DE_LU Day-Ahead Prices Jul 16: Evening Peak up to 341 EUR/MWh
For tomorrow's day-ahead market, we expect the following:
π Daily pattern: On this Thursday, prices open at β¬167/MWh and hold a high overnight floor of β¬145β160/MWh. They rise to a morning peak of β¬185.75/MWh at 08:00, then ease to a midday low of β¬61.66/MWh at 15:00. From there prices climb steeply to an evening peak of β¬340.99/MWh at 20:45 before easing to β¬161/MWh by midnight.
π Range and spread: The forecast spans an intraday spread of approximately β¬279/MWh, from the midday trough of β¬61.66/MWh at 15:00 to the evening peak of β¬340.99/MWh at 20:45.
π»The forecast records no negative prices across any of the 96 quarter-hours.
β‘Elevated evening block: The evening block holds above β¬225/MWh for nine consecutive quarter-hours between 19:30 and 21:45.
Review of last week's forecasts

π Metrics: The forecasts averaged a solid MAE of β¬10.08/MWh and an RΒ² of 0.920 across the week. Sunday 12 July was the clear outlier at β¬24.44/MWh MAE, while five of seven days were below β¬11/MWh MAE.
π Maximum: Peak timing was strong in the second half of the week, with perfect quarter-hour hits on Monday 13 and Tuesday 14 July. Peak values were matched with an average absolute miss of β¬17.10/MWh.
π Minimum: The forecast timed the actual minimum within 30 minutes on six of seven days, including one perfect hit on Sunday. Minimum values were matched with an average absolute miss of β¬17.32/MWh, dominated by a single ββ¬62.09/MWh overshoot on Sunday.
β‘ Negative prices: Negative prices occurred in 50 quarter-hours, and the forecast correctly identified 30 of them (60%), with zero false alarms all week.
Interested in testing these forecasts for free or got feedback?
Just send us an email:
We at Telescope Energy are building a new forecasting platform for the energy sector.
Our platform offers state-of-the-art point and probabilistic forecasts for prices, generation and loads in >30 European markets.
In our weekly blog & newsletter we publish day-ahead electricity price forecasts from one of our frontier ensemble models.