MA

RIN

E A

CCID

ENT

INV

ESTI

GAT

ION

BRA

NCH

AC

CID

ENT

REP

OR

T

VERY SERIOUS MARINE CASUALTY REPORT NO 32/2014 DECEMBER 2014

Combined report on the investigations of the

sinking and abandonment of the DUKW amphibious passenger

vehicle

Wacker Quacker 1in Salthouse Dock, Liverpool

on 15 June 2013

and the

fire and abandonment of the DUKW amphibious passenger

vehicle

Cleopatraon the River Thames, London

on 29 September 2013

Extract from

The United Kingdom Merchant Shipping

(Accident Reporting and Investigation)

Regulations 2012 – Regulation 5:

“The sole objective of the investigation of an accident under the Merchant Shipping (Accident

Reporting and Investigation) Regulations 2012 shall be the prevention of future accidents

through the ascertainment of its causes and circumstances. It shall not be the purpose of an

investigation to determine liability nor, except so far as is necessary to achieve its objective,

to apportion blame.”

NOTE

This report is not written with litigation in mind and, pursuant to Regulation 14(14) of the

Merchant Shipping (Accident Reporting and Investigation) Regulations 2012, shall be

inadmissible in any judicial proceedings whose purpose, or one of whose purposes is to

attribute or apportion liability or blame.

© Crown copyright, 2014You may re-use this document/publication (not including departmental or agency logos) free of charge in any format or medium. You must re-use it accurately and not in a misleading context. The material must be acknowledged as Crown copyright and you must give the title of the source publication. Where we have identified any third party copyright material you will need to obtain permission from the copyright holders concerned.

All MAIB publications can be found on our website: www.maib.gov.uk

For all enquiries:Marine Accident Investigation BranchMountbatten HouseGrosvenor SquareSouthampton Email: maib@dft.gsi.gov.ukUnited Kingdom Telephone: +44 (0) 23 8039 5500SO15 2JU Fax: +44 (0) 23 8023 2459

CONTENTS

GLOSSARY OF ABBREVIATIONS AND ACRONYMS

SYNOPSIS 1

SECTION 1 - FACTUAL INFORMATION WACKER QUACKER 1 3

1.1 Particulars of Wacker Quacker 1 and accident 31.2 Background 41.3 Narrative 41.4 Environmental conditions 131.5 The Yellow Duckmarine 13

1.5.1 Company history and structure 131.5.2 Thefleet 16

1.6 Thecrew 181.6.1 Master 181.6.2 Driver 18

1.7 Thevehicle 181.7.1 TheDUKW 181.7.2 Wacker Quacker 1 21

1.8 Regulatoryrequirements 231.8.1 General 231.8.2 Stabilityandsurvivabilityrequirements 26

1.9 Buoyancy foam 271.10 Post-accidentinspections,testsandtrials 28

1.10.1 MAIBandMaritimeandCoastguardAgencyinspections 281.10.2 Metallurgicalanalysisofhullplating 321.10.3 Stabilitytestsandfloodingtrials 34

1.11 The Yellow Duckmarine safety management system 391.11.1 SafetyPolicyandGeneralSafeOperationalProceduresmanual 391.11.2 Emergency preparedness 411.11.3 Passenger safety announcements 411.11.4 Pre-splashdown safety checks 421.11.5 Life saving appliances 421.11.6 Personalflotationdevices 42

1.12 The Yellow Duckmarine maintenance management system 441.12.1 Planned maintenance routines 441.12.2 Unplannedbreakdownrepairs 44

1.13 Vehicle surveys and inspections 451.13.1 MaritimeandCoastguardAgency 451.13.2 VehicleandOperatorServicesAgency 45

1.14 Industryguidanceandinstructionsforsurveyors 461.15 Passengerquestionnairesandfeedback 461.16 Similarincidentsandaccidents 46

1.16.1 TheYellowDuckmarine 461.16.2 Beatrice 501.16.3 Miss Majestic 501.16.4 DUKW 34 531.16.5 LessonslearnedduringWorldWar2operations 54

SECTION 2 – FACTUAL INFORMATION CLEOPATRA 55

2.1 Particulars of Cleopatra and accident 552.2 Background 562.3 Narrative 562.4 Environmentalconditions 622.5 LondonDuckTours 622.6 Thecrew 62

2.6.1 Master 622.6.2 Tourguide 63

2.7 Cleopatra 632.7.1 Generaldescription 632.7.2 Hull 632.7.3 Propulsiondrivetrain 642.7.4 Ruddercontrol 642.7.5 Brakesystem 642.7.6 Enginecompartment 642.7.7 Engine compartment extraction fan controls 712.7.8 Passengercompartment 712.7.9 Engine fuel oil system 712.7.10 Electricaldistributionsystem 74

2.8 Fire-fightingarrangements 752.8.1 Fire-fightingappliances 752.8.2 Enginecompartmentshut-down 75

2.9 Lifesavingappliances 782.10 Vesselsurvivability 78

2.10.1 Buoyancyfoam 782.10.2 AdditionofbuoyantfoamfollowingthesinkingofWacker Quacker 1 802.10.3 Typesofbuoyancyfoam 80

2.11 Fireinvestigation 822.11.1 Overview 822.11.2 Theexternalhull 832.11.3 Crewcabareaandpassengercompartment 832.11.4 Enginecompartment 832.11.5 Enginecompartmentcoolingairventilationvoidspaces 852.11.6 Centralvoidspace 862.11.7 Hydraulicsteeringsystem 862.11.8 Enginefueloilsystem 862.11.9 Electricaldistributionsystem 892.11.10 Hand-operatedbandbrake 892.11.11 Brakesystemairreservoirs 89

2.12 Externalfireinvestigationsupport 892.12.1 BureauVeritasFireScienceDepartmentfireinvestigationreport 892.12.2 MetallurgicalanalysisofCleopatra’s drive shaft and UJs 91

2.13 Safetymanagementsystem 912.14 Vehicle maintenance 93

2.14.1 Maintenancemanagementsystem 932.14.2 Planned maintenance 932.14.3 Unplannedbreakdownrepairs 95

2.15 Surveysandinspections 952.16 Emergencyresponse 95

2.16.1 ThamesRIBExperience 952.16.2 RoyalNationallifeboatInstitution 95

2.16.3 Police 962.16.4 LondonFireBrigade 96

2.17 Passengerfeedback 962.18 Similaraccidents 97

2.18.1 MAIBdatabase 972.18.2 Elizabeth 972.18.3 Mistress Quickly 972.18.4 WorldWar2 98

SECTION 3 - ANALYSIS 99

3.1 Aim 993.2 Overview 993.3 The sinking of Wacker Quacker 1 99

3.3.1 Overview 993.3.2 The hull failure 993.3.3 The fouling of the propeller 1013.3.4 MaterialconditionofWacker Quacker 1andTheYellowDuckmarinefleet1013.3.5 Maintenancemanagement 102

3.4 ThefireonboardCleopatra 1023.4.1 Overview 1023.4.2 Seatofthefire 1023.4.3 Firedevelopment 1023.4.4 Sourceofignition 1033.4.5 Failureoftheuniversaljoint 1053.4.6 MaterialconditionofCleopatraandLondonDuckToursfleet 106

3.5 Vesselsurvivability 1063.5.1 Regulatorycompliance 1063.5.2 Theuseofbuoyancyfoam 1063.5.3 Buoyancy foam calculations 107

3.6 Passengerandcrewsurvivability 1083.6.1 Riskofentrapment 1083.6.2 Theriskofdrowning 1083.6.3 TheIrishmodel 110

3.7 Roleoftheregulators 1113.7.1 Effectiveness of survey and inspection regimes and unscheduled interventions 1113.7.2 Instructions,guidanceandtraining 1123.7.3 Sharingofinformation 112

3.8 Roleoftheoperator 1133.9 Emergency preparedness 113

3.9.1 General 1133.9.2 Emergency preparedness – Wacker Quacker 1 1143.9.3 Emergency preparedness – Cleopatra 115

3.10 Emergency response – Wacker Quacker 1 1153.10.1 Raisingthealarm 1153.10.2 Theabandonment 1153.10.3 Shorebasedemergencyresponse 116

3.11 Emergency response – Cleopatra 1163.11.1 Raisingthealarm 1163.11.2 Firecontainmentandfire-fighting 1173.11.3 Theabandonment 1183.11.4 Shorebasedemergencyresponse 118

SECTION 4 - CONCLUSIONS 119

4.1 The sinking of Wacker Quacker 1 1194.2 ThefireonboardCleopatra 1194.3 Common safety issues 120

SECTION 5 - ACTION TAKEN 122

5.1 MAIBactions 1225.2 Actionstakenbyotherorganisations 122

SECTION 6 - RECOMMENDATIONS 124

FIGURES

Figure 1 - FirstDUKWamphibiousvehiclesusedtocarrypassengersintheWisconsinDells,USA

Figure 2 - Wacker Quacker 4sinkinginSalthouseDock,Liverpool,on30March2013

Figure 3 - AmphibiouspassengervehicleWacker Quacker 1

Figure 4 - TheYellowDuckmarineamphibiouspassengervehiclesightseeingtour route

Figure 5 - Crew positions during sightseeing tour

Figure 6 - Wacker Quacker 1’sfailedattempttoexitSalthouseDockat1205on15 June 2013

Figure 7 - Wacker Quacker 1’s passenger and crew seating arrangements

Figure 8 - SplashdownintoSalthouseDock

Figure 9 - Abandonshipandrescuesequence

Figure 10 - Police divers’ underwater survey

Figure 11 - PhotographstakenbypoliceaftertherecoveryofWacker Quacker 1

Figure 12 - Wacker Quacker 4 and Wacker Quacker 8beforeandaftertheirconversion to carry 30 passengers

Figure 13 - Propeller shaft v-strut support

Figure 14 - OriginalDUKWhulldesignandcompartmentlayout

Figure 15 - Liverpool DUKW side curtains

Figure 16 - Wacker Quacker 1bilgepumpcontrols

Figure 17 - Wacker Quacker 1 hull drain plugs

Figure 18 - Roadtilttest

Figure 19 - FoamremovedfromWacker Quacker 1, Wacker Quacker 2 and Wacker Quacker 8

Figure 20 - Wacker Quacker 8driver’scabinstrumentationandelectricalwiring

Figure 21 - External patch repairs to the hull of Wacker Quacker 1

Figure 22 - V-strut suspended from Wacker Quacker 1’s propeller shaft

Figure 23 - Sectionsofpropellershafttunnelandafttransversebulkheadcutaway and despatched for metallurgical analysis

Figure 24 - Analysisofpropellertunnelplatingandweldedrepairs

Figure 25 - MAIBfoaminsertiontrialsandfloodingreconstruction

Figure 26 - Hull preparations and foam insertion process

Figure 27 - Wacker Quacker 1floodingreconstruction

Figure 28 - TheYellowDuckmarinelifejacketsandbuoyancyaids

Figure 29 - SimilarhullfailurediscoveredonboardWacker Quacker 2 two weeks earlier

Figure 30 - Hullrepairscarriedoutbygaragestaff

Figure 31 - SinkingofDUKWMiss Majesticon1May1999

Figure 32 - Collisionbetweentugboat/bargeCaribbean Sea/The Resource and DUKW 34 on 7 July 2010

Figure 33 - World War 2 operations manual

Figure 34 - LondonDuckToursamphibiouspassengervehicleCleopatra

Figure 35 - ThewaterbornesectionoftheLondonDuckTourssightseeingroutes

Figure 36 - Engine compartment cooling air vents on a similar London Duck Tours vehicle

Figure 37 - Abandonshipandrescuesequence

Figure 38 - SimplifiedillustrationofCleopatra’s drivetrain

Figure 39 - Typicalslidingsplineddriveshaftanduniversaljointarrangement

Figure 40 - Braking system air reservoirs

Figure 41 - OriginalDUKWenginecompartmentcoolingairsystem

Figure 42 - Modifiedenginecoolingairflowandenginecompartmenthotairextraction

Figure 43 - Cleopatra’s engine exhaust pipe

Figure 44 - Typical London Duck Tours DUKW passenger compartment and crew cablayout

Figure 45 - Boarding platform and ladders

Figure 46 - SimplifiedillustrationofCleopatra’s fuel supply system

Figure 47 - TypicalLondonDuckToursDUKWenginecompartmentfixedfire-fightingsystem

Figure 48 - Emergencyfuelshut-offvalvesandfiredamperreleasehandlesinasimilar vehicle

Figure 49 - BCTQLtd’s2008buoyancyfoamcalculations

Figure 50 - Buoyancy foam packed around the drivetrain in the centre void space of a London Duck Tours DUKW

Figure 51 - Buoyancy foam packed into the engine compartment ventilation void spaces

Figure 52 - Firedamagedpassengercompartmentandcrewcab

Figure 53 - Firedamagewithintheenginecompartmentandstarboardventilationvoid space

Figure 54 - Firedamageunderthecrewcab

Figure 55 - Areaofdeepestburnwithincentralvoidspaceaftoftheautomaticgearbox

Figure 56 - Burntthroughsectionofflexiblehose

Figure 57 - Greasecontaminatedfoamremovedfromaroundthebandbrake

Figure 58 - Laboratoryanalysisoftheforwarddriveshaftuniversaljoint

Figure 59 - Tightly packed foam in central void space

Figure 60 - Hull corrosion within Wacker Quacker 1’s aft void space

Figure 61 - Firedevelopment

Figure 62 - The last of the passengers to escape from Wacker Quacker 1’s passenger compartment

Figure 63 - DublinDUKWsoperatingwithexternalbuoyancytubesandopencanopy

Figure 64 - MemberofthepublicjumpingfromPredator 3 to help passengers in the water

TABLES

Table 1 - SpacesidentifiedbyBCTQaslocationstofitbuoyantmaterial

Table 2 - VolumesandlocationsofbuoyancyfoaminsertedpriortoMAIBfloodingtrials

ANNEXES

Annex A - The Yellow Duckmarine vehicle drivers’/masters’ safety checklists

Annex B - The Yellow Duckmarine 2013 pre-departure and pre-splashdown safetybriefs

Annex C - AnexampleofanearlierversionofTheYellowDuckmarinepre-splashdownsafetybrief

Annex D - The Yellow Duckmarine maintenance and inspection record sheets

Annex E - DesignplansfortheLondonFrogCompanystrengthenedpropellershaft tunnel

Annex F - DesignmodificationscarriedoutbyLondonDuckToursLtdfollowingthe suspension of its operations after the Cleopatrafire

GLOSSARY OF ABBREVIATIONS AND ACRONYMS

A - Ampere

AIS - AutomaticIdentificationSystem

APV - Amphibiouspassengervehicle

BCTQ - Burness Corlett Three Quays Ltd

BV - Bureau Veritas

°C - Degrees Celsius

CCTV - Closed-circuit television

COIF - Certificateofinitialfitnessforroaduse

DfT - Department for Transport

DPA - DesignatedPersonAshore

DSC - Digitalselectivecalling

DSMCode - DomesticPassengerShip(SafetyManagementCode) Regulations2001

GPS - Globalpositioningsystem

kg - kilogram

kPa - kilopascal

kt - knot

kW - kilowatt

LDT - London Duck Tours Ltd

LFB - LondonFireBrigade

LSA - Lifesavingappliance

m - metre

m3 - cubicmetres

MCA - MaritimeandCoastguardAgency

MGN - MarineGuidanceNote

mm - millimetre

mph - miles per hour

MSN - MerchantShippingNotice

MTB - Marinetransferbox

MTL - MaterialsTechnologyLtd

N - Newton

NTSB - NationalTransportationSafetyBoard

NVIC - NavigationandVesselInspectionCircular

PFD - Personalflotationdevice

PLA - PortofLondonAuthority

PSV - PublicServiceVehicle

RIB - Rigid-hulledinflatableboat

RNLI - RoyalNationalLifeboatInstitution

rpm - Revolutionsperminute

t - tonne

TYD - The Yellow Duckmarine

UJ - Universaljoint

V - volts

VHF - VeryHighFrequency

VOSA - VehicleandOperatorServicesAgency

VTS - VesselTrafficServices

WQ - Wacker Quacker

WW2 - World War 2

TIMES: AlltimesinthisreportareUTCunlessotherwisestated

1

SYNOPSIS

Atabout1553on15June2013,theamphibiouspassengervehicleWacker Quacker 1 sankbythebowduringasightseeingtourinSalthouseDock,Liverpool.Thevehicle’s31passengersand2crewabandonedintothewaterandeitherswamashoreorwererecoveredwithoutseriousinjurybythecrewsofthreerecreationalnarrowboats.Wacker Quacker 1hadsufferedseverefloodingthroughtwoholesthathadbeentorninitshullwhenitspropellerwasfouledbyadiscardedcartyre.

Atabout1150on29September2013,afirebrokeoutonboardtheamphibiouspassenger vehicle CleopatraduringasightseeingtourontheRiverThames,London.Its28passengersand2crewwerealsoforcedtoabandonthevehicleandwererecoveredfromthewaterwithoutseriousinjurybythecrewsandpassengersofthreecommerciallyoperatedrigid-hulledinflatableboats.

Wacker Quacker 1wasthesecondamphibiousvehicletosinkinSalthouseDockwithina3-monthperiod;onbothoccasionsthevehiclesdidnothavethequantityofbuoyancyfoamrequiredtoprovidethemandatedlevelofresidualbuoyancy.Followingthe sinking of Wacker Quacker 1,theMaritimeandCoastguardAgencydiscoveredthattheLondon-basedamphibiouspassengervehicleswerebeingoperatedwithabouttwothirdsofthebuoyancyfoamrequiredtomeettheUK’sdamagedsurvivabilitystandard.Asaresult,thevehiclesweretemporarilytakenoffthewatertoallowtheoperatortofittheadditionalfoamneededtoensureitsvehiclesremainedafloatinthefullyfloodedcondition.Inordertoachievethis,additionalunsecuredandunprotectedbuoyancyfoamwastightlypacked around the engine compartment and in the void spaces under the passenger deck. Thisincreasedtheambienttemperaturesinandaroundthevehicles’enginecompartments.

The investigation concluded that:

• Prior to the sinking of Wacker Quacker 1,theLiverpoolandLondonoperators(TheYellowDuckMarineandLondonDuckTours,respectively)hadfailedtomeettheUK’smandatedbuoyancystandard;whentheLondonDuckToursdid,itintroducedthecircumstancesthatledtothefireonboardCleopatra.

• TheLiverpoolbasedvehicleshadbeenpoorlymaintainedandtheirmaterialconditionhadbeenallowedtoprogressivelydeterioratetoanunsafeleveloveraprolongedperiodof time.

• Inbothinstances,thecrewhadlittletimetoco-ordinatetheevacuationprocessandtheconfinednatureofpassengerspacesmadeitalmostimpossibleforthemtocontrolorassist the passengers.

Factorscontributingtotheaccidentsincluded:

• Inrespecttobuoyancy,theoperatorshadnottakenappropriateactiontoaddressthelessonslearnedfromprevioushighprofileaccidentsandfromrecentsimilarhazardousincidents.

• InLiverpool,thepassengersandcrewwerenotadequatelypreparedtodealwiththeemergency situation.

2

• TheMaritimeandCoastguardAgency’speriodicsurveyandinspectionregimes,andtheirunscheduledinterventionshadbeenineffective.

• TheMaritimeandCoastguardAgencysurveyorswithresponsibilityforamphibiouspassengervehicleshadnotbeenprovidedwithappropriateinstructions,guidanceortraining.

TheinvestigationalsoidentifiedthattheDepartmentforTransport’sMaritimeandCoastguardAgencyandVehicleandOperatorServicesAgencybothhadlong-standingconcernsoverthesafeoperationofthesevintageamphibiousvehicles.ButtheydidnotsharetheirconcernsorknowledgeeffectivelybeforeorfollowingthesinkingofthetwoLiverpoolvehicles.ThefireonboardCleopatra could easily have occurred on the road and thereforecomeunderthejurisdictionofthelandbasedregulator.

InresponsetoMAIBrecommendationSB3/2013,theMaritimeandCoastguardAgencytookactiontotemporarilysuspendamphibiousoperationsontheRiverThames.Ithasalsoprovidedguidancetoitssurveyorstohelpthemverifythevolumeoffoamfittedintothehullsofamphibiousvehicles,andhascarriedoutaninternalinvestigationintotheperformanceofitsLiverpoolandOrpingtonMarineOffices.LondonDuckTourshasmodifieditsvehiclestoprovideanalternativemethodofsatisfyingthedamagedsurvivabilitystandard,andhassubsequentlybeenpermittedtoresumeitswaterborneoperations.TheYellowDuckmarinehasbeenputintoadministrationandthenorth-westTrafficCommissionerhaswithdrawnitsPassengerServiceVehiclelicence.

RecommendationshavebeenmadetotheMaritimeandCoastguardAgencyandtheDriverandVehicleStandardsAgencywiththeaimofimprovingcrossagencyinformationflow,thestandardofAPVsurveysandinspectionsandimprovingtheindustry’soperationalpractices.LondonDuckToursLtdhasbeenrecommendedtocarryoutathoroughassessmentoftherisksintroducedbyanyfuturemodificationstoitsvehicles.

3

SECTION 1 - FACTUAL INFORMATION WACKER QUACKER 1

1.1 PARTICULARS OF WACKER QUACKER 1 AND ACCIDENT

SHIP PARTICULARSVessel’s name Wacker Quacker 1

Flag UKCertifying authority MaritimeandCoastguardAgencyVessel type Class V passenger vesselVehiclechassisnumber 353-19572Vehicle type DUKWamphibiouspassengervehicleYearofbuild 1945(convertedtoAPVin2001)Registeredowner Pearlwild LtdTrading name The Yellow DuckmarineConstruction SteelHull length 9.45mLength overall 10.43mWidth overall 2.55mHeightofcanopy(fromtheroad) 4.1mOnroadweight 8.3tVOYAGE PARTICULARSPort of departure SalthouseDock,Liverpool,EnglandPort of arrival SalthouseDock,Liverpool,EnglandType of voyage Sightseeingtour,inlandwaterwaysManning 2MARINE CASUALTY INFORMATIONDate and time 15June2013atabout1553Type of marine casualty or incident VerySeriousMarineCasualtyLocation of incident SalthouseDock,LiverpoolPlaceonboard Propeller shaft tunnelInjuries/fatalities NilDamage/environmentalimpact NilShipoperation SightseeingtoursVoyage segment ArrivalExternal & internal environment Daylight, dry and clear, sheltered waters.Personsonboard 33(2crew+31passengers)

4

1.2 BACKGROUND

Amphibiouspassengervehicles(APV)arewidelyusedthroughouttheworldbytourismandleisurecompaniestodeliverlandandwaterbornesightseeingtours.ThemodernamphibioussightseeingtourconceptoriginatedintheUSAwhen,in1946,anAmericanbusinessmanboughtasurplusUSarmyDUKW1amphibiousvehicleandmodifiedittocarrypassengers(Figure 1).Overthefollowing60years,theamphibioussightseeingtourconceptevolved,withfleetsofconvertedDUKWs,otherex-militaryvehiclesandavarietyofspecificallydesignedcommercialAPVsoperatinginmanymajorcitiesacrosstheworld.

In2000,theAmericanbusinessmodelwasreplicatedintheUKwhenthefirstAPVsightseeingtourswereintroducedontheRiverThames,London,bytheLondonFrogCompanyLtd.Ayearlater,asimilaroperationwasstartedinLiverpool.By2013, four converted DUKWs, Wacker Quacker 1(WQ1),Wacker Quacker 2 (WQ2),Wacker Quacker 4(WQ4)andWacker Quacker 8(WQ8),werebeingoperatedinLiverpool’sSouthDocks2complexbyPearlwildLtdunderthetradingnameofTheYellowDuckmarine(TYD).

On30March2013,WQ4 sank (Figure 2)whileundertowinSalthouseDock,Liverpool.Therewasnooneonboardatthetime,andthesinkingwasinvestigatedbytheMaritimeandCoastguardAgency(MCA).Duringitsinvestigation,theMCAdiscoveredthatthehullsofTYD’svehiclesdidnotcontaintheappropriatequantitiesofbuoyancyfoamrequiredtokeepthemafloatwhenflooded.TheMCAimmediatelysuspendedTYD’swaterborneoperationsandinstructedthecompanytoinsertanagreed volume of additional foam. TYD engaged the services of an independent firmofnavalarchitects,BurnessCorlettThreeQuaysLtd(BCTQ),tocalculatehowmuchfoamwasneededtomeettheMCA’srequirements.Sixweekslater,havingsatisfiedtheMCAthatithadmetitsbuoyancystandards,TYDresumeditsoperations on the water.

1.3 NARRATIVE

Atabout0900on15June2013,themasteroftheDUKWWQ1 (Figure 3), and his driver,arrivedforworkatTYD’sfleetmaintenancegarageinLiverpool,England.Thecompany’s mechanics, having carried out their daily maintenance routines the night before,presentedWQ1 for inspection. The master and driver examined WQ1 and completed the vehicle drivers’/masters’ safety checklist (Annex A). They then drove WQ1fromthegaragetoTYD’spassengerpick-uppointonGowerRoad,adjacenttoSalthouseDock(Figure 4).

WQ1 was one of three TYD vehicles operating that day and was scheduled to undertakeseventoursbetween1000and1830.At0950,themasterstartedtoboardthepassengersforhisfirsttour.Tenminuteslater,WQ1 departed with 30 passengersonboard.Theroadtriparoundthecitycentreandthewaterbornetourthroughtheinterlinkedsoutherndockscomplextookabout55minutestocomplete.During each trip, the master sat next to the driver and assumed the role of the tour guide (Figure 5).

1 ADUKW(Pronounced“duck”)isanamphibiouslandingvehiclethatwasdesignedtotransportmilitarypersonnelandsuppliesfortheUSarmyduringWorldWar2.TheacronymDUKWindicatesthatthevehiclewasdesignedin1942(D),itisanamphibian(U)andhasbothfront-wheelandrear-wheeldrivecapability(KandWrespectively).

2 Liverpool’sSouthDockscomplexismadeupofsixinterlinkeddocks(SalthouseDock,AlbertDock,WappingDock,QueensDock,CoburgDockandBrunswickDock)thathavesignificantheritagevalueandhavebeenredevelopedspecificallyforthetourismtrade.

5

Figu

re 1

:FirstD

UKWamphibiousvehiclesusedtocarrypassengersintheWisconsinDells,U

SA

6

WQ4undertowinSalthouseDock

Figure 2: Wacker Quacker 4sinkinginSalthouseDock,Liverpool,on30March2013

ImagescourtesyofRogerODoherty

7

Figure 3: AmphibiouspassengervehicleWacker Quacker 1

Figure 4: TheYellowDuckmarineamphibiouspassengervehiclesightseeingtourroute

8

Atabout1204,havingcompletedthewaterbornesectionofthethirdtouroftheday, the driver attempted to drive WQ1uptheSalthouseDockslipway.Duringhis approach, the driver increased the vehicle’s propeller speed and attempted to engage six-wheel drive. When WQ1’s wheels made contact with the slipway the vehiclebegantoclimboutofthewater.Thedriveralmostimmediatelyrealisedthatthe engine was losing power (Figure 6). Concerned that the engine might stall, the driverabortedhisattempttoexitthedockandallowedWQ1torollbackintothewater.Heputthepropellerasternandmanoeuvredbackwardstomakeroomfora second attempt. The second attempt was successful and the passengers were driventoGowerStreet,wheretheyweredisembarked.

At1454,WQ1setoffonitsfifthtourofthedaywith31passengersonboard.Fourofthe passengers were children, the youngest of whom, a 2 year old girl, was sitting on hermother’slapatthebackofthevehicle(Figure 7).At1529,WQ1 arrived at the SouthDockscomplexandstoppedontheroadoppositetheSalthouseDockaccessslipway.Fifteensecondslater,thedrivermanoeuvredWQ1 to the top of the slipway, sounded the vehicle’s horn and then drove it into the water (Figure 8).

