Naval Architecture Engineering Case Studies
Technical5 Case StudiesUpdated Sep 2026
Marine design problems seldom stay in one place. A shaft seal leaks because a clearance was never recalculated after a propeller change. A hull block misses its slot because a fit-up gap was accepted on the day shift. Each of the five cases here begins with a component or process that did not behave as the paperwork promised. The analysis draws on the same four sources: design constraints, operating data, quality tolerances, and maintenance records. Read them as a set, because the disagreements between those sources are where most of the useful work sits.
Case Study 1
Propeller shaft seal failures on a coastal ferry
Pellham Reach Ferry Company, an operator of six passenger vessels on a short coastal route with a two-person engineering office ashore.
The situation
The Corran Rose ferry had been in passenger service for about four years when its stern tube seal began leaking at a rate the crew could no longer ignore. The vessel has twin screws and an oil-lubricated stern tube, and the seal is a multi-lip elastomer design fitted at the original build. In the first eighteen months the seal was replaced once. Over the following year it was replaced twice more, each time after the engine room logged rising bilge intake and a climb in stern tube oil water content. The operator asked the design office to explain why a seal rated for the duty kept failing.
The challenge
The easy explanation was the seal supplier, but the replacement parts carried valid conformity documents and the shop had fitted them correctly. The harder question concerned shaft movement. Alignment records showed the shaft drifting toward the aft bearing after each dry-dock period, and the tail shaft clearance had never been rechecked after the propeller was swapped for a larger unit. Nobody had connected the two until the third failure.
Analysis
The first pass at root cause treated the seal as a consumable with a random failure rate. That framing does not survive the records. The design basis assumed shaft deflection within a fixed envelope, but the bearing clearance table still used the mass of the original propeller. Operating data showed the vessel spending a larger share of its hours at the higher load setting, which raised the bending moment felt at the seal. Maintenance logs also recorded two shim changes at the aft bearing without a fresh alignment check. Taken together, the sources point to a geometry problem that no seal could absorb indefinitely.
Teams grounded in Naval Architecture Engineering tend to examine the tolerance stack before blaming any single part, because a shaft seal is only as good as the bore and the shaft surface it runs against. Here the bore tolerance was within specification at delivery, yet nobody had recalculated the permissible runout after the propeller change. The corrective options are a heavier aft bearing with a verified clearance, a seal designed for higher eccentricity, or a realignment with new limits written into the maintenance plan. Each option carries a different downtime and cost profile, and the comparison should rest on the operating data rather than the seal's catalog rating.
- Original deliverySeal and bearing clearance set against the original propeller mass.
- First seal replacementFace wear found; alignment not rechecked before the refit.
- Propeller changedLarger unit fitted; permissible shaft runout not recalculated.
- Second seal replacementShims added at aft bearing; bilge intake trend not escalated.
- Root cause reviewGeometry problem identified; realignment and bearing change approved.
Actions taken
- Recalculate stern tube clearance and permissible shaft runout using the larger propeller mass and the revised load profile before any further seal work is scheduled.
- Realign the shaft line at the next dry-dock period, taking laser readings in cold and warm conditions, and log the offsets against the recalculated limits.
- Fit the higher eccentricity seal only if the recalculated runout exceeds the original design margin, and only after the aft bearing clearance has been verified in writing.
- Add a rule to the maintenance plan that any shim change at the stern tube triggers a full alignment check and an oil water sample within the same shift.
Results
The yard realigned the shaft and fitted a heavier aft bearing at the next dry-dock period. The seal then ran a full operating season without a leak alarm, and bilge intake returned to its earlier baseline. Oil water content stayed below the action threshold at every monthly sample. The operator kept the original seal design, avoiding a supplier change. The shaft clearance record now goes to the engineering office after every propeller or load change. The realignment and bearing work cost about a third of the three seal replacements.
Questions for discussion
- Which of the four records would you have examined first, and what would its absence have cost the diagnosis?
- When does an original design margin stop applying after a component changes, and who should own that decision?
- What evidence would you need before accepting a supplier as the cause of repeated seal failure?
Case Study 2
Hull block welding line loses throughput as defects climb
Corvane Hull Works, a mid-sized yard building steel workboats and fishing vessels, with two assembly bays and one gantry crane.
