The prediction of the electricity and gas consumption is a complex problem that involves socio-economic aspects as well as the strong influence of the weather. We present our experience with electricity and gas consumption forecast for E.ON Energie, a.s., the Czech subsidiary of the German E.ON company.
We quantify the relationship of the various meteorological elements to the electricity consumption during the year. The power consumption depends on a day of the week and four different tariff periods. It is highest in cold winter months due to the increased requirements for heating and lighting. From meteorological parameters, air temperature and sunshine have the greatest impact on the consumption. In very cold and cloudy winter days, the power consumption in the E.ON Energie, a.s. distribution zone reaches 52 GWh per day. If on another day air temperature is the same but the sky is clear, the power consumption drops on more than 1 GWh per day. The greatest uncertainty in the power consumption forecast in winter is thus related to the correct estimates of cloudiness. In addition, the Christmas season strongly affects the final forecast already from mid-December until the end of the first decade of January. In summer, the biggest errors in the consumption forecast are found during very hot days. The most critical factor is to determine the air temperature threshold when air-conditioning units are switched on and used in a large extent.
The gas consumption forecast is a simpler task since we predict only the daily sum. The principles are similar as in the case of the electricity consumption prediction. There is a dependence on a day of the week, but no dependence on a tariff period. Air temperature is a key meteorological factor. The most difficult moment to predict is a response of customers to large day to day temperature changes. Short inertia of the gas consumption needs to be consider in the final forecast, i.e., there are no sudden jumps in the gas consumption when air temperature changes rapidly. Considering cumulative sums of air temperature over several recent days seems to be a useful strategy to solve this issue. The importance of the gas consumption forecast is highest in winter when daily consumption rates are up to six times higher than in summer. The start and end of the heating season are the most complicated moments, since the customers’ behavior plays stronger role than at any other time of the year.