Engineering / InfrastructureLong (years)Detectability: Moderate

Design-life assumptions for subsea components

A subsea system was maintained on the basis that key components would meet a 25-year design life with standard maintenance.

“Silence is not stability.”

Back to archiveDon’t end up here

Decision summary

Year
2015
Failure mode
Design-life as a narrative: the design basis outlived the operating reality.
Silent failure window
2–4 years: performance remained “within spec” until failures clustered.

The original logic

Vendor qualification data, past field performance, and a conservative design basis suggested low probability of early-life failures; spares were expensive and logistics were non-trivial.

Key assumptions

  • Duty cycles would match the design basis (pressure/temperature cycles, actuation frequency).
    Confidence at decision: Medium
    Expected lifetime: 3–5 years
  • Field conditions would remain within corrosion/erosion allowances with the existing chemical program.
    Confidence at decision: High
    Expected lifetime: 2–3 years
  • Vendor qualification data remained representative of installed population (no hidden lot variability).
    Confidence at decision: Medium
    Expected lifetime: 5+ years

What changed

Operations gradually increased cycling due to market-driven dispatch patterns. A supplier change introduced subtle material variability, and the chemical program drifted as teams optimized for cost and availability.

Outcome

A cluster of component failures required an unplanned intervention campaign and re-qualification work, with a multi-quarter production impact.

Early warning signals (missed)

  • Rising actuation counts (telemetry) beyond the original duty basis
  • A creeping increase in “minor” maintenance tasks on related equipment
  • Subtle shifts in failure modes across vendor lots in warranty/returns data

How AssureAI would have helped

  • Make duty-cycle alignment an owned assumption with automated telemetry checks.
  • Tie procurement/vendor-lot changes into the same decision record and require explicit re-validation.
  • Track “near-miss maintenance” frequency as drift evidence.

Non-obvious lessons

  • Design life is conditional; operations quietly rewrite the conditions.
  • Supply chain changes are technical changes.
  • A long horizon needs short-horizon validation checkpoints.
Design-life assumptions for subsea components — Decision Graveyard