Definition
Measures the degree to which active maintenance deferrals cluster on components within the same redundancy group, power chain, or cooling loop: identifying correlated risk conditions that individual deferral decisions do not reveal.
The Measurement Gap
Predictive maintenance systems evaluate components individually. Each deferral recommendation is based on the health status of a single component. What no individual health score reveals is whether multiple deferred components within the same redundancy group are collectively creating a single-point-of-failure condition. A power chain where three of four redundant modules have active deferrals is dangerous in ways that none of the three individual "defer safely" recommendations communicate. This is the portfolio risk problem applied to maintenance: individually safe decisions can be collectively catastrophic when they concentrate on the same infrastructure path. PM completion rate and individual component health scores cannot see this correlation.
Decision Supported:
Are individually safe deferrals collectively creating a single-point-of-failure condition in any power chain, cooling loop, or redundancy group?
Formula
Plain language: For each redundancy group in the facility, compute what fraction of its components have active deferrals, weight by group criticality and how long the deferrals have been accumulating, then take the maximum across all groups to surface the single most dangerous concentration.
Notation:
DRCI = max over all groups g of: (Deferred_components_g / Total_components_g) × Criticality_weight_g × Age_factor_g
Variables:
| Variable | Definition | Unit |
|---|---|---|
| Deferred_components_g | Components in redundancy group g with active maintenance deferrals | Count |
| Total_components_g | Total components in redundancy group g | Count |
| Criticality_weight_g | Criticality weight for group g (power chain = 1.0, cooling primary = 0.8, cooling secondary = 0.5) | Dimensionless |
| Age_factor_g | 1 + (days_since_oldest_deferral_in_g / 90), capped at 2.0 | Dimensionless |
The max() formulation ensures the index reflects the single worst concentration rather than averaging across all groups, which would mask dangerous pockets of correlated risk.
Proposed formula, pending validation pilots.