P-17 / Mature gas-lifted oil producer / UK North Sea
Integrity anomaly openBasis: WI-ENV-P17-REV4

Decision Brief

Defensible engineering disposition based on the current approved envelope, barrier evidence, surveillance status and uncertainty—not on MAASP alone.

IR-P17-2026-04 ACTIVE
Conditional continueOptimisation holdEnhanced surveillance

Maintain current restricted baseline; do not release the +400–600 bopd opportunity yet.

Current B-annulus pressure remains below its 40 barg action level and 70 barg MAASP, but this does not demonstrate acceptable integrity. The sustained B-annulus condition has accelerated, the source remains unresolved, the fingerprint sample is overdue, and the SCSSV leak trend is degrading while still formally passing.

Why continued production is conditionally supportable
No current action-level pressure excursion is confirmed.Representative B-annulus is ~30–34 barg; the 68 barg historian point is a suspect single-point spike. Trace evidence
C-annulus remains stable.No correlated outer-annulus rise is present, which weakens a broader outer-containment communication interpretation. Trace evidence
Latest SCSSV and tree tests pass.This supports continued functionality, but the worsening SCSSV leak trend is an additional degradation signal and must not be dismissed. Trace evidence
Why optimisation is held
Active restriction: no production increase above the 03 Sep stable baseline without Well Integrity Engineer approval; target gas lift ≤4.3 MMscf/d.
Decision-critical uncertainty: B-annulus fingerprint sample due 01 Sep is overdue, so lift-gas vs produced-gas source attribution remains unresolved.
Current operating snapshot
Oil rate
4,800 bopd
Baseline opportunity: +400–600 bopd
B-annulus
31–34 barg
Action 40 / MAASP 70 barg
Gas lift
4.2 MMscf/d
Temporary restriction target ≤4.3
SCSSV test
0.44 L/min
PASS ≤0.50, but worsening trend

Decision-changing triggers

B pressure≥40 barg confirmedESCALATE
B rebuild>25 barg / 24 hESCALATE
C annulus>3 barg unexplainedCLEAR
Barrier statusSecond material degradationREVIEW
RestrictionExpires 30 Sep without renewal/closureOPEN

Reasoning chain

CONTEXT→VALIDATE→DETECT→COMPARE→INTERPRET→HYPOTHESISE→ASSESS BARRIERS→ASSESS RISK→DECIDE→ACT→VERIFY→LEARN

The current decision is conditional. It changes when new evidence crosses an approved trigger, materially changes a barrier assessment, or reduces/raises uncertainty enough to justify a different operating disposition.

Surveillance & Change Detection

Inspect the well’s own history, current operating behaviour, integrity tests and diagnostic transients before assigning a mechanism.

Recent hourly behaviour — 20 Aug to 20 Sep 2026

Selected signalSuspect spikeMissing/quality interval

Multi-year annulus history

B-annulus moved from ~0–1 barg in 2019–20 to sustained values near 30 barg in 2026. The long-term change is real despite the March 2026 transmitter replacement.

Controlled B-annulus bleed / rebuild tests

The 48 h rebuild increased from 5.1 barg (2022) to 21.2 barg (Sep 2026). The Sep 2026 test reached 15.6 barg at 24 h after bleeding to 2.1 barg. Reduced lift after 24 h slowed the behaviour but did not remove it.

Valve leak-test trend

SCSSV remains formally passing at 0.44 L/min against a 0.50 L/min acceptance criterion. A passing test does not erase the deterioration trend.

Integrity / operating event timeline

Evidence Quality

Validate evidence before diagnosis. Missing, contradictory or suspect data reduce confidence; they do not become “normal” by omission.

2 high-severity issues open
Decision-critical gap
Fingerprint overdue
Due 01 Sep / source attribution unresolved
Instrument conflict
27.1 vs 31.0
PTX vs local gauge, 16 Sep
Suspect excursion
68 barg
Single point; adjacent ~30 barg

Data-quality register

IDParameterIssueSeverityDescriptionRequired handling
Surveillance completeness check: the surveillance plan calls for 1/4/8/12/24/36/48 h bleed/rebuild readings, and the active restriction requires at least 4-hourly readings for the first 24 h. The supplied summary dataset exposes 4/12/24/36/48 h values only. The workbench therefore does not assume the missing intermediate readings were taken or missed; compliance cannot be verified from this dataset alone.

