Lisco Gloria: The Drencher That Gave No Water

The Lisco Gloria is the case where detection worked and the ship was lost anyway. A trailer fire on the upper deck was reported within a minute; the master started the drencher three minutes later, and it delivered no water. The BSU found a pump valve control probably set to 'manual', and the fallback that would have worked was never tried.

This matters because the argument for earlier detection is an argument about what the crew can do with the minutes it buys. On the Lisco Gloria those minutes existed — the BSU puts the window in which the garage-area fire could still have been controlled at 'only a few minutes' — and they were spent on a pump that was not delivering, a sprinkler pipe that had parted in the engine room, and a bridge that believed the problem was electrical. The drencher chain, not the detection chain, is where this ro-pax fire became a 199 m constructive total loss.

What happened

A refrigerated trailer on the largely open upper deck caught fire at midnight, an hour and a half out of Kiel, and the ship burned for two weeks.

• The Lithuanian-flagged Lisco Gloria (IMO 9212151, 20,140 GT, 199 m, built Szczecin 2002, ABS class, owned by DFDS Lisco) sailed from Kiel for Klaipėda on the evening of 8 October 2010 with 203 passengers and 32 crew — 235 people.
• Trucks, trailers and cars were parked on the upper deck (deck 6), which is a covered garage area under the superstructure forward and an open weather deck aft. Reefer trailers were plugged into the ship's power and inspected on rounds by the duty crew member.
• A few minutes before midnight the duty crew member smelled burning on his way back from the aft end of the deck; at about the same time the smoke detectors raised a fire alarm for the garage area on the bridge. At 23:58:59 the crew member reported a fire on the first truck on the starboard side, lane 8 — the reefer trailer he had inspected first.
• The fire broke out between the cab and the trailer, jumped to the trailer alongside within a very short time, spread horizontally through the garage area and then aft along the open deck. Tyres and fuel tanks exploded along the whole length of the ship.
• The master ordered evacuation at 00:09:09, eleven minutes after the report. All 235 were taken off by lifeboat and raft and recovered by ships in the area, chiefly the ferry Deutschland; 28 were injured, 23 treated in hospital, none seriously.
• The fire took about 2.5 hours to spread down to deck 4 and several more to reach the superstructure. It could not be extinguished at sea; the ship drifted into Danish waters, was towed to Munkebo, and made fast on 22 October — two weeks after ignition. She was declared a constructive total loss. The cause was narrowed to a technical or electrical fault in the reefer unit, the truck or the ship's supply, and could not be determined further because of the destruction.

The first eleven minutes, from the VDR

The bridge did almost everything the fire plan asks for, in the right order, inside four minutes — and none of it produced water on the deck. The BSU reconstructed the sequence from the voyage data recorder:

• 23:58:59 — duty crew member reports fire on lane 8 by radio; the officer of the watch has the smoke alarm for the garage area and is finding it on CCTV.
• 23:59:20 — master called; 23:59:37 — master on the bridge; 23:59:38 — order to cut power on the upper deck; 23:59:56 — deck crew ordered to the upper deck.
• 00:00:05 — ventilation confirmed off. 00:00:55 — duty engineer ordered to isolate the trailer sockets.
• 00:01:45 — report that the upper deck cannot be entered for smoke. 00:01:54 — master starts the drencher for the upper deck from the bridge.
• 00:03:27 — master notices the drencher is not delivering. 00:03:37 — duty engineer in the engine control room ordered to start it from there.
• 00:05:41 — master notes the drencher still dry despite open section valves. 00:08:53 — sprinkler alarm: the accommodation sprinkler has tripped at 100 °C.
• 00:09:09 — order to evacuate. 00:13:23 — Bremen Rescue Radio broadcasts the mayday relay.

Why the drencher gave no water

Most probably because the drencher pump's valve control was set to 'manual', so starting the system from the bridge opened the section valves but never ran the pump. The BSU is explicit that it believes the drencher 'was actually fully operational on the night of the accident' as hardware; the compromise was a control setting. It could not confirm this after the fire, but notes that in that state the pump's readiness lamp in the engine control room would probably not have been lit — a sign that was there to be read before the voyage, not during it.

What turned a control setting into a total loss was the second event. Activation of the accommodation sprinkler, or the pressure surge that came with it, parted a pipe connection near the engine control room and began an uncontrolled flooding of that space and the compartments below. The duty engineer — the one person who could have fixed the drencher — now had a stability threat in front of him and went to the sprinkler room on deck 5 to deal with it. Nobody told the bridge, so the master could not send the chief engineer to relieve him. By the time the engineer reached the sprinkler room the deck had been burning for more than ten minutes and the route back from the superstructure was already impassable without breathing apparatus.