Thedriverfollowedthecompany’snormalroutethroughtheinterlinkedSalthouse,WappingandAlbertdocks(Figure 4). At1546,thedrivermanoeuvredWQ1 out of AlbertDockandbeganhisapproachtowardstheSalthouseDockslipway.At1548,WQ1’swheelsmadecontactwiththeslipwayandthevehiclebegantoclimboutofthewater.Again,thedriverlostpowerandhadtoaborthisattempttoexitthewaterand allowed WQ1torollbackintothedock.

AsthedriverattemptedtomanoeuvreWQ1 into position to make a second attempt toclimbtheslipway,athudwasheardtowardsthebackofthevehicleanditsdriveshaftsstartedtovibrate.ThedriveralertedthemastertotheproblemandtheydecidedtomaketheirwaytoCoburgDock,whichhadaslipwaywithalesssteepincline.

Figure 5: Crew positions during sightseeing tour

Master’sposition(tourguide)Driver’s position

9

1204

:37

WQ

1 ap

proa

ches

slip

way

WQ

1 fro

ntw

heel

s la

nd o

n sl

ipw

ayW

Q1losespoweronslipway,driveraborts

atte

mpt

to e

xit d

ock.

1205

:29

WQ

1 ro

lls d

own

slip

way

and

man

oeuv

res

aste

rnW

Q1

2nd

appr

oach

to s

lipw

ay1206:41

WQ

1climbsoutofS

althouseDock

Figu

re 6

: Wac

ker Q

uack

er 1’sfailedattempttoexitS

althouseDockat1205on15June2013

10

Figu

re 7

: Wac

ker Q

uack

er 1

's p

asse

nger

and

cre

w s

eatin

g ar

rang

emen

ts

Master’s/tourguideseat

Pas

seng

er c

ompa

rtmen

t

Driv

er’s

sea

t

Bow

com

partm

ent h

atch

cov

er

Eng

ine

com

partm

ent h

atch

cov

er

Raisedplatform

Boa

rdin

g pl

atfo

rm

Steps

2

22

22

22

3 42

22

22

Thetworedseatswerefittedtoshow

whereQueenElizabethIIandthe

DukeofEdinburghsatwhenthey

tookatripthroughAlbertD

ockon

board

WQ

1 in

201

2Motherw

ithher2yearoldchild

sat o

n he

r lap

Fatherofthe2yearoldchild

Familyofsixthatincludeda

70 y

ear o

ld g

rand

fath

er a

nd

two

child

ren

aged

7 a

nd 1

1

11

As WQ1driftedastern,thedriverwasunabletodisengagethewaterpropellerdriveshaft,andtheenginestalled.Themastertriedtohelpthedriverbykickingthegearlevers.Atthesametime,passengersatthebackofthevehiclesawwaterfloodinginfromundertheirseatsandbegantoshouttothecrewmenatthefront.Initially,themastercontinuedtohelpthedriverwhowasaskinghimwherethecompanymobilephonewas.Withthewaterlevelrisingrapidlythepassengersbegantoscreamandshoutevenmore.Thepassengersontheupperlevelatthebackofthevehiclegrabbedlifejacketsfromaboxunderoneoftheaftbenchseatsandbegantopassthemforward.Themasterwenttothebackofthevehicletohelpdistributethelifejacketsandgiveinstructionstothepassengers.

Within 1 minute, the water level within the passenger space had reached the top oftheforwardfacingseatsandsomeofthepassengersbegantopanic.Membersofthepubliconthedocksideandtheownersofseveralrecreationalnarrowboats,whichwereberthedinthedock,begantorealisethattherewasaseriousproblemonboardWQ1.Asthetouristsbegantogatheratthewater’sedge,severalnarrowboatslefttheirmooringsandwenttohelp.

Themastertriedtocalmhispassengersandtoldthemtoputtheirlifejacketsonandremainattheirseats.Mostofthepassengerscouldnotseeorhearthemasterand started to evacuate through the windows (Figure 9).Withwaternowbeginningto lap over the vehicle’s side coaming into the passenger space, more and more passengersleaptintothewater.A70yearoldmalepassenger,whohadclimbedoutofhiswindowandontotheroof,begantolaunchthevehicle’sbuoyantfloats.

Figure 8: SplashdownintoSalthouseDock

12

Figu

re 9

:Abandonshipandrescuesequence

Vehicleflooding

Finalattempttore-starttheengine

Pas

seng

er p

ushi

ng o

n a

win

dow

Passengerlaunchingbuoyantapparatus

PassengerrecoveredonboardP

reda

tor

Pas

seng

ers

swim

min

g as

hore

2yearoldchildplacedontonarrowboat

Mem

berofpublicfrom

Pre

dato

rjum

psinto

wat

er

13

At1553,afterabout2minutes’flooding,thebowofWQ1sank.Asitdidsothemasterjumpedofftheboardingplatformatthebackofthevehicleandintothewaterandthedriver,whowasmomentarilysubmerged,hadtoescapethroughtheforwardport side window and swim to the surface. By now, most of the passengers were inthewaterandmostwerenotwearinglifejackets.Asthefirstofthenarrowboats,Predator 3,arrivedonthescene,oneofitsoccupantsleaptfromitsbowintothewatertohelprescuethepassengers.Membersofthepublicwhohadgatheredonthe dockside and on the access slipway, alerted the emergency services and started tothrowlifebuoysintothedock.

WQ1’sbowhadhitthedockbedbutitssternremainedafloat.Thelastpassengertoleavethevehiclewasthemotherofthe2yearoldchild.Shehadpassedherdaughterthroughtheportaftwindowtoherhusband,beforeclimbingoverthesterndoorontotheaftboardingplatform.Mostofthepassengerswererescuedfromthewater,ortransferredfromthesternofthevehicleontothecanalnarrowboatsthatcametotheirassistance.Someofthestrongerswimmersswamtotheslipwayandwalked ashore.

TYD’soperationsmanagerranfromherofficetoWQ8, which was in the process of boardingpassengers,andalerteditsmaster.TheoperationsmanagerreturnedtoherofficeandWQ8wasdrivenawayfromGowerRoad.Bynow,paramedics,policeofficers,ambulancesandfireengineswereonscene,andasearchandrescuehelicopterwasintheair.At 1556,TYD’soperationsmanagerranfromherofficetothe scene.

Initiallytherewassomeconfusionoverthenumberofpersonsthathadbeenonboard,andwhenaheadcountwascarriedoutninepassengerswereunaccountedfor.Someofthemissingpassengershadalreadybeentakentohospitalforassessment and the others had returned to their hotels to change their clothes. Confirmationthatallpassengersandcrewhadbeenrescuedandweresafewasachieved at 1745.

Later in the evening, police divers carried out an underwater inspection (Figure 10) of WQ1 and then attached recovery slings to the vehicle’s rear wheel axle. Oncerecoveredbackontotheslipway,WQ1wasinspectedandphotographedbyaMerseysidePolicesceneofcrimeofficer.Thepolicediscoveredacartyrewrappedaround the propeller and found two large holes in the vehicle’s propeller shaft tunnel shell plating (Figure 11).

1.4 ENVIRONMENTAL CONDITIONS

Itwasadryday,thewindwasslightandthewatersinthedockswereflatcalm.Theambientairtemperaturewasabout20ºCandthewatertemperaturewasabout15ºC.Liverpool’sSouthDockswasfilledwithsaltwatertakenfromthetidalRiverMersey,andthedepthofwaterwhereWQ1 sank was approximately 4.5m.

1.5 THE YELLOW DUCKMARINE

1.5.1 Company history and structure

In2001,LiverpoolDuckToursLtdlauncheditsamphibioussightseeingtoursaroundthestreetsofLiverpoolcitycentreandthroughthecity’shistoricSouthDockscomplex.In2003,PearlwildLtdpurchasedLiverpoolDuckToursLtdandstartedto

14

Starboardaftwheels

WQ1’sbowonthedockbed

Passengeraislefloorplateslifted

Figure 10: Police divers’ underwater survey

15

Car

tyre

aro

und

prop

elle

rH

oles

in p

rope

ller t

unne

l

Figu

re 1

1:PhotographstakenbypoliceaftertherecoveryofW

acke

r Qua

cker

1

16

tradeunderthenameTheYellowDuckmarine.Inthemonthsleadinguptotheaccident,thecompanyemployed35staffandwasoperatingupto28tourseachday.

Thecompanystructurecomprisedanoperationsdepartment,whichwasledbytheoperationsmanager,andanengineeringdepartment,whichwasledbythedirectorofengineering.Theoperationsteamwaslocatedinanofficenexttothetourcompany’sticketofficeintheAlbertDockbuildings(Figure 4). The operations team included the company’s nominated transport manager3, its sales and marketing managers, and its drivers and crew.

The engineering team worked from the company’s vehicle maintenance garage locatedabout1milefromSalthouseDock.Thedirectorofengineeringoversawtheday to day implementation of the company’s maintenance management system and was TYD’s designated person ashore4.Thegarageoperatedfrom0800-2000sevendays a week using a two shift system, 4 days on and 4 days off. Each shift was led byanengineeringmanagerandhadthreemechanics.

ThemajorityofTYDemployeeshadroadtransportbackgroundsandnoneofthemanagement team or technical staff had any commercial maritime experience prior tojoiningthecompany.

1.5.2 The fleet

TYD’sDUKWswerecertifiedbytheMCAtooperateonCategoryB5 waters as Class V6 passenger vessels. Pearlwild Ltd also held a Department for Transport (DfT)TrafficCommissioners’StandardNationalLicencethatauthorisedTYDtooperateitsDUKWsaspublicservicevehicles7 on the road. Each vehicle was certifiedtocarryupto30passengersandwasrequiredtohaveacrewoftwo.Inaccordance with local arrangements, each vehicle was also permitted to carry up to twobabesinarms8.

WQ1 and WQ2werebothconvertedtoAPVsin2001andwerealmostidentical.WQ8 and WQ4wereconvertedtoAPVsbyacompanyonthesouthcoastofEnglandin2000andwereoriginallyoperatedinGlasgow,Scotland.WhentheyfirstjoinedTYD’sfleetin2006and2009respectively,theirpassengerspaceandcanopy design were different to that of WQ1 and WQ2 and they could only carry 24 passengers.Theircanopieswereoriginallymadefromwoodandhadaretractablesun roof (Figure 12). WQ4 and WQ8 were later converted to carry 30 passengers andthewoodencanopieswerereplacedwithsteelones,similartothosefittedtoWQ1 and WQ2.

3 TheappointmentofaprofessionallycompetenttransportmanagerwasastatutoryrequirementunderthePublicPassengerVehiclesAct1981.Thenominatedtransportmanageristhepersonresponsiblefortheeffective and continuous management of the transport activities of the company.

4 Thedesignatedpersonashoreisthepersonwhoprovidesthelinkbetweenship’sstaffandshorestafftoverifythe implementation of the company’s safety management system.

5 UK Category B inland waters – wider rivers and canals where the depth of water is generally 1.5m or more and where the significant wave height could not be expected to exceed 0.6m at any time.

6 ClassVpassengervessel-shipengagedonlyonvoyagesinCategoryA,BandCwaters.7 Publicservicevehicle–avehiclewithmorethaneightpassengerseatsusedtotransportpeopleforprofit.8Babes in arms – children under the age of 1 year old.

17

Retractablesunroof

WQ8 converted to carry 30 passengers

WQ4

WQ8certifiedfor24passengers

Figure 12: Wacker Quacker 4 and Wacker Quacker 8beforeandaftertheirconversiontocarry30passengers

18

1.6 THE CREW

1.6.1 Master

ThemasterwasBritishandwas60yearsofage.HehadworkedforTYDfor2yearsandheldabespoke,operation-specificBoatmasters’Licence.Hehadbeena master for a year, and prior to that he was a crewman and tour guide. He did not holdapublicservicevehicle(PSV)drivers’licenceandthereforewasnotqualifiedtodrive WQ1 on the road.

1.6.2 Driver

The driver was British and was 51 years of age. He had worked for TYD for 4 monthsandheldaPSVdrivers’licence.HehadbeenlearningtodrivetheAPVsonthe water under the supervision of a master for 3 months and was working towards gaininghisBoatmasters’Licence.PriortojoiningTYDhedrovecommercialbuses,coaches and heavy goods vehicles.

1.7 THE VEHICLE

1.7.1 The DUKW

MilitaryDUKWsweredesignedduringtheWorldWar2(WW2)forthepurposeofmakingbeachlandingsandtransportingtroopsandequipmentawayfromabeachhead.Althoughmechanicallyrugged,hullconstructionwassimplifiedforthesake of an accelerated production schedule and the anticipated short life expectancy ofthevehicles.Over21,000DUKWswerebuiltintheUSbetween1942and1945,themajorityofwhichweretransportedoverseas.

TheDUKWinitsmilitaryconfigurationwasbuiltonaGeneralMotors’2½ton6x6truckchassis(3axlesand6wheels)andhadanoverallhulllengthof9.45m.Itwasequippedwitha6-cylinderGMC270gasoline(petrol)engineandhada5-speedmanualgearbox.Thevehiclehada2-mancrewandcouldcarryapayloadofupto2.3tofstoresor25fullyequippedcombattroops.Ithadamaximumspeedofabout50mphonlandandjustover5ktsonthewater.

Thevehiclehadathreeblade,clockwiserotating,phosphor-bronzepropelleranda single spade-type rudder. The propeller shaft was supported externally in the waterpropellertunnelbyacaststeelv-strut(Figure 13). The vehicle was left-hand drive and on land its front two wheels were steered in a conventional manner. The rudderwasconnected,bywirecables,tothemainsteeringwheelandmovedasthevehicle’s two front wheels were turned.

Thebasichullwasofweldedconstructionandwasfabricatedfromacombinationof2mm,2.4mmand2.8mmthickcold-rolledcarbonsteelplates.Thevehiclecomprisedfivemaincompartments:bow(forward)compartment,enginecompartment,driver’scompartment(crewcab),cargocompartmentandstern(aft)compartment.Thedriver’scompartment,bowandsterndecks,andthesidecoamingweremanufacturedassubassembliesandwereboltedtothebasichull(Figure 14). The original design took into consideration the interaction of road stressesonatruckchassisaswellasthewaterbornestressesonthehull.Interiorbulkheadsandreinforcementribsprovidedextrarigidity(Figure 14),butthehullhad

19

Propeller tunnel

V-strut

Afttransversebulkhead Propeller tunnel

V-strut shoes

Figure 13: Propeller shaft v-strut support

Bulkhead stiffeners

Sidecoaming

20

Afttransversebulkhead

Whe

el a

rche

s

Bow

Stiffeningribs

Car

go c

ompa

rtmen

tD

river

’s c

ompa

rtmen

t

Eng

ine

com

partm

ent

Bow

co

mpa

rtmen

t

Pro

pelle

r tun

nel

Driv

e sh

aft h

ousi

ngs

Flexiblerubberboot

Figu

re 1

4:OriginalDUKWhulldesignandcompartm

entlayout

Ext

erna

l driv

e sh

aft h

ousi

ng

Pressedsteelrubbingrails

21

toflexwiththechassistopreventitfailingduringroaduse.Pressedsteelrubrailswereweldedforwardtoaftalongtheoutsideofthehull(sidesandbottom)toprotectthe shell plating from contact damage and to provide additional structural integrity.

The hull had three drive shaft tunnels: one for the propeller, one for the front wheel-axle; and one for the rear wheel-axles. The tunnels housed the hull penetrations for the propeller shaft and the wheel-axle drive shafts. The hull penetrationforthepropellershaftwassealedusingastandardmarinestuffingbox.The sealing of the hull penetrations for the wheel-axle drive shafts presented the designerswithamoredifficultchallengeasthedriveshaftsneededtomoveupanddown due to the action of the vehicle’s suspension system. Tosolvetheproblem,the drive shafts were encased within watertight housings. The drive shaft housings permittedthesizeofholeinthehullnecessarytoprovidetherequiredverticalmovement and elongation of the drive shaft. The drive shaft housings were sealed at bothendsusingflexiblerubberboots(Figure 14).

Thehullhadtwotransversebulkheads.Theafttransversebulkheadseparatedthe cargo compartment from the stern compartment and the forward transverse bulkheadseparatedthecrewcabfromtheenginecompartment.Theseinternalbulkheadsaddedstructuralstrengthbutwerenotwatertight.Anywaterthatenteredthehullwasfreetotravelthroughoutitslength.Inordertoremovewaterthatmightenterthehulloverthedeckcoamingsorthroughhullpenetrations(duetosealfailuresorhulldamage),thevehiclehadtwopropellershaft-drivende-watering/bilgepumpsandanemergencyhandoperatedbilgepump.Thelargercapacityshaft-driven pump, known as the Higginspump,wascapableofdischarging200gallons(1,100litres)ofwaterperminute.Thecapacitiesofthesmallershaft-drivenpump and the hand operated pump was 50 and 25 gallons per minute, respectively. Tooperatetheshaft-drivenpumps,thepropellerdriveshafthadtobeengaged;thehigher the revolutions, the greater the pumping rate.

TheDUKWwasoriginallyfittedwithsixdrainplugs:3x76mmdiameterhulldrainplugsand3x38mmdiameterdriveshafthousingdrainplugs.Theseallowedthecrew to drain water and oil from the hull after exiting the water, and also provided accessforgreasingsomeofthedriveshafts’universaljoints(UJ).Inadditiontothedrainplugs,fourhulldrainvalveswerefittedtolaterDUKWmodels.

1.7.2 Wacker Quacker 1

WQ1wasbuiltin1945andin2001wasconvertedforuseasanAPV.Itwas9.45mlong(hulllength),2.55mwide,hadaroadheightof4.1mandweighed8.35t.InitsAPVconfigurationitcouldcarryupto30seatedpassengersandwasmannedbyacrew of two.

WQ1waspoweredbya6-cylinder,4-stroke,naturallyaspirated,Bedford330dieselenginewithamanualgearbox.Thechassisandbasichullwereoriginalandits drivetrain arrangements allowed the driver to manually select four or six-wheel drive,andengagethewaterpropellerdriveshaftaheadorastern.Infour-wheeldrive, power was delivered to the rear two axles, when six-wheel drive was engaged power was also delivered to the front axle.

MuchofWQ1’s hull plate thicknesses were as original:

• 2.8mmatthebowandonthebottomatthestern;• 2.4mmonthebottominthecentre;and• 2mm on its sides.

22

Thepropellerandwheel-axledriveshafttunnelswereoriginallyformedfrom2.8mmand 2.4mm steel plate respectively.

The vehicle’s original cargo compartment and stern deck was converted to provide the passenger compartment. The passenger compartment was decked with steel plates.Thedeckwasnotwatertightand,althoughbolteddown,theplatescouldbeliftedtoallowaccessformaintenance.Buoyantmaterialwasrequiredtobefittedinthevoidspacesbelowthepassengerdeckandinseveralotherlocationswithinthevehicle.

Thehullpenetrationsforthevehicle’swheel-axledriveshaftshadbeenmodifiedtoallowtheremovalofthedriveshafthousingsandrubberbootarrangements.The original side coaming around the cargo compartment was extended aft to the transomandawatertightaccessdoorandboardingplatformwasfittedatthestern.Aboardingladderwasfittedontheportsideoftheboardingplatform.

Elasticated cord

Sidecurtainsraised

Zip & velcro release

Sidecurtainslowered

Securinghooks

Figure 15: Liverpool DUKW side curtains

23

Asteelcanopy,withasolidPerspexsunroofandtransparentplasticsidecurtains,was welded to the top of the passenger space side coaming. The side curtains were designedsothattheycouldberolledupandsecuredintheopenpositionorclosedusingzips(Figure 15).ThezipswereconnectedtoindividualwindowpanelsusingVelcrosothatthewindowscouldbepushedopeninanemergency.Elasticcords,whichcouldbestretchedoverhooksontheoutsideofthesidecoaming,werefittedtothebottomofthesidecurtainstohelpholdthemdownwhileontheroad.

Thepassengercompartmenthad12forward-facing2-personbench-typepassengerseatsarrangedforwardtoaftintworowsof6alongitslowerlevel,withacentreaisle.Therewerealsotwoinwardfacing3-personbench-typepassengerseatsontheraisedplatformatthebackofthecompartment(Figure 7).

WQ1hadrecentlyhaditshighcapacityshaft-driven(Higgins)bilgepumpremoved,andwasbeingoperatedwithfourelectrically-drivenbilgepumps.Thebilgepumpoperating switches were located on the driver’s control panel (Figure 16) and couldbesettotheoff, manual or automatic position. When in automatic mode, the bilgepumpswerestartedandstoppedbytheirindividualfloatswitches.Awarningbuzzersoundedwheneverthebilgepumpsoperated.Therewasalsoanemergencyhand-poweredbilgepumppositionedonthestarboardsideofthepassengercompartmentbehindthecrewseats.Inaddition,sevenexternaldrainpipeswithscrew-threadedplugswerefittedtothebottomofthehull(Figure 17). These were usedinsteadoftheoriginalbottomplugstodrainwaterfromthehullwhenonland.

TYD’svehicleswerenotfittedwithanynavigationalaidsanddidnotcarryradioequipment.Eachmastercarriedacompanymobilephonetoallowthemtocommunicatewiththecompany’soperationsofficeandgarage.

1.8 REGULATORY REQUIREMENTS

1.8.1 General

Amphibioussightseeingtoursareconsideredtobeanicheindustrybyboththemaritimeandroadregulators.ThemodificationofaDUKWforuseasanAPVpresentsauniquechallengeasthevehiclemustmeetbothlandandmaritimeregulatoryrequirements.Anymodificationsmadefromtheoriginaldesignmusttakeintoconsiderationtheimpactonboththelandandwateraspectsoftheoperation.

TooperateontheroadasaPSV,theconvertedDUKWswereclassedashistoricalvehiclesandwererequiredtohaveaDfTCertificate of Initial Fitness for Road Use (COIF).AspartoftheCOIFprocess,theDUKWswereinspectedbytheDfT’sVehicleandOperatorServicesAgency(VOSA)9andweresubjectedtoastandardtilttest (Figure 18).

WhenTheLondonFrogCompanyfirstapproachedtheMCAin1998,theMCAtreatedtheDUKWAPVsasnewvessels.Assuch,theMCArequiredownerstocomplywiththelatestconstructionandstabilitystandardssetoutintheMerchantShipping(PassengerShipConstruction:ShipsofClassesIIItoVI(A))Regulations1998.ThestandardsrequiredbytheseregulationswereexplainedintheMCA’sMerchantShippingNotice(MSN)1699(M).

9 Atthetimeofinitialcertification,theCOIFinspectionswerecarriedoutbytheDfT’sVehicleInspectorate.InApril2014,VOSAandtheDfT’sDrivingStandardsAgency(DSA)mergedtoformtheDriverandVehicleStandardsAgency(DVSA).

24

Bilge pump control switches

Crewcabconsole

Starboardforwardbilgepumpdischarge Port aft pump

discharge

WQ1 operating with 2bilgepumpsrunning

Figure 16: Wacker Quacker 1bilgepumpcontrols

25

Port aft drain plug Drain plugchains

Figure 17: Wacker Quacker 1 hull drain plugs

Figure 18: Roadtilttest

26

In2005,theMCAheldameetingatitsheadquarterstoreviewitsinitialapprovalprocessfortheconvertedWW2DUKWsandotherAPVs.Atthetime,anotherDUKW(intendedforuseinPlymouth,England)hadbeensubmittedforapprovalbutthelocalMarineOfficehadconcernsoveritshullconstruction;inparticular,itshullplatethickness.AccordingtotheMCA’sInstructions for the guidance of surveyors:

Unless justified by the use of special materials with supporting information to the satisfaction of MCA HQ, hull thickness of less than 3mm will not be accepted.

TheaimofthemeetingwastoestablishwhatapprovalstandardsshouldbeappliedtofutureAPVs,andwhatactionshouldbetakenwithregardtoexistingDUKWs.ItwasestablishedatthemeetingthattherecordsheldonfileattheMCA’sMarineOfficeswerelessthansatisfactoryandcouldnotbeusedtoconfirmthattheoriginalDUKWAPVswereproperlyapprovedwhentheywerefirstallowedtooperateintheUK.Followingthemeeting,MCAheadquartersinstructeditsMarineOfficestoensurethatallfutureAPVs,andmodificationstoexistingAPVs,wereapprovedtoappropriateclassificationsocietyrules.ThismeantthatfutureconvertedDUKWswouldrequireminimumhullbottomandsideplatethicknessesof3.5mmand3.0mm respectively. Existing vessels were allowed to continue to operate on the understandingthatspecialmeasureswouldbeputinplace;suchasfrequentinspectionsbyoperators,annualhullthicknesschecksbysurveyors,andtheapplicationofepoxycoatings.AprogrammewasalreadyinplacetoincreasethehullplatethicknessoftheLondonbasedvehiclesto3mm.

When Pearlwild Ltd purchased WQ8in2005,theMCA'sstancethatallnewlyconvertedDUKWshadtocomplywithappropriateClassificationSocietyrules,ledtheLiverpoolMarineOfficetorefusetocertifythevehicle.ThecompanyappealedagainsttheMCA'sdecisiononthegroundsthattherewerenoclassrulesforAPVsanditsnewlyacquiredvehiclehadpreviouslybeenissuedaClassVpassengercertificate.In2006theregulatoroverturneditsdecisionandcertifiedWQ8.Similarlyin 2009, WQ4wascertifiedtooperateonthewaterinLiverpool'sSouthDocks.

In2010,theMCApublishedMSN1823(M),Safety Code for Passenger Ships Operating Solely in UK Categorised WatersandMSN1824(M),EU Directive 2006/87/EC (as amended) – Laying Down Technical Requirements for Inland Waterway Vessels. The mandatory safety code was developed in consultation with surveyors and the marine industry, and the technical standards superseded those setoutinMSN1699(M).Althoughmandatory,theDUKWswerenotrequiredtomeetthemorerobusttechnicalstandardssetoutinMSN1823(M)astheyonlyappliedtonewships(thosethatdidnothaveavalidpassengercertificateonthedatethecodecameintoeffect).

In2012,PearlwildLtdacquiredafifthDUKWbut,asithadnotbeenpreviouslycertifiedtooperateasaClassVpassengervessel,thevehiclewasrequiredtomeetthelatestconstructionstandardssetoutinMSN1823(M).Unabletomeetthe new standard, the company did not convert the DUKW to carry passengers and continued to operate with four vehicles.

1.8.2 Stability and survivability requirements

MSN1699(M)setoutthestabilityrequirementsforundamagedvessels(intactstability)andthestandardsofsurvivabilityrequiredforthedamagedcondition.

27

AstheconvertedDUKWshadnointernalwatertightsubdivisionandtheirdeckswerenotwatertight,theywereconsideredtobeopenships.TosatisfythestandardsofsurvivabilityprescribedinMSN1699(M)foraClassVpassengervesseloperating in Category B waters and carrying no more than 50 passengers the DUKWAPVshadtoachievetheheelingteststandarddescribedinSchedule2,Section3,andthebuoyancyteststandarddescribedinSchedule2,Section4.

TomeetthebuoyancytestsurvivabilitystandardtheDUKWswererequiredtohavesufficientbuoyancytoremainafloatwhenfullyflooded.Toachievethis,thevehicle’seffectiveresidualintactbuoyancyhadtobeatleast1.1timesitsintactvolumeofdisplacement(110%buoyancy).

Compliancewiththebuoyancytestsurvivabilitystandardcouldbedemonstratedthroughpracticalfloodingtrialsorbydetailedtheoreticalcalculations.Inadditiontoremainingafloatinafullyfloodedstate,thevehicleswererequiredtoremainuprightandstable.

MSN1699(M)alsostipulatedtheminimumrequirementsfortheconstructionofwatertightbulkheads,bilgepumpingsystems,emergencyelectricalpowerarrangements,fuelsystems,steeringgearandhullpenetrations.Accordingtotheregulations, WQ1wasrequiredtohaveatleastonepoweredbilgepump(whichcouldbeengine-driven)andahandbilgepump.

