Theory 32/Earth, Climate & Energy /Power systems
Reserves Are Sized by the Worst Loss
Reserves are tuned to variability. The standard is written around the worst event.
Frequency containment reserve is sized by the largest credible loss rather than by the variance of imbalance, so two systems with the same variability can need very different reserves depending on their single largest contingency.
01The evidence
Reserve requirements are set by contingency standards, and systems with similar variability differ in requirement with the size of their largest unit.
02The tension
Probabilistic reserve methods optimise against variability, which is appropriate for normal operation and neglects the deterministic contingency that sets the floor.
03The mechanism
Frequency nadir after a loss is set by the lost power and the inertia, not by the distribution of small imbalances. The requirement is therefore a maximum, not a variance.
That makes the largest contingency the determinant, and predicts reserve should track it rather than track variability.
04What this is not
Those methods optimise for typical imbalance. This sets a floor from the largest loss.
05Predictions
- P1Reserve should track largest contingency.
- P2Systems with equal variance should differ by unit size.
- P3Variance should predict poorly.
06Refutation
Reserve tracking variance rather than contingency would favour probabilistic sizing.
07What it changes
It would change reserve procurement.