The untried fallback sat in that same sprinkler room: a hand wheel that cross-connects the drencher to the fire main, bypassing the drencher pump and its valve setting entirely. The BSU could not establish whether the engineer knew the option existed or assumed the pump was already running from the switchboard settings he had made. What is certain is that the drencher was never fed from the fire main, the engineer did not report back to the bridge after clearing the sprinkler problem, and the bridge went on believing the fault was in the power supply. The option of feeding the drencher from the fire main was never discussed, and the master ordered evacuation.

What detection bought, and what spent it

Detection bought roughly three minutes of usable window and the suppression chain spent it on a valve setting. That is the reading that matters for any operator specifying an early-warning layer. The BSU's own judgement is that a working drencher would not necessarily have extinguished the fire — its 14-fire comparison found drenchers extinguished only five of nine fires they were used in — but would 'in all probability' have slowed the temperature rise and smoke build-up in the garage area, kept the fire team in play, and possibly allowed a return to Kiel to fight it alongside. Every one of those outcomes needs water at minute three, not a diagnosis at minute nine.

Two structural facts of the ship made the missing minutes decisive. The upper deck was open aft, so the fire had an unlimited air supply and spread horizontally by flame and radiation across a lane spacing the BSU calls 'normal on all ro-pax ferries'. And an open fire door on the port side let smoke straight into the stairwell, which is why the fire team never got an attack in and why the sprinkler room was unreachable ten minutes on. Neither is unusual. What the Lisco Gloria demonstrates is that on an open ro-ro deck the suppression system's readiness state is as much a part of the detection-to-control chain as the detector itself — an alarm three minutes early is worth nothing to a pump in manual.

What DFDS changed fleet-wide

Almost everything the BSU would otherwise have recommended, before the report was finished. DFDS Seaways' internal review, prompted by this fire and the Pearl of Scandinavia car-deck fire, produced a fleet-wide list that reads as a checklist for the failure chain above: test every drencher section valve from both remote and main control stations; verify the drencher pump and its emergency supply from the fire and emergency fire pumps; mark every valve that must stay open with a 'normally open' sign; make every deck and engine officer competent in normal and emergency drencher operation, retrained at intervals of no more than three months; verify remote release and bridge indication of fire doors; confirm standby generators are connected so the power management system can meet a sudden load; and run fire rounds at least hourly between 22:00 and 06:00.

With the Danish Maritime Authority, DFDS also carried the lessons to the IMO Sub-Committee on Fire Protection as an information paper to FP 55. The BSU's remaining recommendation to the operator was a cultural one: crews must report equipment malfunctions immediately, and resolved defects must be rechecked at ISM audit.

What this means for owners, class and P&I

Audit the suppression chain's readiness state, not just its certificate. A drencher can be class-approved, tested and 'operational' and still be one control setting away from dry — the Lisco Gloria's was. The specific checks DFDS adopted (section valves from every station, pump readiness lamp, fire-main cross-connection drilled, 'normally open' tagging) are the minimum a superintendent should be able to evidence. For class and flag, the case supports the FIRESAFE-era view that fixed-first response on a ro-ro deck depends on a release that is verified end to end, and that the engine-room fallback must be a drilled action, not a hand wheel someone may or may not know about.

For underwriters, the Lisco Gloria is the cleanest counter-example to 'earlier detection would have saved the ship': detection was early, and it changed nothing because the next link was not ready. An early-warning layer that also carries the suppression system's readiness and release confirmation to the bridge closes exactly the information gap the BSU found — the master on this ship never knew why his drencher was dry.

Sources

• 1. Bundesstelle für Seeunfalluntersuchung (BSU) — Investigation Report 445/10, 'Very serious marine casualty: Fire on the ro-ro passenger vessel LISCO GLORIA on 8 October 2010 in the Fehmarn Belt', joint investigation with the Lithuanian Maritime Safety Administration, published 2012: summary; ship and voyage particulars; VDR timeline (Table 5); section 4.3 Fire-fighting (drencher pump valve control, sprinkler pipe failure, fire-main cross-connection); section 4.4 Development of the fire; Table 6 comparison of 14 ferry fires; section 5 Actions taken (DFDS fleet measures, FP 55/12/INF); sections 6–7 Conclusions and safety recommendations.
• 2. IMO — SOLAS Chapter II-2 Regulation 10 and the FSS Code, as cited by the BSU for the automatic sprinkler requirement on passenger ships carrying more than 36 passengers (sprinkler activation at 100 °C on the Lisco Gloria).
• 3. IMO — Resolution A.123(V), Recommendation on fixed fire-extinguishing systems for special category spaces (1967), the drencher basis the BSU discusses in its review of ferry-fire suppression effectiveness.
• 4. Hakkarainen, T. et al. — 'Survivability for ships in case of fire', SURSHIP-FIRE final report, VTT Research Notes 2497 (2009); and Arvidson, M. — 'Large-scale ro-ro deck fire suppression tests', SP Report 2009:29 — the drencher-effect literature the BSU relies on for its conclusion that an operable system reduces peak heat release and fire spread.