A degree day is an indicator of weather as it affects the cooling and heating energy required to maintain a comfortable building temperature. A daily value for heating or cooling degree days is calculated as the difference between the average daily temperature (the 'middle value' between the daily high and the daily low) and the building balance point temperature or BPT (default BPT is 60 degrees F, 16 degrees C).
When the average daily temperature is above the BPT, the result is cooling degree days. When the average daily temperature is below the BPT, the result is heating degree days. The greater the disparity between the BPT and the mean daily temperature, the more energy must be expended to maintain the building at a desirable temperature.
Example 1: Average daily temperature = 80. BPT = 60. Cooling degree days = 20 CDD. (80-60=20)
Example 2: Average daily temperature = 60. BPT = 60. No degree days.
Example 3: Average daily temperature = 30. BPT = 60. Heating degree days = 30 HDD (60-30=30)
In the above examples, Example #2 would require the least energy consumption, and example #3 would require the most. Example #1 would require cooling to maintain the building at a comfortable temperature, while example #3 would require heating.
Someone might ask, "To keep it simple, why not use daily average temperature as an indication of weather severity instead of degree days?" The problem with this approach is that highs and lows will cancel each other out over time, falsely indicating lower heating and cooling needs than actually experienced by the building's mechanical systems. The annual average temperature for Chicago is 49 degrees. Does that mean a building in Chicago needs no cooling system? A warm day (80F average temp) combined with a cold day (40F average temp) average 60F. So do two mild days of 61F and 59F. But in the first case there are 20 CDD and 20 HDD while in the second there are 1 CDD and 1 HDD. Using degree days, it is clear that the relative amount of energy required to maintain a comfortable building temperature during the first set of days is much greater than for the second set of days. But using average temperature, data, both sets of days would appear to be about the same.