All Sectors

Close All Sea Valves: What Freedom II Teaches Fishing Crews About Flood Response

🕑 5 min read words Safety • Incident

Vessel: Freedom II (CN 111) — 15.94m wooden-hulled prawn trawler, Crinan, Scotland
Date & location: 21 February 2024, Firth of Lorn, 11nm south-west of Oban, Scotland
Outcome: Total loss of vessel by foundering. No fatalities — all 4 crew rescued, 2 recovered from the water by the Oban RNLI lifeboat
Human factors: Confirmation Bias · Situational Awareness · Complacency · Communication
Source: MAIB Report 8/2026

7 decision points2 outcome paths — contain the flood / lose the vessel~9 min read + Knowledge Checker

The Pump That Wasn't Pumping

At around 0400 on 21 February 2024, the prawn trawler Freedom II left Crinan to fish the Firth of Lorn. She was 49 years old, wooden-hulled, and carried more installed bilge pump capacity than the regulations required. By mid-afternoon she was on the seabed.

At about 1000, while hauling gear, the skipper felt a strong vibration through the propeller shaft — the kind Freedom II had produced before when the propeller fouled. He pulsed the engine ahead and astern to clear it. It didn't clear. He turned for Oban at slow speed, planning running repairs.

Half an hour or so later, a deckhand went to start the deck wash pump. Nothing came out. The skipper checked the engine room and found the bilge alarms already active, water rising. He concluded the source was the propeller shaft stern seal — the failure he already had in mind from the morning's vibration. He was wrong. The real source was a roughly 1-metre length of unsupported copper alloy pipe, running from a sea cock in the hull to the deck wash pump, mounted rigidly against a flexibly-mounted engine with no flexible coupling in between. Hours of vibration that morning had fatigue-fractured it.

What Should Have Happened Next

MAIB Report 8/2026 is unambiguous on this point: MGN 165(F) instructs crews to shut the relevant sea valve immediately on discovering flooding — and if the source can't be identified, to close all sea valves until the flooding stops. Freedom II's valves had even been modified for access from above the floor plates, precisely so this could be done quickly. No valve was closed. The skipper's diagnosis of the stern seal was never revisited.

As the water rose, the response degraded further. The auxiliary generator wouldn't start because water had splashed onto its control panel. The portable salvage pump's starting cord snapped when it was repositioned. VHF voice calls to the coastguard went unanswered for roughly 45 minutes in an area of known poor signal — because the skipper hadn't activated the boat's DSC distress alert, believing (mistakenly, in the investigation's assessment) that the vessel wasn't yet in "grave and imminent danger."

The Fish Hold Nobody Checked Again

When the Oban RNLI all-weather lifeboat arrived and took Freedom II under tow, two of the four crew were transferred aboard the ALB. The skipper reported that the flooding had slowed. It hadn't — the water had simply spread into the fish hold, which had been checked once, early on, and found dry, then never checked again. The free surface of water spanning both compartments was quietly destroying the vessel's stability. The ALB carried a 47m³/hr salvage pump of its own; it wasn't transferred on arrival because conditions looked too rough, and the decision was never revisited once the vessels reached the sheltered water by the Garvellach Islands.

At 1535, Freedom II lolled suddenly to port and foundered. The skipper and one deckhand went into the water and were recovered by the lifeboat, uninjured. No new MAIB recommendations were issued — the systemic issues (pipework vibration guidance, flood drill standards) were already being addressed following two related investigations, into Opportune (MAIB 3/2026) and Guiding Light/Guiding Star (MAIB 13/2024).

Recommended Actions

No new MAIB recommendations were issued for this specific report — the systemic gaps it exposed were already being addressed through the Opportune (MAIB 3/2026) and Guiding Light/Guiding Star (MAIB 13/2024) investigations. The practical actions the report itself points every fishing crew toward:

When flooding is discovered and the source isn't immediately obvious, close all sea valves first — don't wait to confirm a diagnosis (MGN 165(F), Section 7)
Inspect any seawater suction or cooling pipe connected directly to a flexibly-mounted engine for a flexible coupling or bellows section, and for independent pipe support (MGN 190(F); SAN 29)
Test every bilge pump — including the portable salvage pump's starting cord and fuel — on a weekly basis, not just before survey
Use the DSC distress function as soon as an uncontrolled emergency is confirmed, particularly in known poor-VHF-coverage areas — don't wait for certainty that the vessel is sinking
During any flooding event, check every compartment at regular intervals, including spaces that share a bulkhead with the flooded space — a slowing rate of rise in one compartment can mean water is spreading, not receding

Human Element Analysis

Confirmation Bias

The skipper had experienced propeller-fouling vibration on Freedom II before. When the flood was discovered soon afterwards, he fitted it to a cause he already had in mind — the stern seal — rather than treating the deck wash pump's failure moments earlier as a live clue. Once he'd settled on an explanation, the actions that would have protected against being wrong — closing all sea valves — were never taken.

Situational Awareness

The fish hold was checked once, early, and found dry. It was not checked again until it was nearly full. The skipper read a slowing rate of rise in the engine room as good news; it was actually floodwater spreading into a second compartment, quietly building a free-surface effect that would take the vessel from listing to lolled-over in well under a minute.

Communication

Voice VHF calls into an area of known poor coverage went unanswered for around 45 minutes. A DSC distress alert — which repeats automatically and carries the vessel's identity and position further than voice — would very likely have reached the coastguard sooner. The skipper held off because he judged the vessel wasn't yet in 'grave and imminent danger' — a threshold an uncontrolled, unidentified flood had already crossed.

How This Pattern Repeats

Industry	Incident	The parallel
Aviation	Air France 447, 2009	The crew fixated on an initial, incorrect read of the situation (a stall they didn't recognise as a stall) and didn't systematically re-check the instruments that would have corrected it — the same pattern as sticking with the stern-seal diagnosis while the real source went unchecked.
Offshore	Deepwater Horizon, 2010	A negative pressure test that should have signalled a well control problem was reinterpreted to fit the crew's existing expectation that the well was secure — confirmation bias overriding a live warning sign, with catastrophic consequences.
Nuclear	Three Mile Island, 1979	Operators misread instrumentation consistent with their existing understanding of the fault, and continued acting on that misreading for hours while the actual problem — a stuck-open relief valve — went uncorrected.

See How You'd Handle It

The scenario opens at the moment the deck wash pump produces no water: do you investigate the pump itself, check the bilges first, or head straight for the sea valves? Seven decision points follow, ending at the point where Freedom II's crew ran out of options.

What Every Skipper and Crew Should Take From This

An unexplained flood is a "close all sea valves" situation, not a "diagnose first" situation — MGN 165(F) says so explicitly
Any seawater pipe rigidly connected to a flexibly-mounted engine, with no bellows or flexible coupling, is a latent failure point — check it before it fails at sea
A DSC distress alert doesn't require certainty that the vessel is sinking — an uncontrolled flood with the source unidentified already qualifies
"The flooding has slowed" is not evidence the situation is improving — check every compartment, every time, throughout the emergency
A salvage pump decision made in rough conditions on arrival should be revisited the moment conditions change

Test Your Knowledge

Think you've got this covered? Put it to the test in Crew Connect's free Knowledge Checker — try a quick safety round and see how you score.