The Degree-Day Method: Estimating Building Energy from Weather Data

Long before hourly building simulation existed, engineers estimated heating bills with little more than a weather table and a heat-loss coefficient. That technique — the degree-day method — is the oldest tool in building energy analysis, and it is still one of the most useful. Not because it competes with EnergyPlus, but because it answers a different set of questions: quick estimates, sanity checks and, above all, weather normalization of real utility data.

What a degree day is

A degree day measures how far, and for how long, the outdoor temperature deviates from a base temperature. If the base is 65°F (18.3°C) and a day averages 45°F, that day contributes 20 heating degree days (HDD). A day averaging 85°F contributes 20 cooling degree days (CDD). Sum them over a month or a year and you get a single number describing how much heating or cooling demand the weather imposed: Miami sees roughly 200 HDD a year, Minneapolis over 7,000. The traditional 65°F base isn't arbitrary — it reflects the idea that a typical building doesn't need heat until it's colder than that outside, because occupants, lights and equipment already supply the difference.

The method in one equation

The classic degree-day estimate says seasonal heating energy is heat-loss rate times accumulated temperature difference, divided by system efficiency:

E = (UA × HDD × 24) / η

where UA is the building's overall heat-loss coefficient (envelope conductance plus infiltration), HDD the heating degree days for the period, 24 converts days to hours, and η the heating system efficiency. Two inputs — one describing the building, one describing the climate — and you have an annual heating estimate on the back of an envelope. The same logic, with CDD and equipment COP, gives a rough cooling figure.

The balance point — and why 65°F is often wrong

The method's most important refinement is recognizing that every building has its own balance-point temperature: the outdoor temperature below which it actually starts needing heat. A modern office dense with people, equipment and solar gain may not call for heating until 50°F or lower; a poorly insulated house with few internal gains may need it at 62°F. Using degree days computed at the wrong base can misstate heating demand badly — which is why serious applications use variable-base degree days, computed at the building's estimated balance point rather than the traditional 65°F. The deeper insight survives into modern practice: internal gains offset heating, and buildings with high gains behave very differently from the weather alone.

What the method is genuinely good for

• Weather normalization: was last year's higher gas bill a building problem or just a colder winter? Dividing consumption by HDD — or better, regressing bills against degree days — separates weather from performance. This is the everyday workhorse behind benchmarking comparisons and utility-bill baselines, and it's the simplest form of the whole-facility regression that ASHRAE Guideline 14 formalizes for measuring savings;
• Early-stage estimates: a defensible order-of-magnitude heating figure when all you have is a floor area, a climate and an envelope assumption;
• Sanity-checking simulations: if a detailed model's heating result is wildly out of line with a degree-day estimate, something in the model deserves a hard look;
• Fuel budgeting and tracking: utilities and operators have used degree-day regressions for decades to forecast demand and flag drift in building performance.

Where it breaks down

The degree-day method is a steady-state, envelope-driven picture of a building, and everything it ignores is exactly what dominates modern commercial buildings. It has no hour-by-hour solar gain, no thermal mass storing and releasing heat, no humidity — a serious omission where latent cooling is half the load — no part-load equipment behavior, and no time-of-day anything: schedules, setbacks, demand charges, or the carbon intensity of the grid hour by hour. It cannot see systems that heat and cool simultaneously, economizers, or heat recovery. For a skin-dominated house it can land respectably close; for an internally loaded office tower it is a rough sketch at best. That's why codes and rating systems — ASHRAE 90.1 Appendix G, LEED, Title 24 — require full 8,760-hour simulation: the questions they ask live precisely in the dynamics the degree-day method averages away. Between the two sits the older bin method, which groups hours by temperature band and captures part-load effects, but it too has largely given way to hourly simulation now that computing is free.

The takeaway

Think of the degree-day method as the slide rule of energy analysis: superseded for precision work, indispensable for intuition. Use it to normalize bills, frame early estimates and sanity-check bigger models — and reach for hourly simulation the moment the question involves solar, humidity, schedules, systems or carbon. Knowing which tool the question calls for is most of the skill.

This article is general guidance and reflects information available at the time of writing. Degree-day data sources, base-temperature conventions and code requirements vary by jurisdiction — always confirm the applicable methodology for your project.