The situation
Corvane builds its hulls in prefabricated blocks, and the block welding line is the slowest step between cutting and erection. For most of the program the line completed around fourteen blocks a week. After about three months that figure fell to nine, while the rework rate on finished seams climbed from roughly 4 percent to nearly 11 percent. Overtime was added on two shifts, and output still did not recover. The production manager asked for a process map and a clear view of where the time and the defects were coming from.
The challenge
The welding procedure had not changed, and the welders were the same people who had met the earlier targets. What had changed was the steel supply. A new plate lot arrived with a different surface condition, and the fit-up crew began accepting gaps that the design allowed but the welding procedure did not handle well. Nobody had written a gap acceptance criterion into the shop routine, so each shift applied its own judgment.
Analysis
The process map ran from plate issue through cutting, fit-up, tacking, full welding, and inspection. Cycle times were logged by station for three weeks, which showed fit-up consuming about a third of each block's hours and rising with each gap that needed a shim. Welding time stayed stable. The defect records tie most rework to seam gaps outside the procedure range, and the maintenance log for the tack welding fixtures shows two clamp replacements that were never reflected in the fixture drawings. The variation sources are incoming plate condition, fit-up judgment, and a fixture that no longer matched its drawing.
Three changes were on the table: tighten the plate receiving inspection, rewrite the gap acceptance rule, and re-survey the fixtures. The plate inspection was the most expensive and the least certain, since the new lot already met the mill certificate. Re-surveying the fixtures was cheap and could be finished inside a week, so it went first. The gap rule needed a number: the team set an acceptance limit at the procedure's qualified range, with any seam outside it held for engineering disposition. Each change carries a measurable gate, so progress can be read from the daily record rather than from opinion.
Actions taken
- Re-survey all tack welding fixtures against their current drawings and correct the drawings or the fixtures within one week, logging each deviation found.
- Write a gap acceptance rule at fit-up stating the qualified seam gap, the measurement method, and the escalation step for any seam outside that range.
- Set the acceptance criterion at ninety-five percent of fit-up checks within the qualified gap, measured daily at each station, with a review if two consecutive shifts fall short.
- Hold plate lots from the new supplier for a sample surface check before issue, and release each lot only when the sample meets the surface condition noted in the procedure.
Results
The fixture re-survey found two clamps set about a centimeter off their drawn position. Once corrected, the fit-up crew stopped needing shims on most seams. Within the following weeks the share of checks inside the qualified gap rose to the acceptance level and held there. Weekly output returned to about twelve blocks and rework fell to roughly 5 percent. The plate sample check caught one lot with a surface condition outside the procedure, which was returned before it reached the line.
Questions for discussion
- Which of the three candidate changes would you have ranked first, and what measure would have settled the ranking sooner?
- How should a shop decide which fixture drawing is authoritative when the physical fixture has drifted?
- What would a defensible gap acceptance rule need to say so that two shifts reach the same disposition on the same seam?
Case Study 3
Hazard review of a battery retrofit on a tour boat
Thornwick Brace Marine Consultants, a design office handling conversion and retrofit work for small passenger vessels, with a liaison team of three.
The situation
A tour operator wanted to replace the diesel generator on its eighteen-metre harbor boat with a lithium battery bank and a shore charging connection. The owner's plan put the battery room beneath the passenger cabin, left the existing exhaust routed past the new cells, and placed a charging cabinet on the main deck. The retrofit had already been priced and a contractor booked. The design office was asked to review the drawings before the classification submission, with the goal of having the approach agreed before the next season's bookings open.
The challenge
The drawings did not include a thermal runaway assessment, a ventilation calculation for the battery space, or a fire detection plan that matched the new arrangement. The gas detection layout still reflected the diesel boat, and the charging cabinet had no stated segregation distance from the passenger escape route. The review had to separate drawing gaps, which the designer could fix quickly, from design gaps, which needed a change to the arrangement itself.
Analysis
The hazard list was the first thing to fix. Battery thermal runaway, hydrogen off-gassing during charge, electrical shock at higher voltage, and the loss of propulsion power during a fault were all present, and none had a control assigned to it on the drawings. The design constraints set by the hull and the existing fire main limited where the battery room could go, so the room was moved aft and given its own ventilation duct. Applicable rules for the class and flag of the vessel, along with the recognized standards for lithium cell installation, set the minimum separation, the detection requirements, and the testing of the enclosure.