1.9 BUOYANCY FOAM

UKregisteredClassVDUKWAPVoperatorsusedacombinationoflowandhighdensityclosedcellpolyethylenefoamtoprovidetherequiredlevelofresidualbuoyancy.Thefoamusedwassuppliedinsheetformandwasindividuallycuttofitintonominatedvoidspaceswithinthevehicle’shull.MSN1699(M)requiredownerstoensurethatthebuoyancyfoamwas:

• protected against deterioration;

• adequatelysecuredagainstmovement;and

• installedinsuchawayastoprovidethegreatestpracticablecontributionstothestabilityandsurvivaloftheshipinthefloodedcondition.

TheoriginalbuildspecificationforWQ1’s conversion also emphasised that the foam neededtobetightlypackedandsecuredagainstmovement,andnotbeincontactwith moving parts.

Thevolumeofbuoyancyfoamrequiredtomeetthebuoyancytestsurvivabilitystandardwasdependentonthevehicle’sweightandtheinherentbuoyancyprovidedbyitstyresandotherfixturesandfittings.TheoriginalamountoffoamfittedintotheLiverpoolDUKWsisunknown.Theonlyinformationfoundintheowner’srecordsandtheMCA’sfilesrelatingtobuoyancyfoamrequirements,wasasetofcalculationsundertakenbyBurnessCorlett&Partners Ltd10onbehalfoftheLondonFrogCompanyin1999.ThesecalculationsformedpartoftheLondonbasedcompany’sinitialsubmissionstotheMCAandwerebasedonanumberoftheoretical assumptions and a vehicle weight of 7.2t.

10 BurnessCorlett&PartnersLtdlatermergedwithThreeQuaysMarineServicesLtdtoformBurnessCorlettThree Quays Ltd.

28

FollowingthesinkingofWQ4, BCTQ calculated that WQ2 required9.7m3 ofbuoyantmaterialtobeinsertedintoitshullinordertomeetthedamagedsurvivabilitystandard.Thiswasbasedonaweighbridgeweightof8.3t,passengerandcrewweightof75kgperperson,andanestimatedinherentbuoyancyof2.3t.TheBCTQreportidentified14locations(Table 1) wherethebuoyantmaterialcouldbeinserted.Itsreportalsoexplainedthatitscalculationsweretobeusedforguidancepurposesonly,andthatallvehiclesinthefleetshouldbeaccuratelyweighedandanindividualreservebuoyancycalculatedforeach.

Concerned that 9.7m3 mightnotbesufficienttomeetthesurvivabilitystandard,andwantingtoensurethattheDUKWsremaineduprightinthefullyfloodedcondition,theMCAaskedBCTQtocalculatethemaximumspaceavailablefortheinsertionofbuoyancyfoam.UsingalaserscanofaLondonDUKW(Portia),BCTQidentified22locationswhere,accordingtoitscalculations,about12m3 ofbuoyantmaterialcouldbeinserted.Therevisedcalculationsrequiredanadditional1.1m3 offoamtobefittedinthespacesbelowthepassengercompartmentandcrewcabdecks;almosthalfofwhich was designated to the centre void space. The revised volume in and around theenginecompartmentwasalmostdoubled,from1.56m3to3.06m3.

TheMCAinstructedTYDtopack12m3 ofbuoyancyfoamintoallitsvehicles;thiswastheamounttheMCA’sLiverpoolMarineOfficeunderstoodhadbeenfittedwhenit allowed the company to resume its operations on the water.

1.10 POST-ACCIDENT INSPECTIONS, TESTS AND TRIALS

1.10.1 MAIB and Maritime and Coastguard Agency inspections

Followingtheaccident,MAIBinspectorsandMCAsurveyorsfromtheLiverpoolMarineOfficeexaminedWQ1 and TYD’s other DUKWs at the operator’s vehicle maintenancegarage.TheMCAsurveyorsinspectedthedamageandoversawtheremovalandmeasurementofthebuoyancyfoamfromWQ1, WQ2 and WQ8(WQ4 wasstillbeingrefittedatthetimeanddidnotcontainanybuoyancyfoam).

WQ1’shullwasfoundtocontainabout6m3 ofbuoyancyfoam.Similaramountswerefound in WQ2 and WQ8.Alargeproportionofthefoamremovedwascoatedinoilandgreaseandtherewasevidencethatsomeblocksoffoamhadmeltedasaresultof contact with moving parts (Figure 19).

TheDUKWswerefoundtobeinapoormaterialstate:theyweredirty;theinternalpaint coatings had lifted in many places; large sections of the vehicles’ electrical wiring was unsecured and in poor condition; and many of the instrumentation gaugesfittedtotheconsoleswerefoundtobedefective(Figure 20).

Therewereahighnumberofexternalpatchrepairsanddoublers11 in place on WQ1’s hull. These were particularly evident in the propeller shaft tunnel (Figure 21). There were two large holes in the propeller tunnel plating where the v-strut support shoeshadbeentornfromthehull.Oneholemeasuredabout110mmx140mmandtheotherabout220mmx143mm(Figure 21). The v-strut was hanging freely from the propeller shaft with its pressed steel support shoes still attached (Figure 22). There were several other holes in the hull where the tips of the rotating propeller bladeshadmadecontactwiththetunnelplating.

11 TheMCA'sInstructionsfortheGuidanceofSurveyorsexplainsthat:doubling plates may be acceptable but normally only as a temporary repair in certain circumstances. For example, to increase the strength of plating and stiffeners but not bottom plating (excluding temporary repairs) and only if there is sufficient material in the parent plate to effect a good weld.

29

Meltedfoam

blocks

Recoveredfoam

stackedandmeasured

Figu

re 1

9:Foamremovedfrom

Wac

ker Q

uack

er 1

, Wac

ker Q

uack

er 2

and

Wac

ker Q

uack

er 8

30

Figure 20: Wacker Quacker 8driver’scabinstrumentationandelectricalwiring

Loose wiring under dashboard

Loose wiring in aft buoyancyvoidspace

31

Figure 21: External patch repairs to the hull of Wacker Quacker 1

32

1.10.2 Metallurgical analysis of hull plating

Two sections of the propeller tunnel plating and a section of the aft transverse bulkheadwerecutfromthehullofWQ1 and were sent, along with the propeller shaft v-strut (Figure 23),toTheTestHouse(Cambridge)Ltdformetallurgicalanalysis.Oneofthetunnelsectionsincludedthetwoholeswherethev-strutfeethadbeenattached.Theothersectionincludedlargeareaswhereitwasevidentthatprevioushullrepairshadbeencarriedout.Thesectionoftransversebulkheadincludedtheareawherethev-strutsupportshoeshadbeenattached.

Figure 22: V-strut suspended from Wacker Quacker 1’s propeller shaft

V-strut support shoes

V-strut

33

Sectionsofhullplatingremoved for analysis

Transversebulkhead

Propeller tunnel

V-strut

Figure 23: Sectionsofpropellershafttunnelandafttransversebulkheadcutawayanddespatched for metallurgical analysis

34

AtTheTestHouse(Cambridge)Ltd,specimensofthehullplatingwerecutout,polishedandsubjectedtomicro-structuralexamination,chemicalanalysisandtensile tests. The metallurgical analysis report concluded that:

• ThetwoU-sectionpressedsteelreinforcementribsweldedtotheforwardfaceofthetransversebulkheadhadsufferedextensivecorrosionwastage.

• Theweldconnectingthetransversebulkheadtothepropellertunnelplatinghad largely corroded away.

• The corrosion wastage had occurred sometime prior to the casualty and had compromisedthehull’sstructuralintegrity;inparticularthestructure’sabilityto counteract any torsional stressing induced via the propeller shafting.

• Thepropellertunnelplatinginwayofthesterncompartmentbulkheadandrudderstockhadsufferedwidespreadcorrosionwastage,whichhadbeenaddressedbymultiplephasesoffilletweldedpatchplaterepairs(doublers).

• Evidencesuggestedthatsomeofthehullrepairshadnotbeenfullyeffective,necessitating over-patching at some locations (Figure 24).

• Theoriginalparentshellplatingwaslowstrength,low-carbonrimming12 steel. Thiscontrastedwiththepatchrepairplates,whichappearedtobeofamoremodern low strength fully killed13 steel.

The metallurgical analysis report explained that production of the type of rimming steels seen in the samples was phased out in the 1970s. The report concluded that thebasesteelwaseitheroforiginalconstruction,orveryoldrepairsdatingbacktothe 1970s or earlier.

1.10.3 Stability tests and flooding trials

On30and31July2013,theMAIBusedWQ1 to conduct a series of practical stabilitytestsandfloodingtrialsattheCammellLairdshipyard,Liverpool(Figure 25).InadditiontoMAIBstaffanditsshorecontractors,thetestsandtrialswerewitnessedbyrepresentativesfromtheMCA,TYD,BCTQandLondonDuckToursLtd(LDT).

The aims of the trials were to:

• verifytheresultsofthetheoreticalbuoyancyassessmentsconductedbyBCTQ;

• identify any potential adverse effects associated with the insertion of buoyancyfoamand;

• carryoutareconstructionofthe15June2013floodingincident.

12 Rimmingsteel,alsoknownasdrawingqualitysteel,isatypeoflow-carbonsteelthathasanexcellentsurfacefinishandiscommonlyusedforcold-bendingandcoldforming.

13 Fullykilledsteelhasbeenprocessedsothatthereisnooxygenleftinthesteel.Itisthestrongesttypeofsteeland is normally used in the construction of ships and vessels.

35

Figure 24: Analysisofpropellertunnelplatingandweldedrepairs

Layers of welded patch repairs Plate thicknesses around the holes where v-strutfeethadbeenattached

36

Beforethetrialscouldtakeplace,MAIBinspectorsoversawtherepairofWQ1’s propellertunnelandtheinsertionofbuoyancyfoamintothedesignatedvoidspaces.Asteelinsertplatewasformedandbuttweldedintothepropellertunnel,inplaceofthesteelthathadearlierbeenremovedandsentformetallurgicalanalysis.Two 150mm diameter holes, designed to replicate those found in the hull after the sinking, were cut into the steel plate (Figure 26).Ashortlengthofthreadedsteelpipe,withablankingcap,wasweldedtoeachholetoallowthehulltobemadewatertightfortheinitialstabilitytests.

Thebuoyancyfoamsheetsusedforthetrialswereindividuallymeasuredandweighedbeforeinsertion.Thefoamwasfoundtobeabout5%denser,andtherefore5%heavier,thanthatgiveninitsspecification(33kg/m3).Whenthehighdensityfoam recovered from WQ1wasweighed,itwasfoundtobe20%denser.

Ittookteamsof4men7days(approximately280manhours)toinsertthenewbuoyancyfoamintothehullofthevehicle.Thefoamsheetswereindividuallycuttosizeandtightlypackedintothespacesidentifiedfortheinsertionof9.7m3 ofbuoyantmaterialinBCTQ’sinitialbuoyancyassessmentreport(Table 1). Care was taken not to compress the foam (Figure 26).Inordertoensurethemaximumamountoffoamwasinserted,theoperationalrequirementsofthevehiclewerenottakenintoaccountandfoamwasdeliberatelypackedupagainstmovingparts.Thetotalquantityoffoamfittedwas7.93m3 (Table 2); almost 2m3 less than the calculated amountrequiredtoprovide110%buoyancyand 4m3lessthanthatrequiredbytheMCA.

WQ1beingcranedintoCammell Laird dry dock

Figure 25: MAIBfoaminsertiontrialsandfloodingreconstruction

37

Figure 26: Hull preparations and foam insertion process

Foaminsertionprocess

Repairedpropellertunnel

Floodingplugs

38

Location Volume (m3)Aftbuoyancyspace 2.35Aftportbuoyancyspace 0.38Aftstarboardbuoyancyspace 0.38Portbuoyancyspace 0.84Starboardbuoyancyspace 0.84Centre void space 1.19Port vent space 0.28Starboardventspace 0.28Portsideshellandbox 0.60Starboardsideshellandbox 0.60Forwardbuoyancyspace 1.00Twoaftseatboxes 0.28Twelveseatboxes 0.84Driver and crew seats 0.10Total volume 9.96

Table 1:SpacesidentifiedbyBCTQaslocationstofitbuoyantmaterial

Location Volume (m3)Aftbuoyancyspace 1.6Aftportbuoyancyspace 0.38Aftstarboardbuoyancyspace 0.33Portbuoyancyspace 0.87Starboardbuoyancyspace 0.84Centre void space 1.24Port vent space 0.28Starboardventspace 0.28Portsideshellandbox 0Starboardsideshellandbox 0Forwardbuoyancyspace 1.00Twoaftseatboxes 0.24Twelveseatboxes 0.87Driver and crew seats 0Total volume 7.93

Table 2: VolumesandlocationsofbuoyancyfoaminsertedpriortoMAIB floodingtrials

Theamountoffoamthatcouldbefittedinthreelocationswassignificantlylessthanthequantitiescalculated.Thetotalamountoffoamfittedintheaftbuoyancyspacewas1.6m3,whichwas32%lessthanthe2.35m3 stipulated in the consultant’s report.Hadtheoperationalrequirementsforthesteeringgearbeentakenintoaccount,thequantityoffoamfittedwouldhavebeenfurtherreduced.According

39

totheconsultant’sreport,TYDwasrequiredtofit0.6m3 of foam into the port and starboardsideshellboxpanels.However,themaximumspaceavailableineachofthetwospaceswasnomorethan0.06m3.Althoughtheshortfallsinthecalculatedvolumesoffoamweremainlyattributedtothesethreespaces,hadthevehicle’soperationalrequirementsbeentakenintoaccount,thetotalamountoffoamfittedwouldhavebeensignificantlylessthanthe7.93m3 achieved.

Oncompletionofthebuoyancypackingprocess,WQ1wasweighedtoestablishabaselinebeforeweightswereaddedtoreplicateitsApril2013weighbridgeweight.Weights were then placed on the vehicle’s seats to replicate the passenger and crewweight,andthefueltankwasfilledwithwater.Theremaining2m3ofbuoyancyfoamthatcouldnotbefittedintothedesignatedvoidspaceswasevenlydistributedaround the passenger compartment and secured in place (Figure 27). WQ1 was thencranedintothewaterandaseriesofstabilitytestswascarriedout.Oncethestabilitytestsandaninitialbaselinefloodingtrialhadbeencompleted,theexcessfoamwasremovedandthemainfloodingtrialwascarriedout.Thetrialwasdesignedtosimulatethebestpossiblecircumstancesforthesurvivabilityofthefullyladenvehicleandreconstructthefloodingsequenceexperiencedon15June2013.

Withthecapsremovedfromthetwo150mmfloodingpipes,WQ1 was craned into thedock.Ittookabout5minutesforWQ1tofullyflood(Figure 27). The DUKW assumedastarboardlistbutremainedafloatwiththewaterlevelsettlingabout50mmbelowtheforwardupperedgeofthepassengerspacesidecoaming.WQ1 wasobservedinthisstatefor30minutesbeforeitwascranedoutofthedock.

1.11 THE YELLOW DUCKMARINE SAFETY MANAGEMENT SYSTEM

1.11.1 Safety Policy and General Safe Operational Procedures manual

PearlwildLtdwasrequiredtoproduceasafetymanagementsystemforitsTYDoperationthatcompliedwiththeUK’sDomesticPassengerShip(SafetyManagementCode)Regulations2001(DSMCode).InaccordancewiththerequirementsoftheDSMCode,thecompanyproduceditsSafety Policy and General Safe Operational Procedures manual.

Themanualcomprised15sectionsand3annexesandhadbeenupdatedseveraltimesovertheyearstoreflectchangesincompanypolicy,proceduresandkeypersonnel.Section4ofthemanualsetoutthemasters’responsibilities;Section7includedtheproceduresrelatingtoemergencypreparedness;Section10containedthetextforthepassengersafetyannouncements;andSection14describedthe company’s maintenance management system. The company’s generic risk assessments were attached as an annex to the manual.

40

Fullyflooded

WQ1beingcranedoutofdock

Weights used to simulate passengers & crew

Excess foam

Figure 27: Wacker Quacker 1floodingreconstruction

41

1.11.2 Emergency preparedness

TheproceduressetoutinSection7ofthemanualincludedthosetodealwithemergency situations such as fire, abandon ship, man overboard, collision, and main propulsion or steering failure. The documented procedure for abandon ship was:

• On the command of the Master the crew person is to broadcast to passengers that there is an emergency and they are to prepare to abandon ship.

• The master is to raise the alarm to the Duty Manager via the mobile phone.

• The crew person is to assist passengers to put on lifejackets and release the passenger compartment side curtains.

• Launch floatation rafts – one either side of the vessel and attach to the vessel.

• Launch lifebuoys.

• Order all passengers into the water and to hold onto floatation rafts or lifebuoys.

• Release floatation raft lines.

• Master and crew person into the water, one each side of vessel.

• Check persons in the water.

• Administer first aid as required.

• Maintain lookout for vessels coming to assistance.

• Guide all in the water to the nearest landing point or dock wall ladder/safety chains. [sic]

WQ1’s master and driver had participated in the company’s emergency drills and training programme and had experience of evacuating passengers onto pontoons andtheslipway,buttheywerenotfamiliarwiththeabandon ship procedure describedinthemanual.

1.11.3 Passenger safety announcements

Section10ofTYD’ssafetymanagementmanualincludedtwopassengersafetyannouncements:onetobegivenpriortosettingofffromthepick-uppoint,andonetobegivenwhilestoppedpriortodrivingdowntheslipway(pre-splashdownbrief)(Annex B).

Themanualinstructedthecrewpersontostartthepre-splashdownbriefbymakingajokeaboutgettingoffthevehicle.Duringthesafetybriefthemasterwasexpectedtoexplainthattherewasalifejacketonboardforeveryone;pointoutthelocationofthelifejackets;andtellthepassengersthat ‘in the unlikely event that we need to use them’,amemberofthecrewwillinstructthemontheiruse.Themanualrequiredpassengerstobegiventheopportunitytochoosetowearalifejacketwhileonthewater.Thebriefshouldalsohaveexplainedtheemergencyescaperoutes,

42

themethodofreleasingthesidecurtains,andtheneedtobracethemselvespriortothesplashdown.Accordingtothemanual,thecrewperson’sinitialwaterbasedcommentaryinSalthouseDockshouldincludedetailsaboutthevehicle’slife-savingequipmentandvariousbilgepumpingarrangements.

InearlieriterationsofTYD’sSafety Policy and General Safe Operational Procedures manual (Annex C)thecrewpersonwasrequiredtogiveapracticaldemonstrationofhowtoputonalifejacket.Theearlierbriefsalsomadereferencetobuoyancyfoam.

Thepre-splashdownsafetybriefwasusuallygivenwiththevehicleparkedontheroadoppositetheSalthouseDockslipway.

1.11.4 Pre-splashdown safety checks

FollowingthesinkingofWQ4, the company introduced a procedure for the crew person to check that the hull drain plugs were in place prior to each splashdown. Thiswasdoneduringthemaster’ssafetybrief.ItwasapparentfromtheSouthDocksclosed-circuittelevision(CCTV)footagethatthiswasnotcompletedwhenWQ1 was parked on the top of the slipway at 1530 on 15 June 2013.

1.11.5 Life saving appliances

WQ1carried2,12personbuoyantapparatus14(solidfoam-filledflotationrafts)and4lifebuoys15.Thismetthelifesavingappliances(LSA)requirementsfor32personsassetoutintheMerchantShipping(Life-SavingAppliancesForPassengerShipsOfClassesIIIToVI(A))Regulations1999.Theflotationraftswerestowedontheroofofthevehiclesothattheycouldfloatfreeifthevehiclesankorcapsized.

WQ2carriedthesameLSAasWQ1but WQ4 and WQ8 carried one 20 person flotationraftand4lifebuoys.ThisdidnotmeettheregulatoryrequirementsasitonlyprovidedsufficientLSAfor28people.

1.11.6 Personal flotation devices

TYDwasrequiredtoprovidesufficientpersonalflotationdevices(PFD)16 for everyoneonboardandensurethattheywerereadilyaccessibleandtheirpositionswereclearlyindicated.Upuntil2012,TYDvehiclescarriedsolidfoambuoyancyaids,butatthetimeoftheaccidentWQ1 was carrying manually-activated gas-inflationlifejacketsforitsadultpassengers(passengersweighing32kgandover).Thegas-inflationlifejacketsprovided150Nofbuoyancy.

Thelifejackets (Figure 28)werepackedintransparentplasticbagsandstowedinaboxunderoneoftheinwardfacingbenchseatsatthebackofthevehicle.Solidfoamlifejacketsforchildrenweighinglessthan32kgwerealsoprovided.ThelifejacketsonboardWQ4 and WQ8werestowedinnetsabovethepassengerseats(Figure 28);whentheDUKWswereinspectedbytheMAIBandMCAsomeofthelifejacketswerestillintheirvacuumsealedpackaging.

14 Buoyantapparatus-flotationequipment(otherthanlifebuoys,lifejacketsandbuoyancyaids).15 Thebuoyantapparatusmaybesubstitutedbylifebuoysuptoamaximumof60percentofthevessel’sbuoyantapparatusrequirement,witheachsuchlifebuoybeingsuitabletosupporttwopersons.

16 Personalflotationdevicesaredividedintothefollowingtwomainclasses:thosewhichprovidefaceupin-watersupporttotheuserregardlessofphysicalconditions(lifejackets);andthosewhichrequiretheusertomakeswimmingandotherposturalmovementstopositiontheirfaceoutofthewater(buoyancyaids).

43

Child’sbuoyancyaid Adultlifejacket

Plasticbag

Lifejacketdonninginstructions

Adultlifejacketsinvacuumsealed packaging WQ8

Figure 28: TheYellowDuckmarinelifejacketsandbuoyancyaids

44

1.12 THE YELLOW DUCKMARINE MAINTENANCE MANAGEMENT SYSTEM

1.12.1 Planned maintenance routines

TomeettherequirementsoftheMCAandVOSA,TYD’sgaragestafffollowedanintegratedpassengervesselandPSVplannedmaintenanceregime.Thepaper-basedmaintenancemanagementsystemallowedthemechanicstorecordthemarine and road routines on a common set of maintenance sheets. The mechanics’ daily,weeklyandmonthlymaintenanceroutineswererecordedonthecompany’sA,B and C safety inspection and service sheets (Annex D).

Anydefectsfoundduringtheinspections,andtheactionstakentorectifyormitigatethem,hadtoberecordedonthemaintenancesheets.Themechanicorengineerwho had completed the inspection, and carried out the service and repairs, was requiredtosignadeclarationconfirmingthattheworklistedhadbeencarriedoutandthathe/sheconsideredthevehicletoberoadworthy.

Thecompanyhadalsoprovidedtwodailychecklists:oneforthemaster/driverandoneforthecrewperson/tourguidetocomplete(Annex A) prior to driving the vehicle out of the garage each day.

1.12.2 Unplanned breakdown repairs

AsdescribedinSection1.12.1,themajorityoftheunplanneddefectmaintenancewasrecordedbythemechanicsontheplannedmaintenancerecordsheets.Theengineeringmanagersalsomadecontemporaneousnotesinhandovernotebooksdetailinganyunusualdefectsthathadbeenidentifiedduringtheirshift.

Ifthemastercalledthegaragetoreportaproblemontheroadoronthewater,the engineering manager would deploy a mechanic, or team of mechanics, in the garage’sbreakdownresponsevan.IfavehiclewasimmobilisedonthewaterasecondDUKWorthecompany’srescueboatmightbeusedtoassist.

AreviewofTYD’srecordsestablishedthatthethreemostcommonproblemsonthewater involved leaks due to hull corrosion, loss of steering and the failure to engage six-wheeldrive.Theseproblemsoccurredonaregularbasis.Thefailuretoengagesix-wheeldriveoftenresultedinthevehiclebeingunabletoclimbtheSalthouseDockslipwayandtheneedtoexitthewaterfromCoburgDock.Onthedayoftheaccident, WQ2 and WQ8werebothexitingthewaterfromCoburgDock.

On14June2014,thedaybeforetheaccident,WQ1loststeerageandhadtobetowedalongside.Thebuoyancyfoaminthesterncompartmenthadshifted,causingtherudder’stillerarmtojam.Thegarageteamremovedsomeofthebuoyancyfoamandrepositionedtherest.Accordingtothemaintenancerecords,someoftheadditionalfoamfittedaftertheWQ4sinkingwasalsoremovedfromthebowcompartmentandthecentralvoidspacebecauseithadcausedtheenginetooverheatandhadmadecontactwithrotatingshafts.TheMCAhadbeenmadeaware of the overheating issue and had authorised the removal of some foam from thebowcompartment.

45

1.13 VEHICLE SURVEYS AND INSPECTIONS

1.13.1 Maritime and Coastguard Agency

Asaminimum,theMCA’spassengershipcertificateshadtoberenewedat5-yearlyintervals.Toensurecompliancewiththeapplicablemerchantshippingregulationsforpassengershipconstruction,LSAandcollisionavoidance,theMCAsurveyedeachofTYD’svehiclesannually.ItalsocarriedoutannualinspectionstocheckcompliancewiththerequirementsoftheDSMCode.WQ1 was last surveyed on 25 May2013,followingthemodificationsmadeasaresultoftheWQ4 sinking.

Inordertocheckthewatertightintegrityofthevehicle’shulltheMCAsurveyorscarried out ‘dip’ tests. The dip tests were carried out during each annual survey and involvedtheinternalvisualinspectionofthehullwiththevehicleinthewater.On15February2013,severalpinholeswereidentifiedinWQ1’s hull during an annual surveydiptest.Thesurveyorraisedadeficiencyandinstructedthecompanytorectifythedefectbeforeallowingthevehiclebackonthewater.Thecompany’smaintenance records showed that similar corrosion related leaks were not uncommonandthatthegaragestaffhadfrequentlycarriedouttheirowndiptests.

TheMCAalsocarriedoutrandomincognitoinspectionsofthevehicleswhiletheywereinservice.ThemostrecentincognitoinspectiontookplaceonboardWQ4 on 7May2012.Priortosplashdown,thecrewmadeapointoftellingthepassengersthat a wave of water would enter the vehicle through the door and wash through the passenger space. The surveyor noted that the crew did not close the aft doorway and, when challenged, they confessed that they had not done so for some time anddidnothavethedoorwaysplashguardonboard.Oncompletionofthetour,thesurveyoridentifiedhimselftothemasterandinstructedhimtoceaseoperationsand return to the garage. The surveyor wrote to the owner advising him of his observations.Inhisletter,thesurveyorpointedoutthatthepassengerswerenotgivenalifejacketdonningdemonstration.Healsowarnedthatreferencesmadetothe sinking of the TitanicandthefilmThe Poseidon Adventure might cause offence to some passengers.

1.13.2 Vehicle and Operator Services Agency

VOSAcarriedoutannualinspectionsofTYDvehiclesatitscommercialvehicletestcentreinLiverpool.Italsocarriedoutrandomroadsidevehicleandgaragebasedmaintenance inspections.

Inthe5yearsprevioustotheaccident,VOSAinspectorshadcarriedoutninerandominspectionsofTYDvehicles(8ontheroadand1atthegarage)andhadissuedprohibitionnoticestothecompanyduringsevenofthem.Fiveofthesevenprohibitionnoticeswereissuedforsteeringrelateddefects.Thegaragebasedinspectionidentifiedthatthecompanywasnotcomplyingwiththemaintenanceregimedescribedinitsstatementofintent.TheprohibitionrateforTYDvehicleswas88%,whichwasalmostfivetimesthenationalaverage(18%).Asaconsequence,TYDwassubjecttoaTrafficCommissioners’publicenquiry,whichhadbeenscheduled to take place the week after the WQ1 sinking.

46

1.14 INDUSTRY GUIDANCE AND INSTRUCTIONS FOR SURVEYORS

Atthetimeoftheaccident,therewereseveralothertypesofAPVsbeingoperatedwithintheUK.TherewerealsootherconvertedDUKWsbeingoperatedascodedvessels, which were permitted to carry a maximum of 12 passengers, that had nobuoyancyfoamfitted,anddidnotmeetthemandateddamagedsurvivabilitystandard.TherewasnoindustrybodyoranAPVoperators’codeofpracticetorefertoforbestpractice.

