Hi, I live in a 1960 split level (south-central PA, just north of the zone 4/5 boundary) and am getting ready to do some significant energy upgrading.  Total floor space is around 2200 ft2 on four half levels (not counting attached garage), plus another 500 ft2 in a 1970 addition over crawl space on the back.  Heat is hydronic baseboard with what I believe is the original oil boiler.  Addition has electric resistance heat and a woodstove.

I put together a spreadsheet to do some dirty calculations on where my heat is going and was a bit shocked at the results:  Half of my conductive heat loss is going out through surfaces backed by soil (the slabs under the lower half-level and the slab and walls of the basement half-level).  Maybe my assumptions are off-base?

Here's my basis:
0F outside, 63 inside, 30F in garage, attic, and crawlspace, 40F average soil temperature in contact with slabs and belowground walls
Main level and upstairs walls have R-11 batts for an assembly U factor of 0.095 - 0.097
Lower level (aboveground) walls are 8" combined CMU and brick veneer with cheap paneling on lath (no insulation), U = 0.43
Basement walls (belowground, unfinished) are 8" CMU, U = 0.9
Slabs are U = 1

So my biggest question right now is:  Are these assumptions, particularly 40F soil and U ~ 1 for slabs and buried walls, reasonable?

Thanks,
Will

Heat goes to cold, so if you heat your house to 70o and have uninsulated basement walls & floor, as well as an uninsulated stair down and a door that is neither insulated or weatherstripped, that 70o heat will head towards the coldest place which may very well be your basement walls and floors. And yes, an 8" basement wall has an R value similar to single pane glass, or R-1.

You don't mention air leaks, but that can be up to 50% of heat loss, so a blower test is a necessity before you start doing anything. Fixing air leaks is essential to upgrading a house and can often be done inexpensively.

Thanks Bob, yes I have done a blower door test and have been chasing leaks.  The 50+ year old doors are weathersealed as best I can, but my upgrade plans include replacing doors and windows, as well as adding insulation.  The funny thing is, the bottom two levels are actually the warmest during winter - no doubt a byproduct of keeping that oil guzzler full of 180F water.  It's hard to accept that the warmest rooms have the biggest problems.  Plan (so far) for HVAC is to replace the oil boiler with a gas mod/con (need to find out whether my existing baseboards can support low water temperatures) and to add  minisplits and/or ducted minisplit in the upper two levels to supplement the fossil heat and provide AC.  Once that's done the basement could end up being cold.

For heat loss calculations, does 40F average soil under the slabs sound right, or will I end up with warmer soil in the center?

we figure soil temps at about 50o in central NH. Yes the basement can be cold if you remove the fossil fuel heating systems as I have done this year (I've actually just turned them off to see the results which are workable) My immediate goal is to complete my basement insulation, then add a wood stove which I'll use when I'm working in the basement.

Sounds like you do not plan to insulate the basement floor - that's always an issue due to head room, but at least insulate the walls. We install 4" Thermal in new builds, but even 2" will help a lot, plus spray foam at the sills & rim joist.

The basement walls may very well be R-1 but the ground beyond is considerably better and the year round ground temperature presents a known load.

Manual 'J' ignores the basement load if it is unconditioned and insulation below the frost line makes ROI very questionable below this level unless you are heating the slab.

If you install a Modcon you will miss the heat previously wasted to the basement, most especially directly above the old boiler.

Unless your fin-tube is marginal to begin with, a properly sized Modcon boiler with indirect water heater will be more efficient than the old boiler, but the cost of fuel and efficiency of the equipment proposed has to be considered first.

Thanks Bob I and BadgerBoiler.

I've revised my calcs with 55F ground temperature and the results are a lot more reasonable. They still show a lot of heat loss into the ground but much of that is attributable to the fact that waste heat is keeping the basement a lot warmer than I would chose to heat it. The floor and walls feel cool to touch, not cold. The biggest conductive heat loss - both from the calcs and from feeling surface temperatures - is through the uninsulated structural masonry walls of the exposed lower level. I'm thinking of following the external insulation process outlined here: http://www.buildingscience.com/documents/insights/bsi079-deep-dish-retrofits/ using polyiso against the masonry and EPS on top (trying to avoid XPS for GHG reasons). Also would extend the external insulation up over the upper level (we are planning to replace the aluminum siding and improve the insulation on the upper level anyway). Is this a good approach for my circumstances?

For the existing stick-built walls I assumed R-11 batts with 25% framing, plus another 10% where the batts are ineffective or missing. Appropriate?

My existing baseboards appear to be way oversized. With existing insulation on a design day (0F) I calculate around 220 Btu/hr per foot of fin-tube in the upper and main levels, 535 Btu/hr-ft in the uninsulated lower level (and the space stays warm). With the upgrades I'm planning demand drops to around 100 and 200 Btu/hr-ft (up/down). This takes into account conductive heat loss only; no adjustment for infiltration or internal heat sources.

A 25% framing fraction is typical for 16" o.c. framing. R13s existed in 1960, as did R7 "econobatts" that only filled half of the cavity. Paper facers came in both asphalted & plain kraft flavors, as did aluminum-clad.

Find out if the batts have aluminum facers, which could limit what you put on the exterior. Foil facers (on batts or foam) are true vapor barriers, and you don't want to build a moisture trap.

If it turns out you have half-depth econobatts or sections where the mice have taken away half the fiberglass it's worth squirting some cellulose in there. If you have full depth batts and can take a square foot sample, compare the weight to new R13 or R11s. R11s are a bit more than half the density of n R13's. If the old batts aren't at LEAST as dense as new R13, R11 is the right guesstimate.

Thanks Dana. When I sealed the knee wall between the upper level and the attic over the main level the batts had kraft paper facings and they filled the 2x4 bays, so I'm pretty confident they are vapor-permeable and > R7. If they are R13, the error in overall wall U value due to assuming R11 is the same as changing my WAG of missing/ineffective batts from 10% to 8.3%, so it's probably not worth obsessing about that (and as an Engineer that doesn't come easily to me).