The $3.8 Million Bolt: How One Undertorqued Fastener Destroyed a Ferry Engine

On April 22, 2021, the Washington State Ferries vessel Wenatchee was on a sea trial in Puget Sound near Bainbridge Island, Washington. She had just come out of a months-long drydock overhaul — two of her four main diesel engines completely rebuilt by experienced factory technicians, everything checked and logged. By all accounts, she was ready to carry passengers again.

At 1:31 p.m., running at 100 percent power, a crew member spotted white smoke. Then "red glowing items" started popping out of the engine.

Within three minutes, the crew had closed every watertight door, shut off ventilation, cut the fuel. The fire self-extinguished. All 13 crew members were safe. But when engineers entered the engine room, they found wreckage: a connecting rod and half of a piston scattered across the floor. The No. 3 main engine was destroyed. So was the adjacent No. 4.

Total damage: $3,790,000. The Wenatchee stayed out of service for nine more months, worsening crew shortages and canceling ferry runs across Puget Sound.

The National Transportation Safety Board investigated and, in March 2022, released Marine Investigation Report 22/06 with a clear, confirmed finding: one undertorqued bolt caused everything.

(Sources: NTSB Marine Investigation Report 22/06 | The Seattle Times | The Maritime Executive)

What the NTSB Found

The failure traced back to one specific fastener: a lower basket bolt on the connecting rod assembly attached to the No. 6 crank pin. This bolt — one of several that clamp the connecting rod's bearing cap around the crankshaft journal — was not tightened to the manufacturer's required torque specification during the November 2020 overhaul.

The result was a cascading failure that unfolded in stages.

The undertorqued bolt gradually loosened as the engine ran. As it backed off, the connecting rod assembly lost its tight fit around the crankshaft. That precise fit isn't just structural — it creates the oil film boundary that allows the bearing to rotate without metal-on-metal contact. Lose the fit, lose the oil film.

The bearing started running dry. Friction and heat spiked. The bearing failed. The connecting rod — no longer properly anchored to the crankshaft — subjected the remaining bolts in the set to forces they were never designed to handle alone. They deformed and fractured from tensile overload, one after another. The connecting rod for cylinder 14 broke free entirely and was ejected through an inspection cover with enough force to punch through it. Crankcase gases rushed out. Fire started.

Forensic analysis confirmed it. One intact bolt was recovered from the oil pan; its fastening nut was found separately. It had simply backed off. Every other bolt in the set had been destroyed by the load it absorbed when that first one let go.

What the Correct Approach Looks Like

The Wenatchee's overhaul was performed by experienced, factory-certified technicians. This isn't a story about cutting corners or using cheap parts. It's a story about the critical gap between "feels tight" and is tight.

Connecting rod bolts in large diesel engines like the EMD 710 series carry enormous specified torque values for a reason. The connecting rod-to-piston-pin bolt alone specifies 450 lb-ft (610 Nm). These numbers aren't suggestions — they are the precise preload required to stretch the bolt enough to maintain clamping force through the thermal cycling, vibration, and mechanical loading of engine operation. Miss the spec, and you don't get a slightly weaker connection. You get a connection that will fail. The only open question is when.

The NTSB's recommended corrective action was straightforward: use a calibrated torque wrench, follow the manufacturer's required tightening procedure and specified torque values, and verify completion before closing up.

"Calibrated" is the key word. A torque wrench that has never been checked against a reference standard — or one that has been dropped, overloaded, or stored under tension — can read incorrectly by 20, 30, even 40 percent. At a spec of 450 lb-ft, a 25 percent error means the bolt is torqued to 338 lb-ft. That gap is the difference between an engine that runs for another decade and one that tears itself apart.

In multi-step torque sequences (common in engine assembly), each step serves a specific purpose: seating the component, distributing load evenly, achieving final preload in a controlled sequence. Skip a step or apply the final value in a single pass, and the result can look like full torque on paper while leaving clamping force dangerously uneven across the joint.

The NTSB also noted that this was not an isolated incident. Their investigators had examined several similar marine engine casualties and found the same root cause in each: failure to tighten fasteners to manufacturer-specified torque values.

The Broader Lesson

The Wenatchee story is compelling because nothing about it was exotic. The engine was a common, reliable platform. The technicians were experienced. The fasteners themselves were appropriate for the application — correct grade, correct size, correct type. The failure wasn't in the fastener selection at all. It was in the installation.

This matters for anyone who works with fasteners in mechanical assemblies. A bolt's specified torque value tells you how much preload the joint needs to function as designed. That spec exists because an engineer calculated exactly how much clamping force the bolt needs to generate under load, accounting for the material properties of the bolt, the stiffness of the clamped members, the operating temperature, the vibration environment, and the expected dynamic loads.

When you don't hit that spec, you're not building the joint the engineer designed. You're building a different joint — one whose actual behavior under load was never analyzed or validated.

And in the case of a connecting rod assembly running at 900 rpm inside a 16-cylinder diesel engine, that difference costs $3.8 million and nine months of ferry service.

The Right Fastener — Properly Installed

Fastener failures that make headlines tend to be dramatic: bridges, ferries, aircraft. But the underlying lesson applies whether you're reassembling heavy equipment, torquing a flange connection in a plant, or reinstalling any component that's been through a maintenance cycle.

The right fastener for the job isn't just about the correct grade, material, and size. It's about the correct installation — the right torque, the right sequence, the right tools, and the discipline to verify each one before you close it up.

One bolt at the wrong preload is a failure waiting to happen. It just picks its own timing.