Power Continuity (PWC)
Power to the vault, the cryo bank, and the cores. The one utility nothing critical can survive losing.
Formula
Thresholds & Bands
| Band | Range | State |
|---|---|---|
| Stable | ≥ 99.99 | ok |
| On backup | 99-99.99 | warn |
| At risk | < 99 | crit |
Why This Metric Matters
Power Continuity is the non-negotiable substrate upon which every other PANOPTICON capability depends. Loss of power to BCL-rated containment zones, cryogenic storage vaults, or PANACEA inference cores does not merely degrade operations -- it triggers irreversible failures including biological sample loss, containment breaches, and complete surveillance blackout. Even momentary power interruptions to critical zones can invalidate cold-chain integrity certifications and force costly re-qualification cycles. The Directorate classifies any PWC excursion below 99% as a Continuity-Level event requiring immediate escalation to the Office of the Director.
Threshold Justification
The 99.99% stable-band floor translates to a maximum allowable downtime of approximately 52 seconds per year for critical-zone power, a figure derived from the thermal runaway window of the cryogenic vault systems and the PANACEA core checkpoint-recovery interval. The 99% critical threshold represents the boundary beyond which cascading failures -- including Containment Integrity degradation and Mesh Latency spikes -- have been observed in historical incident data.
Historical Context
The metric was introduced following the Q1 2025 Grid Incident, in which a 14-minute primary-power interruption to the East Wing resulted in partial cryogenic thaw and a 6-hour PANACEA core restart cycle. Prior to that event, power uptime was monitored informally by Facilities without Directorate-level visibility. Post-incident, redundant generation capacity was doubled and the metric was elevated to Tier-1 dashboard status, where it has maintained 99.995% or better for the subsequent four quarters.
Collection Method
Power state is monitored continuously by dedicated metering hardware at every distribution panel serving a critical zone, reporting to the Synaptic Data Fabric at 1-second resolution. Uptime is calculated as a weighted average across all monitored panels, with weights assigned according to zone BCL classification -- BCL-4 zones carry the highest weight. Backup generator and UPS state transitions are logged independently and reconciled against primary-feed telemetry to detect unreported switchover events.
Known Failure Modes
The most significant failure mode is silent UPS degradation, in which battery capacity has deteriorated but the system continues reporting nominal status until an actual transfer event reveals insufficient holdover time. Metering hardware itself can fail closed, continuing to report power-present status during a genuine outage and masking the event from the Synaptic Data Fabric. During planned maintenance windows, coordinated switchovers between primary and backup feeds can produce sub-second gaps that individually fall below reporting resolution but cumulatively erode the uptime figure over long observation periods.