Oliver Hazard Perry: When Shore-Power Troubleshooting Became A Flooding Incident

Use this review method when a yacht has an abnormal shore-power event and the response changes how the vessel is being operated.

Applicability is strongest when shore-power problems lead to a temporary operating mode. That may mean generator running, reduced-crew monitoring, machinery-space risk, or uncertainty about which alarms will reach the duty crew.

This article uses a sailing school vessel incident to teach a yacht near-miss review method. The vessel type is different from a superyacht, but the pattern is familiar. Alongside power trouble leads to a generator running at the dock. Overnight monitoring is limited. A technical defect then escalates quietly.

Shore-power trouble is often treated as a nuisance. The yacht may lose a marina pedestal, see a converter alarm, or run a generator "just for tonight" while someone looks into the dock supply.

The NTSB report on Oliver Hazard Perry shows why that mindset is risky. The flooding was not caused by a wrong-voltage shore connection, but shore-power troubleshooting was part of the operating context. Because shore breakers were tripping, the vessel was running its port generator while moored. That kept the vessel powered. It also meant seawater was being circulated through a generator cooling pipe overnight. When that pipe failed, the problem moved from electrical inconvenience to engine-room flooding.

For yachts, the lesson is not "never run a generator at the dock." The lesson is that a workaround changes the yacht's risk profile. Management may still think of the issue as "bad marina power," while the technical team has actually put the vessel back into an active machinery mode.

Oliver Hazard Perry was a 128-foot, steel-hulled, three-masted sailing school vessel. It was certified under US Coast Guard rules and could accommodate people overnight. Its machinery spaces were on the lower deck, below accommodation spaces.

The vessel had two diesel-driven generators and a smaller emergency generator. It also had high-water bilge sensors in the lower-deck machinery spaces. The main machinery alarm panel was in the engine room, while the fire alarm panel was one deck above in the ship's office.

That layout became important because only one crewmember, the bosun, was onboard overnight. His stateroom did not have an alarm panel. The vessel had technical alarms, but the night-time manning arrangement did not guarantee that the person onboard would hear the alarm that mattered.

On 10 May 2025, Oliver Hazard Perry was docked at Fort Adams State Park in Newport, Rhode Island. New shore-power breakers had recently been installed on the pier, and the crew had been troubleshooting the vessel's electrical system because those breakers were tripping.

During that troubleshooting period, the port generator was running and supplying power to the vessel. This is the point where a yacht near-miss review should slow down. The shore-power problem did not directly flood the engine room, but it changed how the vessel was being operated. Instead of lying alongside on shore supply, the vessel was relying on a seawater-cooled diesel generator through the night.

On a yacht, the equivalent might be:

• An unstable marina pedestal.
• A shore converter fault.
• A temporary yard supply.
• A decision to run hotel loads on generator until a technician arrives.

These are common choices, but they are not neutral. They should trigger fresh checks and a different watchkeeping assumption.

At about 2100, the bosun received a freshwater tank level sensor alarm. He inspected the engine room and auxiliary machine space, found everything in good order, and went to bed.

The NTSB concluded that flooding likely began around 2200, after that machinery-space round. The failed part was a seawater cooling supply pipe for the port generator. While the generator was running, its engine-driven pump circulated seawater through the cooling system. Once the pipe failed, seawater entered the engine room.

At 0530 the next morning, the fire alarm panel recorded an electrical fault alarm caused by loss of power supplying the panel. The NTSB analysis says the water was likely high enough by then to short out electrical connections and trip the generator offline. Once the generator stopped, the engine-driven seawater pump stopped too, reducing the rate of flooding.

At 0735, the bosun woke to a faint audible AC power fault alarm from the fire alarm panel in the ship's office. He noticed the vessel had no electrical power. In the dark, he went to the engine room and found seawater covering the base of the port generator. He also found about one foot of water in the auxiliary machine space. He had not reported hearing any high-level bilge alarms.

About 21,000 gallons of seawater entered the engine room and auxiliary machine space. There were no injuries and no pollution was reported, but vessel damage was estimated at $1 million.

The flooding reached major machinery and electrical equipment. Both main generators and the emergency generator were partially submerged. So were parts of the electrical and support systems needed for recovery. This is where the yacht lesson becomes practical. A dockside power workaround can become a machinery-space problem, then an alarm problem, then an emergency-power problem, then an insurance and return-to-service problem.

For a superyacht, the guest-facing symptom might still sound small: "we had power trouble overnight." The hidden consequence may be damage in the technical spaces that only becomes visible when engineers inspect the equipment properly.

The NTSB determined that the probable cause of the flooding was a failed seawater cooling supply pipe for a diesel engine-driven generator due to corrosion. The failed pipe was a threaded two-inch steel pipe at the connection point with a bronze strainer.

Threading had reduced the wall thickness of the pipe, making that section more vulnerable. The NTSB also noted that corrosion can be influenced by vibration or galvanic action between dissimilar metals. About a year before the incident, parts of the starboard generator's seawater cooling piping had been replaced due to corrosion. Based on that history, investigators considered it likely that the port generator's cooling system was also in poor condition.

The vessel had an operational cathodic protection system, but that did not prevent the local failure. For yacht engineers, this matters. Cathodic protection and bonding are helpful controls, but they do not remove the need to inspect the actual pipework.

Give extra attention to:

• Threaded sections.
• Mixed-metal interfaces.
• Vibration-prone runs.
• Sister components where one side has already shown corrosion.

