What a window U-factor actually costs you
A window's U-factor is the rate at which heat passes through one square foot of it for each degree Fahrenheit of temperature difference, measured in Btu per hour. A U-factor of 0.85 means that on a day 30°F colder outside than in, each square foot leaks 0.85 × 30 = 25.5 Btu every hour. Multiply by area and by the number of degree-hours in your heating season and you have the annual heat loss through the glass. Divide by your heating system's efficiency and by the heat content of a unit of fuel, and you have the fuel it takes to replace that heat.
That chain is the whole calculation. What it tells you, uncomfortably often, is that replacement windows do not pay for themselves through fuel savings. A window is one of the most expensive square feet of building envelope you can buy, and its U-factor — even after replacement — is still four to eight times worse than the wall it sits in. Cutting a bad number to a good number still leaves a bad number, and you paid a wall's worth of money per window to do it.
That does not make replacement wrong. It means the fuel saving should not be the argument. Rot, failed seals, sashes that will not open, single glazing that runs with condensation every winter morning, draughts you can feel from the sofa, and the resale value of a house that no longer needs its windows done — those are the reasons people replace windows, and none of them appear in this calculator. What this page does is stop a salesperson from putting a fuel saving in front of you that the arithmetic does not support.
The degree-day formula, term by term
The formula multiplies five things and divides by two. Take them in order.
Area × ΔU gives the reduction in heat loss per degree-hour. Only the difference in U-factor matters, not either value on its own — replacing 1.10 with 0.55 saves exactly as much fuel as replacing 0.75 with 0.20, all else equal. This is why triple glazing so often disappoints on payback: going from 0.30 to 0.18 saves less than half of what going from 0.85 to 0.30 saved, for more money.
HDD × 24 converts the season into degree-hours. A heating degree day is one degree of temperature deficit sustained for a day, so 5,000 degree-days is 120,000 degree-hours. Degree days are published to a base temperature — 65°F in the United States — which stands in for the outdoor temperature below which a building needs heat. Use the base-65 figure for your own station; a mismatched base is the single easiest way to be wrong by 30%.
÷ η accounts for the fact that you buy fuel, not heat. An 80% AFUE furnace burns 1.25 Btu of gas for every Btu that reaches the room, so a low-efficiency system makes each unit of saved heat worth more money. A heat pump with a seasonal COP of 2.5 does the reverse: it delivers 2.5 Btu of heat per Btu of electricity bought, so the same heat saving buys back far less fuel. Enter a COP as a percentage — 2.5 becomes 250 — and the arithmetic handles it.
÷ E × P converts fuel energy into money. A therm of natural gas is 100,000 Btu by definition. A kilowatt-hour is 3,412 Btu. A gallon of propane holds about 91,500 Btu and a gallon of no. 2 heating oil about 138,500 Btu; those are the conversion figures the U.S. Energy Information Administration uses. Divide your last annual bill by the units consumed to get the price, so that fixed charges are spread across the units rather than ignored.
Two things the formula deliberately leaves out: air leakage and solar gain. A new window normally leaks less air than the 40-year-old unit it replaces, which is a real saving the U-factor does not capture; and glazing that blocks solar heat gain reduces free winter heat, which is a real loss it does not capture either. In a heating-dominated climate these two roughly offset. In a cooling-dominated one, look at the solar heat gain coefficient instead — see the general energy upgrade payback calculator for a framework that handles cooling.
Worked example: 180 ft² of single glazing in a 5,000 HDD climate
Twelve original windows, 180 ft² in total, currently rating about U-0.85. The quote is for low-E double glazing at U-0.28, installed, at $900 a window. Heating is a 90% AFUE gas furnace, gas costs $1.40 a therm, and the location sees 5,000 base-65 heating degree days. The annual heating bill is $1,400.
- Reduction in U-factor. 0.85 − 0.28 = 0.57 Btu/hr·ft²·°F.
- Heat lost now. 0.85 × 180 × 5,000 × 24 = 18,360,000 Btu, or 18.36 MMBtu a year.
- Heat lost after. 0.28 × 180 × 5,000 × 24 = 6,048,000 Btu, or 6.05 MMBtu.
- Heat no longer lost. 18,360,000 − 6,048,000 = 12,312,000 Btu. Equivalently, 0.57 × 180 × 5,000 × 24 gives the same figure directly.
- Fuel no longer burned. 12,312,000 ÷ 0.90 = 13,680,000 Btu of gas, and 13,680,000 ÷ 100,000 = 136.8 therms.
- Annual saving. 136.8 × $1.40 = $191.52, which is 191.52 ÷ 1,400 = 13.7% of the heating bill.
- Project cost. 12 × $900 = $10,800.
- Simple payback. 10,800 ÷ 191.52 = 56.4 years. Over the 25-year service life the fuel saving totals 25 × $191.52 = $4,788, leaving a net position of 4,788 − 10,800 = −$6,012.
Read step 6 and step 8 together. The windows cut the heating bill by nearly 14% — a genuine, visible reduction — and still do not come close to repaying themselves. Both statements are true at once, and a quotation that shows you only the first one is not lying, it is just not finishing the sentence.
How to read the payback figure
Compare the payback against the service life, not against a target. If the payback is shorter than the life you entered, the job repays itself in fuel before it needs doing again, and everything after that year is profit. If it is longer — which is the usual outcome for a whole-house replacement at market prices — the fuel saving is a partial offset against a cost you are incurring for other reasons.