TheMCAandVOSAhadworkedtogetherinthepasttofacilitatethesafeintroductionandoperationofAPVs.However,by2013bothorganisationswereworkinginisolationandneitheremployedstaffwithdesignatedAPVexpertise.

Tofacilitatesurveysandensureconsistency,eachMCAsurveyorwassuppliedwiththeMCA’sInstructions for the Guidance for Surveyors. This was a detailed document that covered the numerous areas surveyors could expect to encounter duringvesselsurvey.Theguidecontainednospecificguidelinesrelatingtoamphibiousvehiclesandnoinformationregardingtheinstallationofbuoyancyfoam.

1.15 PASSENGER QUESTIONNAIRES AND FEEDBACK

Followingtheaccident,theMAIBinterviewedseveralofthepassengersandgatheredadditionalfeedbackusingpassengerquestionnaires.Allthosewhowereinterviewedandthosewhocompletedthequestionnairesraisedconcernsrelatingtothequalityofthesafetybrief.Mostsaiditwastooquick,unclearandnottakenseriouslybythecrew.

Ofthosewhorespondedtothequestionnaire(12passengers),62%thoughtthattherewerenotenoughlifejacketsonboardandhalfsaidtheydidnotknowhowtoput them on or operate them. The father of the 2 year old child struggled to put a buoyancyaidonhisdaughterbecausethewaiststrapwasbuckledshut(Figure 28) andhadnotbeenextended.

Therewasnopositivefeedbackinthequestionnairesrelatingtotheactionsofthecrewand38%ofrespondentssaidthatthecrewpanicked.Duringinterview,several of the passengers empathised with the crew and said that the speed with which WQ1flooded,thelocationofthelifejacketsandtheconfinednatureofthepassengerspacemadeitalmostimpossibleforthemtoco-ordinateandcontroltheevacuation process.

1.16 SIMILAR INCIDENTS AND ACCIDENTS

1.16.1 The Yellow Duckmarine

TheMAIB’sdatabasecontainedrecordsoftwopreviousaccidentsinvolvingTYDDUKWs.ThefirstwasacollisionbetweenWQ2andtheTYD’spassengerboatSkylarkinWappingDockon28January2013;thesecondwasthesinkingofWQ4.

TheTYD’sinvestigationintothecollisionbetweenWQ2 and Skylark found that the skipper of Skylarkhadbeenfoolingaround.Hehaddeliberatelymanoeuvredhisvessel at speed, in close proximity to WQ2, in order to create a wake and induce a rollonboardtheDUKW.

47

WQ4 sank while under tow very close to where WQ1 sank (Figure 2). Prior to thesinking,thevehiclehadsufferedasteeringfailureandhadbeentowedontoapontooninSalthouseDockbyanotherDUKW.Duringtherecovery,thecrewstoppedtheengineandthenrealisedthatwaterwasenteringthebilges.Uptothatpoint, the vehicle’s chain-driven Higginspumphadbeencontrollingthelevelofwaterinthehull.Oncealongside,thepassengerswerelandedandsupportarrivedfromthegarage.Unabletorestarttheengine,theydecidedtousetherescueboattotowthefloodingvehicletotheSalthouseDockslipway,andwhileundertowitsank.When WQ4wasrecoveredfromthedockitbecameapparentthatoneofthehulldrain plugs was missing.

Inadditiontothelackofbuoyancyfoam,theMCAinvestigationintothesinkingofWQ4 highlighted that: the company’s maintenance standards, the crew’s knowledge ofthevehicles’bilgesystems,andtheimplementationofthecompany’sproceduresfellwellshortofreasonableexpectations.Asaresultofitsfindings,theLiverpoolMarineOfficerecommendedthatthefindingsofitsinvestigationreportbetakenintoaccountatacentralpolicylevelandthattheOrpingtonMarineOfficecarryoutafullreviewofthebuoyancyarrangementsonboardtheLDTvehicles.

Duringtheinitialstagesofthisinvestigation,itbecameapparentthatseveralotheraccidentsandincidents,whichhadnotbeenreportedtotheMAIBortheMCA’sLiverpoolMarineOffice,hadoccurredinthemonthsleadinguptothesinkingofWQ1.Notably,inJanuary2013anincidentoccurredthatrequiredpassengerstobeevacuatedfromaDUKWinSalthouseDock,andon1June2013WQ2 suffered a hull failure.

Thefirstincidentoccurredafterasteeringfailure.Insteadofcallingthegarageandarrangingtobetowedtoapontoon,themasterdecidedtoattempttomanoeuvrehis vehicle astern onto the slipway, using his front wheels to steer through the water. Ashisvehicle’srearwheelslandedontheslipwaythevehicleveeredtoport,felloffthe concrete ramp, heeled over, and then wedged itself against the dock wall. The passengers were evacuated onto the dock side through the aft door.

TwoweeksbeforeWQ1 sank, WQ2 suffered a hull failure around one of the propeller shaft v-strut support shoes (Figure 29).Thefailurewasidentifiedbythegaragestaffafterthemasterhadreportedheavyflooding.Thehullplatingaroundthestarboardv-strutfootwasfoundtobeheavilycorroded,andtheaffectedsectionof steel was cropped out and an insert plate was welded in place (Figure 30).Adoublerplatewasfittedontheinsideofthehullandsealedusingasiliconbasedcompound.Thev-strutwasboltedtothehulldoublerplate,butthedoublerplatewasnotconnectedtotheafttransversebulkhead.TheMCAwasnotmadeawareofthe repair.

TYDalsohadseveralincidentsontheroad.InSeptember2010,fourpassengerswereinjuredandtakentohospitalwhenawheelfelloffWQ2 whileitwasbeingdriven along a city centre street.

48

Figure 29: SimilarhullfailurediscoveredonboardWacker Quacker 2 two weeks earlier

1 June 2013

Starboardv-strutfoot

V-strut support shoe

Hole torn in propeller shaft tunnel plate

49

External repair

Resinsealant

Insert

Internalrepair

Siliconesealant

Steelplate

Figure 30: Hullrepairscarriedoutbygaragestaff

50

1.16.2 Beatrice

On31March2001,theLondonFrogCompany’sDUKWBeatrice suffered a hull failure17andstartedtotakeonwaterwhenitspropellerwasfouledbyalargefloatingobject.ThemasterraisedthealarmandthenbeachedhisvehicleonthenorthbankoftheRiver,oppositeLambethFireStation.Beatrice’s 29 passengers were safely evacuatedontoaLondonFireBrigadevessel.

Beatricesufferedsevereunderwaterdamage.Itspropellershaftwasbent,andthev-strutmountingshadbeenrippedoutofposition.Thiscausedthepropellertunnelplatingtoruptureandthehulltostartflooding.Thepassengerswereawareofthewaterenteringthehullandtookituponthemselvestostartdonninglifejackets.Theyhadnotbeengivenalifejacketdemonstrationandfounditdifficulttoputthemon,particularlyasthestrapshadnotbeenextended.

DamagedsurvivabilityandtheuseofbuoyancyfoamwasnotdiscussedintheMAIBinvestigation report of the accident.

AsaresultoftheaccidentTheLondonFrogCompanyreplacedalargesectionofthepropellertunnelplatingonallitsvehicleswithheaviergaugesteel.Itfurtherreinforcedtheareaaroundthev-strutmountingbyweldingandrivetinganadditionalthree layers of 2mm thick plate to the propeller tunnel (Annex E).

The propeller tunnel plating was not strengthened on the Liverpool DUKWs. FollowingthesinkingofWQ4, LDT had written to TYD to offer its help and make some recommendations to improve future safety. The strengthening of the propeller tunnelplatingwasoneoftherecommendationsmadeintheletter.AcopyoftheletterwaspassedtotheMCAatthetime.

1.16.3 Miss Majestic

On1May1999,theDUKWAPVMiss Majestic (Figure 31) sankbythestern,withadriverand20passengersonboard,duringaregularsightseeingtouronLakeHamilton,Arkansas,USA.About7minutesafterenteringthewaterthevehicleflooded,listedtoportandrapidlysankbythesternin19mofwater.Thedriverandsevenpassengersescapedfromthesinkingvehicleandwererescuedbypleasureboatersinthearea.Theremaining13passengers,3ofwhomwerechildren,losttheir lives.

ThemajorsafetyissuesidentifiedintheUSGovernment’sNationalTransportationSafetyBoard’s(NTSB)investigationreport18 related to:

• Vehicle maintenance• CoastGuardinspectionsofMiss Majestic• CoastGuardinspectionguidance• Reservebuoyancy,and• Passengersurvivability.

17 MAIBinvestigationreportnumber3/2001–Report on an investigation of a propulsion failure and subsequent beaching of the Class V amphibious passenger craft Beatrice on 31 March 2001, opposite the River Thames Fire Station, Lambeth.

18 NTSB/MAR-02/01–SinkingoftheAmphibiousPassengerVehicleMiss Majestic, Lake Hamilton, Near Hot Springs,Arkansas,May1,1999.

51

When Miss Majesticwasinspectedaftershehadbeenrecoveredfromthelakebed,theinvestigatorsdiscoveredthatthereardriveshafthousing’saftrubberboothadworkedloosebecauseithadnotbeenfittedcorrectly(Figure 31).

TheUSDUKWswerenotfittedwithbuoyancyfoamandreliedonbilgepumpstoprovidedamagedsurvivability.TheNTSBreportstatedthatthelackofreservebuoyancyneededtokeeptheDUKWafloatinthefloodedconditionwasaflawinitsdesignwhenoriginallyconvertedforpassengerserviceandwasamajorcontributoryfactor.

TheNTSBinvestigationfounddeficienciesintheoperator’smaintenanceproceduresandthestandardofUSCoastGuardinspections.Thereportalsoidentifiedproblemswiththeaccessibilityofthelifejacketsandhighlightedthatthepassengers were not shown how to don them.

View after the accident of the vehicle’sreardriveshafthousingboot

Miss Majestic after its conversion to a commercialamphibiouspassengervehicle

Figure 31: SinkingofDUKWMiss Majestic on1May1999

52

TheNTSBreportconcludedthat:

• TheprobablecauseoftheuncontrolledfloodingandsinkingofMiss Majestic wasthefailureofthevehicleoperatortoadequatelyrepairandmaintaintheDUKW.

• Miss MajesticsankbecausetheDUKWhadnowatertightbulkheadsandnoreservebuoyancy,andbecauseitsHiggins pump failed.

• Thecanopywasamajorimpedimenttothesurvivalofthepassengers.

• TheCoastGuard’sinspectionprogramforMiss Majesticwasinadequateandcursory.

• ThelackofCoastGuardguidanceandtrainingfortheinspectionofDUKWscontributedtotheinadequateinspections.

• Thewearingoflifejacketsbeforethevehicleentersthewater,wherecanopiesare removed, would improve passenger safety.

Priortopublishingitsreport,theNTSBissuedthefollowingrecommendationtotheoperatorsandrefurbishersofAPVs:

Without delay, alter your amphibious passenger vessels to provide reserve buoyancy through passive means, such as watertight compartmentalization, built in floatation, or equivalent measures, so that they will remain afloat and upright in the event of flooding, even when carrying a full complement of passengers and crew.

Onlyoneofthe30APVoperatorsintheUSacceptedthisrecommendation.TheDUKWoperatorsarguedstronglythatcompartmentalizationwouldintroduceunacceptablelevelsofstressinthehullwhenoperatingontheroad,andtheinsertionofbuoyancyfoam wasnotpracticalandwouldintroducemanyproblems.

In2001,theUSCoastGuardpublisheditsNavigationandVesselInspectionCircular19(NVIC)1-01,Inspection of Amphibious Passenger Carrying Vehicles. The aimoftheNVICwastoprovideaguidetobestpracticefortheinspection,operationandmaintenanceofAPVssothataconsistentapproachcouldbetakenacrossallregionsoftheUS.Theguidanceforhullsurveysincludedalistofareasofconcern,one of which was the propeller shaft v-strut and its connection to the hull.

TheNVICalsoprovidedguidanceontheinstallationofbuoyancyfoam.TheCoastGuardacknowledgedthatbuoyancyfoamcouldbeusedtoaugmentsubdivision,butwarnedthatitsinstallationgenerallyaggravatesmaintenanceproblemsbyrestrictingaccess. The guidance also pointed out that, as foam deteriorates with age, periodic samplingisrequiredtodetermineitscondition.

WhentheNTSBreportwaspublishedin2002itreissueditsearlierrecommendationbutprovidedalistofrecommendedinterimmeasuresforoperatorswhohadyettoprovidereservebuoyancy.Theseincluded:

19 TheUSCoastGuardHeadquartersissuesNVICstodisseminaterecommendedpolicy,requirements,procedures,orguidanceforCoastGuardmarinesafetypersonnelandthemarineindustry.

53

• Removalofcanopieswhenonthewater.

• Removalofunnecessaryhullplugsandreductioninsizeofhullpenetrations.

• Installationofindependentlypoweredelectricbilgepumpstosupplementthechain-driven Higginspump(orsimilarcapacitydewateringpump).

• Installationofhighlevelbilgealarms.

• CompliancewiththeprovisionsoftheUSCoastGuard’sNVIC1-01.

When WQ1sank,mostoperatorsoforiginalDUKWAPVsintheUShadimplementedtheinterimmeasuresprescribedintheNTSBreportbuthadnotmetthereservebuoyancystandardsrecommended.

1.16.4 DUKW 34

On7July2010,abarge(The Resource),beingpushedbythetugboatCaribbean Sea,collidedwiththeanchoredAPVDUKW 34ontheDelawareRiverinPhiladelphia,USA.DUKW 34 sank and two of its 35 passengers drowned. The APVhadsufferedanenginecoolingproblemthatthemasterhadinitiallythoughtwasafire.Themasterstoppedhisengineandshutdowntheenginecompartmentbeforeanchoringintheriver.Thecauseofthecollision(Figure 32)wasattributedtoafailuretomaintainanappropriatelookoutbythecrewonboardthetugthatwaspushing The Resource.TheNSTB’sinvestigationreport20 explained that DUKW 34’s master did not fully appreciate the risk of collision when he anchored in a navigation channel.Thereportalsocommentedonthejocularnatureofthemaster’ssafetybriefandthefactthatcontrarytocompanyprocedurehedidnotdonalifejacketduring the demonstration.

20 NTSB/MAR-11/02–CollisionofTugboat/BargeCaribbean Sea/The ResourcewithAmphibiousPassengerVehicle DUKW 34 Philadelphia, Pennsylvania, July 7 2010.

Figure 32: Collisionbetweentugboat/bargeCaribbean Sea/The Resource and DUKW 34 on 7 July 2010

Barge The Resource running over DUKW 34

54

1.16.5 Lessons learned during World War 2 operations

In1944,theUSarmyreviewedthelessonslearnedduringthefirst2yearsofDUKWoperationsinEuropeandthePacificandcarriedoutapracticalstudy.Itsfindingswere used to produce a revised set of standing operating procedures and technical data, which was documented in The DUKW, Its Operation and Uses manual.

Themanualprovidedup-to-dateinformationdesignedtopromotethemostefficientutilizationofDUKWs,andprescribedseveralmodificationstoreducethelikelihoodofvehiclebreakdowns,floodingincidentsandfires.Themodificationsincludedtheretrofittingofbottomplugstowageracksinthedriver’scompartment,andpropellerguards (Figure 33).

Themanualwarnedofthedangersofenteringthewaterwithoutfirstcheckingthebottomplugsareinplace,andinstructedtheDUKWengineerstopaintwarningnotices next to the plug stowage racks. The army had discovered that the DUKWs wereparticularlysusceptibletopropellerfoulingandprovidedinstructionsforthefittingofpropellerguards.

Drain plug stowage and caution notices

Figure 33: World War 2 operations manual

Propeller guard for DUKW

55

SECTION 2 – FACTUAL INFORMATION CLEOPATRA

2.1 PARTICULARS OF CLEOPATRA AND ACCIDENT

SHIP PARTICULARSVessel’s name Cleopatra

Flag UKCertifying authority MaritimeandCoastguardAgencyVessel type Class V passenger vesselVehiclechassisnumber 3536279Vehicle type DUKWamphibiouspassengervehicleYearofbuild 1945(convertedtoAPVin1999)Registeredowner London Duck Tours LtdConstruction SteelHull length 9.45mLength overall 10.35mWidth overall 2.48mHeightofcanopy(fromtheroad) 3.8mOnroadweight 8.66t(2009weighbridgeweight)VOYAGE PARTICULARSPort of departure Lacks Dock, London Port of arrival Lacks Dock, LondonType of voyage Sightseeingtour,tidalriverestuaryManning 2MARINE CASUALTY INFORMATIONDate and time 29September2013atabout1154Type of marine casualty or incident VerySeriousMarineCasualtyLocation of incident RiverThamesnearTowerGardens,Westminster

Placeonboard Enginebay,shafttunnelandpassengercompartment

Injuries/fatalities NilDamage/environmentalimpact NilShipoperation SightseeingtoursVoyage segment Transit

Environment Daylight;fair;lightwinds;ebbtidestream<1.75knots

Personsonboard 30(2crew+28passengers)

56

2.2 BACKGROUND

On20June2013,twosurveyorsfromtheMCA’sOrpingtonMarineOfficeattendedtheLDTfleetmaintenancegaragetodeterminethevolumeofbuoyancyfoampresentintheLDTvehicles.Thebuoyancyfoamwasremovedfromtherandomlyselected DUKW Rosalind and weighed. Having estimated the foam density, the surveyorscalculatedthatthevehiclehadcontainedabout6m3offoam.Asimilarexercise carried out on DUKWs Titania and Elizabeth determined that they containedabout5.5m3and5.8m3 of foam respectively.

FollowingdiscussionswiththeMCA,LDTsuspendeditswaterborneoperationsandundertooktoincreasethelevelsofbuoyancyfoamfittedtoitsvehiclesto9m3.On1July2013,theMCAinstructedLDTtofurtherincreasethevolumeofbuoyancyfoamineachvehicleto10.6m3.

On7August2013,followingtheanalysisoftheresultsofthepracticaltrialsconductedbytheMAIBinLiverpool[Paragraph 1.10.3],theMAIBmadearecommendation(2013/221)totheMCA:

require operators of DUKW passenger vessels in the UK to demonstrate that they are able to provide 110% effective residual intact buoyancy in their vessels, and where buoyancy foam is fitted for this purpose, the quantity installed is measured by volume and the foam does not impede the operation or maintenance of key equipment.

Bythistime,sixofLDT’svehicleshadbeenreturnedtothewater.Cleopatra was re-packedwithfoambetween12and22Augustand,on24August2013,returnedtothewaterwith10.6m3ofbuoyancyfoamonboard.On24September2013,ateamofMCAsurveyorsarrivedattheLDTgaragetowitnessandassesstheinsertionofbuoyancyfoamintoDUKWPortia.

2.3 NARRATIVE

At1110on29September2013,DUKWCleopatra (Figure 34)leftitsSouthbankdeparture point and commenced a standard tour of the Westminster area of London, England.Onboardwerethedriver,thetourguideand28passengers.Uponcompletion of the on-land portion of the tour, the vehicle arrived at Lacks Dock slipway, Vauxhall, where the driver handed control of the vehicle to its master for the river section of the tour. While the driver and master carried out their routine handover inspection, the tour guide gave the passengers a pre-splashdown safety brief,whichincludedademonstrationofhowtodonthepassengerbuoyancyaids.

At1145,themasterdroveCleopatradowntheslipwayintotheRiverThamesandheadeddownriver.AsCleopatraapproachedLambethBridge(Figure 35), the tour guide noticed an unusual smell. He told the master, who then checked thatthevehicle’shandbrakewasdisengaged.Thesmellsoondispersedandthemaster continued on with the tour. When Cleopatra approached Westminster Bridgethemasterturnedthevehiclearoundandcommencedthereturnjourneyupriver.AsCleopatra passed the Houses of Parliament, the tour guide noticed the unusual smell again, and saw wisps of light grey smoke coming from the engine compartment’s cooling air inlet (Figure 36).Ataboutthesametime,someofthepassengersnoticedlightgreysmokerisingfromthefloorclosetothetourguide’sleft leg.

57

Initially,themasterthoughtthatthesmokemighthavebeencausedbyanengineexhaustproblemand,asaprecaution,heangledCleopatra towards the shore. The tour guide then made an announcement to the passengers informing them that there wasaproblemwiththevehicle.Thesmokeincreasedinthickness,darkenedincolourandbegantobillowoutoftheenginecompartment’sstarboardextractionfanvent (Figure 36).Atthispointapassengertowardstherearofthevehicleshouted,“Fire! Fire!”

Figure 34: LondonDuckToursamphibiouspassengervehicleCleopatra

ImagecourtesyofPhilipBisset

58

Figure 35: ThewaterbornesectionoftheLondonDuckTourssightseeingroutes

LambethBridge

LacksDockSlipway

ReproducedfromAdmiraltyChartBA3319-0bypermissionoftheControllerofHMSO

Siteofgrounding

59

Engine compartment forward cooling air outlet vent

Engine compartment cooling air inlet vent

Engine compartment extraction fan vent

Figure 36: Engine compartment cooling air vents on a similar London Duck Tours vehicle

60

At1154:20themastercalledthePortofLondonAuthority(PLA)VesselTrafficServices(VTS)byradioandaskedfortheassistanceofthefirebrigade.BeforeVTScouldacknowledgethecall,theradioonboardCleopatra ceased functioning. Nevertheless,thebriefbutanxiousradiocallhadbeenheardbynumerousrivercraftoperatorsandanunknownsourceadvisedthePLAVTSbyradiothatCleopatra wasonfireclosetotheHousesofParliament.ThetransmissionwasoverheardbyThamesCoastguard,anditimmediatelynotifiedtheLondonFireBrigade(LFB)andtheRoyalNationalLifeboatInstitution’s(RNLI)TowerPierandChiswicklifeboatstations.ThecoastguardthenadjustedthepositionofaWestminsterCCTVcameraandbegantomonitorthesituationonitsvisualdisplayscreen.

The master continued to steer Cleopatra towards the shore until the vehicle’s front wheelstouchedbottombyVictoriaTowerGardens(Figure 35).Onceonthemud,the master kept the propeller engaged ahead to keep the craft pressed against the shore.Asthefireescalated(Figure 37),manypassengersfledinpanictowardstherear of Cleopatra,somegrabbingbuoyancyaidsastheydidso.Otherpassengersclimbedoutofwindowswhentheyrealisedthatthiswastheireasiestandquickestescaperoute.Thepassengerwhohadfirstraisedthealarmsucceededinpartiallyopeningthereardoorbutwasquicklypushedasidebyotherpassengers,separatinghim from his family.

The tour guide physically pushed his way through the passengers with the intention ofloweringtheboardingladdertoassistevacuation.However,theweightoftheescaping passengers standing upon the raised port ladder prevented the guide from untyingthesecuringropeandloweringtheladder.Duetothebarrierpresentedbythepartiallyopendoorontheboardingplatform, peoplehaddifficultyaccessingandloweringthestarboardladder,resultinginmanyhavingtojumpintothewater.Aspeopledispersedfromtheboardingplatform,thestarboardladderwaseventuallylowered and the door fully opened, allowing the last evacuees to descend the stairs to the water.

Themasterwasthelastpersontoleavethevehicle,andonlydidsoafterconfirmingthat the passenger compartment was empty of people.

Within 3 minutes of the master’s incomplete radio transmission, three rigid-hulled inflatableboats(RIB)fromtheThamesRIBExperienceCompany21 arrived on scene. By this time, some passengers had managed to wade to shallower water. SomeofCleopatra’spassengersabandoneddirectlyontotheRIBsandtherestwere recovered from the water (Figure 37). Within 2 minutes of their arrival, all passengersandcrewhadbeenrecoveredontotheRIBs.

At1159,theLFBfireboat,Fire Dart,arrivedatthescene,closelyfollowedbyalandapplianceontheriverbank,adjacenttoCleopatra.Thefirewasquicklybroughtundercontroland,by1201,itwasextinguished(Figure 37).

TherescuevesselscarriedthesurvivorstoLFB’sLambethRiverFireStationpontoonwhereall30weredisembarkedandmetbyambulancestaff.Threeofthepassengersweretakentohospitalforfurthermedicalexamination,butweredischarged later that day.

21 TheThamesRIBExperienceCompanyoperatedhighspeedRIBtoursfortouristsontheRiverThames.

61

Figu

re 3

7:Abandonshipandrescuesequence

ImagescourtesyofD

ailyMailonline©Tony.MargiocchiBarCraftmedia

62

2.4 ENVIRONMENTAL CONDITIONS

Itwasadryday,withamoderate(11-16kts)east-north-easterlywindandaneaptide.ThedepthofwaterbetweenLambethBridgeandWestminsterBridgeathighwater(0915)was5.2m.Atthetimeoftheaccident,theebbingtidalstreamadjacenttotheHousesofParliamentatmid-streamwasabout1.75kts.

2.5 LONDON DUCK TOURS

InNovember2001,theLondonFrogCompanyLtdceasedtradingandwentintoliquidation.InMay2002CapitalFrogsLtd,whohadtakenownershipoftheliquidatedcompany’sassets,re-introducedtheAPVsightseeingtourservice.In2004CapitalFrogsLtdwasrebrandedandchangeditstradingnametoLondonDuckTours.Between2006and2012LDTincreaseditsfleetofDUKWAPVsfrom5to9.LDT’sconvertedDUKWswerecertifiedbytheMCAtooperateduringdaylighthours on Category C22 waters as Class V passenger vessels.

Atthetimeoftheaccident,LDTemployed74people,includingofficestaff,fleetmaintenance operatives, road drivers, river masters, tour guides and freelance staff. ThecompanyhadanofficeandabookingshopinWaterloo,closetotheLondonEye,andafleetmaintenancegarageinBattersea.

LDT’s route covered the Westminster area of London and its guided sightseeing tourlastedabout1hour,whichincluded25minutesontheRiverThames.Thepassengerpick-upanddrop-offpointwaslocatedonChicheleyStreet,directlyinfrontofthecompany’sbookingshop.Thevehiclesenteredandexitedtheriveratthe Lacks Dock slipway, Vauxhall, and cruised as far as Westminster Bridge (Figure 35).

2.6 THE CREW

TheLDTvehicleswerecrewedbytwopeopleatalltimes.Ontheroad,thevehiclewasmannedbyadriver(PSVlicenceholder)andatourguide;ontheriver,itwascrewedbyamasterandatourguide.Thedriverhandedthevehicleovertothemaster at Lacks Dock prior to the vehicle entering the water. During the handover thecrewwererequiredtocarryoutasetofpre-entrychecksanddeliverthesafetybrieftotheirpassengers.

2.6.1 Master

Cleopatra’smasterwasBritishandwas71yearsofage.Hehadbeenawaterman23 and lighterman24ontheRiverThamesformostofhisworkinglife.HehadbeenanLDTDUKWmasterfor18monthsandheldanMCAapprovedTier1Level2Boatmasters’ Licence.

WhenhejoinedLDT,themasterundertookdedicatedfamiliarisationtrainingonDUKWhandlingandhadsincetakenpartinregularemergencydrills,includingfireandabandonment.

22 CategoryCwaters-tidalrivers,estuariesandlarge,deeplakesandlochswherethesignificantwaveheightcouldnotbeexpectedtoexceed1.2matanytime.

23 Waterman - a river worker who transfers passengers across and along city rivers and estuaries in the United Kingdom.

24 Lighterman-acrewmanwholoadsanddischargescargoesonlighterbarges.

63

2.6.2 Tour guide

The tour guide was British and was 27 years of age. He had worked for LDT for 4 months. Prior to working for LDT, he had some experience with watercraft and heldmilitaryqualificationsinthisarea.Duringhis4monthsasatourguidehehadbeengiveninductiontrainingontheDUKWoperationsandhadparticipatedinanemergencydrilllessthan3weeksbeforethefire.

2.7 CLEOPATRA

2.7.1 General description

Cleopatra’s external appearance and passenger compartment layout was similar to that of WQ1.However,asaresultofthedesignmodificationscarriedoutbyLDT,theLondonbasedAPVsweresignificantlydifferenttothoseoperatedinLiverpool.