The documentation was the other half of the problem. The submission package had a single general arrangement drawing and no battery cell datasheet, so the thermal behavior the design depended on could not be checked. Operating data from the existing vessels covered load profile only, not cell behavior under fault. The review recommended four approval gates: a hazard register with named controls, cell and enclosure test reports, a ventilation and detection calculation, and an as-built inspection before the first charter. Each gate closes a specific piece of missing evidence rather than a general concern.
- Hazard identification List thermal, hydrogen, electrical, and propulsion hazards with a named control for each.
- Risk control Assess controls against the hull, fire main, and ventilation constraints.
- Evidence review Check cell datasheets, test reports, and calculations before any drawing is signed.
- Approval gate Pass each gate in sequence and log the sign-off against its register entry.
Actions taken
- Commission a thermal runaway assessment for the proposed cell module and enclosure, including propagation between modules and the venting path to open air.
- Relocate the charging cabinet off the main deck to a position with a documented segregation distance from the passenger escape route, and show the distance on the revised general arrangement.
- Recalculate battery room ventilation and gas detection layout for the new arrangement, using the room volume and the worst credible charge rate, then update the detection plan.
- Hold the contractor's start until the four approval gates are passed in sequence, and record each gate's sign-off against the hazard register entry it closes.
Results
The owner accepted the sequence and delayed the contractor start by roughly six weeks. The relocated charging cabinet cleared the escape route distance, and the ventilation recalculation called for a larger duct than first drawn. The cell supplier provided test reports for the enclosure, and the thermal runaway assessment identified a venting path that needed a second outlet. The first two approval gates closed without comment from the classification reviewer. The final inspection is set for the period before the first charter.
Questions for discussion
- Which hazard on the original list would you have been most likely to miss, and what would have exposed it?
- When the operator's own operating data cannot speak to a fault condition, what other evidence should carry the decision?
- How would you sequence approval gates when a contractor is already booked and the owner wants the schedule kept?
Case Study 4
Survey vessel sea trials slip on steering gear and crane
Osterhill Yard Partners, a builder of research and survey vessels with one covered hall and a project office of six engineers.
The situation
Osterhill was building a fifty-two-metre survey vessel for a research client. The contract milestone was sea trials, and the penalty clause began to apply once the trial date moved more than thirty days. Two problems arrived in the same month. The steering gear castings were delayed at a foundry because of a pattern repair, and the design office issued a revision to the deck crane foundation after a stability check. Each problem alone was recoverable. Together they pushed the trial date toward the penalty threshold. The project manager, Ines, had three recovery options on the table and four days to choose between them.
The challenge
The three options differed in cost and in what they actually recovered. Overtime on the gate fabrication could move the hull outfitting forward a few days without changing the contract. Air freight for replacement castings would cut the gear delay sharply, but it would add cost the budget could not absorb. Accepting the slip cost the least in cash but gave up the client's seasonal sea window.
Analysis
The comparison started from the schedule itself. The critical path ran through the steering gear delivery, which carried six days of float before it touched the trial date, and through the crane foundation, which needed a re-welded doubler and a load test before commissioning. Recovering time on the gear path was worth more than recovering time elsewhere, because it moved the critical path. Quality tolerances on the gear housing ruled out an unchecked substitute casting, so any alternative needed a fresh alignment check. The design constraints on the foundation, set by the crane's load chart, left little room to shorten that task.
The trade-offs were put in numbers. Overtime on gate fabrication cost about 38,000 currency units and moved the trial date by about five days, but most of that time fell on work off the critical path. Air freight cost about 61,000 units and recovered nine days on the gear path, minus one day for the alignment check. Accepting the slip cost the penalty clause, roughly 90,000 units, and recovered nothing. Per day of trial date actually protected, air freight was the most expensive option, yet it was the only one that kept the client's window.
Actions taken
- Authorize overtime on the gate fabrication and hull outfitting for a fixed period, limited to work that sits on the critical path so the spend buys schedule.
- Place the air freight order for replacement steering gear castings only after the substitute pattern is inspected by the foundry, and book the alignment check the same week.
- Re-issue the crane foundation doubler detail with the revised stability loads, and schedule the load test at the first available window so commissioning is not held behind it.
