Heating and cooling degree days in energy analysis
Degree days turn weather into a number you can put next to your energy use. Here is how HDD and CDD are calculated, which base temperature to pick, and how to normalize consumption correctly.
You cannot tell whether a building improved or the winter just got milder unless you account for weather. Degree days are the standard way to do that. They convert outdoor temperature into a single figure that tracks how much heating or cooling a location needed over a period, so you can compare this January against last January on fair terms.
What a degree day is
A heating degree day (HDD) measures how far and how long the outdoor temperature sat below a base temperature, and a cooling degree day (CDD) measures how far it sat above. The daily calculation is straightforward: take the day's mean temperature, usually the average of the high and the low, and compare it to the base.
- If the mean is below the base, that day's HDD equals the base minus the mean
- If the mean is above the base, that day's CDD equals the mean minus the base
- Sum the daily values over a month or a year to get the period total
A day with a mean of 8 C against an 18 C base contributes 10 heating degree days. A day at exactly the base contributes zero. Add the days up and you have a weather index you can line up against your consumption.
The base temperature is a choice, not a constant
The most common base is 65 F, which is 18.3 C, and Canadian practice rounds to a base of 18.0 C. But that number is a convention, not a law of physics. The base is meant to be the outdoor temperature at which internal gains and solar gains offset the building's heat losses, so no heating is needed. For real buildings that balance point typically falls between 55 and 65 F, and a well-insulated modern building may balance lower than an old one.
If you normalize an efficient building against a 65 F base when its true balance point is 60 F, you overstate its weather-driven load and understate its baseload. Picking a base that reflects the building, ideally validated against its own energy signature, is what makes degree day analysis trustworthy.
Using degree days to normalize energy
The workflow is simple in principle. Regress your metered energy for each period against the matching degree days, and you get two useful numbers: a baseload that does not vary with weather, and a slope that tells you how much energy each degree day costs you. Once you have that relationship, you can predict what a normal-weather year should cost and measure real performance against it.
| Metric | What it tells you | Watch out for |
|---|---|---|
| Baseload (intercept) | Weather-independent use: plug loads, lighting, ventilation | Rises quietly when controls drift |
| Slope (per degree day) | Heating or cooling efficiency of the envelope and plant | Sensitive to the base you chose |
| Model fit (R-squared) | How much of your use weather explains | A poor fit means the base or the data is wrong |
For weather normalization to work, three things have to line up: the weather station has to represent the building's location, the degree days have to cover the exact billing period, and the base has to suit the building. Portfolio Manager's own degree days calculator draws on hundreds of stations across the United States and Canada precisely so the weather data matches the site.
Where degree day analysis goes wrong
The math is easy. The data handling is where it breaks. Billing periods that shift by a few days each month, meters that are read on estimates, and mismatched weather stations all corrupt the regression before you see the result. A tidy chart of energy against degree days can hide a base that was never validated or a month of estimated reads. Clean, period-aligned data is the difference between a normalization you can defend and one that quietly misleads.
Get the base right, match the periods, and pick the right station, and degree days become one of the most reliable tools in energy analysis. Skip those steps and the same tool will point you in the wrong direction.
Frequently asked questions
What base temperature should I use for degree days?
The conventional base is 65 F (18.3 C), and 18 C is standard in Canada. The physically correct base is the building's balance point, where internal and solar gains offset heat losses, which for many buildings falls between 55 and 65 F. Validating the base against the building's own energy signature gives the most reliable results.
How are heating degree days calculated?
Take each day's mean temperature, usually the average of the daily high and low. If the mean is below the base temperature, that day's heating degree days equal the base minus the mean. Summing the daily values over a month or year gives the period total.
What is the difference between heating and cooling degree days?
Heating degree days accumulate when the daily mean is below the base temperature and estimate heating demand. Cooling degree days accumulate when the mean is above the base and estimate cooling demand. Most climates produce both across a year.
Why does weather normalization with degree days sometimes fail?
Common causes are a weather station that does not represent the building, degree days that do not match the exact billing period, estimated meter reads, and a base temperature that does not suit the building. Any of these weakens the regression and produces a misleading result.
Weather Normalization: Why Year-Over-Year Energy Comparisons Mislead
A warm winter can hide a failing retrofit, and a cold one can erase a real gain. Here is how heating and cooling degree days work, how weather normalization is done, and when it matters for benchmarking and measurement and verification.
How to improve your ENERGY STAR score
Your ENERGY STAR score compares your building to its peers on a 1 to 100 scale. Here is what actually moves it, where the quick points are, and why clean data comes first.