Cleopatrahadan88kWIvecoBeta120,6-cylinder,naturallyaspirateddieselengineandanAlisonAT545automaticgearbox.Itwaspropelledthroughthewaterbyasingle,4-bladed,clockwiserotatingpropeller,andhadasinglespade-typerudder.

Thecontrolsandinstrumentationforbothroadandrivertransitswerepositionedinfrontof,andadjacentto,themaster/driver’sseat.Thefloorplatesaroundthedriver’sseat had several open penetrations to accommodate the drivetrain gear levers and steering linkages.

Cleopatrawasequippedwithaveryhighfrequency(VHF)digitalselectivecalling(DSC)radioforcommunicationwiththePLA,LDTandothervessels.Navigationonthewaterwascarriedoutvisually.Thevehiclewasequippedwithanautomaticidentificationsystem(AIS),interfacedwithaglobalpositioningsystem(GPS),whichalloweditspositiontobemonitoredremotelybythePLA,LDTandotherAISenabledcraft.ThevehiclewasalsofittedwithalandbasedtrackingsystemthatallowedLDTstaff to monitor its position on the road.

2.7.2 Hull

Cleopatra’sbasichulllayoutwassimilartothatoftheoriginalDUKWandLiverpoolAPVs,buttheminimumshellplatethicknessoftheLDTvehicleshadbeenincreasedto3mm.TheshellplatinginthepropellershafttunnelhadbeenstrengthenedfollowingthelessonslearnedfromthehullfailuresufferedbyDUKWBeatriceinMarch2001(Paragraph 1.16.2).

ItwasastatutoryrequirementforaccommodationspacesonboardClassVpassenger vessels such as Cleopatra,tobeseparatedfrommachineryspacesbygas-tight,insulatedfireprotectionbulkheads.Thisrequirementwashighlightedin1999byBurnessCorlett&PartnersLtdinitsinitialdesignsubmission.Duetoperceivedtechnicaldifficulties,theDUKWoperatorswereexemptedfromtherequirementontheunderstandingthattheopeningsforthevehicles’driveshaftsandcontrollinkageswouldbekepttoaminimum.

64

2.7.3 Propulsion drivetrain

AsimplifiedillustrationofCleopatra’s propulsion drivetrain arrangements is shown in Figure 38.Theoutputdrivefromtheautomaticgearboxwasconnectedtothemarinetransferbox(MTB)byaslidingsplinedshaftviatwoType2030universaljoints(UJ)(Figure 39).ThemainoutputshaftfromtheMTBwasconnectedtotheroadtransferbox.Anarticulateddrivewastakenfromtheroadtransferboxtodrivetheforwardwheelswhileafurtherdrivewastakenaft,throughaband-typehandbrakeassembly,todrivetheaftertwosetsofwheels.Theforwardwheelscouldbeindependently engaged and disengaged from the master’s position to provide a four orsix-wheeldriveconfiguration.

ThepropellershaftwasbroughtintousejustpriortoenteringthewaterbymanuallyengagingtheMTBfromthemaster’scontrolposition.Thepropellershaftwassupportedinternallyalongitslengthbyaplainjournalbearingandapropellerthrustbearing.

The maximum engine speed while on the road was approximately 2200rpm. When thevehiclewasinuseontheriverfirstgearwasmanuallyselectedontheautomaticgearboxandtoachievenormalcruisingspeedtheenginewassetatapproximately1200rpm.TheresultantspeedoftheoutputshafttotheMTBwasapproximately650rpmwithapropellershaftspeedofabout850rpm.

2.7.4 Rudder control

TheLDTfleethadbeenfittedwithelectro-hydraulicsteeringgear.Therudderwascontrolledusingajoysticklocatedonthedashboardnexttothesteeringwheel.Itcouldalsobecontrolledmanuallyinanemergencyfromthebackofthe vehicle, using a handwheel. The emergency steering system had a 25-litre hydraulic oil header tank, which was positioned on the port aft end of the passenger compartment.

2.7.5 Brake system

Thewheelbrakeswereair-assistedandwereoperatedbydepressingthefootpedalat the master’s control position. The engine-driven air compressor charged three, 20-litreairreservoirstoapproximately10bar.Thereservoircylinderswerelocatedunderthepassengercompartmentdeckbehindthemaster’sseat(Figure 40). Flexiblepipesconnectedthecylinderstothefootpedaloperateddistributionvalve,andfromtheretothebrakes.

2.7.6 Engine compartment

Cleopatra’senginecompartmentlayoutwaslastmodifiedin2012;thiswastoaccommodatethevehicle’snewIvecoBeta120engineandAlisonautomaticgearbox.Themodificationsincludedtheenlargementofthedriveshaftholeintheforwardtransversebulkhead,therepositioningoftheengineradiatorandflowalterations for the engine and engine compartment cooling air system.

Aswithmostroadvehicles,Cleopatra’s engine was water cooled and its cooling watertemperaturewasregulatedbyanaircooledradiatorandthermostaticby-passvalve. Traditionally, the radiator is positioned at the front of a road vehicle and air is forcedthroughitbythevehicle’sforwardmotion.TheDUKWradiatorwasencased

65

1 Differential gearboxes

Universal joint couplingsNote: Revolutions relate to “on water” operation.

4-bladed propeller(850rpm)

V-strut support bearing

Stern gland

Thrust bearing

1

Gland

Band brake

Plain bearing

Road transfer box

Marine transfer box

Splined driveshaft 650rpm

Alison AT545 automatic gearbox

Iveco Beta 120 engine (1200rpm)

Rear transmission box

1

1

Wheel

Propeller shaft

Figure 38: SimplifiedillustrationofCleopatra’s drivetrain

66

Splineddriveshaft

Universaljoint

End cap and needle bearings

Figure 39: Typicalslidingsplineddriveshaftanduniversaljointarrangement

67

inasteelhullandthecoolingairhadtobeforcedthroughitbyanengine-drivenfan.The vehicle also had a keel water cooler, which was externally mounted under the bowtosupplementtheengine’saircooledradiator.

OntheoriginalDUKW,coldairwasdrawnintotheenginecompartmentthroughanopendeckgratinginthecargocompartmentbehindthecrewseats(Figure 41). Thecoolingairpassedunderthedriver’scab,acrosstheenginecompartment,andthroughtheradiatorintothebowcompartment.Thehotairthenpassedthroughtheportandstarboardcoolingairpassagesandexhaustedtoatmospherethroughthesideventilationgrilles.Inadditiontocoolingthewaterintheradiator,thecoolingairalso cooled the engine compartment. Earlier DUKW models had an auxiliary cooling airinletflapfittedattheaftendoftheenginecompartmentaccesshatch.Thiscouldbeopenedtoprovideextracoolingairinhotclimates.

When the cargo compartment was converted into the passenger space the original maincoolingairinletgratingwascoveredupandtheenginecompartmentandbowcompartmenthatchcoversweremodifiedtoaccommodateaftfacingcoolingairsupply and exhaust vents (Figure 36). During their 13 years in operation, the cooling airflowsontheLDTDUKWshadbeenalteredseveraltimes.Atthetimeoftheaccident,theoriginalenginecompartmentcoolingairpassageshadbeenremovedfrom Cleopatra. The cooling air was drawn into the compartment through its aft hatch cover vent and, having passed through the radiator, exhausted to atmosphere throughtheforward(bowcompartment)hatchcovervent.

Inordertopreventthebuild-upofhotairattheafterendoftheenginecompartment,electrically-drivenextractionfanswerefitted(Figure 42). The hot air was drawn into theportandstarboardventilationvoidspacesandexhaustedtoatmospherethroughthe vehicle’s original side vents.

Flexiblepipes

20 litre air reservoirs

Figure 40: Braking system air reservoirs

68

Themainengineexhaustpipewasrouteddownthestarboardsideoftheenginecompartment.Itpassedthroughtheupperpartofthestarboardventilationvoidspaceandterminatedonthedeckatthestarboardforwardcornerofthecrewcab.Thesectionoftheexhaustpipewithintheenginecompartmentwaslaggedbutthesection that passed through the void space was not (Figure 43).

Bow

com

partm

ent

Auxilliarycoolingairinletflap

Starboardcoolingairoutletgrille

Coo

ling

air i

nlet

gra

ting

Bow

com

partm

ent h

atch

cov

erE

ngin

e ra

diat

orP

ort c

oolin

g ai

r pas

sage

Por

t coo

ling

air o

utle

t gril

leC

argo

com

partm

ent

Figu

re 4

1:OriginalDUKWenginecompartm

entcoolingairsystem

69

Airextractionfan

Eng

ine

com

partm

ent v

entil

atio

n vo

id s

pace

Coo

ling

air o

utle

t and

hot

air

extra

ctio

n

Coolingairinletandflow

throughtheengineradiator

Simplifiedillustration

Figu

re 4

2:Modifiedenginecoolingairflow

andenginecompartm

enthotairextraction

70

Engine compartment

Lagged section of the engine exhaust pipe

Unlagged section of the engine exhaust pipe

Cleopatra’sstarboardventilationvoidspacefollowingthefireandresidualfoamremoval

Figure 43: Cleopatra’s engine exhaust pipe

71

2.7.7 Engine compartment extraction fan controls

Theextractionfanswereconfiguredforautomaticstartandmanual-overridecontrol.Thestarboardextractionfanwaswiredtostartautomaticallywhentherudder’selectro-hydraulic steering system was started in readiness for river operations. The portextractionfanwasconfiguredtostartiftheenginecoolingwatertemperaturereached95ºC.Themastercouldindependentlystarttheportextractionfanfromhisseatedpositionifnecessary.Itisunderstoodthatthestarboardfanwasrunningandtheportfanwasstoppedwhenthefirebrokeout.

2.7.8 Passenger compartment

Cleopatra’s passenger and crew seating arrangements (Figure 44) were similar to those of WQ1. The side curtains were also similar in design and secured in the closedpositionbyVelcro,butdidnothavetheadditionalelasticatedcords.Theforwardwindscreenandsidewindowsadjacenttothemaster’sandtourguide’sseatsweretoughenedglassinlinewithPSVrequirements.Thesteelfloorplatesunderthecrewseatswerecoveredbyasyntheticcarpet.

Accessandegresstothepassengercompartmentwasbywayofaboardingplatformatthebackofthevehicleandanaftwatertightdoor.UnliketheLiverpoolDUKWs,theboardingplatformhadtwofoursteppedladders,oneontheleftandoneontherighthandsideoftheboardingplatform.Eachladdercouldberaisedorloweredbymeansofarope,whichwassecuredtoasmallsetofcleatsontheaftbulkhead(Figure 45).Duringroadjourneys,oneladderwasrequiredtobeleftintheloweredposition,whereasontheriverbothladderswerestowedintheirraisedposition.Generally,forroadsideembarkationanddisembarkation,thevehiclewasparkedontheleftsideoftheroadandtheleft(port)ladderwasused.Tofacilitatethis,thedoorwassecuredat90ºtotheaftbulwark(Figure 45).Thedoorcouldalsobeopenedthrough180ºandsecuredflatagainstthebulwarktoallowaccessandegressviatherightsideladder.

Priortothefire,thetourguidewasseatedfacingthedriver,withhislegsinthecompanionway so that he could interact and maintain eye contact with passengers. The vehicle’s side curtains were secured in the open position.

2.7.9 Engine fuel oil system

Cleopatrahada160litrefueltank,whichwaslocatedintheaftcompartmentunderthestarboardinboardfacingpassengerseat.Theenginefuelsupplyandspillreturnlineswerefittedtothetopofthetank.Bothlinesranforward,underthefloorplating,downthestarboardsideofthevehicle.Thespilllinepasseddirectlyintotheenginecompartmentwhilethesupplylinewasroutedunderthecrewcabdashboardtotheportsideofthevehiclebeforepassingintotheenginecompartment.Underthedashboard,aflexiblebraidedpipehadbeenusedtoconnectthecoppersupplylinetoaball-typeemergencyshut-offvalve.

Theengine-drivenfuelliftpumpdrewthefuelthroughacombinedwaterseparator/filterandafinefilterbeforedeliveringittothefuelinjectionpump.Thespillreturnlinewasdirectlyconnectedtothefuelinjectionpump.Whentheenginewasrunning,thefuelsupplylinebetweenthefuelstoragetankandthefuelliftpumpinletwasunder negative pressure; after the lift pump, the system was under positive pressure (Figure 46).

72

Passengerbuoyancyaids

Figure 44: Typical LondonDuckToursDUKWpassengercompartmentandcrewcablayout

73

Vehicleboardingplatform

StarboardladderPort ladder

Stairwaysecuringcleat

Passenger compartment door in its closed position

Passenger compartment door open - port ladder in use

Figure 45: Boarding platform and ladders

74

2.7.10 Electrical distribution system

Cleopatrahadtwoseparate24Velectricaldistributionsystems:oneforstartingthe engine and one to provide power to the vehicle’s auxiliary systems. Each was suppliedwithpowerfromtwo,12V,55ampere-houracidgelbatteriesconnectedinseries.Eachsetofbatterieswasre-chargedbythesameengine-drivenalternator,butthesystemscouldnotbecross-connected.

Theenginestartingbatteriesandmainbatterycontactorwerelocatedinanenclosureunderthefront,portsidepassengerbenchseat.Inadditiontoenginestarting,thesystempoweredthedashboardinstrumentation,lightingandthefuelsupply valve solenoid through fused protected circuits.

Eng

ine

com

partm

ent

Fuel

tank

160

litre

s

Fuel

em

erge

ncy

shut

-off

valv

e

Flex

ible

fuel

pip

e

Fuel

inje

ctio

n pu

mp

Fuel

lift

pum

p

Wat

er s

epar

ator

Filt

er

Insi

de p

asse

nger

co

mpa

rtmen

t

Ext

erna

l fue

l em

erge

ncy

shut

-off

valv

e

Spi

ll lin

e

Fuel

sup

ply-

vacu

um c

ondi

tions

Fuel

filte

r

Fuel

sup

ply-

pres

sure

co

nditi

ons

Figu

re 4

6:SimplifiedillustrationofC

leop

atra

’s fu

el s

uppl

y sy

stem

75

Theauxiliarybatterieswerelocatedinanenclosureunderthefront,starboardsidepassengerbenchseat.Amanually-operatedswitchwasfittedtoisolatethebatteriesfromtheelectricaldistributionsystemandthebatterieswereprotectedfromhighcircuitcurrentconditionsbya100Afuselocatedinthebatteryenclosure.Thebatterypowercableswereroutedunderthepassengerdecktoensuretheywerewell away from sharp edges and were secured to the deck angled-support structure. Theauxiliaryservices,whichincludedtheVHFradio,passengerannouncementsystem,electro-hydraulicruddercontrols,bilgepumpsandlighting,wereprotectedbyappropriatecapacityspade-typefuses.

2.8 FIRE-FIGHTING ARRANGEMENTS

2.8.1 Fire-fighting appliances

Cleopatracarriedthree9litrefoamandone2kgdrypowderportablefireextinguishers.Thefoamextinguisherswerestowedonthestarboardsideofthevehicle:oneinfrontofthetourguide’sseat,onebehinditandoneatthebackofthepassengercompartment close to the access door. The dry powder extinguisher was stowed on theportsidebehindthedriver’sseat.

Theenginecompartmenthadafixedfire-extinguishingsystem,whichcompriseda9litrefoamextinguisherconnectedtodistributionpipeworkfittedwiththreedischargenozzles(Figure 47).Thesystemwasactivatedbyamanual-pulllever,whichwaslocated on the left hand side of the master’s seat.

Noneoftheportablefireextinguisherswereusedduringthefireandtheenginecompartment’sfixedfire-extinguishingsystemwasnotactivated.

2.8.2 Engine compartment shut-down

Inthecaseofanenginecompartmentfire,thefuelsupplytothecompartmentcouldbeisolatedbyclosingoneofthevehicle’stwoemergencyfuelshut-offvalves.Oneoftheemergencyshut-offvalveswaslocatedunderthecrewcabdashboardandcouldbeclosedbythecrewfromtheirseatedpositions.Theotherwasfittedoutsidethevehiclebelowtheportsideofthewindscreen(Figure 48).

Theenginecompartmentcoolingairinletcowlingandtheportandstarboardventilationvoidspaceventswerefittedwithremotelyoperatedfiredampers.Thefiredamperscouldbeclosedbythecrewfromtheirseatedpositionsbypullingindividualwire-pullhandles (Figure 48).Themaincoolingairoutletontopofthebowcompartmenthatchcoverwasnotfittedwithafiredamperandcouldnotbeclosedintheeventoffire.TheLiverpoolDUKWshadasimilarcoolingairinletarrangementfittedtotheirenginecompartmenthatchcoversbuttheydidnothavefiredampers.

Neitherofthefuelshut-offvalveswereclosedbeforetheabandonment.Thecrewattemptedtoclosetheportandstarboardventilationvoidspacefiredampersbuttheywereunabletoreleasethesteel“R”clipsthatsecuredthewirepullhandlesbecausetheyweretoohot.Themasterwasunawareofthecoolingairintakefiredamperandtherefore did not attempt to close it. During the post-accident inspections of the wire pullarrangementsonotherLDTvehicles,MAIBinspectorsfoundthatseveraloftheRclipsweredifficulttoremoveundernormalconditions.

76

Buoyancy foam

Foamdistributionpipe

9 litre foam extinguisher

Figure 47: TypicalLondonDuckToursDUKWenginecompartmentfixedfire-fightingsystem

77

R-clip

External fuel shut-off valve

Fuelshut-offvalve

Enginecompartmentventilationfiredamper release handles

Figure 48: Emergencyfuelshut-offvalvesandfiredamperreleasehandlesinasimilarvehicle

78

2.9 LIFE SAVING APPLIANCES

Cleopatracarried:2buoyantapparatus(1x20personand1x12person),4lifebuoys,30adultbuoyancyaids,2infantbuoyancyaids,and2manually-activatedgas-inflationlifejackets.

Thebuoyancyaidswereprovidedforthepassengersandtheinflatablelifejacketswereforthecrew.Theadultandinfantbuoyancyaidswereofthesolidfoamtype,theyprovided100Nand50Nofbuoyancyrespectively,andwerestowedinracksabovethepassengerseats.Thecrewlifejacketsprovided150Nofbuoyancyandwerestowedclose to their seats.

Thebuoyancyaidswerestoredinrobusttransparentplasticbags,manyofwhichweretightlyfittedand,accordingtothepassengerfeedback,weredifficulttoremove(Figure 44). Prior to Cleopatra enteringtheriver,passengersweregivenabuoyancyaid donning demonstration in accordance with the company’s operating procedures. Donninginstructionswerealsoprovidedonthebackofthepassengerseats.

Themajorityofthepassengersabandonedwithbuoyancyaids,howeverfewhadthemsecuredproperlyandsomesimplyheldontothem.One4yearoldchildabandonedwithoutabuoyancyaid.

2.10 VESSEL SURVIVABILITY

2.10.1 Buoyancy foam

TheinitialbuoyancycalculationscarriedoutbyBurnessCorlett&Partners Ltd in 1999 (Paragraph 1.9),werebasedonseveraltheoreticalassumptions.Thecalculations were intended to provide an indication of how the converted DUKW couldbeconfiguredtosatisfythedamagedsurvivabilityrequirementssetoutinMSN1699(M).Otherthanthoseprovisionalcalculations,LDTandtheMCAwereunabletoprovideanyrecordsrelatingtotheoriginalagreedvolumesanddistributionofbuoyancyfoamfittedwithinthehullsoftheLondonbasedDUKWs.

InNovember2008,theMCAraisedconcernsaboutprogressiveincreasesintheweightoftheLDTDUKWsasaresultofseveralstructuralmodifications.Inresponse,LDTinstructedBCTQtore-assessthebuoyancyrequirementsforitsvehicles.Duringtheprocess,theMCAremindedLDToftheneedtotakeaccountofthepositionofthevehicles’driveshaftsandgearboxlinkageswheninstallingfoam,andexplainedthattheuseofpre-formedblocksoffoamwouldbepreferabletotheexisting method of cutting and layering individual sheets of foam.

BCTQ used DUKW PortiatoreassessthebuoyancyfoamrequirementsfortheLDTfleet.Basedonavehicleweighbridgeweightof8.82tandapassengerandcrewweightof80kgperperson,itwascalculatedthat12.22tofbuoyancywasneededtoprovidethestatutory110%buoyancyforthevehicleinitsfullyladencondition.BCTQestimatedthatthevehicleitselfprovided3.17tofinherentbuoyancy,andconcluded that 9.05m3ofbuoyancyfoamneededtobeinserted.

UsingtheoriginalresultsfromBurnessCorlett&PartnersLtd'scomputermodeltohelpcalculatetheavailablespacewithinthevehicle’shull,BCTQidentified13separate locations that it estimated, collectively, could hold 10m3 of foam (Figure 49). The BCTQ report re-emphasised the need to ensure that the foam in the central void space was securely fastened to prevent it making contact with the rotating drive shafts.

79

Figu

re 4

9:BCTQ

Ltd’s2008buoyancyfoam

calculations

Cen

tre v

oid

spac

e

80

TheMCAconsideredPortiatoberepresentativeoftheLDTvehiclesandaccepted9m³asthestandardvolumeoffoamfortheLDTfleet.Accordingtoweighbridgerecords, Cleopatraweighed160kglessthanPortiain2009.In2012,DUKWElizabeth,withaweightof9.16t,wastheheaviestvehicleintheLDTfleet(340kgheavier than Portia).

2.10.2 Addition of buoyant foam following the sinking of Wacker Quacker 1

TheMCA’sdecisioninJune2013toincreasetheLondonDUKWs’buoyancyfoamrequirement,from9m3to10.6m3,wasbasedoncalculationscarriedoutbyitsinternalstabilityunitfollowingthesinkingofWQ1.TheincreasedrequirementmeantthatLDThadtoidentifynewlocationstoinstallfoam,andtofitincreasedvolumesoffoam,wherepossible,intothespacespreviouslydesignatedinBCTQ’s2008report.Toensurethemaximumvolumeoffoamwasfittedinthehullspacesbelowthe passenger compartment deck, LDT packed additional foam into the central void spaceabove,below,andonbothsidesofitsvehicles’rotatingdriveshaftsandUJs(Figure 50).

Oncethesespaceswerefull,thecompanypackedfoamintootherspaces,suchas the passenger compartment and the engine compartment’s ventilation void spaces (Figure 51). The foam in the cooling air void spaces was packed close to theunlaggedengineexhaust,andgapshadbeenlefttocreatechannelstoallowairfromtheenginecompartment’sextractionfanstoflowoutofthesidevents.Toprevent the foam moving within the ventilation void spaces, the garage mechanics glued the individual pieces of foam together.

The insertion of additional foam within the central void space, the engine compartment and its ventilation void spaces also increased the level of thermal and acousticinsulationsurroundingtheenginecompartment.Priortofittingtheextrafoam,itwasnotunusualforthepassengercompartmenttobecomeoverlywarm,requiringthemastertoswitchonthecompartment’scoolingfans.Followingthefittingoftheadditionalfoam,itwasnotedthatthepassengercompartmentwasmuch cooler and the fans were seldom used.

2.10.3 Types of buoyancy foam

ThebuoyancyfoamusedonboardCleopatra was a mix of:

• singlesheetand3-plylaminate,closedcell,low-density,non-flameretardantPlastazoteLD18polyethylenefoam;and

• singlesheethigh-density,closedcell,flame-retardantMicrolenPE-30XFRpolyethylene foam.

Thelow-densityfoamwaspackedintothepassengercompartmentbulkheads,aroundseatingandunderthepassengercompartmentfloorplatesinwayofthe propulsion drive shafts. The high-density foam was packed into the engine compartment and in its ventilation void spaces.

Thelow-densityfoamhadameltingpointof107ºCandaflashpointof>300ºC;themeltingpointofthehigh-densityfoamwasunknownbutitsburningpoint25 was estimatedtobe430ºC.

25 TheSpanishmanufactureroftheproductconfirmedthattheterm“burningpoint”,referredtointheMSDS,isthesameas“flashpoint”.

81

Universaljoint

Lowdensitybuoyancyfoam

Shardsofgreasecontaminatedfoam

Universaljoint

Figure 50: Buoyancy foam packed around the drivetrain in the centre void space of a London Duck Tours DUKW

82

2.11 FIRE INVESTIGATION

2.11.1 Overview

MAIBinspectorsandmembersofLFB’sFireInvestigationTeamconductedajointexamination of Cleopatra,atLDT’sfleetmaintenancegarage,duringtheperiod29-30September2013.

Ventilation space Outletvent

Extraction fan housing

Figure 51: Buoyancy foam packed into the engine compartment ventilation void spaces

83

Adetailedvisualexaminationwasconductedofallareasofthevehicle.Initially,allexternalandnormallyaccessiblespacesandcompartmentswereexamined.Oncompletion,thepassengercompartmentfloorplateswereliftedtoexposethebuoyancyfoam,andallowaccesstothevehicle’sdrivetrainarrangementsandauxiliarysystemslocatedinthevoidspacesbelow.

Itwasapparentfrominsidethepassengercompartmentthatthefiredamagewasconfinedtothecrewcabareaandthefronttworowsofpassengerseats.Thebuoyancyfoaminthecentralvoidspacewascarefullyremoved,startingatthehand-operatedbandbrakeassemblyandworkingforward.Thefoamunderthecrewcabandthefronttworowsofpassengerseats,eithersideofthecentralvoidspace,was also removed.

Theinitialfindingsindicatedthattheseatofthefirewaslocatedwithinthecentralvoidspace,beneaththecrewcab,immediatelyaftoftheenginecompartment’snon-continuoustransversebulkhead.Detailedexaminationandexternallaboratorysupportwasnecessarytodeterminetheprobablesourceofignition.

2.11.2 The external hull

Theouterhullpaintworkinthevicinityoftheenginecompartment’sstarboardventilation void space, and in the area where the engine exhaust pipe exited onto thedeck,hadbeenpartiallyconsumed.Lightsmoke-damagewasevidentontopofboththeengineandbowcompartmentaccesshatchesbuttherewasnoevidenceofpaintblistering.Thewindscreenandthesidewindowsadjacenttothetourguide’sseatweremissingbutthoseadjacenttothemaster’spositionremainedintact.Thepaintontheportsideofthevehiclewasunmarkedbyfireorsmokedamage.Theenginecompartmentairintakeandtheportandstarboardventilationvoidspacefiredampers were in the fully open position.

2.11.3 Crew cab area and passenger compartment

Thefirehadtotallyconsumedthemaster’sandtourguide’sseats,leavingonlythesteelframework,andhadpartiallyconsumedthefirsttworowsofthepassengerseatsandthecarpetedfloorcoveringinthisarea(Figure 52).Thedashboardwasseverelyfire-damagedwiththeareainfrontofthetourguide'sseatbeingtheworstaffected.

Therewassignificantcombustiondebrisandheatdamageinthecentresectionofthecrewcabfloorplates.Thedamageinthisareawasfarmoreseverethaninanyother area of the passenger compartment.

Thebuoyancyfoamfittedtothehull,outboardofthestarboardforwardpassengerseatshadalsobeeninvolvedinthefireandhadbeenpartiallyconsumed.Therewasalsowiderangingheatdamagetotheportandstarboardsidecurtains.

2.11.4 Engine compartment

Theundersideoftheenginecompartmentaccesshatchhadbeenfiredamagedandwascoatedwithsoot,aswasthecompartment’safttransversebulkhead.Theengine’sfuelfilter/water-separatorbowlhadpartiallymelted.Theengineandautomaticgearboxoillevelswerecheckedandfoundtobeatthecorrectrunninglevels.

84

Thestarboardextractionfanhadsufferedextensiveheatdamagewithmostofitsplastic grille consumed. The fan had fallen from its fastenings and was found on the wheel arch (Figure 53).Theportextractionfanremainedintactbutitsplasticprotective grille had suffered minor heat damage and was covered in soot deposits.

