Degradation Model Fidelity Score (DMFS)

GMI-RE-005Renewable EnergyInfrastructure-DependentPattern O-1: Confidence Calibration

Definition

Ratio of the storage optimizer's projected cumulative capacity fade, given the dispatch decisions it actually made, to the capacity fade measured in periodic reference capacity tests. Detects an optimizer trading battery life for near-term revenue on a degradation model that no longer matches the cells.

The Measurement Gap

Battery energy storage systems present a unique optimization problem: every dispatch decision simultaneously generates revenue and consumes battery lifespan. The dispatch optimizer embedded in every BESS installation contains a degradation model that estimates how much lifespan each charge-discharge cycle costs. That degradation estimate feeds directly into the revenue optimization math. If the model says a cycle costs 0.02% of total capacity, the optimizer will approve any arbitrage spread that exceeds the projected lifespan cost. If the actual cost is 0.04%, the optimizer is destroying value twice as fast as it calculates.

The measurement gap is the absence of a structured comparison between what the optimizer's degradation model predicts and what actually happens to the battery. Most operators track overall capacity fade through periodic reference capacity tests. They also track revenue from dispatch decisions. But they do not connect these two data streams to answer the question: is the optimizer's internal model of degradation costs accurate?

Without DMFS, an operator can show that revenue is strong and capacity is declining, but cannot determine whether the revenue was worth the capacity consumed. A BESS generating excellent arbitrage revenue while its degradation model understates actual fade by 30% is effectively borrowing from future capacity to pay today's returns. The books look good until the battery reaches end-of-warranty capacity thresholds years earlier than projected.

The decision DMFS supports is whether the revenue optimizer is consuming battery lifespan faster than it projects, meaning the revenue captured today is coming at an unrecognized cost to tomorrow's capacity and warranty position.

Formula

After each measurement period, compare the optimizer's projected cumulative capacity fade (based on the dispatch decisions it actually made) against the actual measured capacity fade from periodic reference capacity tests. The ratio tells you whether the optimizer's internal degradation model is optimistic, pessimistic, or calibrated.

Formula (Notation):

DMFS = Projected_Capacity_Fade / Actual_Capacity_Fade

Variable Definitions:

Variable Definition
Projected_Capacity_Fade the degradation model's cumulative projected capacity loss (percentage) based on actual dispatch history over the measurement period
Actual_Capacity_Fade measured capacity loss (percentage) from reference capacity tests conducted at the beginning and end of the measurement period

Edge Cases and Measurement Boundaries:

Capacity tests are typically infrequent (quarterly or semi-annually), so DMFS can only be computed at test intervals, not continuously. Temperature-dependent degradation should be segmented by season because summer cycling stress may be underestimated while winter cycling is well-modeled. DMFS measures the cycle-aging component only; calendar aging is a separate baseline that should be accounted for before computing the ratio.

Confidence Grading (required alongside every DMFS value):

Grade Basis Interpretation
A: Empirical 5+ observed events in measurement window provide direct tail observation Measurement grounded in observed outcomes; highest confidence
B: Hybrid 1-4 observed events plus Monte Carlo simulation with Bayesian updating Blends observation and simulation; moderate confidence
C: Simulated Zero observed events; derived entirely from Monte Carlo simulation Modeled, not observed; treat as indicative, not validated

A DMFS of any value at Grade C is a simulation that agrees with itself. It tells you the model is internally consistent but not that it matches reality. Grade C scores should trigger proactive stress-testing against historical near-miss events and cross-asset analogues rather than acceptance as calibrated.

Proposed formula, pending validation pilots.

At a Glance

Unit
Ratio
Trend Direction
Toward 1.0
Reporting Frequency
Per capacity test interval
Category
Model Calibration and Prediction Trustworthiness
← Weather Deferral Accumulation Index (WDAI)Forecast Horizon Decision Alignment (FHDA) →

Assess Your Organization

The AI-GPS Assessment evaluates your AI governance maturity and recommends KPIs like this one for your implementation roadmap.

Take the Free AI-GPS Assessment