GREYHAVEN / UK NORTH SEA / SYNTHETIC DATA

Production Loss Atlas

90 days of production. One governing constraint at a time.

01 APR — 29 JUN 2026

Production across the hub

Daily averages · kboe/d. All timestamps are synthetic local platform time.
Technical PotentialOutlook (dashed)ActualPhysical loss area

Loss attribution

Volume in selected dates · kboe · click a constraint to inspect its events

Boundary assumptions

Default accounting uses Loss_Event_Truth. Alternatives test operational timestamp uncertainty; they do not overwrite delivered production or claim a different physical history.

Later start leaves the early deficit unassigned. Later handoff moves the existing 6.5 kboe/d deficit from export to compression; it never sums both. 13:30 is a deliberately conservative attribution sensitivity, contradicted by HIST-012 as the physical handoff.

Calculation and source policy

Each half-open interval [start, end) has one governing constraint. Volume = loss rate × elapsed hours ÷ 24. Intervals are split at midnight, range boundaries and assumption boundaries. Daily physical loss = (Technical Potential − Actual) × 1 day. Outlook shortfall = (Outlook − Actual) × 1 day.

Truth intervals control physical accounting; historian anchors event sequence; register and logs are reconciled, not blindly summed. Child records are diagnostic only. Daily averages do not provide sub-day measured rates. Replay shows a benchmark reconstruction, not a continuous historian trace. No independent well-level allocation is inferred from the bad GHV-04 meter.

Recurring losses / investigation priorities

Selected dates · volume ranking. Episode counts merge the three 18 May phases.
K-201B recurrence after service is the key effectiveness test. Export pressure and export pump losses share a path, but are not assumed to share a root cause.

Register reconciliation

Inspect register entries in this date range

Event reconstruction

Select an episode to synchronise process, evidence, attribution and actions.

· 18 MAY

Evidence tracks

Initiator, mechanism, underlying cause

Evidence ≠ certainty

Action ownership and effectiveness

Source record explorer

Full retrospective source snapshot. These records may contain facts learned after the replay cursor. Workbook values are preserved; derived views are identified in the atlas.