Crewcabarea

Passenger compartment

Figure 52: Firedamagedpassengercompartmentandcrewcab

85

2.11.5 Engine compartment cooling air ventilation void spaces

ThestarboardvoidspacehadsufferedseveredamageconsistentwiththeimagesofthesmokeandfireseenonCCTVandvideofootageevidence.Themajorityofthehigh-densityflame-retardantbuoyancyfoamfittedintheareahadbeenconsumed.Thatremainingwasextensivelyburnt (Figure 53). There was no evidence that the firehadmigratedintotheportvoidspace.

Port ventilation extraction fan

Starboardventilationvoid space

Remnantsofhighdensitybuoyancyfoamwithintheventilation void space

Remainsofthestarboardventilation extraction fan

Figure 53: Firedamagewithintheenginecompartmentandstarboardventilationvoidspace

86

2.11.6 Central void space

Beforethecrewcabandpassengercompartmentdeckplatingwasremoved,the forward section of the central void space was examined from the engine compartment.Itwasapparentfromviewingtheareathroughthenon-continuoustransversebulkheadthattheareaaroundthedriveshaftconnectingtheautomaticgearboxtotheMTBhadsufferedconsiderablefiredamage.

The internal hull coating under the plates was largely consumed, and that remaining wasblistered.Thenon-flame-retardant,low-densityfoam,whichhadbeenfittedaroundthedriveshaft,hadbeentotallyconsumedbythefire(Figure 54) with most oftheremainingexposedfoamsurfaceshavingmelted.Whenthefloorplateswereremovedfromabovethedriveshaftitbecameevidentthatthedamagewasmorepronouncedonthestarboardsideoftheshaft.

Thesurfaceofthesplineddriveshaftwascoveredinwhatappearedtobemoltenfoam.TheshaftwasintactandstillconnectedtotheautomaticgearboxandMTBbyitsUJs,andtheshaftwasabletobeturnedbyhand.Thedriveshaft,completewithitstwoUJs,wasremovedfordetailedanalysis,thefindingsofwhicharedetailedatParagraphs 2.12.1 and 2.12.2.

Oncetheshafthadbeenremoved,theareaofdeepseatedburnwasfullyrevealed.ThesiteofdeepestburnwasfoundtobeunderthepositionoftheforwardUJ,whichhadbeenconnectedtotheoutputflangeoftheautomaticgearbox.Apithadformedinthefoaminthisareaandthefoamhadbeenalmosttotallyconsumed(Figure 55).

Theoillevelsinthemarineandroadtransferboxeswerecheckedandfoundtobeat their correct running levels.

2.11.7 Hydraulic steering system

Althoughtheheadertankfortheelectro-hydraulicsteeringsystemwaslocatedattherearofthevehicle,itslevelwascheckedtodeterminethepossibilityofoilleakagefeedingorcontributingtothefire.Thelevelwasfoundtobenormal.

2.11.8 Engine fuel oil system

Asectionofflexiblehose,locatedunderthedashboard,whichhadconnectedthefuelsupplylinetotheinternalemergencyfuelshut-offvalve,hadbeencompletelyburntthrough(Figure 56).Thelinewasabovethelevelofthefueltankandwouldnothavebeenunderpressureatthetimeofthefire.TheflexiblehosedidnotmeettheMCA’sminimumfireteststandards.

The engine fuel supply and spill pipework was pressure tested to 70kPa; the testpressurewasheldsuccessfullyfor10minutes.Allconnectionsfromthefuelinjectionpumptothefuelinjectorswerecheckedandfoundtobetightandtherewas no evidence of fuel leakage. Both emergency fuel shut-off valves were found to beopen.

87

View

look

ing

aft t

hrou

gh

the

non-

cont

inuo

usbulkhead

Crewcabdeckplates

Alisonautom

aticgearbox

Splineddriveshaft

View

look

ing

forw

ard

with

thecrew

cabdeckplates

lifte

d

MTB

Figu

re 5

4:Firedam

ageunderthecrewcab

88

Alisonautomaticgearbox

View looking forward towards the engine compartment with the drive shaft removed

Figure 55: Areaofdeepestburnwithincentralvoidspaceaftoftheautomaticgearbox

Figure 56: Burntthroughsectionofflexiblehose

Fuelisolationvalve Flexiblehose

89

2.11.9 Electrical distribution system

Theauxiliarybatterieshadsufferedlightheatdamagetotheircasings,cabling,plasticterminalcovers,andsecuringstraps.Theprotective100Afusewasfoundtohaverupturedbutallconnectionswerefoundtobetight.Theinsulationonthecablesthatranunderthefloorplatesclosetothecentralvoidspacehadbeenconsumed.Therewassignificantdamagetothewiringloomunderthedashboard.

2.11.10 Hand-operated band brake

Noneofthelow-densityfoamremovedfromaboveandaroundthebandbrake,orthatundertheoutputshaftfromtheroadtransferboxwasfiredamaged.Thebandbrakeshowednosignsoffiredamageoranyevidenceofoverheating.SignificantgreasedepositswerefoundontheremovedfoamandinthevicinityofthebandbrakeconnectingUJs (Figure 57).

2.11.11 Brake system air reservoirs

Thefloorplatesandfoamcoveringtheairbrakereservoirswereremoved.Whiletherewasnofiredamageinthisimmediatearea,theflexiblepipesconnectedtothefootoperatedairdistributionvalve,adjacenttothefootpedal,hadbeenconsumedbythefire.Therewasnocompressedairchargeremaininginthereservoirs.

2.12 EXTERNAL FIRE INVESTIGATION SUPPORT

2.12.1 Bureau Veritas Fire Science Department fire investigation report

Inordertodeterminethemostlikelysourceofignition,thepreliminaryfiresceneexaminationidentifiedaneedto:

• Determinetheignitabilitycharacteristicsofsamplesofcontaminatedandnon-contaminatedlow-densityandhigh-densitybuoyancyfoam.

• ConductavisualexaminationofthesplineddriveshaftassemblyandtheUJsthathadconnectedtheautomaticgearboxoutputshafttotheMTB.

• Establishiftheenginecompartment’sstarboardextractionfanwasdamagedby,orwasapotentialcauseofthefire.

Toachievethis,LFBcontractedtheLondon-basedBureauVeritas(BV)FireScienceDepartmenttoundertakeaseriesofthelaboratorytestsandexaminations.

TheBVFireScienceDepartmentreport26 found that:

• Theedgesofboththelowandhigh-densityfoamcouldbeeasilyignited,especiallythosewhichhadbeencontaminatedwithoilorgrease.

• Onceaflamewasestablisheditspreadreadilyacrossthefoam.

• Smallersectionsofbothfoamtypesweresignificantlyeasiertoignitethantheirbulksurfaces.

26 Fire Investigation – London Duck Tours, River Thames, near the Palace of Westminster - LFB/13-595 dated 9 December 2013.

90

• Thehigh-densityflame-retardantfoamproducedmuchmore,anddenser,blacksmokethanthelow-densitynon-flame-retardanttype.

• The surface of the drive shaft and its UJs was covered in melted foam.

• Therewasnoobservablephysicaldamagetothedriveshaftassembly.

• Theenginecompartment’sstarboardextractionfanhadnopre-existingelectricaldefectsandwasnotcausaltothefire.

Bandbrake

Foamremovedfromaroundthebandbrake

Excess grease

Excess grease

Figure 57: Greasecontaminatedfoamremovedfromaroundthebandbrake

91

Thelaboratoryreportconcludedthat:

“There was no obvious ignition source in the vicinity of the buoyancy foam, and therefore the most likely cause of the fire was the action of the rotating drive shaft (or other moving parts) on the oil contaminated surfaces of the buoyancy foam blocks.”

2.12.2 Metallurgical analysis of Cleopatra’s drive shaft and UJs

FollowingBV’svisualexaminationofthedriveshaftassembly,theMAIBcontractedHampshire-basedMaterialsTechnologyLimited(MTL)toconductdetailedmetallurgicaltestingofthesplinedsteelshaftanditsUJs.MTLwastaskedto:

• X-ray,dismantleandexaminethedriveshaftandbothUJsfordefectsandsigns of overheating.

• Examinetheneedlebearingrollersandcontactsurfacesforwear.

• DeterminethetypeofgreaseusedtolubricatetheUJs.

TheMTLlaboratoryreportfoundthatbothUJshadsufferedwhatappearedtobefrettingcorrosionindicativeofmisalignedin-serviceloading.ItwasevidentfromtheextentofthewearonthesurfacesoftheUJjournalbearingsthatthiswouldhavebeenpresentpriortothefire.Theexcessivewearhadcausedcircumferentialcrackingtooneofthejournalcapscontainingtheneedlebearingrollers(Figure 58),andresultedinelevatedtemperaturesofbetween300-400°C,asconfirmedbybluingoftheneedlebearingrollers.

FouriertransforminfraredspectroscopyanalysiswascarriedoutonthesamplesofburntgreaseontheexternalsurfacesoftheUJsandthatcontainedwithinthebearingcaps.TheresultshadsimilarcharacteristicstotheexemplarCastrolHeavyDuty Plus grease.

TheMTLreportconcluded:

“… the overheated bearings, indicated by the evidence of bluing on the needle bearings, can be considered to be a possible cause of ignition where the environment is suitable and where flammable materials with the appropriate properties/characteristics are in contact or in close proximity…”

2.13 SAFETY MANAGEMENT SYSTEM

InaccordancewiththerequirementsoftheDSMCode,LDThadproducedanOperationalProceduresManualandhadidentifieditsmanagingdirectorasthecompany’sDPA.Themanualincludeddedicatedsectionsforreactingtofireandinstructionsforabandonment.Italsostatedtherequirementforcrewtraining,including the need for emergency drills.

Section16oftheOperationalProceduresManualdescribedtheactionstotakewhenafirewasidentified.Thecrewareinstructedto:

1. Check on the passengers and raise the alarm.

92

2. Put out a call on VHF channel 14.

3. Tackle the fire.

4. In the event of a fire in the engine bay, close the air vents and turn on the automatic emergency fire extinguisher.

5. Steer the craft to minimise the effect of the wind on the fire and warn other craft (a series of short blasts).

6. IF THE FIRE IS A DANGER TO PASSENGERS AND CREW THEN PREPARE TO ‘ABANDON SHIP’.

Figure 58: Laboratoryanalysisoftheforwarddriveshaftuniversaljoint

Wearonjournals

Wear induced pitting on surfaceofneedlebearingrollers

Bluing oxidation on needle bearingrollers

Fracturedbearingcap

93

Sections17and20ofthemanualdescribethe"Mayday"distressprocedureandabandonshipproceduresrespectively.TheguidancegiveninSection17statedthat,if in the opinion of the master it is an extreme and absolute emergency, the crew should:tunetheVHFradiotochannel16;waitforagapintransmission;transmitadistresscall.LDT’sabandonshipprocedurewas:

1. Make a ‘MAYDAY’ distress call on VHF channel 16.

2. Order all passengers and crew to don lifejackets.

3. Order second crew member to climb onto the upper deck (i.e. the roof of the passenger compartment) and pass down the buoyancy apparatus to the rear of the vessel. If time is of the essence then instruct that the three skull craft be thrown overboard.

4. Assemble passengers and ensure that less able passengers are paired up with more able passengers.

5. Disembark passengers and crew onto buoyant apparatus using the rear steps of the vessel if it is safe to do so.

6. If by opening the rear bulkhead door it would aggravate the position by increasing the volume of water being taken on board then instruct that the emergency exits are used.

7. If buoyant apparatus is full then assist remaining passengers into water.

8. Take a head count and keep a constant check of numbers.

9. Maintain watch for other vessels.

10. STAY CALM AND KEEP UP MORALE.

2.14 VEHICLE MAINTENANCE

2.14.1 Maintenance management system

SimilartotheTYDoperationinLiverpool,theLDTvehicleswereinspected,servicedandmaintainedin-housebythecompany’sgaragestaff.Thegaragemechanicsfollowed a maintenance management regime that was designed to comply with bothMCAandVOSArequirements.Paperrecordsweremadeoftheplannedandunplanned maintenance carried out on each vehicle.

2.14.2 Planned maintenance

Eachmorning,thecompany’sengineerswererequiredtoensurethatallvehicleswereroadandriver-worthybeforeallowingthemtoleavethegarage.Beforedeparture,thedriverhadtocomplete,signanddateacombinedvehicle and vessel checklist.

94

Aspartoftheweeklymaintenanceroutines,thegaragemechanicshadtoinspectandmanuallygrease/lubricatethedrivetrainbearingsandUJs.Theyalsocheckedtheoillevelsintheengine,automaticgearbox,transferboxesandsteeringsystem.This was last carried out on Cleopatraon23September2013,6dayspriortotheaccident.

Thevehicles’internalUJswerelubricatedwithCastrolPremiumHeavyDutyPlusgreasewhichhadaflashpointof232ºC.TheexternalbearingsandUJswerelubricatedwithLintonLubricants’ProtexBluegrease(flashpointof>200ºC).

To allow the garage mechanics to carry out their weekly maintenance tasks, they had to remove the tightly packed foam from each DUKW’s central void space (Figure 59). The need to remove and then reinstall the foam to grease the drive shaftbearingsandUJsmeantthatthetimetakentocompletethevehicles’weeklymaintenanceroutinesincreasedsignificantlyandrequiredthevehiclestobetakenout of service for a full day.

When the greasing procedure was demonstrated using DUKW Desdemona, it was notedthatthefoamsectionsweredifficulttoremoveandwereincloseproximitytotherotatingshaftassemblies.SimilartoCleopatra, there was extensive evidence of greasecontaminationofthefoam.Itwasalsonotedthatthepositionsofeachfoam

Figure 59: Tightly packed foam in central void space

Greaseimpregnatedcrumbsoffoam

Foamcontactwith shaft

95

blockorsheethadnotbeenmarkedandforcehadtobeappliedtoreinstatethefoam.Oncerefitteditwasnotpossibletoidentifyifthefoamwasincontactwiththerotating parts.

2.14.3 Unplanned breakdown repairs

The unplanned maintenance work was also recorded on the paper record sheets. Followingtheinsertionofadditionalbuoyancyfoam,thegaragestaffexperiencedseveralrelatedissuesthatrequiredunplannedmaintenanceaction.

DuringtheinspectionoftheLondonDUKWs,itwasobservedthatthefoamfilledspacescontainedfineshardsor“crumbs”offoam(Figures 50 and 59).Mostofthecrumbswerecreatedwhenthegaragestaffcuttheindividuallysizedblocksoffoaminsitu.Itwasalsoevidentthat“crumbs”hadbeencreatedbytherubbingactionbetweentheblocksoftightlypackedfoamduringtheremovalandrefittingprocess,andasaresultofcontactbetweenthefoamandtherotatingdrivetrain.Thefoamcrumbshadatendencytobuildupinthebilges,andhadoccasionallyblockedthebilgepumpsuctions.

2.15 SURVEYS AND INSPECTIONS

TheLDTwaterborneoperationwasoverseenbytheMCA’sOrpingtonMarineOffice,anditssurveyorsappliedasimilarsurveyandinspectionregimetothoseinLiverpool.UponnotificationofanMCAsurveyorinspection,thegaragestaffpreparedthevehicle(s)byremovingfoam,asnecessary,frompre-selectedlocations.Thisenabledthesurveyortogainaccesstoandinspectareassuchasthehullandthedrivetrainarrangements.Oncompletionofsurvey,andtypicallyfollowingthe surveyor’s departure, the foam was reinserted.

The most recent survey of Cleopatra tookplaceon3October2012.Duringthesurvey,itwasidentifiedthatthevehiclewasoverdueastabilityheeltest,whichwassubsequentlycompletedwithinthe4monthsallottedbythesurveyor.ThetotalquantityandconditionofbuoyancyfoamwasnotcheckedbytheMCAatthesesurveys and inspections.

2.16 EMERGENCY RESPONSE

2.16.1 Thames RIB Experience

ThamesRIBExperiencewasoperatingitsthreeRIBs-Ultimate, Experience and Adrenaline - closebywhenthefirebrokeout.TheskippersofthethreeRIBsheardCleopatra’srequestforassistanceandimmediatelywenttoitsaid.AllofCleopatra’s passengersandcrewwererescuedfromthewaterbeforeFiredartandtheRNLIlifeboatsarrivedonscene.

2.16.2 Royal National lifeboat Institution

ThenearestRNLIstationtoWestminsterwasatTowerPier.Thelifeboatfromthisstationwasunabletoimmediatelyassistasitwasattendinganotherincident.Inviewofthis,theChiswicklifeboat,whichwasseveralmilesupstreamwastasked.

96

Towerlifeboathastilycompleteditsearlierdeploymentmissionandwasonthesceneofthefireat1206.Chiswicklifeboatarrivedat1208.ThelifeboatscarriedoutariversearchforpossiblecasualtiesuntilitwasconfirmedthatallCleopatra’s occupants were safely on shore.

2.16.3 Police

Atthetimeoftheaccident,theMarinePoliceUnithadonevessel,MP2, on duty. However,thiswasattendingthesameincidentastheTowerPierRNLIlifeboat.MP2 initiallybecameawareoftheincidentfromoverheardradiotransmissions.Thiswaslaterconfirmedbythepolice’scomputeraideddispatchnotificationservicefromLambethpolicestation.Anairbornepolicehelicopterwasnotifiedoftheincidentanddiverted to the scene. The helicopter arrived on scene at 1200 and provided a live video link direct to the coastguard.

2.16.4 London Fire Brigade

Inadditiontoconventionallandbasedfireappliances,theLFBhadtwofireboats,onedutyvesselandonereserve.ThesewerestationedatLambethFirePontoonabout400mfromthefirescene(Figure 35).

ThefireboatFiredartwasnotifiedoftheincidentbythecoastguard.Aroundthesametime,amemberofthepublicmisinformedtheLFBcontrolofabusfireneartheHousesofParliamentandalandbasedappliancewastaskedtoattend.

Firedartarrivedonsceneat1159,quicklyfollowedbythelandbasedappliancetaskedtotheerroneousbusfire.Afterextinguishingthefire,Firedart towed Cleopatra away from the scene.

2.17 PASSENGER FEEDBACK

Aspartoftheinvestigationandinformationgatheringprocess,theMAIBpostedquestionnairestothe28passengers’declaredaddresses.Fourteenpassengers(50%)respondedbycompletedquestionnaireorwrittennarrative,whileafurther6(21%)communicatedtheirthoughtsbytelephonetoMAIBinspectors.

Alloftherespondersstatedthatasafetybriefandbuoyancyaiddonningdemonstrationhadbeengivenbytheguide.Theyalsocommentedonthelight-heartednatureofthesafetybrief.However,15%oftherespondersfeltthatthepassengersdidnotgivethebrieftheattentionitdeserved.Someofthepassengersthoughtthiswaspossiblyduetothejocularnatureofitsdelivery.

Alltherespondersexpressedconcernaboutthebuoyancyaidsbeingstowedinstrongpolythenebags.Nineteen(70%)oftheresponderssaid,oncetheyhadremovedthebuoyancyaidsfromthebags,theyhaddifficultyreleasingthestraps;5(19%)gaveup and entered the water without them.

Amixedresponsewasgiventothecrew’sactions,withsomerespondersthinkingthecrewdidlittletoassist,whereasothershighlypraisedtheirpositiveactions.Mostcommentatorsreferredtothelevelofpanicamongthepassengersfollowingthefire,andrecognisedthedifficultythecrewhadtryingtocontrolsuchasituation.

97

2.18 SIMILAR ACCIDENTS

2.18.1 MAIB database

TheMAIB’sdatabasecontainedrecordsof52previouslyreportedincidentsinvolvingLDTvehicles.Ofthose,29wereattributedtomechanicalandelectricalfaults,2resultinginfiresand3requiringthepassengerstobeevacuatedontoothervessels.The records also indicate that 9 of the mechanical failures were related to drivetrain issueswith2ofthemspecifyingUJfailures.Threecollisionsand6groundingshavealsobeenreported.

ManyoftheaccidentslistedontheMAIBdatabasewerenotreportedtotheMAIBbythecompany.AccordingtoMCArecords,itwasawareofanadditionalsixincidentsbetween2010and2013thathadnotbeenreportedtotheMAIB,whichresultedinLDTDUKWsbeingtowedofftheriver.Ofthose,fourweretheresultofmachinerybreakdownsandonewasafouledpropeller.

On17May2013,theLFBfireboatFireflashself-launchedinresponsetoafireonboardDUKWDesdemona.Atthetime,Desdemona was ashore at the Lacks Dock slipwayandthefirehadbeenextinguishedbythetimethefirecrewarrived.TheincidentwasnotreportedtotheMAIB,butoverheatingofthevehicle’shandbrakewasunderstoodtobethecause.

2.18.2 Elizabeth

On12July2013LDT’sDUKWElizabethwastowedfromtheRiverThamesfollowingthefailureofonetheUJsconnectingtheautomaticgearboxoutputtotheMTB.AccordingtoLDT’smaintenancerecords,thiswasanuncommonoccurrence.Elizabethwasnotinserviceatthetimeoftheaccident,ithadrecentlybeenpackedwithadditionalfoamandwasbeingusedtodelivercrewtrainingpriortobeingputbackintoservice.TheincidentwasreportedtotheMAIBbythePLAand,followingarequestformoreinformation,LDTadvisedtheMAIBthatitsinternalinvestigationhadidentifiedthattemperatureswithintheenginecompartmentandsurroundingareashadincreasedfollowingitsventilationmodificationsandtheinsertionofadditionalbuoyancyfoam.Asaresult,itwasconsideredlikelythattheUJhadoverheatedandrundryoflubricant.

ItwastheDPA’sunderstandingthat,asaresultofthisincident,newlysourcedhigh-temperatureresistantgreasewasbeingusedacrosstheDUKWfleettolubricatetheUJs.However,thisinvestigationhasidentifiedthatthereplacementhigh-temperature grease had the same properties as that previously used.

2.18.3 Mistress Quickly

At1125on5June2008,LDTDUKWMistress Quickly suffered a mechanical breakdownontheRiverThames.Initially,aknockingnoisewasheardcomingfromtheenginecompartment,afewminuteslatertheengineseized.ThemasterimmediatelyreportedtheincidenttotheThamesVTSanddroppedanchor.ThiswasobservedbythecrewofanLFBfiretender,anditwassubsequentlymanoeuvredalongside to provide assistance.

98

Mistress Quickly’smasterobservedsmokeandsparkscomingfromtheenginecompartment vents and immediately isolated the fuel supply line and closed the inletventfiredamper.Theninepassengersandtwocrewthenevacuatedontothefiretender.TheLFBcrewclimbedonboardtheDUKWandoperateditsfixedfireextinguishingsystem.Whentheyopenedtheenginecompartmenthatchthefirewasstillburning,sothefirefightersextinguisheditusingtheirownportableextinguishers.

Followingtheincident,theChiefInspectorofMarineAccidentswrotetoLDTinstructingittoensurethatmastersofitsvesselsarefullyacquaintedwith,andregularly exercised in the emergency procedures contained within the company’s operational procedures manual.

2.18.4 World War 2

DUKWs have always had engine and engine compartment cooling issues, particularlyinhotclimates.TheUSarmy’sThe DUKW, Its Operation and Uses manualwarnedagainstthedangersofrestrictingthecoolingairflowtoandfromtheenginecompartment.Themanualspecificallywarnedagainstputtinganythinginthe forward compartment or within the air passages, pointing out that the heat of the engine cooling air plus the heat of the exhaust piping is almost certain to start a fire.

99

SECTION 3 - ANALYSIS

3.1 AIM

Thepurposeoftheanalysisistodeterminethecontributorycausesandcircumstancesoftheaccidentasabasisformakingrecommendationstopreventsimilar accidents occurring in the future.

3.2 OVERVIEW

The causes and circumstances of WQ1’shullfailureandsubsequentsinking,andthefireonboardCleopatrawillbeconsideredinthissection.Theunderlyingcontributoryfactorsassociatedwiththeseadverseoutcomeswillbediscussedandthevehicleabandonmentsandemergencyresponseswillbeanalysed.

ThebodyofevidencetosupporttheanalysiswascomprehensiveandincludedvideoandCCTVfootage,witnessaccounts,scientificanalysisofphysicalevidence,practicaltrialsandreconstructions,anddocumentedrecords.TheMAIBalsoreceivedahighlevelofco-operationandexpertinputfromanumberofotherAPVoperatorsanddesignersfromwithintheUKandabroad.

3.3 THE SINKING OF WACKER QUACKER 1

3.3.1 Overview

WQ1sanksuddenlybythebowinSalthouseDock,Liverpool,because:

• Itsufferedmassivefloodingthroughtwolargeholesinitsouterhullplatingbelowthewaterline,andthevehicle’selectricpoweredbilgepumpsdidnothave the pumping capacity needed to cope with the rate of water ingress.

• Oncefullyflooded,WQ1didnothavesufficientresidualbuoyancytoremainafloat.

The holes in WQ1’s hull were created as a result of the forces generated when its propellerbecamefouledbyanolddiscardedcartyre.WQ1’s 31 passengers and 2 crewabandonedthevehicleandeitherswamashoreorwererescued.Noonewaskilledorseriouslyinjured.

3.3.2 The hull failure

When WQ1’s propeller picked up the car tyre, the resultant forces were transferred to the vehicle’s hull through its propeller shaft v-strut. The structural strength of the hullwasnotsufficienttoabsorbtheforcesgenerated,allowingthev-strutfeettoberippedfromtheirmountings.Thiscreatedtwolargeholesbelowthewaterlineinthepropeller tunnel plating.

The v-strut was designed to support the propeller shaft and transfer the transverse forces acting on the propeller to the vehicle’s hull. The internal v-strut support shoes were designed to spread the load and transfer dynamic forces acting on the v-struttothevehicle’safttransversebulkhead.However,theweldsthatfastenedtheinternalshoestotheafttransversebulkheadstiffenershadcorrodedaway,ashadtheweldsbetweenthebaseofthebulkheadandtheoutershellplating(Figure 60).

100

This meant that the full weight of the forces acting on the hull through the v-strut wasbeingtakenentirelybytheshellplating.Thepropellertunnelplatingitselfhadsufferedseverecorrosionwastageandthereforehadbeensignificantlyweakened.

Figure 60: Hull corrosion within Wacker Quacker 1’s aft void space

Starboardafttransversebulkheadv-strutshoestiffener

Afttransversebulkhead

Propeller tunnel

101

RegardlessofWQ1’s material condition, the DUKW’s hull plating was never designedtoabsorbsuchforces.ThispointwasidentifiedbytheUSarmy’srecommendationtofitpropellerguards(Paragraph 1.16.5) and reinforced when the propeller of DUKW BeatricewasfouledontheRiverThamesin2001(Paragraph 1.16.2). Had TYD learned lessons from previous accidents, or acted on the advice of LDT after the sinking of WQ4, and reinforced the propeller tunnel plating, it is possiblethatthefoulingofthepropellerwouldnothaveresultedintherupturingofthe vehicle’s hull.

3.3.3 The fouling of the propeller

WQ1’spropellerdriveshaftseizedshortlyafterthevehiclehadrolledbackintothewaterfollowingthedriver’sabortedattempttoclimboutofSalthouseDock.Thepolicediversfoundseveraldiscardedcartyresandotheritemsofdebrisonthedockbedatthebottomoftheconcreteslipwayandclosetothesunkenvehicle.

ItislikelythatthetyrewasdisturbedfromthedockbedbyWQ1’s wheels as it rolled backdowntheslipway,andwaspickedupbythepropellerasthedrivertriedtomanoeuvre the vehicle astern.

PropellerfoulingwasaforeseeablehazardthatwouldhavebeenidentifiedhadtheriskstoTYD’swaterborneoperationsbeensystematicallyassessed.Thelikelihoodofthehazardbeingrealisedwouldhavebeensignificantlyreducedifstepshadbeentakentoperiodicallyinspectandcleardebrisfromthebottomoftheslipways.Thefittingofapropellerguardmightalsohavereducedthelikelihoodofthetyrefoulingthe propeller.

