in a 3.5" cavity, R13, so it's important to know the cavity depth to calculate it correctly.

A typical 2x4 framed house with a 25% framing fraction, half-inch wallboard on the interior, R13s in the cavities, half-inch plywood and 5" clapboards comes in at about R10-R11 "whole wall=-R" or U0.09-U0.10. So at a temperature difference of 65F (68F indoors, +3F outdoors) it'll be losing (0.09 x 65F=) 5.9 BTU/hr per square foot, to (0.10 x 65F=) 6.5 BTU/hr per square foot. If you strip the clapboards down to the sheathing it'll lose about R0.5 of "whole wall-R" down to say R9.5, but then adding an inch of polyiso or closed cell spray foam brings it up to R15.5, or U0.065, which reduces the wall losses to (U0.065 x 65F=) 4.2 BTU/hr per square foot of wall.

That's quite a bit less loss than a standard 2x4 wall, it's about 30% less, an important factor when calculating the load.

If you're stripping the older wood siding down to the plank sheathing it may be cheaper and better to put 1.5" or 2" rigid polyisocyanurate foam board sheathing over it rather than 1" of spray polyurethane. It can be pretty cheap if using recliaimed roofing polyiso from commercial building re-roofing or demolition. There are many building materials salvage yards or dedicated foam reclaimers out there sell it at 25-35% of the price of virgin-stock foam (eg: https://toledo.craigslist.org/search/sss?query=rigid+insulation ). Roofing polyiso has a labeled R-value of about R5.5/inch, but derate it to R5/inch for this application (so, 1.5" foam board would be good for at least the code-prescribed R7.5, and 2" would be good for at least R10, even old stuff that spent the last 25 years on the roof of a commercial building.) Some new polyiso is labeled R9-R10 at 1.5", but I'd still derate that to R5.5/inch for dew point control in this stackup.

Are the newer double-panes low-E, or clear glass? This too matters. A clear glass double-pane window runs about U0.5-U0.6, but typical 10-20 year old low-E double pane runs U0.33-U0.37, and a 2010 or newer version is more likely to come in around U0.30-U0.32. A U0.6 double pane loses fully TWICE the amount of heat of a U0.30 double-pane, so getting that at least sorta-right matters too. Also, run the heat load only on the "after upgrades" picture of any window changes, and assume any new replacement windows will be U0.32 or lower (U0.32 is current code-minimum). In new construction it's usually cost effective to specify a low-E window with two low-E coatings, on surfaces #2 & #4, which will usually come in around U0.24-U0.26 for the whole window assembly. It's an upcharge from a code-minimum U0.32 window but not a huge upcharge (the way a triple-pane would be), but it's a difference you'll be able to feel when it's 0F outside. In much colder climates (such as the upper peninsula of Michigan) the coating on surface #4 increases the amount of window condensation in winter by quite a bit, making the upcharge to triple panes worth considering, but it's not a big deal in your location. Something like the insulated glass units from Cardinal using LoĒ-180 + LoĒ-i89 combination with fairly high solar gain would be a good choice for your location, but just about any U25-ish window with a solar heat gain coefficient (SGHC) of 45 or higher would be pretty good. Most of the big midwestern window manufacturers have dual low-E glass as an option.