The losses represent mainly the energy dissipated due to the Joule effect during transmission over the high and extra high voltage network. They depend mainly on consumption, the generation plan and cross-border exchanges.
The system loss rates are included between 2 and 3.5% of consumption, depending on the seasons and time of day. On average, the rate comes to 2.5%, which represents about 11.5 TWh (TeraWatt-hours) per year.
Each week, RTE draws up loss forecasts for the coming week. These forecasts are established on the basis of predicted consumption, generation and exchanges. They are made available to those concerned on the Web site on Thursday (12 a.m.). These forecasts are refined two days before so as to take into account the meteorological conditions. These forecasts are refined on day D-2 according to the latest meteorological forecasts. They are then used to establish the delivery programmes sent to suppliers with whom RTE has signed energy exchange contracts.
Actual losses are calculated as the difference between injections (generation injected + physical imports) and extractions (consumption extracted + physical exports) at the terminals of the public transmission system.
The calculation process is aligned with the flow reconstitution process. The losses for a given month M are calculated at the deadlines M+3, M+6 and M+12. Newly available data are published on the 15th of each month.
E.g. on 15/09/2007, the values M+3 from May 2007, M+6 from February 2007 and M+12 from August 2006 are published on the website.
They are made available to players in the tables below.
Daily loss forecasts are established at least two days before they are published, based on the latest available consumption predictions, and an estimate of the losses that have occurred on previous days. These forecasts are then used to establish energy delivery programmes.
Weekly loss forecasts are established on the basis of predicted consumption, generation and cross-border exchanges. These forecasts are updated every Thursday. They cover the period from Saturday to the following Friday and serve as a basis for the elaboration of daily forecasts. Since 08/02/2006, weekly forecasts have been established on the basis of actual temperature predictions, rather than normal seasonal temperatures, as was previously the case.