3.3.4 Material condition of Wacker Quacker 1 and The Yellow Duckmarine fleet

Followingitssinking,WQ1wascloselyexaminedbyMAIBinspectorsandMCAsurveyorsandwasfoundtobeinanextremelypoormaterialcondition.Internallyits hull was heavily corroded (Figure 60);largesectionsofitselectricalcablingwasfoundtobehangingloose,andwereunprotected;andmuchofitsinstrumentationdid not work. Externally its hull was covered in a patchwork of welded repairs (doublers)thathadnotbeenapprovedbytheregulatoranddidnotmeetbasicmaritimestandards.Itwasalsoevidentfromthecompany’smaintenancerecordsthat WQ1’ssteeringsystemanddrivetrainarrangementswereunreliableandproneto failure.

TheothervehiclesintheLiverpoolfleetwereinasimilarconditiontoWQ1.Itwasapparent that the garage staff had worked long hours to keep the TYD vehicles in service. However, it was also clear that they were under-resourced and WQ1, and therestoftheLiverpoolfleet,hadbeenallowedtodeterioratetoanunsafeconditionover a prolonged period of time.

OperatinginseawaterexposedthehullsofTYD’svehiclestoahighriskofcorrosionandtheirthinshellplatingmadethemparticularlysusceptibletocorrosionrelatedhullfailure.AlthoughthefoulingofWQ1’s propeller triggered the hull failure in this instance, the hull was close to failing under normal operating conditions, as was the case with WQ2 only 2 weeks earlier (Paragraph 1.16.1). WQ1hadbeenpoorlymaintained,anditshouldnothavebeenoperatingonthewater.

102

3.3.5 Maintenance management

TYD’spaper-basedmaintenancemanagementsystemwasdesignedtosatisfytherequirementsofboththemaritimeandtheland-basedregulators.Itallowedthe company’s garage mechanics to record their scheduled inspections, planned maintenance routines and defect repairs on daily, weekly and monthly record sheets.

The records showed that the DUKW masters often reported excessive levels of floodingandthatthemechanicsfrequentlydiscoveredandrepairedholesinthevehicles’ hulls. The types of hull failures the garage staff were discovering should havebeenreportedtotheMCAandtheMAIB.Thehullrepairsshouldhavemetrecognisedmaritimestandardsandshouldhavebeenoverseenorapprovedbytheregulator.

The maintenance management system successfully captured the day to day issues thegaragestaffhadtodealwith.Thenumberofmachinerybreakdowns,hullfailuresandfloodingincidentsinthemonthsleadinguptothesinkingofWQ1 was remarkablyhigh.ThesinkingofWQ4andthewarningsissuedbytheMCAfollowingits most recent incognito inspection, meant the company was fully aware of the waterbornerisksitwasrequiredtomanage.Despitethis,TYDdemonstratedthatitwaspreparedtotakepassengersontothewaterinseverelysub-standardvehicles.

3.4 THE FIRE ON BOARD CLEOPATRA

3.4.1 Overview

Cleopatrasufferedamajorfireduringthewaterbornesectionofanamphibioussightseeingtour.Thefirequicklyspreadtothepassengercompartment,engulfingits28passengersand2crewinthickblacksmoke.ThisledtotherapidabandonmentofthevehicleintotheRiverThames.Noonewaskilledorseriouslyinjured.

3.4.2 Seat of the fire

Theexaminationofthefiredamagedvehicleandburnpatternsindicatedthattheseatofthefirewaslocatedinthecentralvoidspace,belowthecrewcab,immediatelyaftoftheenginecompartmenttransversebulkhead(Figure 55). This assessmentwassupportedbywitnessaccounts.

Smokewasfirstseencomingfromthecoolingairinletontopoftheenginecompartmenthatch,quicklyfollowedbysmokerisingfromunderthefloorplatesinthecrewcabarea.Theareaofdeepestburnwasaroundtheautomaticgearboxoutputshaft’sUJ,whichwasconsistentwiththelocusofthefirebeingdirectlyinthisarea.

3.4.3 Fire development

Oncethefireunderthefloorplateshadbecomeestablished,thetightlypackedgreasecontaminatedbuoyancyfoamprovidedareadysourceoffuel.Asthefoammeltedintothefirethevapoursreleasedwouldhavereadilyignited,addingtotheintensityofthefire.

103

Theenginecompartment’sstarboardextractionfanandtheengineradiatorfanwouldhavedrawntheflamesandhotgasesintotheenginecompartmentthroughtheholesinthetransversebulkhead(Figure 61).Fromhere,thestarboardextractionfanforcedthesmokeandflamesthroughtheenginecompartment’sstarboardventilationvoidspace,pre-heatingandsubsequentlyignitingthebuoyancyfoamcontainedwithinit.Again,thisissupportedbywitnessaccountsthatdescribedthesmokeasbeinginitiallywhite/greyandthenturningblack.Thisisconsistentwiththefirelaboratory’sfindingsthat,whenburning,thelow-densityfoam(packedinthecentralvoidspace)andthehigh-densityfoam(packedintheenginecompartment’sventilationvoidspaces)producedlightgreyandblacksmokerespectively.

Thesmokeandflamesexitedthevehiclethroughthestarboardventilationvoidspace outlet vent. The fan would have continued to run until its wiring failed or its casingmelted.Afterwhich,thehotgasestravellingthroughthevoidspacewouldhave created a chimney effect; drawing cold air into the engine compartment throughtheventsinthehatchcoverstofeedthefire.Thefireunderthefloorplatesalsoburntthroughtheflexibleairhosesthatsuppliedthecompressedairtothefootbrakedistributionvalve.Thesubsequentreleaseofthecompressedairwouldhaveaddedtotheintensityofthefireforashortperioduntilthethreeairreservoirsemptied.

Asthefiredeveloped,itspreadtothepassengercompartment,causingthesidecurtainsandforwardsectionofseatingtobeconsumed.Theflexiblefuelhoselocatedunderthedashboardburntthroughbecauseitdidnotmeettherequiredfireteststandard.Thiswouldhavecausedasmallamountofdieselfueltofeedthefireuntil the fuel supply line emptied.

Thelargeholeinthetransversebulkheadbelowthecrewcaballowedthefiretospreadquicklytothestarboardventilationvoidspace.HadtheconvertedDUKW'spassengerspaceandenginecompartmentbeenseparatedbyaninsulated,gas-tightbulkhead,inaccordancewiththeregulatoryrequirements,thefirewouldnothavespreadsoquickly.

3.4.4 Source of ignition

IndeterminingthecauseofthefireonboardCleopatraanumberofpotentialignitionsources were considered against the circumstances prevailing at the time. These included:

• Contactbetweenthebuoyancyfoamandtheunlaggedsectionoftheengineexhaustintheenginecompartment’sstarboardventilationvoidspace.

• Overheatingofthebandbrake.

• Leaks from the engine fuel oil system, the hydraulic oil steering system and drivetraingearboxes.

• Electrical short-circuit, overload and component defects.

Thefireinvestigationdeterminedthatnoneoftheabovewerecontributory.

104

Figu

re 6

1:Firedevelopment

1-S

eatofthefireinthecentralvoidspaceunderthe

crew

cabdeckplates

2-F

irespreadthroughthenon-continuousbulkhead

intothestarboardenginecom

partm

entventilationvoid

spac

e1

2 2

3

3-T

hefiredevelopedinthepassenger

com

partm

ent f

rom

the

vent

ilatio

n vo

id s

pace

and

he

at tr

ansf

er th

roug

h th

e de

ck p

late

s

105

Self-heating27ofthegreaseonthebuoyancyfoamandheatgenerationasaresultoffrictionbetweenthefoamandthedriveshaftarrangementswerealsoconsidered.

LaboratoryanalysisoftheUJs(Paragraph 2.12.2) used to connect the splined driveshafttotheautomaticgearboxandtheMTBdeterminedthatbothhadsufferedexcessivewear.Inparticular,thelaboratoryreportidentifiedthatthejournalcapsupportingtheneedlerollerbearingsononeofthearmsfittedtotheforwardUJhadfracturedcircumferentiallyandthattheneedlerollerbearingshadsufferedfrompre-existingoverheating.Thereportidentifiedthattemperaturesofbetween300and400°Cwouldberequiredtocausethebluingthatwasevidentontheneedlerollers.Thesetemperatureswouldhavebeensufficienttoignitethefoam,particularlythefineshardsofgreasecontaminatedfoamthatwerepresentatthetime.

Havingconsideredalltheevidence,theMAIBfireinvestigationconcludedthatthemostlikelysourceofignitionwascontactbetweenthegreasecontaminatedbuoyancyfoamandtheoverheatingUJthatconnectedtheMTBdriveshafttotheautomaticgearboxoutputshaft.Regardlessofthemechanismofignition,itisalmostcertainthatthepackingofunsecuredandunprotectedbuoyancyfoamaroundmoving machinery in the central void space introduced the circumstances that led to thefire.

3.4.5 Failure of the universal joint

Thepackingofadditionalbuoyancyfoamaroundthedrivetrainandwithintheventilationvoidspacesmadeitmoredifficulttocarryoutbasicgreasingroutines.Italso changed the characteristics of the environment within the engine compartment and in the spaces surrounding it. The foam created an insulating effect that had increasedtheenginecompartmenttemperatureandreducedtheflowofairoverthe shaft lines. This reduced the dissipation of heat from the central void space. This phenomenon was evident from the reduced temperatures experienced in the passengercompartmentandwasacknowledgedbyLDTwhenDUKWElizabeth suffered a similar UJ failure a couple of months earlier (Paragraph 2.18.2).

The temperature increase around the drivetrain would have reduced the viscosity of the grease within the UJs, and increased the likelihood of the grease deteriorating or beingexpelledfromtherotatingUJs.ThemostcommoncausesofUJfailuresarelackoflubricationandshaftmisalignment.BothwillcausetheUJtooverheatandturnblue.Thejointsstarttosqueakandcausevibrationthatincreaseswithspeed,beforetypicallyfailingwithaloudclunk. This was the case with Elizabeth and would almost certainly have happened on Cleopatrahadthefirenotbrokenout.

LDTwasunderpressuretomeetthedamagedsurvivabilitystandardandgaveinsufficientattentiontotheadverseeffectsandtherisksassociatedwithpackingfoam close to the unguarded drivetrain and unlagged exhausts. The DUKW operator chosetoaccepttheincreasedmaintenanceburdenandcontinuedtooperateitsvehicleswithanelevatedriskofmechanicalfailureandfire.

27 Self-heatingoflubricationproductsonopencellinsulationisarecognisedproblemandtypicallyoccursinareaswithahighambienttemperature.

106

3.4.6 Material condition of Cleopatra and London Duck Tours fleet

AfterCleopatra’sfire,theLDTfleetwassubjecttoasimilarlevelofscrutinytothatofthevehiclesinLiverpool.Ingeneral,theLDTvehicleswerefoundtohavebeenwellmaintainedandinasoundmaterialcondition.Itwasalsoevidentthatthecompanyhadinvestedsignificantresourcestoimprovetheoperationalperformanceandcorrosion resistance of its vehicles.

3.5 VESSEL SURVIVABILITY

3.5.1 Regulatory compliance

TomeetthedamagedsurvivabilitystandardsetoutinMSN1699(M),theUK’sClassVDUKWswererequiredtohavesufficientresidualbuoyancytoremainafloatanduprightinthefullyfloodedcondition.Toachievethis,theLiverpoolandLondonbasedoperatorsundertooktoinsertbuoyancyfoamintothehullsoftheirvehicles.Inadditiontoproviding110%buoyancy,theoperatorswererequiredtoensurethatanybuoyantmaterialusedwassecuredagainstmovementandprotectedagainstdeterioration.

FollowingtheMCA’sinterventioninLondonin2008,BCTQcalculatedthatabout9m3ofclosedcellfoamwasneededtoprovide110%buoyancyonboardtheLDTDUKWs. Despite this, and the sinking of WQ4 and WQ1, LDT continued to operate itsvehiclesontheRiverThameswithlessthan70%ofthecalculatedquantityofbuoyancyfoamrequiredtostayafloatinthefullyfloodedstate.LDTonlytookstepstoaddresstheshortfalloncethelong-standingnon-compliancehadbeenidentifiedbytheMCA.

The evidence compiled during this investigation overwhelmingly showed that the UKoperatorshadnot,andcouldnotachievethemandateddamagedsurvivabilitystandard.HadoneofLDT’sDUKWssufferedasimilarleveloffloodingtothatofWQ1,itwouldprobablyhavesunkequallyasfast.

3.5.2 The use of buoyancy foam

The military DUKW was designed over 70 years ago for mass production during WW2 and was expected to have a very short operational life. DUKWs had an open hull with no watertight segregation and relied wholly on high capacity de-watering pumpstosurviveincidentsofflooding.

InitsreportintothesinkingofMiss Majestic in 1999 (Paragraph 1.16.3),theNTSBconcludedthatithadbeenaflawinthedesignofDUKWs,whentheywereoriginallyconvertedforuseasAPVs,nottoprovideadequatereservebuoyancytoensurethevehiclesremainedafloatinthefloodedcondition.DespiteitsrecommendationtoUSoperatorstoretrospectivelyprovidereservebuoyancythroughpassivemeans,themajorityofUSoperators(andothersacrosstheworld)haveyettosatisfythesurvivabilityrequirementsthathavealwaysbeenpresentwithintheUK.ThisisnotanindicationthattheregulatorsandoperatorsintheUS,andothercountries, are willing to accept lower standards and a higher level of risk, more an acknowledgementthatthesevintagevehiclescannoteasilybemodifiedtomeetmodernsurvivabilitystandards.

107

DUKWshaveaveryhighweighttosizeratioand,asopenboats,ahighvolumeofbuoyancyfoamisneededtosatisfythedamagedsurvivabilitystandard.Insertingsuchlargequantitiesofbuoyancyfoamintroducesmanyproblemsfortheoperatorsandforsurveyors.Specifically,itimpedesaccessforroutinecleaning,maintenanceand inspection, and increases the risk of water entrapment and resultant corrosion. The use of individually cut foam sheets, rather than the previously recommended pre-formedsolidfoamblocks,compoundstheproblemasthefoamtakeslongertoremoveandrefit,andincreasestheriskofmovement.

3.5.3 Buoyancy foam calculations

OncetheMCA’sinvestigationintothesinkingofWQ4establishedthattheLiverpoolDUKWsweredeficientofbuoyancyfoam,TYDsoughttheassistanceofBCTQ.BCTQwascontractedtoprovideareservebuoyancycalculationforatypicalLiverpoolDUKW,andidentifylocationsandapproximatethespaceavailabletofitbuoyantmaterial.BCTQfollowedasimilarprocesstotheoneithadadoptedinLondonin2008,andestimatedthat9.7m3ofbuoyancyfoamwasneededforan8.3tvehicle.

The sinking of WQ1sosoonaftertheMCA’sinterventionwasofgreatconcern.ItcastdoubtonthevalidityofthetheoreticalcalculationsconductedbyBCTQandthepracticalitiesofinsertingandmaintainingtheprescribedquantitiesofbuoyancyfoaminthehullofaDUKW.WhentheMCAdiscoveredthattheLDTvehicleswerealsodeficientofbuoyancyfoam,LDTinitiallyincreasedthequantityoffoaminitsvehiclesto 9.05m3;thevolumecalculatedbyBCTQin2008.Thiswas0.65m3 less than that calculatedforan8.3tLiverpoolDUKW;however,theweightofsomeoftheLondonvehicleshadincreasedtoover9t.ThesediscrepanciesledtheMCAtocarryoutitsowncalculationsandinstructLDTtofurtherincreasethevolumeoffoamto10.6m3.

ItwasclearfromthepracticaltrialsandreconstructionconductedbytheMAIBinJuly 2013 (Paragraph 1.10.3) that:

• Thecalculatedvolumeoffoamrequiredtoprovide110%buoyancycouldnotphysicallybefittedintothelocationsdesignatedintheBCTQreport.

• The insertion of 7.93m3 of foam into the designated locations kept WQ1 afloat28butrenderedthevehicleinoperable.

• TheclosedcellbuoyancyfoamusedbyTYDwasbetween5and20%denser(heavier)thanitsspecificationindicatedandthereforecalculatingthevolumeoffoambyweightcouldnotbereliedupon.

• Atbest,WQ1couldonlyhavehadabout7m3ofbuoyancyfoamonboardwhentheMCAalloweditbackontothewater.

• The assumption that TYD’s DUKWs had the space to insert 12m3ofbuoyancyfoam was wholly unrealistic.

Despitethesefindings,andMAIBrecommendation2013/223(Paragraph 2.2), LDT continuedtostuffincreasedquantitiesoffoamintoitsvehicleswithoutthepotentialadverseconsequencesbeingproperlyassessed.

28 AsevidentfromBCTQ’sbuoyancyassessmentsforbothTYDandLDTvehicles,theinherentbuoyancyofeachconvertedDUKWcanvarysignificantly,andthereforetheresultsoftheMAIB’sreconstructionshouldnotbeusedasabenchmarkforothersimilarvehicles.

108

TheoreticalcalculationshadindicatedthattheUK’sconvertedDUKWswereabletomeetthedamagedsurvivabilitystandardthroughtheinsertionofbuoyancyfoamalone.However,historicalevidenceandpracticaltrialsconductedbytheMAIBdemonstrated that they could not.

3.6 PASSENGER AND CREW SURVIVABILITY

3.6.1 Risk of entrapment

When WQ1 sank, the driver and several of the passengers were still inside the vehicle.Fortunately,whenthebowofthevehicletouchedbottom,thesternremainedabovewaterandtheywereabletoescapethroughthesideandsternwindows.Thelastpassengertoescape,themotherofthe2yearoldchild,climbedoverthepassengercompartmentaccessdoorontotheboardingplatform(Figure 62).

TheNTSB’sMiss Majestic reportidentifiedthatthevehicle’scanopyhadprobablyhampered the escape of many of those who died. The guidance provided in the USCG’sNVIC1-01containedseveralrecommendationsaimedatreducingtheriskofentrapment.ThisledAPVoperatorsintheUS,andinothercountries,toredesign their vehicles’ canopies, windscreens and side curtains, and improve their procedurestohelpincreasepassengersurvivability.

WQ1 and Cleopatrabothhadsolidcanopyroofsandtransparentplasticsidecurtains.Haditbeenrainingatthetimeoftheaccidents,itislikelythatthesidecurtainswouldhavebeenclosed.Thiswouldhavemadeitsignificantlymoredifficulttoescapefromthepassengerspace.Similarly,hadthewaterinSalthouseDockbeendeeper,ortheadditionalfoamnotbeenfittedthepreviousmonth,thelikelihoodofentrapmentwouldhavebeenhigher.

3.6.2 The risk of drowning

Toassistthepassengers’survivalinthewatertheDUKWswereequippedwithPFDs,buoyantapparatusandlifebuoys.AccordingtotheTYDandLDTabandonshipprocedures,thecrewwererequiredtohelpthepassengerstodontheirPFDsandlaunchthebuoyantapparatusbeforeevacuatingtheirvehicles.Duringbothaccidents,theconfinednatureofthepassengercompartmentssignificantlyhamperedthecrews’abilitytoassistthepassengersandcontroltheabandonmentprocess.Inbothcases,thecrewdidnothavethetimetolaunchthebuoyantapparatus and most of the passengers, some of whom could not swim, and crew wereforcedtoenterthewaterwithoutaPFD.

ThedecisionconcerningwhentowearorinflateaPFDinanenclosedorsemi-enclosedboat,isaregulartopicofdebate.ItisobviousthatthemosteffectivewayofensuringthatpassengersarewearingcorrectlydonnedPFDswhentheyabandonaboatistoinsistthattheywearthematalltimesonthewater.However,itmightbearguedthatthiswouldfurtherincreasetheriskofpassengerentrapment;itmightalsodeterpotentialcustomersfromgoingonanamphibioussightseeingtrip.

AsthepassengerswereexpectedtodontheirPFDsbeforeevacuatingintothe water, and TYD offered the opportunity to wear one while on the water, the argumentagainstinsistingthatallpassengerswearaPFDduringthewaterbornesectionoftheDUKWtoursappearstobeweak.

109

Motherofthe2yearoldchildevacuatingthe passenger compartment carrying a child’sbuoyancyaid

Figure 62: The last of the passengers to escape from Wacker Quacker 1’s passenger compartment

Fatherofthe2yearoldchildpassingher to the driver

110

BothTYDandLDTDUKWshavebeenevacuatedsuccessfullyonanumberof occasions over the years following relatively minor accidents and incidents. However,thenatureofthecurrentUKDUKWfleetsmeansthatseriousfireandfloodingeventsgivetheoccupantsverylittletimeinwhichtoassimilatethedanger,preparethemselvestoabandonship,andtoactuallyabandon.WearingPFDsthroughoutthewaterbornejourneywouldsignificantlyreducethetimeneededtoprepareforabandonmentshoulditbecomenecessary,butonlyifthisdoesnotresult in an increased risk of entrapment.

3.6.3 The Irish model

InIreland,APVoperatorshavebeenpermittedtooperatevintageDUKWswithouthavingtoprovideanyresidualbuoyancy.Tomitigatetheconsequencesofseriousflooding,theIrishregulatorrequiredtheoperatorsto:

• Fitexternalbuoyancytubes(Figure 63) designed to slow the sinking process andmakethevehiclesinkbodily.

• Retractthecanopyroofandopenthesidecurtainspriortoenteringthewater.

• RequirepassengersandcrewtowearPFDswhileonthewater.

Passengerswearingbuoyancyaids

Openroofandopensides

Externalbuoyancytubes

Crewman

Figure 63: DublinDUKWsoperatingwithexternalbuoyancytubesandopencanopy

111

• Provideafastrescuecraft,rescuecrewandaninflatableliferaftattheslipway.

• Limit operations to a non-tidal area.

Thisapproachfocusedonpassengersurvivabilitybyreducingtheriskofentrapmentanddrowning,ratherthanvehiclesurvivability,andintroducedseveraloftheinterimmeasuresrecommendedbytheNTSBfollowingthesinkingofMiss Majestic (Paragraph 1.16.3).TheIrishmodeldemonstratesthatopentoppedAPVscanbeoperated successfully in similar weather conditions to those experienced in the UK, andthatpassengersarewillingtowearPFDs.

3.7 ROLE OF THE REGULATORS

3.7.1 Effectiveness of survey and inspection regimes and unscheduled interventions

ItwasapparentfromMCArecordsthatthemaritimeregulatorhadlong-standingreservations over the safe operation and seaworthiness of DUKWs. Despite these concerns, and the maritime regulator’s periodic survey and inspection regime and its unscheduled interventions, the Liverpool and London DUKWs continued to operate foroveradecadewithoutcomplyingwiththemandateddamagedsurvivabilitystandard.

Prior to the sinking of WQ4 inMarch2013,surveyorspaidlittleornoattentiontotheconditionorquantityofbuoyancyfoamduringtheirperiodicinspections.This was particularly apparent in Liverpool, where the originally agreed volume offoamrequiredtomeetthestandardwasunknown.Thefoamthatwasinplacewas typically removed prior to the surveyors’ arrival at the operators’ maintenance garagesanditwasrefittedaftertheyhadleft.Thefoamthathadbeenremovedwasnotmeasuredanditsconditionwasnotassessed.FollowingWQ4’s sinking, the MCA’sLiverpoolMarineOfficeinvestedsignificantresourcestooverseeandverifyTYD’sbuoyancyfoampackingprocess.Havingdoneso,itisdifficulttoexplainhowits surveyors concluded that 12m3offoamhadbeenpackedintothevehicles.

AlthoughtheMCA’sinterventiondidnotachieveitsaimofkeepingafloodedvehicleafloatandupright,fittingmorefoamintothevehiclesdidslowthesinkingprocessand helped keep the stern section of WQ1’scanopyabovewater.Thisgavethepassengersandcrewvitalextrasecondstoabandonthevehicleandprovidedaplatform to keep the 2 year old child out of the water (Figure 62). Those two positive outcomes almost certainly helped to save lives.

Buoyancy issues aside, it was clearly apparent from external visual inspection thattheLiverpoolDUKWshadbeenpoorlymaintained,andoncloserinspectionitwasobviousthattheywerenotfitforuseaspassengercarryingvesselsonthewater.TheLiverpoolDUKWshadbeenallowedtodeteriorateprogressivelyoveraprolongedperiodoftime.ThisgoesbeyondnicheAPVspecificissuesandraisesseriousconcernsabouttheoveralleffectivenessoftheMCA’ssurveyandinspectionregimeappliedtoAPVsoperatinginLiverpool.

HadtheMCA’ssurveyandinspectionregimeofDUKWAPVsbeenmoreeffective,theimpactoffittingtherequiredamountoffoamtoachieve110%buoyancyontheoperabilityofthevehicleswouldhavebeenimmediatelyapparentandalternative

112

solutionssought.Giventhatittookaround13yearstodiscoverthatbothTYDandLDT’sDUKWswereoperatingwithinsufficientbuoyancy,itisextremelyfortunatethat they did not suffer more serious accidents during this period.

3.7.2 Instructions, guidance and training

TheMCAdidnotprovideAPVspecificinstructions,guidanceortrainingtoitsmarinesurveyors.TherewasnoprescribedmethodforverifyingthequantityofbuoyancyfoamonboardanoperationalDUKW,orguidanceonhowitshouldbesecuredandprotected.

ThelackofguidanceandthepressureplacedontheMCAbytheoperators,resultedinareactiveapproachbeingtakentosolvealong-standingregulatorynon-conformity.Inshort,theMCAandtheDUKWoperatorsbecamefixatedwithtrying to stuff the necessary volumes of foam into the vehicles’ hulls without having properlyassessedtheconsequences.

HadtheMCAsurveyorshadadequateinstructionsandguidancetoreferto,amoreconsistentapproachtothesurveyandinspectionofDUKWsmighthavebeenachieved.Furthermore,itislesslikelythatTYDwouldhavebeenallowedtoresumeoperations on the water following the sinking of WQ4, and LDT might not have beenpermittedtopackunsecuredandunprotectedfoamsoclosetotherotatingdrivetrain.

3.7.3 Sharing of information

TheLiverpoolMarineOfficewrotetoMCAheadquartersfollowingthesinkingofWQ4andrecommendedthattheLondonMarineOfficebeinstructedtocheckthelevelsofbuoyancyintheLDTvehicles.ThisinformationwasnotacteduponandtheLDT vehicles were allowed to continue to operate unchecked.

VOSAhadsimilarlong-standingconcernsaboutthesafeoperationofvintageDUKWstothoseoftheMCA.ThefireonboardCleopatra occurred on water and thuscameundertheregulatoryjurisdictionoftheMCA;itcouldjustaseasilyhavehappened on the road 30 minutes earlier, or later, and therefore have come under theregulatoryjurisdictionofVOSA.However,theroadregulatorwasnotconsultedduring the foam packing process.

Neitheroftheregulatorshadadesignatedamphibiousvehiclepointofcontactorindustryspecificexpertise,andtherewerenoarrangementsinplaceforthetwoDfTagenciestosharetheirknowledgeorconcerns.TheAPVindustryhastocomplywithbothlandbasedandmaritimeregulations,andcompliancewithoneisoftenatoddswiththoseoftheother.Insuchcircumstances,compromisesintheformofequivalenciesneedtobeagreed.However,itbecameapparentduringtheinvestigationthattheAPVindustryoftenfounditdifficulttoidentifyfocalpointswithinthetwoDfTagenciesandtoobtainaclearandconsistentarticulationoftheirregulatoryobligations.

Whileitmightbeacceptableforhistoricvehiclestooperateonthehighwayundercertain conditions, the nature of the marine environment is such that more modern safetystandardsneedtobeapplied.Indeed,followingtheMarchioness disaster, LordJusticeClarke’sThamesSafetyreviewconcludedthat,ingeneral,newsafetyregulationsshouldbeappliedtoallvessels,whetherneworexisting.Giventhe

113

potentialfortheretobetensionbetweenthesetwoapproachesthereisaneedfor the regulators involved to cooperate to agree a co-ordinated approach for the guidanceofAPVoperators.

