The $3.8 Million Bolt: One Undertorqued Fastener Set a Ferry on Fire

On April 22, 2021, the M/V Wenatchee — a 460-foot Washington State Ferries vessel built to carry 202 vehicles and nearly 1,800 passengers — was running post-maintenance sea trials in Puget Sound near Bainbridge Island, Washington. The engines were being loaded up in sequence, the way the manufacturer recommends to seat new piston rings after an overhaul. Everything was going fine.

Then, at 1:31 in the afternoon, running at 100 percent power, a crew member spotted white smoke. Then "red glowing items just popping out" of an engine. Within three minutes, the crew had sealed watertight doors, killed ventilation, and cut fuel to the engine room. When the smoke cleared and engineers could safely enter, they found shattered components scattered across the floor — a connecting rod and half of a piston had blasted through an inspection cover and ignited hot pressurized crankcase gases.

No one was injured. But the damage to the vessel came to $3,790,000.

The National Transportation Safety Board investigated and published Marine Investigation Report 22/06 in March 2022. Its conclusion was blunt: the probable cause was "insufficient tightening (torqueing) of a lower basket bolt during the recent engine overhaul."

Not a defective bolt. Not the wrong bolt. A bolt that was never tightened to specification.

What Went Wrong

The Wenatchee's No. 3 main engine — an EMD 710 series diesel, one of the most widely used marine powerplants in the world — had been overhauled by experienced factory technicians during the winter maintenance period. The overhaul included new pistons, connecting rods, liners, and bearings. It was completed without any unusual reports.

During sea trials, the connecting rod assembly on the No. 6 crank pin — serving cylinders 6 and 14 — began working loose. Forensic analysis traced it back to a single lower basket bolt that had never been torqued to the manufacturer's specified value. Without the right clamping force, vibration gradually backed the nut off. Once the fastener loosened, the connecting rod lost its precise alignment with the crankshaft. The bearing lost lubrication. Heat built rapidly. The connecting rod let go with enough force to punch through an inspection cover and spray hot gas and burning oil into the engine room.

One intact bolt from the crank assembly was later recovered in the oil pan. Its nut was found separately — it had backed completely off during operation. All the other bolts in the set were found deformed and broken, destroyed by the forces that followed once the first bolt failed.

The NTSB noted this was not a one-off. Similar fastener-related engine casualties caused fires aboard the Carnival Liberty cruise ship in 2015, the Nenita bulk carrier in 2016, and the Red Dawn offshore supply vessel in 2017. Undertorqued fasteners on marine engines are a recurring problem, not a freak accident.

What the Right Installation Looks Like

The right fastener was used on the Wenatchee. Correct size, correct grade, correct material. None of that was the issue.

The failure was in the final step — torquing the bolt to the value the manufacturer specified. And this is the part of fastener installation that gets skipped, estimated, or assumed more often than any other.

Here is what correct installation of a critical fastener requires:

Use the specified torque value. Manufacturers publish torque specifications for every fastener in a critical assembly. Those numbers are engineering calculations, not guidelines. They account for thread pitch, bolt diameter, material strength, clamp load requirements, and the forces the joint will face in service. When a bolt is undertorqued, the joint doesn't have the clamping force it was designed around. When it's overtorqued, the bolt stretches past its yield point, weakens, and becomes a candidate for fatigue failure.

Use a calibrated torque wrench. A wrench that hasn't been calibrated recently can read off by enough to produce an undertorqued fastener without the technician ever knowing it. In high-stakes assemblies, calibration isn't optional.

Verify completion. The NTSB specifically recommended that maintenance personnel verify all torque requirements have been completed before signing off on a job. Checklists exist because humans skip steps, especially on long, complex overhauls where a single bolt among dozens can get overlooked.

The NTSB made one other point worth underlining: overtorquing is its own failure mode. More force is not safer. A bolt that has been stretched past yield is already compromised, whether or not it looks intact. The answer is hitting the specified value — not the maximum possible.

The Broader Lesson

The Wenatchee is a marine engine story. But the principle applies wherever fasteners hold critical assemblies together under vibration, thermal cycling, and load: in manufacturing equipment, heavy vehicles, structural frames, industrial machinery, and anywhere bolted joints carry real consequences if they let go.

"Good enough" torque is one of the most common shortcuts in maintenance and assembly work. Workers hand-tighten and move on. Torque wrenches go uncalibrated. Multi-step sequences get rushed. In most cases, nothing happens — the joint is close enough to hold. But the gap between "close enough" and "correct" is exactly where failures like this one live.

The Wenatchee's crew acted fast and no one got hurt. That is genuinely fortunate. But both the No. 3 and the adjacent No. 4 engines had to be fully rebuilt, along with additional damage throughout the engine room — all of it tracing back to a bolt that never reached its torque spec.

The Right Fastener Is More Than the Right Part

It would be easy to look at the Wenatchee incident and chalk it up to human error in a specialized trade. But the lesson is simpler than that, and it applies everywhere:

The right fastener for the job means the right grade, the right material, the right size — and installed correctly to specification. Torque is part of the specification. The fastener is not installed until it's torqued. And "snug" is not a torque value.

A bolt that looks tight and a bolt that is tight are two different things. On a ferry in Puget Sound in 2021, the difference was $3.8 million.

Source: NTSB Marine Investigation Report 22/06 — Engine Room Fire aboard Ferry Wenatchee, National Transportation Safety Board, March 2022.