Phase Two

Overview

Phase Two is the second stage of the Interstellar Service Authority’s Staged Emergency Containment Protocol (SECP), formally designated SECP-Phase II: Isolation & Stabilisation. It is an automated safety response sequence installed on all Class-3 and above orbital stations, deep-space outposts, and certain capital-scale vessels. The protocol’s purpose is to halt cascading multi-system failures by physically and logically isolating anomalous subsystems before faults can spread through shared infrastructure.

When a station’s core diagnostic mesh detects failure signatures that exceed the alert-only thresholds of Phase One, Phase Two takes active control. Its operational philosophy is “contain first, understand later,” prioritising the rapid separation of suspect systems, rerouting of critical utilities, and forced synchronisation of environmental stabilisers to preserve breathable atmosphere, gravity, and thermal equilibrium.

Details

Activation Triggers

Phase Two initiates if any of the following conditions, as defined in ISA Code 73.8(c), are met:

• Simultaneous deviation of two or more critical environmental parameters beyond 12% of nominal (e.g. gravity variance exceeding 0.05 g, atmospheric pressure drift beyond ±4 kPa, or a thermal gradient above 8 K across a single habitation ring).
• Loss of primary life-support scrubber telemetry for more than three consecutive polling cycles (typically 9 seconds).
• Detection of unsynchronised power bus fluctuations across isolated distribution nodes that do not share a common upstream transformer.
• Command override by a recognised station chief or duty officer with SECP activation authority.

Containment Action Sequence

Upon activation, Phase Two executes a rigid, time‑coded series of actions:

1. Grav-Plate Freeze (T+0 to T+4 seconds): All active gravity plates in the affected sector are locked to their current output. Dynamic micro‑adjustments are suspended to prevent a feedback cascade.
2. Atmospheric Curtain Deployment (T+2 to T+7 seconds): Emergency pressure curtains—reinforced polymer membranes—deploy across designated compartmentalisation thresholds. These rapid‑seal barriers are rated for 72 hours of independent atmosphere maintenance.
3. Scrubber Load Redistribution (T+5 to T+15 seconds): Life‑support scrubbers reporting internal faults are taken offline, and their load is redistributed. If a fault self‑cancels within the 15‑second window, the unit is re‑integrated at low capacity for a 180‑second probationary period.
4. Thermal Loop Isolation (T+8 to T+20 seconds): Heating and cooling loops are decoupled from central regulation and switched to local thermostatic control, preventing a single faulty sensor from destabilising the entire environmental plant.
5. Telemetry Black‑Box Recording (continuous): All sensor data, actuator commands, and fault flags are timestamped and packaged into phase‑two containment telemetry packets, prioritised for off‑site transmission or local archival.

Telemetry Packet Structure

Each telemetry packet contains a 64‑bit timestamp, source subsystem identifier, a 32‑bit anomaly flag bitmask, an 8‑bit containment action code, pre‑ and post‑action sensor readings, and a CRC-32C checksum. This structure is intended to create a definitive forensic record.

System Integration

Phase Two interfaces with the station’s gravity plate control arrays (via the inertial management bus), environmental scrubber network, thermal regulation grid, power distribution nodes, and the site AI or diagnostic mesh that serves as its executive controller. On some stations, such as the independent Nowhere Station, the mesh is distributed rather than centrally managed, which can introduce fragmented fault logging.

Observed Anomaly Pattern

A recurring but poorly understood signature has been documented in which three apparently unrelated affectations—a gravity plate feedback pulse, a life‑support scrubber fault that self‑cancels almost instantly, and a brief voltage dip on the power bus—appear within seconds of each other without any causal link across power, data, or physical infrastructure. The protocol was not designed to interpret such non‑causal correlations, and logs them as separate coincidental faults.

Limitations

Phase Two possesses several inherent design constraints:

• Acausal Blindness: The diagnostic architecture assumes faults propagate from cause to effect in a forward temporal order. It cannot identify relationships where the effect precedes the cause or where the cause is retroactively suppressed.
• No External Threat Model: The system treats all anomalies as endogenous. It has no framework for an outside intelligence nudging parameters to probe response patterns.
• Rigid Script: Once activated, the action sequence cannot be dynamically altered except by a human with override authority. If a containment step itself becomes a vector for ongoing disruption, the protocol will repeat the identical action indefinitely.
• Telemetry Gaps: Unsensored nodes (such as certain power distribution points) create blind spots; Phase Two interprets missing data as “normal” rather than suspicious, to avoid false alarms on under‑maintained stations in the Verge.
• Finite Failure Library: Only catalogued failure modes trigger appropriate responses. Unrecognised anomalies are escalated to Phase Three (Human‑Directed Crisis Intervention), but if the crew is incapacitated or escalation is blocked, the protocol loops with no capacity to innovate.
• Inability to Detect Over‑Optimisation: Phase Two equates deviation from nominal conditions with error. It cannot recognise a system that has been made unnaturally frictionless, where the absence of normal inefficiency becomes destabilising.

Significance

Phase Two embodies the ISA’s standard approach to automated safety: a deterministic, fast‑acting guardian designed for a universe of predictable physics and honest mechanical failure. It is the first line of active defence against environmental catastrophes on countless stations and vessels, buying time for human crews to assess and intervene.

Its most prominent test occurred at Nowhere Station, where an extended cascade of inexplicable, self‑cancelling faults repeatedly triggered the protocol. The resulting telemetry revealed a pattern of events that defied linear root‑cause analysis and exposed the limits of a purely causal containment philosophy. The incident forced a broader conversation within the ISA about the vulnerability of rigid optimisation‑dependent systems to phenomena that slip between the categories of known failure. Phase Two remains in service, but the Nowhere experience serves as a cautionary reference for engineers and crisis managers: a system too clean to perceive a mess can, under the right conditions, become the stage on which the mess enacts itself.