3.8 ROLE OF THE OPERATOR

RegardlessofthechallengespresentedtotheMCAandVOSAbythisnicheindustry, the onus to comply with the UK’s regulations and deliver a safe sightseeing experienceliesfirmlywiththeoperator.BothTYDandLDT shouldhaveidentifiedthattheywereunabletosafelyfittherequiredvolumeofbuoyancyfoamintothespacesdesignatedbytheirnavalarchitect.Thisshouldhavebeenhighlightedtotheregulatorsandanalternativesolutionshouldhavebeensought.Operatorsshould not rely on regulators to provide through-life assurances that their vehicles remaincompliantandsafe.ItwouldbetotallyimpracticaltoexpectMCAsurveyorstooverseetheremoval,measurementandrefittingofbuoyancyfoamoneachvehicleduringtheirperiodicinspections;theMCAisnotresourcedtodoit,andtheoperatorswouldnotbepreparedtopaythemtodoit.

Inadditiontoneverhavingmetthedamagedsurvivabilitystandard,TYDandLDTwereslowtoreacttothesafetyissuesidentifiedfollowingpreviousaccidentsandincidents.SomeofthecircumstancesthatledtothesinkingofWQ4 and WQ1 and thefireonboardCleopatrahadbeenidentified70yearsearlierduringWW2andinmore recent accident investigation reports, some of which involved multiple fatalities.

FollowingthesinkingofWQ4, LDT wrote to TYD and offered its assistance and advice,butitwasapparentthatlittlehadbeendonepriortothattoworktogetherintheinterestsofpassengersafety.ItwasevidentduringthisinvestigationthatsomeoftheUKbasedAPVoperatorswerefiercelycompetitiveand,despitetheirgeographical separation, treated each other’s operations as commercial competition.

Amphibioussightseeingtoursareverypopularwithtouristsandthereappearstobeamarketfortheindustrytoexpand.However,theaccidentsinLiverpoolhadasignificantimpactonthecredibilityandreputationofotherAPVoperatorswithintheUKandabroad.ThiswasparticularlysoinLondonastheLDTvehicleshadavery similar appearance to those in Liverpool and the common perception was that bothcompaniesoperated70yearoldWW2vehicles.Asasmallnicheindustry,itisparticularlyimportantthatAPVoperatorsworktogethertopromotebestpractice,improvevehicleandpassengersafetyandre-buildpublicconfidence.

3.9 EMERGENCY PREPAREDNESS

3.9.1 General

Inordertominimisetheconsequencesofamarineaccident,avesselanditspassengersandcrewneedtobepreparedtodealwithavarietyofemergencysituations.VesselsarepreparedthroughdesignandtheprovisionofLSAandothersafetyequipment.Vesselownersandoperatorspreparetheircrewsbyprovidingthem with guidance and procedures, and through the delivery of training. To ensure traininghasbeeneffectiveandemergencyproceduresarefullyunderstood,ships’crews should conduct realistic emergency response drills on a regular periodic basis.

114

Ships’crewspreparetheirpassengerstodealwithemergenciesthroughtheprovisionofinstructions,demonstrationsand,insomecases,thewearingofPFDs.TYDandLDTpreparedtheirpassengersthroughthedeliveryofverbalsafetybriefsandtheprovisionofPFDdonninginstructionsonthebacksofthepassengerseats.LDTalsogaveapracticalPFDdonningdemonstration.

3.9.2 Emergency preparedness – Wacker Quacker 1

ThelifejacketsonboardWQ1 were stored out of sight, in plastic packaging, under a seatatthebackofthevehicle.Thestrapsonthechildren’sbuoyancyaidshadnotbeenfullyextended.Thismeantthattheparentsofthe2yearoldwereunabletoputoneontheirdaughterbeforeevacuatingthepassengerspace.WQ1’sbuoyantapparatuswasstowedonthecanopyroof.ThecapacityoftheLSAonboardWQ4 and WQ8waslessthanthemaximumnumberofpersonsthevehicleshadbeencertifiedtocarry.

ThecrewonboardWQ1hadbothreceivedtrainingandhadparticipatedinemergencydrills,buttheabandonshipdrillscarriedoutbythecrewwerenotinaccordance with the procedure in TYD’s safety management system. During the drills the crew practised evacuating the passengers onto pontoons at the edge of thedocksratherthanintothewater.Themasterhadexperiencedproblemsonthewater in the past and on those occasions he had evacuated his passengers onto pontoons.Itwasevidentthatthemasterandhisdriverwerenotfamiliarwiththeprocedureforabandonmentonthewaterandthereforetheirtraininganddrillshadnotbeeneffective.

Priortoenteringthewaterinthebuild-uptotheaccident,WQ1 was parked on the roadatthetopoftheslipwayforjust15seconds.Thisindicatesthatthesafetybriefwaseitherextremelyshortorwasgivenontheroadpriortoarrivingatthedocks.ItwasapparentfromthefeedbackprovidedtotheMAIBbythepassengersthatthesafetybriefgivenbythecrewhadnotbeentakenseriously,andwasineffective.TheMCAhadrecentlywrittentoTYDlistingitsconcernsaboutthestandardofthecompany’spre-splashdownsafetybriefsandinstructedittoprovidealifejacketdonningdemonstration.However,thecompanycontinuedtodeliveritssafetybriefsinajocularmanneranddidnotshowpassengershowtodontheirlifejackets.Subsequently,themajorityofthepassengersdidnotknowwheretheirlifejacketswerestowed,howtoputthemonorhowtoinflatethem.

InearlieriterationsofTYD’sSafety Policy and General Safe Operational Procedures manual,thepre-splashdownsafetybrief(Annex C) was more comprehensive and included practical demonstrations. Passengers might have an expectation that they aregoingtobeentertainedbythecrewduringtheirtour,andmightnotwanttolistentoasafetybrief.Nonetheless,itisessentialthattheyaregiventhebestopportunitytounderstandthesafetyproceduresandknowhowtoputonanduseaPFD.Inordertoachievethis,thesafetybriefshouldbeclearandconciseandbedeliveredwith conviction and authority.

ItisclearthatWQ1,itscrewandtheirpassengershadnotbeenadequatelypreparedtodealwiththeemergencysituation.Thelifejacketswerenotreadilyavailable,manyofthepassengersdidnotknowhowtodonthemorinflatethemandthecrewwerenotfamiliarwiththeabandonshipprocedure.

115

3.9.3 Emergency preparedness – Cleopatra

AswasthecaseinLiverpool,Cleopatra’s crew had little time to react to a rapidly escalating situation. However, the vehicle, its crew and their passengers had beenbetterpreparedtodealwithit.Itwasclearfromthepassengerfeedbackthat,althoughgiveninalight-heartedmanner,thesafetybriefwasmoreeffectivethanthatgivenonboardWQ1, and the passengers were shown how to don theirbuoyancyaids.Further,thebuoyancyaidsonboardCleopatra were readily accessibleinracksabovethepassengers’heads.However,thebuoyancyaidshadbeenstowedinplasticbagsthatweredifficulttoremoveandtheirstrapshadnotbeenfullyextended.TheobviousadverseimplicationsofthishadpreviouslybeenpointedoutintheMAIB’s2001Beatrice report (Paragraph 1.16.2).

ThecrewonboardCleopatrahadbothreceivedtrainingandhadparticipatedinemergencydrills.AccordingtoLDT’sabandonshipprocedurethecrewwereexpectedtolaunchthebuoyantapparatusandthenseatthepassengersontopofthem.Thebuoyantapparatuswasnotdesignedtosupporttheirratedcapacityofpeopleoutofthewater.Althoughitmightbeappropriatetoplacesmallchildrenorinjuredandweaksurvivorsontopoftherafts,attemptstoseatallthepassengersonthemwouldprobablybeunsuccessfulandwoulddelaytheabandonmentprocess.

3.10 EMERGENCY RESPONSE – WACKER QUACKER 1

3.10.1 Raising the alarm

The crew of WQ1wereunconcernedwhentheDUKWfailedtoclimbtheslipwayastheyhadexperiencedthisproblemearlierinthedayandonseveraloccasionsinthepast.Thedriverandthemasterthoughtthevehiclehadagearboxproblemandfocusedtheirattentiononresolvingit.Theydidnotinformtheoperationsofficeorexplain to the passengers what had happened.

When WQ1begantofloodthevehicle’selectricbilgepumpsstartedautomaticallyandsomeofthepassengerstriedtoalertthecrew.Initially,thecrewappearedtoignoretheshoutsfromthebackofthevehicleandthebilgepumpalarms.Again,they were used to water entering the hull and might have initially underestimated the rate of ingress.

When the crew realised the severity of the situation, the master made his way to the backofthevehicle.Thedriverwasunabletofindthemaster’smobilephoneandthereforedidnotalerttheoperationsofficeoremergencyservices.Fortunately,theincidentoccurredinaverybusypartoftheSouthDockscomplexwhereimmediateassistancefrombystanderswasavailable.

3.10.2 The abandonment

AsWQ1 sank,themajorityofthepassengersanditstwocrewenteredthewater,manywithoutlifejackets.Thosewhohadthemwereeitherunabletoputthemonproperlyorcouldnotinflatethem.Somepassengersswamashore,butthemajoritywererescuedfromthewaterbythepeopleonboardthreerecreationalnarrowboatsthathadbeenberthedinSalthouseDock.

116

WQ1’s crew might have thought that they had time to co-ordinate and control the abandonmentprocessfollowingtherecentadditionofbuoyancyfoam.However,vitaltimewaslosttryingtopassthelifejacketsforwardtothepassengersatthefrontofthevehicleandhavingtoremovethemfromtheirpackaging.Instinctively,thepassengersusedtheirowninitiativeandabandonedWQ1 with or without theirlifejackets.Someoftheweakerswimmersandsmallerchildrenwereabletoholdontothebuoyantapparatusthathadbeenlaunchedearlierbya70yearoldpassengerwhowasgrandfathertotwoofthechildrenonboard.

WQ1’screwwereunabletoeffectivelyassistthepassengersbecauseoftheconfinednatureofthevehicleandabasiclackofemergencypreparedness.AccordingtoTYD'sabandonshipprocedure,thecrewwereexpectedtolaunchthebuoyantapparatusandhelpthepassengersdontheirlifejacketspriortoabandonment.Hadthecrewbeenabletocalmandcontrolthepassengers,andattemptedtofollowthecompany'sabandonshipprocedure,itislikelythatliveswouldhavebeenlostbecauseofthespeedatwhichthevesselsank.Giventhecircumstancesofthisaccident,thecompany'semergencyresponseproceduresprovedtobeunrealistic.

3.10.3 Shore based emergency response

Within10minutesofthealarmbeingraisedbymembersofthepublictheemergencyservices(police,fireandambulance)wereonsiteandhadtakencontrolof the incident. By this time, the passengers and crew had either swum ashore or beenrecoveredontothenarrowboats.

Duringtherescuephase,TYDdidlittletoassistanditsrescueboatremainedalongsideinSalthouseDock.Therewasnonominatedmusterpointandsomepassengersdispersedfromthescenebeforeaheadcountwascarriedout.Althoughtherewerenomajorinjuries,ittookthepoliceseveralhourstoaccountforeveryone.Hadthecompanypreparedamajorincidentresponseplanandcarriedoutregulardrillstochallengeandtestit,itscontributiontotherescueeffortmighthavebeenmoresignificant.

Theeffortsofthenarrowboatcrewswereinstrumentalinthesaferecoveryofmanyof those in the water, particularly as some were struggling to, or could not swim. TheyoungmanwhojumpedintothewaterfromthebowofPredator 3 (Figure 64) to assist those struggling in the water was fully aware of the risk he was taking, and shouldbecommendedforhisbravery.

3.11 EMERGENCY RESPONSE – CLEOPATRA

3.11.1 Raising the alarm

Atthefirstsightofsmokethetourguideadvisedpassengersofapossibleproblem,and the master steered CleopatratowardstheshoreinthebeliefthatCleopatra had anexhaustproblem.Thesmokesoonescalatedtofire,andjustoveraminutelaterthemasterattemptedtoraisethealarmusinghisVHFradio.Withinseconds,radiocommunication was lost and it took the intervention of an unknown third party to informVTSthatCleopatrawasonfire.

117

Intheeventofavesselbeingingrave and imminent danger29, emergency protocol requiresastrickenvesseltotransmitadistressalert.InCleopatra’s case this did nothappen.Themaster’sverbaldistresscallwasnotprecededbythesignalword“Mayday”andwasnottransmittedonchannel16,thededicatedVHFchannelforsafety alerting.

ThedistressprotocolforvesselsfittedwithDSCenabledradiosisfortheradiooperatortosendaninitialalertbyDSCbeforecommencingvoicecommunications.OnmostradiostheDSCfunctionisactivatedbyapushbuttonthatneedstobehelddepressedforbetween3and5seconds.ThebasicDSCdistressalertthenautomaticallytransmitsthevessel’sidentification,itspositionandtimeoftransmissionwithoutanyfurtherinputfromtheradiooperator.Althoughitistemptingandcomfortingtouseverbalcommunicationsinsuchasituation,thesameinformationasgivenbyDSCcannotbeverballytransmittedin3to5seconds.Additionally,erroneouspositionalinformationisoftengivenduringverbalcommunicationsbyoperatorsunderpressure.Inmanycircumstances,thesevitalearlysecondscanmakethedifferencebetweenasuccessfulrescueandthelossoflife.

3.11.2 Fire containment and fire-fighting

Althoughthefirehadstartedunderthedeckplatesinthecrewcabarea,theearlyindicationswerethatthecrewweredealingwithaproblemwithintheenginecompartment.WhenitbecameobviousthattherewasafireonboardCleopatra, themastertookthedecisiontodrivehisvehicleontothenearestriverbank.Onceaground, the master’s initial priority was to use the engine to keep his vehicle

29 Theunambiguousinternationallyrecognisedinterpretationofavesselindistress.

Figure 64: MemberofthepublicjumpingfromPredator 3 to help passengers in the water

118

pressedontotheriverbank.Whentheenginewasstoppedthecrewwereunabletoclosetheventilationvoidspacefiredampersbecausethe‘R’clipshadbecometoohotandtheheatandsmokegeneratedbythefirequicklyforcedthemtomoveawayfromthecrewcabarea.Theenginecompartment’sfixedfire-extinguishingsystemwasnotactivatedandnoneofthevehicle’sportablefire-extinguisherswereusedtofightthefire.

Thiswasaseriousfirethatspreadquickly,anditisforjustsuchemergenciesthatfiredrillsareconducted.Inthisinstance,onboardfire-fightingwouldhavebeenonlypartiallyeffectivebecausethefirewasnotintheenginecompartment.Therefore,themaster’sdecisiontodrivehisvehicleagroundprovedtobeasoundoneandsignificantlyreducedtherisktolifeduringtheabandonment.Nonetheless,itwouldhavebeenappropriateforthecrewtoisolatethefuelsupplies,closetheengineventfiredampersandoperatetheenginecompartment’sfixedfire-extinguishingsystemassoonaspossible.

3.11.3 The abandonment

WhenthefirebrokeoutonboardCleopatramostofthepassengersgrabbedbuoyancyaidsandmovedtowardsthebackofthevehicle,awayfromtheflames.Thetourguidewasunabletoassistorcontrolthepassengersandhadtoforcehiswaypastthemtogettothesterndoor.Unabletoopenthedoor,somepassengers,understandably,abandonedintothewaterwithoutgrabbingordonningabuoyancyaid.

Thespeedatwhichthefireescalatedgavelittleornotimeforthecrewtopreparethepassengersandthevehicle’sLSAforacontrolledabandonment.SimilartotheabandonmentofWQ1inLiverpool,thepassengersonboardCleopatra were forced to act on instinct and exit the passenger space under their own initiative. Had the sidecurtainsbeendown,theconditionswithinthepassengerspacewouldhavedeterioratedevenmorerapidlyanditwouldhavebeenmoredifficulttoescape.

3.11.4 Shore based emergency response

Within2minutesofthefirebreakingout,Cleopatra’spassengersandcrewhadbeenrecoveredfromthewaterbythecrewofthreepassengercarryingRIBs.LFB’sfireboat,Firedart,andashore-sidefireappliancewasonthescenewithin5minutes,andthefirewasquicklyextinguished.Thepoliceandambulanceservices,andtheRNLIwerealsoquicklyonthesceneandsoonestablishedthatallpassengersandcrewhadbeenrecovered and were safe.

Thelocationandtimeoftheaccidentwasextremelyfortuitous.Hadthefireandsubsequentabandonmenthappenedathighwater,orinmid-river,theoccupantswouldhavehadtoabandonintodeeperwater,whichwouldhaveincreasedthelikelihoodofseriousinjuryandlossoflife.Althoughtheemergencyservicesrespondedquickly,thecommendableactionsofthoseonboardtheThamesRIBExperienceboatswasfundamental to the safe recovery of Cleopatra’s passengers and crew.

119

SECTION 4 - CONCLUSIONS

4.1 THE SINKING OF WACKER QUACKER 1

1. WQ1sankbecauseitsufferedrapiduncontrolledflooding,anddidnothavethemandatedlevelofresidualbuoyancyneededtokeepitafloatinitsfullyfloodedcondition. [3.3.1]

2. WQ1begantofloodwhentwoholesweretornintoitshullbytheforcesgeneratedwhenitspropellerwasfouledbyadiscardedcartyre.Itbecamefullyfloodedbecauseitselectricpoweredbilgepumpswereunabletocopewiththerateofwateringress. [3.3.1]

3. Had TYD learned lessons from previous similar hull failures or followed the advice of LDT and reinforced WQ1’s propeller shaft tunnel plating, the hull might not have failed. [3.3.2]

4. WQ1wasinanextremelypoorcondition.Itshullwasheavilycorrodedandwascoveredwithapatchworkofunapprovedsubstandardrepairs,itwasmechanicallyunreliable,andregularlyfailedtoclimboutofthedock.Thevesselwasnotseaworthyanditshouldnothavebeenoperatingonthewater. [3.3.4]

5. TYD'svehiclessufferedahighnumberofmechanicalbreakdowns,hullfailuresandfloodingincidentsduringthemonthsleadingtothesinkingofWQ1. The operatorwasawareofthepotentialconsequencesoffloodingbutcontinuedtotakepassengers onto the water in unseaworthy vehicles. [3.3.5]

6. TheMCA’ssurveyandinspectionregimeinLiverpoolprovedtobeineffective;thematerialconditionoftheTYDfleetwasallowedtoprogressivelydeterioratetoanunsafe condition over a prolonged period of time. [3.7.1]

7. WQ1,itscrewandtheirpassengershadnotbeenadequatelypreparedtodealwiththeemergencysituation.Thelifejacketswerenotreadilyaccessible,manyofthepassengersdidnotknowhowtodonthemorinflatethemandthecrewwerenotfamiliarwiththeabandonshipprocedure. [3.9.2]

8. TYDdidlittletoassisttheemergencyresponseeffort,itsrescueboatwasnotusedand a passenger headcount was not carried out. [3.10.3]

4.2 THE FIRE ON BOARD CLEOPATRA

1. ThelocusofthefireonboardCleopatrawasunderthecrewcabfloorplatesinthehull’s central void space. [3.4.2]

2. Thebuoyancyfoamthathadrecentlybeenpackedintothecentralvoidspaceprovided a ready source of fuel. [3.4.3]

3. Hadaninsulatedgas-tightbulkheadbeenfittedbetweenCleopatra'saccommodation areas and engine compartment, in accordance with passenger ship constructionstandards,thefirewouldnothavespreadsorapidlythroughtheenginecompartment to the passenger space. [3.4.3]

120

4. Thepackingofunsecuredandunprotectedbuoyancyfoamaroundthevehicle'sengine compartment and drive shafts introduced the circumstances that led to the fire.[3.4.4]

5. Themostlikelysourceofignitionwascontactbetweengreasecontaminatedbuoyancyfoamandanoverheatingdriveshaftuniversaljoint. [3.4.4]

6. LDTgaveinsufficientattentiontotheadverseeffectsofpackingbuoyancyfoamclose to moving machinery and unlagged exhausts, and allowed its vehicles to operatewithanelevatedriskofmechanicalfailureandfire. [3.4.5]

7. Themaster’sdecisiontodrivehisvehicleagroundprovedtobeasoundoneandsignificantlyreducedtherisktolifeduringtheabandonment.[3.11.2]

8. Cleopatra,itscrewandtheirpassengerswerebetterpreparedtodealwithanemergencysituationonthewater.However,thespeedatwhichthefireescalatedgavelittleornotimetoconductacontrolledabandonment.[3.11.3]

9. Hadthefireandsubsequentabandonmenthappenedathighwater,orinmid-river,thelikelihoodofseriousinjuryandlossoflifewouldhavebeensignificantlyhigher.[3.11.4]

4.3 COMMON SAFETY ISSUES

1. Foroveradecade,littleattentionwasgiventobuoyancyfoambytheoperatorsandthe regulators, and the UK’s Class V DUKWs were operated without having met the mandateddamagedsurvivabilitystandardssetoutinMSN1699(M).[3.5.1]

2. HadaLondonDUKWsufferedasimilarleveloffloodingtothatofWQ1, it would probablyhavesunkequallyasfast.[3.5.1]

3. The UKs converted DUKWs could not safely meet the regulatory standards for damagedsurvivabilitythroughtheinsertionofbuoyancyfoamalone.[3.5.3]

4. ThecombinationofalackofguidanceandinstructionsandpressurefromtheDUKWoperatorsledtoareactiveapproachbeingtakenbytheMCAtoalong-standingregulatorynon-conformityandafixationwithstuffingfoamintorestrictedspaces. [3.7.2]

5. ThefireonboardCleopatra occurred on water and thus came under the regulatory jurisdictionoftheMCA.Itcouldjustaseasilyhavehappenedontheroad,andthereforehavecomeundertheregulatoryjurisdictionofVOSA.[3.7.3]

6. TheMCAandVOSAbothhadlong-standingconcernsaboutthesafeoperationoftheconvertedDUKWs,buttheydidnotsharetheirconcernsandknowledge,orcommunicate with each other to address common issues and minimise regulatory conflicts. [3.7.3]

7. Theconfinednatureofthepassengercompartmentmadeitextremelydifficultfor the crew to control and assist the passengers and increased the likelihood of entrapment. [3.11.3]

121

8. ItwasextremelyfortunatethatallonboardWQ1 and Cleopatrawereabletoevacuateintothewaterunharmed.Inbothcasesthepassengerswereforcedtoacton instinct and exit the vehicles under their own initiative. Had:

• thequantityofbuoyancyfoamonboardWQ1notbeenincreased2monthsbeforetheaccident;

• thedepthofwaterinSalthousedockbeengreater;

• Cleopatra'smasterbeenunabletorunhisvehicleaground;

• itbeenrainingonthedayofeitheraccident(andhencethesidewindowsbeingclosed);or

• the crews of WQ1 and Cleopatrafollowedtheircompany'sabandonshipproceduresbycontrollingthepassengersandpassingoutlifejacketsbeforeleaving their vehicles;

theriskofentrapmentandthelikelihoodoflossoflifewouldhavebeenconsiderablyhigher. [3.10 and 3.11]

122

SECTION 5 - ACTION TAKEN

5.1 MAIB ACTIONS

The Marine Accident Investigation Branch has:

• IssuedsafetybulletinSB3/201330,containingrecommendationS2013/233,whichrecommendedtheMCAto:

• In addressing recommendation 2013/221, ensure that the means used by DUKW operators to achieve the required standard of buoyancy and stability for their vessels does not adversely impact on their safe operation. Furthermore, these vessels should not be permitted to operate until satisfactory levels of safety can be assured under all feasible operating conditions.

5.2 ACTIONS TAKEN BY OTHER ORGANISATIONS

The Maritime and Coastguard Agency has:

• CarriedoutaninternalinvestigationintothefileandrecordmanagementinitsLiverpoolandOrpingtonMarineOfficeswithparticularattentiongiventoMCAoversightofDUKWAPVs.Asaresultofitsfindingsithas:

• Taken steps to improve the standards of record management at, and the sharingofinformationbetweenitsMarineOffices.

• RequiredAPVoperatorstosubmitdocumentaryevidencetodemonstratecompliancewiththeirstatutoryobligations.

• Producedamethodologyforensuringthecorrectquantityofbuoyancyfoamis in place.

• Commissioned an independent DUKW regulatory compliance study and outlined a phased approach to the development and implementation of a mandatory technicalstandardforamphibiouspassengervehicles.

• EnforcedtheregulatoryrequirementtofitaninsulatedfireprotectionbulkheadbetweentheenginecompartmentandaccommodationspaceonexistingDUKWs.

• HasworkedwithLDTtodevelopa‘referenceDUKW’standardand,asaresult,hasapprovedaseriesofdesignmodificationsandauthorisedthecompanytoresumeitsoperationsontheRiverThames.

DfT Traffic Commissioner has:

• WithdrawnPearlwildLtd’sPassengerServiceVehiclelicence.

30 SB3/2013–The sinking of the DUKW amphibious vehicle Wacker Quacker 1 in Salthouse Dock, Liverpool on 15 June 2013, and the fire on board the DUKW amphibious vehicle Cleopatra on the River Thames, London on 29 September 2013.

123

Pearlwild Ltd has:

• PlacedTYDintoadministrationandisnolongeroperatingAPVs.

London Duck Tours Ltd has:

• ModifieditsDUKWstosatisfythedamagedsurvivabilitystandardbyusingacombinationofbuoyancymaterialandcompartmentalisation(Annex F). The modificationsinclude:

• Theinstallationofaninsulatedgastightandwatertighttransversebulkheadbetweentheenginecompartmentandthepassengerspace.

• Mountingthepassengerseatsontopofindividualairtightaluminiumbuoyancytanks.

• Replacementofthesteeldeckplateswithaluminium.

• Theprovisionofdedicatedvoidspacesfortheinsertionofbuoyancybytherelocation of the fuel tank and the ancillary machinery and services under the passengercompartmentfloorplates.

• BoxinginthedrivetrainjournalbearingsandUJs.

• The replacement of the original fuel lines with approved solid drawn and flexiblesteelpipework.

• Changes to the engine and engine compartment cooling air system.

• Theprovisionofafiredamperfortheforwardenginecompartmenthatchcooling air vent.

124

SECTION 6 - RECOMMENDATIONS

The Maritime and Coastguard Agency and the Driver and Vehicle Standards Agency are recommended to:

2014/153 Identifysinglepointsofcontactforamphibiousvehicleissuesandputprocesses in place to allow them to work together, in consultation with the industry, to explore potential cross agency synergies, identify regulatory conflictsandagreeacoherentapproachtothesurveyandcertificationofnewandexistingamphibiouspassengervehicles.

The Maritime and Coastguard Agency is recommended to:

2014/154 Provideamphibiousvehiclesurveyguidanceandinstructionstoitssurveyors.

2014/155 Workwithindustrytodevelopanamphibiousvehicleoperators’codeofpractice.

2014/156 Ensure that measures to reduce the risk of passenger entrapment and improvethelevelsofpassengersurvivabilityareincludedinitsproposedtechnicalstandardforamphibiouspassengervehicles.

2014/157 RequireexistingDUKWoperators,whichmaychoosetorelyontheinsertionofbuoyancyfoamtomeettherequireddamagedsurvivabilitystandards,todemonstratethroughriskbasedanalysisthatthefoamdoesnotadverselyaffect the safe operation of the vehicles.

London Duck Tours Ltd is recommended to:

2014/158 Usethesafetylessonsidentifiedinthisreporttotakefurtheractiontoensure,asfarasisreasonablypracticable,itsvehicles,crewandpassengersarebestpreparedtodealwithemergencysituations.Inparticular,attentionshouldbegiven to:

• ThereadinessanduseofPFDs:thepracticalitiesofthecurrentarrangementsshouldbereviewedandconsiderationgiventorequiringallpassengerstowearPFDswheneverDUKWsarewaterborne.

• Establishingappropriateandachievableemergencyprocedures:theseshould include the marshalling of passengers, alerting potential responders andabandonment.

• Development of effective training drills.

• Enginecompartmentshut-downandfire-fighting.

• Lowering the risk of passenger and crew entrapment: assess in particular whether the current canopy arrangements are appropriate.

Safetyrecommendationsshallinnocasecreateapresumptionofblameorliability

Marin

e Accid

ent R

epo

rt