Three situations genuinely do produce short paybacks, and it is worth checking whether you are in one. The first is a single-glazed house in a severe climate: at 8,000 HDD with electric resistance heat, a ΔU of 0.7 on 250 ft² saves enough each year that the numbers change character. The second is a partial job: replacing only the six worst windows costs a sixth as much and captures most of the loss, because heat loss is proportional to area and the worst windows are usually the biggest. The third is a window you were replacing anyway because it has failed — in that case the relevant cost is not the whole price but the upcharge for the better U-factor, which is often a few tens of dollars a unit and pays back in single-digit years. Enter that upcharge as the cost per window and the answer changes completely.
Treat simple payback for what it is: undiscounted and un-escalated. It ignores the time value of money, which lengthens the true payback, and it ignores fuel-price inflation, which shortens it. Over 20-plus year horizons those two are the same order of magnitude, so simple payback is a fair first pass. What it is not is an investment return; if you want to compare against other uses of the money, look at home improvement ROI and at the equity a renovation creates, which are the terms in which most window jobs are actually justified.
Annual heating-bill saving per 100 ft² of window
| Reduction in U-factor | 2,000 HDD | 4,000 HDD | 6,000 HDD | 8,000 HDD |
|---|---|---|---|---|
| 0.20 | $14.93 | $29.87 | $44.80 | $59.73 |
| 0.30 | $22.40 | $44.80 | $67.20 | $89.60 |
| 0.40 | $29.87 | $59.73 | $89.60 | $119.47 |
| 0.55 | $41.07 | $82.13 | $123.20 | $164.27 |
| 0.70 | $52.27 | $104.53 | $156.80 | $209.07 |
Read the largest cell in this table — $209 a year for a 0.70 U-factor improvement across 100 ft² in a severe climate — against what 100 ft² of installed replacement window costs, and the shape of the payback problem is immediate.
What this calculator does not include
- Air leakage. A properly installed new window is normally tighter than an old one. Infiltration can be a meaningful share of a leaky house's heat loss, and this calculator prices none of it.
- Solar heat gain. Low-E coatings that cut winter heat loss also cut free winter solar gain. In a sunny heating climate with south-facing glass, the net saving can be noticeably less than the U-factor arithmetic suggests.
- Cooling. The degree-day term here is heating only. In a cooling-dominated climate, the solar heat gain coefficient drives more of the bill than U-factor does.
- Trim, plaster and paint. A window job usually leaves interior trim and exterior siding to make good. Ask whether the quote includes it; if not, price it into your renovation budget.
- Rebates and tax credits. Utility rebates and any federal or state credit reduce the installed cost directly. Subtract them from the per-window figure before entering it.
- Comfort, noise and condensation. Real, valuable, and not monetisable from a U-factor. A room that stops having a cold zone by the window becomes usable in winter, which no fuel bill records.
Where the U-factor number comes from
U-factors on North American windows are measured under NFRC 100, the National Fenestration Rating Council's procedure, and reported on the NFRC label attached to the unit. The label figure is the whole-window U-factor — it includes the frame, the sash and the edge-of-glass zone, all of which perform worse than the middle of the glass. Manufacturer literature sometimes quotes a centre-of-glass number instead, which is always lower and always flattering. If a quote gives a U-factor without saying it is the NFRC rating, ask.
The ENERGY STAR windows programme sets maximum U-factors that differ by climate zone, so a unit that qualifies in the South will not qualify in the North. Certification is based on NFRC-rated values, which is why the two appear together on the label.
Where windows sit among envelope upgrades
Rank envelope work by dollars per unit of heat saved, and windows almost always come last. Air sealing is first — it is cheap, it addresses a loss path that is often larger than the glazing, and it needs no permits. Attic insulation is second: it covers a large area at a low cost per square foot and the R-value gains are dramatic, which is why attic insulation payback is usually measured in single-digit years. Wall insulation follows where it can be retrofitted. Windows come after all of those.
If the existing windows are sound, a storm window or a low-E interior panel gets a large share of the U-factor improvement for a small share of the cost, which shifts the payback arithmetic by roughly the ratio of the two prices. If they are not sound, the replacement decision has already been made on other grounds and the fuel saving is a bonus rather than a justification.
One last framing that helps: enter the upgrade rather than the replacement. If you have decided to replace anyway, the question is not whether U-0.28 beats U-0.85 — it is whether U-0.20 triple glazing beats U-0.28 double glazing for the extra couple of hundred dollars a unit. Put the price difference in the cost field and the ΔU between the two candidate products in the U-factor fields, and the calculator answers the question you are actually facing. Pair the result with a like-for-like comparison of the quotes, because window bids differ in scope far more than they differ in glass.
Key terms
- U-factor
- The rate of heat flow through a window per square foot per degree Fahrenheit of temperature difference, in Btu per hour. Lower is better. It is the reciprocal of R-value, so U-0.25 is R-4.
- Heating degree day (HDD)
- A measure of heating demand: for each day, the amount by which the mean outdoor temperature falls below a base temperature, summed over the year. The U.S. convention uses a 65°F base.
- AFUE
- Annual Fuel Utilization Efficiency — the seasonal ratio of heat delivered to fuel energy bought, for a furnace or boiler, measured under the U.S. Department of Energy test procedure.
- SHGC
- Solar Heat Gain Coefficient — the fraction of incident solar radiation a window admits. Low SHGC helps in cooling climates and hurts in heating ones.
- Therm
- 100,000 Btu, the unit natural gas is normally billed in. One kilowatt-hour is 3,412 Btu, so a therm is about 29.3 kWh of energy.