f97.beERIK SCHMIDT’S WORKSHOP
← All terms

Glossary · Energy industry & energy transition

Degree days and heating degree days

Degree day number Heating degree days heat demand Heating limit temperature building weather adjustment Heating load heat pump

Degree days and heating degree days are indicators of how strongly the weather in a given period forces heating. They apply to a location, a region or the site of a building. They are calculated from daily mean outdoor temperatures and two fixed values: a reference temperature and a heating limit. Only heating days are counted, that is, days whose daily mean temperature is below the heating limit. For each heating day a temperature difference is formed, and these values are added up over the period. The unit is the kelvin-day (Kd or K·d).

Definitions: Which convention is meant?

In Germany, the degree-day figure G20/15 is widespread. The first number stands for an assumed room temperature of 20 °C, the second for a heating limit of 15 °C. On each heating day, the difference between 20 °C and the daily mean outdoor temperature is formed. The German Weather Service (DWD) calculates the degree days for the official climate factors from daily mean values of its measuring stations according to guideline VDI 3807. In addition, there are heating degree days in the narrower sense, usually called G15. Here the heating limit itself serves as the reference value, so the calculation is 15 °C minus the daily mean. Eurostat, the EU's statistical authority, uses a third variant: on days with a daily mean of up to 15 °C, the difference to 18 °C is counted; on warmer days, zero.

An example shows the differences: a day with a 5 °C daily mean yields 15 Kd under G20/15, 10 Kd under G15 and 13 Kd under Eurostat. A day at 16 °C does not count in any of the three variants. With G20/15 and with Eurostat, the daily value jumps at the heating limit: a day just below the limit immediately yields roughly 5 or 3 Kd respectively, while a day just above yields zero. The terms Gradtagszahl, Heizgradtage and heating degree days are often used interchangeably in everyday speech, but they refer to different calculation methods. Figures from different conventions must not be compared directly. Every evaluation should therefore state the reference temperature, heating limit, period, data source and the way the mean was formed.

The unit is easily misleading. A kelvin-day is neither a temperature nor an amount of energy, but a temperature difference multiplied by a duration. The longer and colder a heating period is, the higher the sum. A mild winter produces lower values and a cold one higher values, even if the same building is heated unchanged. Only together with a building property does an amount of energy result, for example with the heat loss in watts per kelvin. Mathematically, watts per kelvin times kelvin-days times 24 hours gives watt-hours. This is only a rough physical framework, since sunshine, internal heat sources and controls shift the result considerably.

Distinction from heat demand, consumption and heating load

Degree days do not measure the heat demand, only the influence of the weather. How much heating energy is actually consumed additionally depends on many factors: the building envelope, ventilation, control technology, user behaviour, internal heat gains, solar radiation, the efficiency of the heating system and the mode of operation. Two buildings at the same location have the same degree-day figure but can consume very different amounts.

The heating load is also something different. It is a power, usually given in kilowatts: the heat output a building needs at a defined low outdoor temperature. It is important for sizing heat generators, heating surfaces and heat pumps. The degree-day figure, by contrast, is a sum over a period and is mainly used to assess annual and monthly consumption. A building can have a high heating load and still consume little in a mild year.

The difference between demand and consumption is equally important. Demand is calculated using model assumptions about building, climate and use, whereas consumption is measured. Degree days do not automatically connect the two levels. They provide a correction factor for the temperature influence. Anyone who infers the energy quality of a building from a single year of consumption without reference to the weather mistakes a weather effect for an efficiency characteristic.

Weather adjustment and climate factors in Germany

The most important application is weather adjustment, also called weather normalisation. Here, the consumption for space heating is multiplied by the ratio of a reference degree-day figure to the actual degree-day figure of the period under consideration. The principle is simple: if the actual degree-day figure is smaller than the reference, the year was milder and the adjusted consumption is higher. If it is larger, the year was colder and the adjusted consumption is lower. In this way, higher consumption in a cold year is not hastily judged a deterioration, and lower consumption in a mild year is not judged an efficiency gain.

