175

Annex 4

ANNEX 5

MAIB’s interpretation ofthe results of themodel experiments

The model experiments made an invaluable contribution to the investigation, revealing thevulnerability of even a well-found vessel such as Gaul to an encounter with a group of verylarge breaking waves.

Model experiments provide a satisfactory way to predict performance and behaviour of thefull-scale vessel, although there are often inherent limitations. This is because themodelling is of necessity a simplification of the vessel and the environment, and the resultsmust therefore, take this into account. Model experiments in steep breaking waves are theonly viable way of predicting the ship’s response to extreme seas.

The model used possessed greater stability (scaled) than the full-size vessel, simply becauseof unavoidable effects of scale. For example, the 1 millimetre thick bulwarks on the modelscale to 46 millimetre full size, about six times the real thickness. The range of stability ofthe model was determined by experiment to be almost 180°, which meant that it wascompletely self-righting. This was not realistic as the full size vessel was calculated to have arange of stability of about 120° (Figure 10 in Annex 2). However, the calculations for thefull size vessel ignored many minor items of buoyancy such as the winches and box sectionmasts. If these items had been included, the calculated range of stability of the vessel wouldhave increased, but not to 180°. The range of stability of the vessel probably lay between120° and 130°. This would of course be rapidly degraded by flooding through non-closeableopenings.

As a consequence of its exaggerated range of stability, the model would not have beenknocked-down to as large an angle as the full size vessel, and would have recovered morequickly. In effect, a knock-down on the full size vessel would have been worse than impliedby the model experiments. For this comparison it has been assumed that both the modeland full size vessel are completely sealed against water ingress, which was also unrealistic.However, the experiments carried out with doors and hatches opened, showed that waterwould not have flooded in fast enough during a knock-down to have prevented a partialrecovery of the vessel.

Also, the model was not fitted with the minor ventilators on the starboard side, which ithas been estimated would have increased the rate of flooding to the factory and otherspaces during a knock-down. Again, this factor would have allowed the model to recoverbetter than the full size vessel.

176

Marine Accident Report 4/99 ‘Gaul H 243’

The effect of cargo and gear shifting was also not represented on the model, nor theblockage of freeing ports from loose gear sliding across the trawl deck. Both these effectswould have been critical to the survival of the vessel.

The model experiments are indicative of what could have occurred with the vessel, butprobably exaggerated its ability to have withstood, and to have recovered from, a knock-down. Therefore, the data gleaned from the model experiments, when scaled for the full-size vessel, could be optimistic. However, the calculated large range of stability of the vesselstrongly indicates that she would have made a partial recovery from the knock-down.

Nevertheless, in spite of inherent limitations of the experiments, it is concluded that if thevessel had been struck by a large breaking wave it would have been rolled to an extremeangle and this would have compromised its safety.

177

Figures

FIGURES 1–46

179

Life raft containers

Stairs down

Wooden door Man overboardlifebuoy stowageP & S

Wooden door

Life raft containers

Figure 1: Gaul – bridge deck layout

Radio room

Chartroom

Ship’swheel

Flaglocker

Controlconsole

Haul/shootconsole

Han

dra

ilWooden door

180

Figure 2: Kurd – bridge interior looking to starboard

1

2

3

45

6

11

10

12

7

8

9 KEY:

1 Ship’s wheel

2 Rudder angle indicator

3 Voice pipe

4 Periscope for binnacle compass

5 Searchlight handle

6 Voice pipes

7 Radar set

8 Gyro compass repeater

9 Steering tiller for docking

10 Stairwell

11 Chart room

12 Control console

181

4

Figure 3: Kurd – bridge interior (looking to port over control console)

2

3

6 5

1

KEY:

1 Haul/shoot console

2 Fire alarm panel

3 VHF telephone handset

4 Autopilot controls

5 Chart room

6 Radio room

182

KEY:

1 Autopilot control

2 Paper plotter

3 Radar

4 Gyro compass repeater

5 Elac echo sounder

6 Echo sounder

7 Propeller pitch control lever

Figure 4: Kurd – bridge interior (control console, looking forward)

