Want to be sure I am doing this right: For the load, Basement = 1700 SF with 3" EPS under it at R12.6 (walls R20) and design temp of 68deg (using -4F normal temp) Garage = 700 SF with 3" EPS under it, same design temp 1700 x 0.08 x (68-(-4)) = 9800 load 700 x 0.08 x 72 = 4000 - figure garage doors will be less R than the walls we'll make this 5500 Total load = 15.5k with a 10% fudgefactor = 17k -4 might be too low a temp though. most of the time the average outside temp will be way higher. But the load of 15-17k is this right?

Without knowing how much of the wall are there is above grade, or whether R20 is continuous insulation vs center-cavity on some studwall or studwall+ foam stackup there's no telling what the heat load is.

If you're planning on heating the place with a slab, with a heat load much north of 7 BTU/foot R12.5 EPS would be a bit on the skimpy side for sub-slab R. If the design heat load really comes out more like 5 BTU/ft of radiant slab or less it's fine.

Still, nothing about your calculation makes any sense. Yes, the U-factor of R12.6 slab is about 0.08, but the ground underneath the slab will never be -4F (it'll probably be 50F or better averaged over the whole slab) and the temperature of the bottom of slab where it's losing heat through the EPS will be well above 68F if it's heating a 68F room.

The heat loss of the basement will be primarily be from the above-grade portion of the walls, windows & doors, but since the bottom of the slab may be 85F or better the heat loss through the slab-EPS is still a real fraction, even with 55F dirt. Air leakage through the garage door is a big error factor that's hard to have any accuracy with.

Once you have an inch of "real" (EPS, XPS) insulation under the typical basement slab diminishing returns are my main focus. Once the ground is "charged" in the fall the ground reaches thermal equilibrium and voila. Much depends on the qualities of the substrate, which is why we insulate them all and why the snake-oil people promoting "bubble-foil" and "blanket" ..."insulation" get away with outright fraud.

In renovation I would rather have foam the rim joist than put one sheet of EPS on a basement wall. If you do insulate basement walls (I used Thermax in my own renovation), you could stop at the frost line and have a hard time seeing the fuel savings as you go down the wall and under the floor.

In a recent renovation on Cedar Lake in Minneapolis a 1955 brick ranch with tuck-under garage ended up with heat loads in the single digits for the basement and garage, with foamed rim joist and 2" on the wall plus, R11 bats to the interior. The design water temperature for the new slab, over 2" of XPS was 75°F. Yes I know it will never come on Dana...

No Surf, not the right approach at all. The delta T you are using would only apply to above ground wall area, but would NOT apply to a concrete slab floor or to below ground basement wall areas. In addition to knowing the heat loss of the walls, you also need to have a separate breakout for the concrete floor heat loss in order to properly design a hydronic radiant floor heating system. Furthermore, the concrete floor heat loss is NOT normally based on the floor area, but is normally based on the floor perimeter. So you might want to check out our heat loss analysis and hydronic radiant floor heating design software/instructions to get a better handle on how this works. The hydronic radiant floor heating design software will also provide the recommended minimum floor insulation R-value based on your design input parameters.

In Surfsup's Chicagoland zone 5 location code-min IRC 2012 style is R15 slab-edge for heated slabs, and R15 c.i. for basement walls too.

http://publicecodes.cyberregs.com/icod/irc/2012/icod_irc_2012_11_sec002.htm

BSC's recommendation for sub-slab across the full slab in zone 5 is R10, even in unheated slabs. The rationale isn't entirely heating energy use, but summertime moisture as well. See Table 2, p10:

http://www.buildingscience.com/documents/reports/rr-1005-building-america-high-r-value-high-performance-residential-buildings-all-climate-zones

I agree he's probably going to be OK with R12 as long as he keeps the garage door closed and doesn't have a bunch of window area raising the heat load, and the slab temp with it. But in other threads of his he has talked up the "modern design" with lots of glazed area (and correspondingly high heat loads), but he really hasn't provided even a fraction of the necessary info to estimate the heat load here.