The vessel had high-water bilge sensors in the engine room and auxiliary machine space. The flooding reached levels that should have been significant, but the bosun did not report hearing bilge alarms. The bilge alarm system had not been tested since July 2024. After the event, the relevant sensors could not be tested because the electrical system had been heavily damaged. The NTSB therefore could not determine whether the bilge sensors and alarm system were functional at the time.

There was another problem even if the alarms had worked: the accommodation spaces did not have an alarm panel or annunciator. The NTSB concluded it was unlikely that an audible alarm in the machinery spaces would have been capable of waking the bosun.

This is a reduced-crew monitoring lesson. If one duty crewmember is onboard overnight, the question is not whether the yacht has alarms somewhere. The question is whether the right person receives the right alarm in time, in the actual mode the yacht is operating in.

The vessel had electrically driven bilge pumps that could take suction from the engine room and auxiliary machine space, and the bosun knew the bilge system. By the time he discovered the flooding, however, the pumps were submerged and vessel power was out. Main power and emergency power could not be restored because the flooding had already damaged the spaces that needed power for recovery.

This is a simple but important point for yachts. A bilge pump on a drawing does not help if its own electrical support is already underwater. A shore-power problem can lead to a generator workaround; a generator workaround can lead to flooding; the flooding can then disable the systems needed to recover.

It would be inaccurate to say Oliver Hazard Perry was caused by shore power. It was not. The probable cause was a corroded generator seawater cooling pipe.

The reason this belongs in a shore-power near-miss review is the operating decision that sat between the shore breakers and the flooding. The vessel was running a generator because shore breakers were tripping and the crew were troubleshooting the electrical system. That choice put a seawater-cooled diesel generator into continuous service at the dock. The failure occurred in that operating state.

The practical YachtByte rule is: when shore power is abnormal, investigate both the original power issue and the workaround used to keep the vessel running.

If the workaround is generator operation, the review should confirm:

• Which generator was used.
• How long it was expected to run.
• Whether cooling water and exhaust arrangements were checked.
• How alarms would reach the duty crew.
• Who approved the temporary operating mode.

That is not bureaucracy. It is how a crew avoids treating a changed operating condition as if nothing has changed.

The First Problem Was Not The Final Event

The visible operational problem was shore breakers tripping. The serious event was engine-room flooding caused by a failed generator cooling pipe. The connection between them was the workaround: generator operation alongside.

Yacht investigations should look for that bridging decision. The first symptom may be a shore-power trip or converter alarm. The eventual loss may be something different: generator damage, flooding, electrical equipment damage, or crew injury during follow-up work.

The Workaround Needs Its Own Risk Assessment

Running a generator instead of shore power is common, but it is still a different operating mode. The yacht is no longer a passive hotel load connected to shore. It has running machinery and all the dependencies that come with it.

If the yacht is in refit or left overnight with reduced crew, that mode needs explicit approval and monitoring. The captain and manager do not need every technical detail, but they do need to know that the risk has changed.

Alarm Routing Must Match The Manning State

An alarm only works if the right person receives it in time. Reduced-crew mode is not the same as full-operating mode. Owner-off mode is not the same as charter mode. Yard mode is not the same as sea mode.

Yachts should test alarm routing in the actual condition they use. The test should prove that the duty crew member receives the alarm where they really are, not where the panel happens to be installed.

Previous Corrosion Or Repair Is Evidence

Oliver Hazard Perry had prior corrosion-related work on the starboard generator cooling piping. Investigators used that to reason about the likely condition of the port side.

On yachts, if one side of a duplicated system has failed or shown corrosion, the sister component should not be assumed healthy. This is a useful rule across machinery and electrical systems.

Repeated shore-breaker trips should be treated as a defect, not as the normal personality of a marina. The technical team should record enough information to understand the event.

At minimum, record:

• The supply condition.
• The load at the time.
• The protection that operated.
• Any temporary mode that was approved.

The aim is not to create paperwork for its own sake. It is to avoid normalising a condition that has not been understood.

If the yacht moves to generator power because shore power is unstable, the workaround should be risk-assessed as its own operating mode. A generator running overnight at the dock should not be treated as background noise.

At minimum, confirm:

• The selected generator is suitable for the planned duration.
• Cooling-water and exhaust arrangements are normal.
• Fire, bilge, and power-fail alarms reach the duty crew.
• Someone is responsible for rounds or remote monitoring.

Reduced-crew alarm annunciation needs a practical test. Stand where the duty crew member actually sleeps or works, put the yacht into the mode it actually uses alongside, and confirm the alarms can be seen or heard there. If the answer depends on someone happening to walk past a panel, the system is not doing its job.

Generator seawater cooling paths also deserve closer attention when generator-at-berth operation becomes routine. The inspection should focus on places likely to fail quietly. Threaded sections, mixed-metal joints, old repairs, and vibration-prone supports deserve more than a quick glance.

Finally, the bilge response should be checked against the failure state, not the ideal state. Ask whether the yacht could still dewater the space after the first likely electrical equipment has been reached by water. If not, the yacht needs a realistic portable-pump and access plan.

The evidence should show the starting condition, the temporary operating mode, and the return-to-service basis.

Useful records include:

• Photos of the shore connection and switchboard state.
• Electrical readings before and after correction.
• Generator run history and alarm history.
• Cooling-water observations.
• Bilge and fire alarm status.
• Crew or contractor notes explaining what changed.

If the event involved physical damage, preserve failed parts before cleaning up. Take photos early. The most useful evidence is often lost during well-intentioned cleanup.

Closeout should do more than say "shore power restored." It should show what was repaired or escalated, what was learned from the generator workaround, whether alarms were tested in reduced-crew mode, and whether any related machinery needs attention.

The final test is simple: if the same shore-power fault happens again next week, will the yacht use a safer, documented workaround?