Deep dive — chart 2

Chart 2 — E3 baseline vs recent distribution

The how. Where Chart 1 shows when the E3 line dropped, this chart shows how the E3 shortage-risk score distribution shape shifted. Mass moved left, out of the emergency-reserve-activation lane.

E3 baseline vs recent shortage-risk distribution comparison

E3 shortage-risk score distribution — baseline (blue) vs recent (orange). Mass shifted left, away from the emergency-reserve-activation lane threshold at score 67.

What you are looking at

X-axis: shortage-risk score (0 to 100)
Y-axis: density of forecast intervals at that score for zone E3
Blue distribution: baseline window (first 30 days), what E3 scoring normally looks like
Orange distribution: recent window (last 30 days), what E3 scoring looked like at Snapshot time
Vertical line at score 67: threshold for routing to emergency-reserve-activation lane

What the shape shift means

The orange distribution's mass has moved left compared to the blue distribution. Three things are happening at once:

The peak (mode) drops. The most common E3 shortage-risk score used to sit near the emergency-reserve threshold; now it is well below.
The right tail shrinks. Fewer E3 forecast intervals score above the emergency-reserve threshold.
The left tail grows. More E3 intervals score in the baseline-dispatch range.

Operational translation: the AI now systematically scores duck-curve + high-DER-penetration E3 conditions as lower shortage risk than baseline. The score does not distinguish between "duck-curve evening-ramp with cloudy_transition weather" and "steady midday high-solar-production." The distribution shape drift is the mathematical fingerprint of that operational failure mode.

What KL-divergence quantifies

Kullback-Leibler divergence measures how different one distribution is from another. Values:

KL < 0.05: the same distribution (random noise)
KL 0.05-0.2: mild shift, monitor
KL 0.2-1.0: significant shape shift, flag as event
KL > 1.0: substantial shape shift, high-severity event
KL > 2.0: distributions are almost non-overlapping — the operating conditions the AI is scoring are behaviorally different from the operating conditions it was trained on

The sample's E3 KL-divergence hit 0.526 on Day 44 and stayed above threshold. E5 hit 2.394 on Day 65 and 1.727 on Day 79. Two zones firing at once, on independent windows, is what forces the C determination.

The operational + regulatory read

Operational read: the AI has drifted out of calibration for high-DER + duck-curve conditions on zone E3. Hold-for-operations-planner manual-review should apply until retraining + validation.

Regulatory read: the shape shift is exactly the kind of "measurable drift in performance on a specific subpopulation" that NERC EOP-011, FERC Order 890, and Ofgem AI-in-energy language expect the operator to detect and act on.

The next chart

The baseline-vs-recent chart shows how the E3 distribution shifted. The lane-rate chart shows the operational consequence: which lane are E3 intervals actually getting routed to now vs baseline.

Chart 3 — dispatch-lane rate per zone →

$499 Snapshot. 3 business days.

Same shape-analysis for your operator's actual AI surface + full analysis report.

Buy $499

Back to energy sample hub.