For energy consumption certificates in Germany, the procedure is bindingly regulated. According to the information portal of the Federal Office for Building and Regional Planning, which now refers to the Building Modernisation Act (GModG), the measured consumption for space heating from at least the last 36 months must be climatically adjusted. The basis is § 82 paragraph 3 in conjunction with paragraphs 4 and 5. The DWD climate factors must be used. These compensate for two things at once: the weather of the period and the location of the site. All consumption values are converted to the reference location Potsdam.

The DWD calculates the climate factors for rolling periods of twelve months each and for every postcode area. The climate factor is the quotient of the mean annual degree days of the Test Reference Year (TRY) Potsdam, stated as 3666.8 Kd, and the current annual degree days at the location. For consumption certificates issued from 1 May 2014, Potsdam rather than Würzburg is the reference location. A worked example: a building consumes 15,000 kWh for space heating at a location with 3300 Kd. This gives a climate factor of about 1.11 and an adjusted value of around 16,670 kWh. According to the DWD, the climate factors are also based on provisions of the Heating Costs Ordinance (Heizkostenverordnung).

Adjustment is also important outside energy certificates. It is used for energy reports, building management, operating cost analyses, energy-saving guarantees and the data basis of municipal heat planning. Without it, renovations are easily misjudged. A new heating control, for example, can be effective even though absolute consumption rises because of a colder winter.

Limits of the method

Degree-day correction works best for buildings whose consumption depends strongly and regularly on the outdoor temperature. It becomes inaccurate with large internal heat loads, changing use, vacancy or greatly altered ventilation. Hot water hardly depends on the outdoor temperature. Therefore only the share for space heating is adjusted. In the methodology of the European efficiency database Odyssee, a share that is independent of the climate is split off for this purpose. Anyone who adjusts the entire gas or district heating consumption across the board, although a large part is accounted for by hot water, distorts the result.

A fixed heating limit of 15 °C is also a strong simplification. The outdoor temperature at which heating is actually required depends on the building. Well-insulated houses with a lot of solar and internal heat often need heating only at significantly lower temperatures, while uninsulated older buildings need it earlier. A uniform limit facilitates comparisons but does not represent every building physically. Moreover, degree days capture only air temperature, not wind, sunshine or humidity.

Relevance for heat pumps and power grids

When space heating is electrified with heat pumps, weather-dependent heat demand becomes weather-dependent electricity consumption. With air-to-water heat pumps, a second effect is added: their coefficient of performance falls as it gets colder outside. On cold days they therefore need more electricity for the same amount of heat than on mild days.

Degree days above all show how strongly temperature acts over a season or several days. They do not replace an analysis of hourly loads. Two months with the same degree-day figure can have very different load profiles. Short frost peaks burden grids and power plants differently than a uniformly cool phase, especially when they coincide with little wind and solar power. In the distribution grid, what matters is how many heat pumps in a neighbourhood draw high power at the same time. Blocking periods, buffer storage, the thermal inertia of buildings and tariff rules play a role here. A high degree-day figure implies a higher heating energy demand, but not a particular grid peak.

Typical misinterpretations

  • Degree days are not a property of a building. They describe the weather at a location according to a chosen calculation convention. How strongly this weather shows up in consumption is determined by the building.

  • Values according to G20/15, G15 and Eurostat cannot be compared without conversion. This also applies to values with different data sources or reference periods.

  • Falling degree-day figures due to climate change do reduce the mean demand for space heating. However, individual cold spells remain decisive for heating load, grid peaks and reserves. The growing cooling demand is described separately by cooling degree days.

  • Weather adjustment does not replace a missing measurement. Nor does it correct an unclear split between space heating, hot water and process heat.

Used correctly, degree days and heating degree days separate the weather from building properties and mode of operation. Their strength lies in comparing periods and locations, not in fully explaining energy consumption. For the power system, they show how strongly electrified heat is shaped by cold periods, without replacing an analysis of power peaks.

Sources

Related glossary entries

Edited by eCMS on 12.11.2026 05:05

© 2026 Erik Schmidt. All rights reserved. Quotations with a link to the source are welcome.