Heating degree days:
formula and worked example

A temperature index built for energy

Average temperature alone is a poor guide to heating bills. Two months can have the same mean while one has a steady chill and the other swings between mild and bitter. Heating degree days solve this by counting only the cold: each day contributes how far its mean temperature fell below a base, and warm days contribute zero.

Because heating demand is roughly proportional to how far outdoor temperature sits below the point at which a building needs heat, summing the daily values gives a number that tracks fuel or electricity use. Energy managers use it to compare winters, check whether a building is performing better than last year, and size equipment.

The idea is old and simple, which is part of its appeal. It converts a stream of daily temperatures into a single seasonal severity figure that anyone can compare with the same figure from another year or another town.

The daily formula

• Base temperature: the outdoor temperature below which the building typically needs heating. Conventions vary by country and building type.
• Mean temperature: usually the average of the day's highest and lowest readings, or the mean of hourly readings.
• Result: degree days for one day. Add the daily values to get a week, month or season.

The zero floor matters. A day warmer than the base adds nothing, and it must not subtract from colder days elsewhere in the month.

Where hourly data is available, some methods sum hourly differences and divide by 24, which is more accurate on days with large swings. For most building work, the daily mean method is adequate and is what published degree day tables use.

Worked example: base 18, mean 11

A building uses a base temperature of 18 degrees, and one winter day had a mean of 11 degrees.

1. 1Base temperature: 18.
2. 2Mean temperature: 11.
3. 3Difference: 18 − 11 = 7.
4. 4The value is positive, so the day contributes 7 heating degree days.

If a 30-day month averaged 7 such degree days each day, the monthly total would be 210. Compared with a milder month at 120, you would expect roughly 75% more heating energy in the colder one, all else equal.

The same arithmetic with a mean of 20 would give 18 − 20 = −2, which the floor turns into zero. That day adds no heating demand under this method, even though it is nominally warmer than the base by two degrees, and it is correct that it adds nothing.

How the totals are used

The number has no meaning alone, but it is powerful for comparison. A building that used 40,000 kWh in a winter with 2,000 degree days and 36,000 kWh in a winter with 1,800 degree days used the same 20 kWh per degree day. A change in that ratio signals a real change in performance, not weather.

• Weather normalisation: divide energy use by degree days to compare periods fairly.
• Forecasting: scale expected demand with a forecast of degree days.
• Benchmarking: compare similar buildings across locations.

A practical way to use it is to draw a simple chart of monthly energy against monthly degree days. If the points fall roughly on a line, the building behaves predictably, and the slope is its energy per degree day. Points far above the line suggest a fault, an unscheduled use or a controls problem.

Why the base temperature is touchy

Raising the base from 18 to 19.8 takes the day's result from 7 to 8.8, a gain of 1.8 degree days, or about 26% from a 10% change in the base. Since the day's difference is already small, a modest shift in the base changes the answer by a large relative amount. Pick a base that matches the building and keep it fixed.

Slips when working with degree days

• Averaging the month's temperature first and then applying the formula, which hides warm days that should count as zero.
• Using a different base for each site and then comparing totals directly.
• Mixing Celsius and Fahrenheit data and bases.
• Treating degree days as energy, when they are only an index to be paired with a measured building response.
• Using the heating formula for summer cooling needs.

Applying it to your own building

Collect daily mean temperature for the period, choose a base that matches your building, sum the daily results, and divide metered energy by the total. After a winter or two you will have a figure for energy per degree day that is a dependable baseline for spotting problems or measuring the effect of an upgrade.

If hot water or cooking is on the same meter, the line will not pass through zero, because those loads exist even on warm days. That intercept is the base load, and separating it from the weather-driven part gives a clearer view of where savings are available.