Definition
GDCR measures the ratio of actual generator runtime and loading (driven by AI demand response participation and controlled testing) to the runtime and loading assumptions built into the preventive maintenance schedule, detecting the hidden risk where generators enrolled in demand response programs accumulate operating hours and stress cycles at rates that far exceed the PM schedule's design basis.
The Measurement Gap
Generator preventive maintenance programs are designed around standby generator usage, typically 1–2 hours per month of loaded runtime for monthly testing, plus occasional emergency use. OEM maintenance intervals for oil changes, filter replacements, governor calibration, and cooling system service are calibrated to that duty cycle. When facilities enroll generators in AI-managed demand response programs, those generators may accumulate 10–20 hours per month of loaded runtime across multiple demand response events. The generator's PM schedule has no mechanism to detect this increased runtime: the CMMS tracks hours against a fixed PM interval, but that interval was set assuming standby duty, not semi-regular demand response participation. The generator ages at a multiple of the PM schedule's assumption while the maintenance program remains unchanged.
Decision Supported:
GDCR determines whether generator runtime under AI demand response participation has grown so far beyond the PM schedule's design basis that maintenance intervals need to be recalculated before the next PM cycle, preventing premature component failure in the asset most critical to resilience during utility outages.
Formula
Plain language: Multiply the ratio of actual runtime to the PM schedule's assumed runtime by the ratio of actual average load to the PM schedule's assumed load. A result of 1.0 means the generator is operating exactly as the PM schedule assumes. A result of 5.0 means the generator is accumulating stress at five times the rate the PM schedule was designed for.
Notation:
GDCR = (Runtime_actual / Runtime_schedule_basis) × (Load_actual_avg / Load_schedule_basis)
Variables:
| Variable | Definition | Unit |
|---|---|---|
| Runtime_actual | Total loaded operating hours per month: sum of hours the generator operated at ≥30% of rated capacity, including all demand response events, controlled tests, and emergency dispatches | Hours/month |
| Runtime_schedule_basis | Monthly runtime assumption built into the current PM schedule (typically 1–2 hours/month for standby generators; document this assumption explicitly) | Hours/month |
| Load_actual_avg | Average load percentage during all runtime hours in the period | Percentage of rated capacity |
| Load_schedule_basis | Average load percentage assumed in the PM schedule (typically 50–75% for test loads; document explicitly) | Percentage of rated capacity |
Compute GDCR per generator unit: AI demand response may exercise some generators preferentially, creating non-uniform duty accumulation across the fleet. Track loaded hours only (≥30% of rated capacity) to exclude monitoring and warm-up periods. Document the PM schedule basis assumptions explicitly; if they are not documented, that gap must be resolved before GDCR is meaningful.
Proposed formula, pending validation pilots.