A heating degree day is a measure of how much heating a day required, and it makes two winters comparable.
The definition
Take a base temperature, conventionally 65°F, on the assumption that a house needs no heating above it because internal gains cover the difference.
For each day, take the average outdoor temperature. If it is below the base, the difference is that day's degree days.
HDD for a day = max(0, 65 − average outdoor temperature)
A day averaging 45°F contributes 20 HDD. A day averaging 70°F contributes none. Sum them over a season.
What the numbers look like
Annual heating degree days run from a few hundred in the far South to over 9,000 in the northern Plains and interior Alaska. Most of the populated United States falls between 2,000 and 7,000.
They are published by NOAA by station and by month, and utilities frequently print them on the bill precisely so customers can make the comparison below.
The useful calculation
fuel per degree day = fuel used ÷ HDD for the same period
That single number normalises out the weather.
If it is roughly the same as last year, the house and the furnace are behaving the same way and the bill went up because the winter was colder.
If it has risen, something has changed — a furnace losing efficiency, a duct that has come apart, insulation disturbed, a window left open, or a change in how the house is used.
If it has fallen, an improvement is working, and this is how to verify that insulation or air sealing actually did what it was supposed to.
Using it to check a claim
This is the honest way to evaluate a retrofit.
Take fuel use per degree day for the season before, and for the season after. The difference is the improvement, with the weather removed.
Comparing raw bills instead is why so many efficiency improvements appear either miraculous or useless depending on what the weather did.
The base temperature caveat
65°F is a convention, not a property of a particular house.
A very well insulated house with substantial internal gains may not need heat until it is much colder outside, so its true base is lower and the standard HDD figure overstates its heating requirement. A leaky house has a higher base.
For year-to-year comparison of the same house this does not matter, because the same convention applies to both sides. It matters when comparing houses.
Where it does not apply
Degree days do not capture wind, which drives infiltration, or solar gain, which offsets heating on clear days.
They also say nothing about the cooling season, which has its own cooling degree day measure with the same construction.
The practical routine
Record the fuel used and the degree days for each billing period, from the bill or from NOAA.
Divide. Keep the number.
After two seasons there is a baseline, and after that the calculation tells you within a month whether something has changed — long before a failing furnace announces itself any other way.
Sizing uses a different number
Degree days describe a whole season and are used for estimating fuel consumption. They are not what a furnace is sized against.
Sizing uses the 99% winter design temperature — a single cold condition — because the furnace has to meet the load on the coldest hours, not on the average. A mild climate with a very long heating season can have more degree days than a harsh one with a short winter, and still need a smaller furnace.
Mixing the two is a common error in DIY sizing discussions.
Where to get the data
NOAA publishes monthly and seasonal degree day totals by weather station, and many utilities print the period's degree days directly on the bill alongside the consumption.
Several free services will produce a custom series for a location and a chosen base temperature, which is useful for a well-insulated house whose true base is below 65°F.