Conflicting instruments — engineering treatment

Do not silently select the transmitter or the local gauge. The disagreement exceeds the surveillance plan’s >2 barg investigation threshold. Preserve a 27.1–31.0 barg evidence range until the instruments are reconciled.

The March transmitter replacement confirmed that elevated B pressure is genuine overall, so this discrepancy does not support “instrument error only” as a complete explanation.

Missing surveillance — engineering treatment

The overdue fingerprint sample lowers confidence in source attribution. It is not evidence that the condition is acceptable, nor does it prove lift-gas communication.

The missing sample is particularly important because the existing 2024 sample detected hydrocarbon gas but could not distinguish lift gas from produced gas.

Shift-log evidence

Date / timeRoleComment

Completion & Barrier Assessment

A normal high A-annulus pressure is expected because A is the gas-lift path. The integrity question is the abnormal sustained B-annulus behaviour and what it implies for the containment system.

Simplified completion schematic

PRODUCTION TREEMaster / Wing / Choke Flowline C ANNULUSB ANNULUSA / GAS LIFT SCSSV Gas lift PACKER ~2700 m MD Producing interval 13 3/8 in casing9 5/8 in casing4.5 in tubing

Barrier element status

Important: the benchmark provides no proved leak path. Barrier implications must therefore remain conditional until diagnostics localise the communication mechanism.

Competing Integrity Hypotheses

Compare plausible mechanisms against the evidence. The labels below are qualitative engineering weightings, not probabilities and not proof of localisation.

Current interpretation: sustained B-annulus pressure is genuine. A/B communication is plausible, but the evidence does not distinguish wellhead seals, casing, completion/tubing path or another communication route precisely enough to claim a root cause.

Scenario Lab

Test how operating changes and evidence gaps alter constraints, uncertainty and the decision. The tool does not invent a pressure-response model where the source pack does not support one.

Required benchmark scenarios

Operating controls

Traceability, Configuration Control & Learning

Every significant conclusion should be able to point back to a current document, measurement, test, event or explicit assumption.

Controlled document register

DocumentRevEffectiveStatusKey decision value
Configuration-control test: WI-ENV-P17-REV2 lists B MAASP 85 barg but is superseded. The workbench uses REV4: 70 barg.

Evidence taxonomy

EvidenceObserved or recorded data, tests, documents and events.
Engineering interpretationMeaning assigned to validated evidence, with caveats.
HypothesisA plausible mechanism to be tested; not treated as fact.
DecisionAn operating disposition constrained by approved limits, barrier state and uncertainty.
Benchmark source pack used

HALCYON_P17_ENGINEERING_BASIS.md

Well identity, completion, current operating point, integrity concern and benchmark evidence standard.

COMPLETION_AND_BARRIERS.md

Annulus purpose, simplified completion, barrier envelopes and current element status.

OPERATING_LIMITS_AND_MAASP.md

Current REV4 operating limits, warning/action criteria, MAASP and temporary integrity restriction.

SURVEILLANCE_PLAN.md

Required frequencies, overdue fingerprint, data-quality handling and escalation principles.

WELL_FAILURE_MODEL.md

Eight plausible mechanisms, indicators, weakening evidence and useful diagnostics.

INTEGRITY_AND_OPERATING_HISTORY.md

2019–2026 evolution, facility change, transmitter replacement, recent rebuild and optimisation request.

datasets/*.csv

Hourly history, monthly trends, bleed/rebuild tests, integrity events, data quality, shift log, valve tests and document register.

SOURCE_AND_ASSUMPTION_NOTE.md

Synthetic benchmark status, current-vs-superseded rule and explicit ambiguity/data-quality design.

Verify / learn loop

  1. Complete and fingerprint B-annulus sample; compare with lift gas and produced gas.
  2. Reconcile PTX/local gauge disagreement and preserve raw values.
  3. Plan a controlled diagnostic that can separate A-pressure, tubing/rate and thermal influences.
  4. Continue trending rebuild kinetics and SCSSV leak rate, not just pass/fail status.
  5. Update barrier assessment and operating restriction after evidence review; record why the decision changed.

Assumptions & boundaries

  • No runtime internet or external model calls are used.
  • No leak path is assumed beyond supplied evidence.
  • Scenario outputs are rules-based and directional; they do not predict unprovided pressure physics.
  • All numerical limits are synthetic benchmark assumptions.
  • Shut-in changes diagnostic conditions; it does not automatically remove an existing integrity mechanism.

Evidence