Spray foam and HVAC size: a heating-load calculation for an Erie house

The question

A furnace or heat pump is sized from the heating load: how many Btu per hour the house loses on a design-cold night. Air sealing and insulation change that number. For an ordinary older house in Erie, how much does a closed-cell spray foam retrofit change it, and what does that mean for the equipment?

The house and the inputs

A two-story colonial, 25 × 40 feet, 2,000 square feet above a full basement. The retrofit is the work we do most in older homes: seal the attic floor and top it up, and spray closed-cell foam on the rim joists. The walls and windows stay as they are.

Input	Before	After	Where it comes from
Design temperatures	70°F in, 11°F out	same	Erie's 99% heating design temperature (ASHRAE 2025); Manual J uses the 99% value
Walls, 1,870 sq ft net	2×4, R-11 fiberglass, U-0.093	unchanged	Cavity and stud paths, 25% framing
Windows and doors	300 sq ft at U-0.50, 40 sq ft at U-0.40	unchanged	Older double-pane windows
Attic floor, 1,000 sq ft	About R-19 left, U-0.048	Bypasses sealed with closed-cell, topped up to R-49, U-0.020
Rim joists, about 100 sq ft	Bare 2×10 and sheathing, U-0.26	3 in. closed-cell, U-0.040	About R-21 for the 3 in. of foam (one HFO product lists R-14 at 2 in. and R-26 at 3.5 in.)
Air leakage	10 ACH50	6 ACH50 (40% less)	LBNL's ResDB equates 10 ACH50 with a normalized leakage of 0.55; detached U.S. houses average about 0.61, and older ones leak more. See the range below

Method

Each surface loses heat at its U-factor times its area times the temperature difference (59°F). Air leakage is converted from the blower-door number to a design-condition airflow (ACH50 × 24,000 cubic feet, the basement included, ÷ 60 ÷ 15) and costs 1.08 × cfm × 59°F. That is the core of an ACCA Manual J heating calculation. Manual J software adds the basement, duct losses and its own leakage method; those add to both columns, so they are left out to keep the comparison clean.

A technician in a safety vest crouches beside test gauges and cables inside a framed house bathed in red light — A blower door test at NIST's Net-Zero Energy Residential Test Facility in 2011. The test measures how much air a fan has to move to hold the house at a 50-pascal pressure difference with the outdoors; that airflow per hour, divided by the house's volume, is the ACH50 number used in this study.Photo: NIST, public domain

Results

Part of the heating load	Before	After
Walls (unchanged)	10,250 Btu/h	10,250 Btu/h
Windows and doors (unchanged)	9,790 Btu/h	9,790 Btu/h
Attic floor	2,850 Btu/h	1,160 Btu/h
Rim joists	1,550 Btu/h	240 Btu/h
Air leakage	16,990 Btu/h	10,200 Btu/h
Total	41,430 Btu/h	31,640 Btu/h

Design heating load of the example Erie house, before and after the retrofit: 41,430 Btu/h falls to 31,640 Btu/h (24% less). Air leakage falls from 16,990 to 10,200 Btu/h; the attic and rim joists fall from 4,400 to 1,400 Btu/h; walls and windows are unchanged.

Air leakage was the biggest single part of the load, about 41% of it, and it is where closed-cell foam does most of its work: at the rim joists and the attic floor, two of the places older houses leak.

How much the leakage drops decides the result:

Air leakage after	Cut	Heating load after	Load reduction
7.5 ACH50	25% (ENERGY STAR's modeling assumption)	34,190 Btu/h	17%
6 ACH50	40%	31,640 Btu/h	24%
5 ACH50	50%	29,940 Btu/h	28%

A blower-door test before and after settles which row a real house lands in.

What it means for the equipment

Heat pump. NEEP's cold-climate guide says to match the heat pump's capacity at the design temperature to 100 to 115% of the heating load (the DOE's guidance: size to the design heating load, with slight oversizing acceptable). Before the retrofit that is about 41,000 to 48,000 Btu/h at 11°F; after, about 32,000 to 36,000. The difference is roughly 10,000 Btu/h, close to one nominal ton, usually one equipment size.
Furnace. ENERGY STAR's design rules allow a heating-only furnace of 100 to 140% of the load. Before, that tops out near 58,000 Btu/h of output; after, near 44,000.
Cooling. The same attic and air-leakage changes cut the cooling load too, but this estimate doesn't calculate it. Erie County is heating-dominated (RESNET lists a heating-to-cooling degree-day ratio of 2.9), so the heating load sets a heat pump's size.
Money. The smaller unit costs less up front, and a unit matched to its load runs longer, steadier cycles. In This Old House's 2026 national averages for central air, a 4-ton system averages $13,136 and a 3-ton system $8,654.

What the numbers do and don't say

It is a worked estimate, not a Manual J report. A real calculation adds the basement, ducts, the house's measured leakage and room-by-room loads. The difference between the two columns is the useful part.
The walls set a floor. With the walls and windows unchanged, they are 63% of the load after the retrofit. When walls are open, in a gut renovation or new construction, closed-cell in the walls takes the load down further.
Do the envelope first. NEEP's heat pump guide says it directly: "Planned weatherization work should be done first if possible, to reduce HVAC size and cost." Then size the new equipment to the house as it is.

Savings vary. Find out why in the seller's fact sheet on R-values. Higher R-values mean greater insulating power.

Assumptions: steady state at the 99% design temperature; series-parallel wall U-factor with interior and exterior air films, siding R-0.6, sheathing R-0.5, drywall R-0.45, studs at R-1.25 per inch; attic floor with air films and drywall; design infiltration from ACH50 ÷ 15; no internal or solar gains in the heating load, as Manual J does.