Electricity consumption depends on many factors, such as season, day of the week, and also weather or temperature. If a winter month is comparatively mild, electricity can be saved here. If the weather is rather cool, savings efforts must be intensified. For this reason, the current deviation of the average temperature from the long-term normal temperature is shown in the first graphs as an indication of whether the respective month was particularly warm or rather cool. In addition, the weather-adjusted savings can be displayed using the mouse pointer. However, temperature is not the only factor influencing final consumption.

The peak load hours correspond to the hours that, in total, at least 10% of all hours between the beginning of December and the end of March showed the highest electricity consumption in the past.

The peak load hours for Switzerland are determined on the basis of historical time series of Swissgrid on final electricity consumption (in each case the last 5 winter periods). Based on the historical time series, fixed time windows (hour and day of the week) are defined for each winter period over the months in which load should be reduced, i.e. electricity should be saved. The excess or reduced consumption is shown for each customer group. A model based on smart meter data and machine learning determines the contributions of the individual customer groups.

Disclaimer Data

The calculation of excess or reduced consumption is based on statistical data from the SFOE and modelled data. The statistical data is available for the second preceding month. Since this data is not available for the previous month, modelled data must be used. The SFOE is working on making measured data available in a more timely manner. The weather adjustment of the figures is based on models of the SFOE and has a range of uncertainty that is not stated for simplicity. The customer groups result from a machine learning-based model that processes smart meter data from seven distribution grid operators and extrapolates to the whole of Switzerland. The data should accordingly be regarded as a national estimate. The daily over- or under-consumption per customer group in the current month are based on the same machine learning model and daily processed and extrapolated data. Weather and other effects are also included here. The data must be understood as an indicative value. For the calculation of the peak load times, the data from Swissgrid's "Energy Overview" are used. Experience shows that these data repeatedly have quality problems that are only corrected over time by the electricity industry (subsequent reports). Therefore, significant over- or underestimations may occur at times. The contributions of the individual customer groups to the savings are based on a model and must be understood as indicative. Switzerland-wide measured values of the consumption of these customer groups are not available. Since the definition of peak load hours is based on analyses of historical time series, it is possible that in the current year, due to the many influencing factors, peak loads actually occur at different times than predicted.