Three years after delivery, someone will ask why

The phone call comes years after the project closed. A flag state investigator working a casualty wants to understand the basis for a stability calculation. A P&I club is reconstructing the decision chain behind a ballast system that failed on its fourth voyage. An owner is pursuing a warranty claim against a machinery supplier and needs the design office to confirm what performance envelope was specified during the newbuild. In each case, the question is the same: what did your office decide, when did you decide it, and who signed off? The assumption embedded in every one of these inquiries is that you have the record.

Post-delivery accountability for design offices arrives through several distinct channels, each with its own timeline and evidentiary standard. Class society periodical surveys — the five-year special survey, the intermediate survey — occasionally surface anomalies that require explanation of original design intent. Structural condition assessments on aging hulls will sometimes trace a question back to a scantling decision made at the newbuild stage. Flag state administrations investigating incidents under the ISM Code can issue requests for technical documentation that extend to design correspondence, not just approved drawings. These are not hypothetical scenarios. Any office that has worked on deepwater vessels, high-value passenger ships, or complex naval architecture projects has likely received at least one such inquiry in the past decade.

The correspondence record is the raw material of any credible response to these inquiries. Not the final approved drawings — those survive reliably, held by class and owner alike. Not the technical reports — those are usually recoverable. The correspondence: the thread in which a structural engineer explained to the owner's representative why the frame spacing was changed from the original tender specification. The email in which the chief designer documented the decision to accept a reduced safety margin on a piece of deck equipment pending further sea trial data. The reply from the flag state surveyor agreeing to a specific interpretation of a rule ambiguity. This is the material that makes the difference between a coherent explanation and a gap in the record. It is also the material most likely to be missing.

Project servers get decommissioned. This is entirely routine — the project is complete, the office moves on, the storage costs are unnecessary. What follows decommissioning varies: some offices archive to tape or compressed formats, others migrate to newer storage systems, some simply delete. The engineers who managed the correspondence are often the same engineers who eventually leave, taking institutional memory with them and, sometimes, local copies of files that were never properly archived to begin with. A senior engineer who spent two years working a complex propulsion redesign and then moved to another firm has no obligation to preserve the correspondence record. In practice, that correspondence often exists only in the email account that was deactivated when they handed in their access badge.

The evidentiary standard applied in post-delivery inquiries is not sympathetic to operational reality. A P&I club reconstructing a casualty does not accept "the project server was decommissioned in 2022" as an answer. A flag state investigation does not adjust its expectations based on how a design office manages its email infrastructure. The office is expected to produce the record. If the record cannot be produced, the absence of documentation becomes itself a fact in the investigation — one that rarely reflects well on the office. The design decisions may have been technically sound and properly agreed. Without the correspondence, there is no way to demonstrate that they were.

The gap between the duration of liability and the durability of records is specific and measurable. Classification society approval records are retained indefinitely, and post-incident investigations — flag state, P&I, or judicial — can reach back to original design documentation without a fixed limitation period. A ship built in 2018 that enters a major casualty investigation in 2031 brings with it thirteen years of potential exposure. Standard project server retention practices — where they exist at all as formal policy, rather than informal habit — rarely extend beyond five years after project close. The correspondence from the critical design review phase, the builder's specification negotiations, the class society technical queries, the owner's approval chain: all of it is likely gone before the ship completes its first decade of service.

The industry has spent considerable effort improving the quality of delivered documentation packages — the technical file that transfers to the owner at handover. This is useful. It does not address the problem. The technical file documents what was decided. The correspondence record documents why, and by whom, under what constraints, with what reservations noted and what alternatives considered and rejected. These are not the same thing. The first survives. The second is what an investigator or a court actually wants, and it is precisely what most design offices cannot reliably produce three years after the project closed.