I'm not at all surprised that with 2" of foam + R11s on the studwall an R10 basement slab in MN would be in single digits, maybe even low single-digits, with very low water temp requirements. But IIRC from other threads his is a walkout basement, and he has been talking going code min (or sub-min) R for his above grade wall assemblies (R19s in 2x6, when IL state code spells out R20 cavity fill min for 2x6, R13 + 5 for 2x4 construction in his area.) Even with R20 basement walls he could have a real heat load for the basement.

Surf also likes to do ROI analyses to ascertain the financial merits of the various construction design options. This is indeed the right approach.  A ROI analysis would be appropriate to sort out the appropriate level of insulation too. IRC insulation code-min requirements for heated slabs typically result in 20-30% floor downward heat loss of the total heat gain provided by a hydronic radiant floor heating system. While many contractors might be OK with this, a homeowner who is building and planning to stay in their home for more than a couple years might want to do this ROI analysis and target less than 10% floor downward heat loss. Surf clearly needs to sharpen his pencil and do a proper heat load loss analysis BEFORE designing a hydronic radiant floor heating system.

Posted By sailawayrb on 19 Sep 2013 10:48 AM
Surf also likes to do ROI analyses to ascertain the financial merits of the various construction design options. A ROI analysis would be appropriate to sort out the appropriate level of insulation too. IRC insulation code-min requirements for heated slabs typically result in 20-30% floor downward heat loss of the total heat gain provided by a hydronic radiant floor heating system. While many contractors might be OK with this, a homeowner who is building and planning to stay in their home for more than a couple years might want to do this ROI analysis and target less than 10% floor downward heat loss. Surf clearly needs to sharpen his pencil and do a proper heat load loss analysis BEFORE designing a hydronic radiant floor heating system.

TESTIFY, brother sailaway!

I've encouraged him to do some serious upgrading to the building envelope before investing the solar thermal he has already bid out too. Between the glass sliders and all the windows he was looking at something on the order of 25,000 BTU/hr of heat loss on the glass alone.  You can still have pretty modern architecture without gia-normous glazing fractions, and every square foot of code-max glass that gets replaced by even code-min wall is a big step in the right direction, costing a lot less than the glass it replaces.

Agreed and I am sure he appreciates your expertise and guidance. No one wants to live in a cave (well…maybe some folks do), but it is always good to keep in mind that even the best windows are an order of magnitude worse than the worst ceiling/wall assemblies relative to heat loss performance. Likely more BTUs are lost via windows and oversized buildings (i.e., increased area exposed to the delta T) than anywhere else. One should always try to take advantage of passive solar heating to some extent by placing the majority of windows on the equator facing wall and having minimal or even zero windows on the other walls. Obviously this approach has to be blended/traded with the building architectural style, one’s life style, and one’s long-term housing operational budget, but this should always be in the back of one’s mind…and one should feel some sense of pain/shame every time one adds a window to a design that doesn’t generate any heat gain in the Winter and is NOT fully shaded in the Summer…

IIRC his shading factors give him almost no direct passive solar gains in winter due to buildings on the abutting property on the southern side, with some direct gains in summer.

Once you're at ~U0.35 or less almost all windows, even on the north side will gain more total BTUs than they lose in the winter, but the high overnight and cloudy day losses of overglazed house raise the peak heating loads sky high, and the sunny day gains raise the peak cooling loads too. The key is finding the right balance for all site factors. You don't have to live in a cave to have both low peak loads and low overall energy use. Most PassiveHouses have high glazing fractions, but carefully selected and VERY high performance windows, and mostly south facing, with optimal shading factors that won't be found on his site.

The heat loss for windows and sliders was indeed ~24k. I am going to probably drop 3 sliders and put in some windows instead. This will reduce the loads but marginally compared to the whole home load. I get passive solar in the upstairs bedrooms...most E and W facing windows/doors will have tree shade in summer and leafless tree sun in the winter. W side will get a bit more summer sun. I have an internal shading system for the large window specified.