- Notify the client in writing of the revised trial date and the basis for it, so the penalty clause position is documented before the next progress meeting.
Results
The overtime and air freight were approved, and the foundry shipped the castings in the fourth week of recovery. The crane doubler passed its load test on the first attempt. The trial date moved eleven days rather than the thirty that would have triggered the penalty. The added cost came to about 99,000 currency units, and the client accepted about half of it as a change order tied to the revised crane loads. The substitute casting's alignment check found a clearance two millimeters out, corrected before trials.
Questions for discussion
- How would you test whether the five days recovered by overtime were on the critical path before approving the spend?
- When does a change order to the client make sense, and what evidence would the yard need to support one?
- What would have changed your recommendation if the penalty clause had been triggered at fifteen days rather than thirty?
Case Study 5
Cutting fuel use on an offshore support vessel fleet
Saltmarsh Offshore Services, an operator of four offshore support vessels on fixed charter schedules, with a small technical office and a dockside maintenance team.
The situation
Saltmarsh ran four support vessels on long charter contracts, and fuel was the largest variable cost after crew. Management set a target of reducing fuel burn per operating hour without reducing bollard pull, station-keeping performance, or the maintenance interval on the propulsion plant. Two vessels were also due for hull coating and deck machinery work in the same period, which gave the operator a chance to look at material use and waste alongside energy. The technical office was asked to shortlist options and compare them on lifecycle and operational measures.
The challenge
The options did not compare easily. A hull coating change saves fuel over time but needs dry-dock days, a deck machinery swap may add mass, and heat recovery depends on engine loads the vessels rarely sustain. Each option also touched something the classification surveyor checks, whether that was shaft alignment, coating thickness, or the service interval. A saving on one measure could quietly cost performance on another.
Analysis
The operating data set the starting point. Station-keeping made up most of the fleet's hours, so a saving that only showed at transit speed would be worth little. The hull coating gain depended on how the surface held up in the first months after dry-dock, and the specification fixed a thickness tolerance the shop had to hold. Maintenance records showed the propulsion plant running cleanly, so any option had to leave the service interval intact. Lifecycle measures added materials, coating waste, and dry-dock days over the remaining asset life.
The bar scores run from 1 to 10 and combine four judgments: fuel saving at the station-keeping profile, lifecycle material and waste, performance margins, and maintenance interval. The scores are the team's reading of the measured data, not a formula, so the ranking is a shortlist rather than a verdict. The deck machinery swap looked strong on fuel at first, yet it scored lowest overall because its added mass cut the bollard pull margin on one vessel. Coating and propeller restoration fit the existing dry-dock visit, so they form the core of the recommendation.
Actions taken
- Specify the hull coating to the thickness tolerance in the shop record, and schedule the application for the first dry-dock window on the lead vessel.
- Restore the propeller surfaces at the same dry-dock visit, measuring blade profile and pitch before and after so the fuel gain is traced to a recorded change.
- Run a six-month heat recovery trial on one vessel with a logged engine load profile, and proceed only if sustained loads reach the level the trial sets.
- Drop the deck machinery swap from the current scope, and re-quote it only with a mass budget that keeps bollard pull above the contract minimum on every vessel.
Results
The coating and propeller work went ahead at the first dry-dock window, and the lead vessel's fuel burn per operating hour came in under the target over the first full charter period. Maintenance records showed the service interval unchanged. The heat recovery trial is still running, and early readings suggest the sustained load threshold will be met on only one vessel. The deck machinery swap was withdrawn, which avoided a mass problem that would have surfaced at the next bollard pull test. Waste from coating removal was logged by weight and sent through the approved contractor.
Questions for discussion
- Which measure in the scoring would you weight most heavily for a fleet on long charters, and what would change if the charter pattern shifted?
- How would you separate a fuel gain that comes from the change itself from one that comes from a different load pattern in the same period?
- When a saving on one measure threatens a margin on another, what is the right point to stop and redesign rather than accept the trade?
Learning Resources
Technical CASE STUDIES
China, India, United States, Indonesia, Pakistan, Nigeria, Brazil, Bangladesh, Russia, Japan, Mexico, Philippines, Vietnam, Egypt, Turkey, Germany, United Kingdom, France, Italy, Canada, South Korea, Australia, Spain.