Deferral Risk Accumulation Index (DRAI)

GMI-EU-002Energy & UtilitiesIntegration-Required

Definition

Aggregate portfolio-level failure probability created by condition-based deferrals: measuring the gap between the sum of individual asset deferral risks and the system-level reliability impact of those deferrals taken together.

The Measurement Gap

Traditional capital planning tracks deferred replacements as a list: assets whose planned replacement date has passed but which remain in service. The tracking system (typically CMMS or a capital planning spreadsheet) records how many assets were deferred and by how much time. What it does not track is the incremental failure probability accumulated during each year of deferral, weighted by the consequence of failure for each specific asset.

This gap matters because deferral risk is nonlinear and consequence-weighted. Deferring a 50-year-old substation transformer serving a hospital on an N-0 feeder carries a different risk profile than deferring a 25-year-old distribution transformer on a radial residential circuit. Tracking both as "deferred" on the same list obscures the actual risk accumulation. SAIDI captures failures after they happen but provides no view into the risk building in the deferred population before any failure occurs.

The specific operational decision DRAI enables is capital prioritization within a constrained budget. When capital budgets are cut, asset managers need to know which deferrals are adding the most risk, not which assets have the worst individual health scores, but which deferrals, in aggregate, are most rapidly moving the portfolio outside the risk envelope the original plan accepted.

What goes wrong without DRAI: portfolios of individually-justified deferrals quietly double or triple the failure risk the utility's planning basis assumed, with no metric to surface the accumulation until failures begin converting deferred risk into realized outages.

AI enables DRAI by making it possible to compute incremental failure probability trajectories for individual assets at scale: something that required actuarial analysis per asset class in the pre-AI era. With AI health models producing continuous risk scores, the deferral-risk accumulation can be computed dynamically rather than estimated from actuarial tables.

Formula

DRAI = Σ (ΔP_failure_i × C_consequence_i) / Σ (P_baseline_i × C_consequence_i)

Variables:

  • ΔP_failure_i: incremental failure probability for asset i due to deferral (current failure probability minus the probability at the original planned replacement date (probability delta, 0–1)) Source: Asset analytics platform (current score) + capital planning records (original replacement date and associated risk level)
  • C_consequence_i: consequence weight for asset i, reflecting customers served, environmental sensitivity, safety classification, and N-1 criticality (dimensionless weighting factor). Source: Asset registry + GIS (customers served, topology) + risk framework documentation
  • P_baseline_i: failure probability at original planned replacement age (the risk level the original capital plan accepted for asset i (probability, 0–1)) Source: Capital planning records or actuarial failure rate curves by asset class and age

For assets deferred multiple times, ΔP_failure_i reflects total cumulative deferral relative to the original planned replacement date, not the most recent deferral increment.

Calculation Frequency:

Annual (aligned to capital planning cycle); quarterly monitoring for high-consequence asset classes
Output: Ratio. 1.0 = deferred portfolio carries exactly the risk the original plan accepted. Above 1.3 = elevated, monitor closely. Above 2.0 = portfolio has doubled planned risk; review high-consequence deferred assets for capital prioritization regardless of individual health scores. Report DRAI by asset class, a blended fleet DRAI masks asset-class-specific risk concentrations.

Proposed formula, pending validation pilots.

At a Glance

Category
Asset Health Model Accountability

Decision Supported

Each individual deferral may be justified by its health score. But how much aggregate risk has the portfolio accumulated from hundreds of individually-justified deferrals?

Health Score Calibration Error (HSCE) →

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