External wall will be R21 HP batt 2x6 framed with 1.5" polyiso on the exterior. This is roughly another R10 (6.5*1.5=9.75) This puts my exterior wall at approx R25 continuous. (R21*.75+10)

With 3" EPS I have re-calculated the slab/wall/door heat loads for the garage and basement to total 20k using 3" EPS foam under the slab, for Basement and Garage, respectively:

Slab: 4860, 2140
Walls: 6250, 5620
Doors: 0, 1530 (no doors in basement, garage doors are entered as R10)

Adding them all I get approx 20k

I do appreciate GREATLY the advice here. I am limited by the lot orientation, cramped Chicago housing, curb appeal, budget, etc etc etc...Yes, I'm not going to build what I would consider an ugly, dark, dungeon of a home but I am trying to maximize what I can do with teh dollars I have. I know it is never "enough" but ultimately the budget reigns king. I am putting in a metal roof, great windows and doors, doing exterior polyiso, and fighting my architect, contractors, inspectors, the village, etc along the way by proposing MORE energy efficiency and doing my best not being in the trade and researching it at night after a 10-11 hour day of work and travel. What was I thinking building a home? LOL

With R1 allowance for gypsum + sheathing + siding, the whole-wall R is about right, ~R25, which is a U-factor (1/R) of 0.04 BTU/hr per square foot per degree-F.

If you're using cheap heat load tool, be sure that it's not mis-interpreting R25 as a center-cavity number, which would yield a whole-wall R of only ~R16, a U-factor of 0.063, instead of 0.40, which would overestimate the wall losses by about 35%.

I still don't get how the slab has such a high loss. The slab loss will be pretty constant, drifting only very slowly with average outdoor temp, largely unaffected by the 99% outside design temp, rising a bit with increased slab-temp requirements. It's not a simple equation, but there is no way that the slab losses are going to be even half the wall losses. Even an uninsulated (except for slab-edge) probably wouldn't be losing more than 6 BTU/ft, and with R12 under it radiant slab it would be about an order of magnitude less. A measurable fraction of the total basement loss, but nowhere near even 15% of the total.

Looking over some of my heat load notes from my own ~1500' basement, with R18 on the walls, wall losses are about 1600 BTU/hr @ 0F outdoor temps, window & door losses are ~1400 BTU/hr @ 0F, and I did a wag of 5BTU/ft for 7500 BTU/hr the uninsulated slab. (Losses vary in different parts of the basement, based on IR thermometer spot-checks, but it's about right.) Even if you have twice the amount above grade wall that I do your wall numbers are still on the high side.

BTW: No way are the garage doors truly R10. Many (most?) garage door manufacturers inflate performance numbers, but are usually (not always) careful to not step over the FTC fraud line in how they word it. Even those listing funny high R values don't always spell out that it's a center-panel number, but usually have an accompanying U-factor. See page 4 of this document:

http://www.overheaddoor.com/garage-doors/Documents/brochures/thermacore-collection-brochure.pdf

Note they spec an R value of 12.71, but a U-factor of 0.24 (=R4.17) for one product, R17.5 with a U-factor of 0.16 (=R6.25) for another. Their published U-factors are what you should be using in the heat load calc, which is indicates where the "whole-assembly" R-value really is.

I was keeping it simple with just load = area x u-factor x dT

I plan to enter allthis into hvac calc this weekend for better number. I am hoping to get r18 doors, so i halved it. I'll post it when i have it.

Right, load = area x u-factor (which is inverse of R-value) x delta T. Just make sure you use the appropriate u-factor and delta T as Dana recommended.

Dana, I am not sure I understand your comment about cheap heat load analysis tools miss interpreting the R-value? Seems like the free heat load analysis tools (like on our website) and the cheaper tools just use the exact R-value that is entered "as is" WITHOUT adjusting it at all. Seems like the more expensive heat load analysis tools tend to use multiple drop down menu selections and are also more likely to adjust the entered R-value within the software...perhaps in manner to make the analysis inaccurate or completely invalid if the user does not fully understand what is occurring within the software. While the free and cheap heat load analysis tools do require the user to be able to accomplish fifth grade math to first calculate the whole-wall R-value of the given assembly BEFORE entering it into the tool in order to obtain the more accurate whole-wall R-value analysis, there is zero ambiguity/uncertainty as to what R-value was used by the software and the analysis. So I believe the risk of having an inaccurate or completely invalid analysis is likely much higher when the user is challenged by fifth grade math and also uses the more expensive heat load analysis tools...which is likely common place because the more expensive tools create better looking outputs that customers perceive as being a better analysis without considering the competency of the user.