Why your imbalance penalty isn't a forecasting problem
Generation accuracy and imbalance cost are different problems with different solutions. Most vendors solve the first and bill you for the second.
Analysis drawn from production forecasting in Georgia and Turkey — imbalance, market reform, and what honest backtesting actually looks like.
Generation accuracy and imbalance cost are different problems with different solutions. Most vendors solve the first and bill you for the second.
What it takes to forecast whether balancing settles above or below day-ahead, and why the honest ceiling is 85%, not 100%.
Quarter-hourly settlement, mandatory aggregator membership, portfolio-level KÜPST. What to prepare now.
There is no native 15-minute numerical weather prediction for Turkey. From 2027, that becomes everyone's problem.
What we learned modelling Georgian run-of-river hydro, and why it transfers to every emerging power market.
Issue-time discipline, rolling-origin retraining and leakage audits: the three questions that separate a real number from a slide.
Want one of these ahead of publication, or a walkthrough of the underlying analysis? Write to us.
Send us twelve months of generation history. We’ll come back with a measured estimate of the imbalance cost and lost revenue your current setup is carrying — before you commit to anything.