Graddagemodeller

Graddagemodeller, or degree-day models, are a class of statistical tools used to estimate and forecast building energy demand based on outdoor temperature. The approach derives from the concept of heating and cooling degree days, which quantify how far daily temperatures are from a base level that triggers heating or cooling needs.

The core idea is to relate energy consumption to temperature-driven demand. A base temperature, Tb, is chosen, and heating degree days (HDD) and cooling degree days (CDD) are defined as HDD = max(0, Tb − T_out) and CDD = max(0, T_out − Tb), where T_out is the outside temperature. A typical graddagemodell expresses energy use E as E = a + b_HDD · HDD + b_CDD · CDD + ε, or in simpler forms as E = a + b · HDD for heating-dominated situations. Coefficients measure how sensitive energy use is to temperature deviations.

Applications include short-term forecasting, benchmarking building performance, and load disaggregation. Models are calibrated with historical energy data and weather data, often using linear regression or robust alternatives. Variants vary the base temperature, include nonlinear terms or separate parameters for heating and cooling seasons, and may add additional weather or occupancy factors.

Advantages of graddagemodeller include simplicity and interpretability, with relatively low data requirements. Limitations arise from assuming a linear relationship between load and degree days, potential mis-specification of Tb, and the exclusion of non-temperature drivers such as occupancy, equipment efficiency, and internal heat gains. Graddagemodeller remain a common first step in energy analysis and policy planning.