7

2

1

3

4

56

KEY:

V = Natural vents to factory

1S, 1P…13P = Location of freeing ports

LKR = Locker

183

Figure 5: Plan of trawl deck (aft)

Looking forwardalong trawl deck

FISHLOADINGHATCH

FISHLOADINGHATCH

STERNRAMPGATES

HATCHTO NETSTORE

GOALPOSTMAST

“A” FRAMEMAST

SPARE BOARDSTOWAGE

STARBOARD

VENTS TOENGINE ROOM

DOOR TO FACTORY

DOOR TO ENGINE ROOMESCAPE

HATCH TOFISH MEAL

HOLD

DOOR TOACCOMMODATION

V

HATCH TOFISH MEAL

HOLD

22" DEEP BOBBIN RAIL P&S

ENGINE ROOM VENT

SPARE BOARDSTOWAGE

PORT

HATCH TOENGINEERS’

STORE

Typical freeing port

V LKR OVER

LKR OVER

TRAWLWINCH

Looking to Port

V

V

Plan and photographs courtesy of BAe Land Sea System

184

Figure 6: Kurd – looking aft along trawl deck

KEY:

1 “A” Frame mast

2 Goalpost mast

3 Net store hatch

4 Fish loading hatches

5 Stern ramp gates1

2

34

5

4

185

Figure 7: Kurd – looking forward along trawl deck

KEY:

1 DF Aerial

2 Main mast

3 Liferafts

4 Door to accommodation

5 Net arena

6 Bobbin rail

7 Door to factory

1

2

33

4

5

6

7

186

Figure 8: Kurd – stern ramp and gates

1

2

KEY:

1 “A” Frame mast

2 Goalpost mast

3 Port trawl door in its stowedposition

4 Hydraulically-operated sternramp gates closed

5 Starboard trawl door in itsstowed position

34

5

187

Figure 9: Arab – Emergency escape door from engine room, located trawl deck starboard.

Figure 10: Arab – trawl deck, starboard. Door to factory

188

1

34

5

2

KEY:

1 Natural ventilation inlet to factory

2 Door to factory

3 Loop used for holdingdoor open

4 Emergency escape doorfrom E.R. (Fig. 9)

5 Stowage rack for sparetrawl doors

189

Figure 11: Arab – trawl deck (aft), looking aft. Fish loading hatches, open

Photograph courtesy of BAe Land and Sea Systems

KEY:

1 Hatch to netstore, open

2 Locking pin

3 Securing clip

4 Goalpost mast

1

2

3

3

3 3

3

4

3

2

190

Figure 12: Arab – factory deck

191

Figure 13: Description and technical specification of Sealion ROV

Reproduced by courtesy of Racal

192

Figure 13 (continued): Description and technical specification of Sealion ROV

193

Figure 14: Sidescan sonar image of wreck

Note: 1) Areas shaded in purple are fishing nets caught in the wreck.

194

Figure 15: Side scan sonar survey tracks

Area of ROV visualinspection on wreck

(see Figure 17)

195

Figure 16: Sidescan sonar images of the seabed around the wreck

196

Figure 17: Area of ROV visual inspection

197

Figure 18: Photo mosaic of starboard side

The photographic mosaic is incomplete in those areas where there was not sufficient detail or the correct perspective provided by the video survey.

198

Figure 19: Photo mosaic of port side

The photographic mosaic is incomplete in those areas where there was not sufficient detail or the correct perspective provided by the video survey.

199

Figure 20: “Beaufort” liferaft container on seabed

200

Figure 21: Damage to outboard face of port funnel

3D interpretationof funnel damage

201

Figure 22: Photo mosaic of transom

Trawl door

Goalpost mast

Steerable Kort nozzle

202

Figure 23: A typical steerable Kort nozzle

Figure 24: Gaul’s Kort nozzle viewed from astern

Photograph courtesy of Kort Propulsion

Figure 25: Photo mosaic of forecastle deck (forward)

203

1st port vent

Figure 26: Photo mosaic of deckhouse front

204

Weathertight door toaccommodation

closed and secured

Bridge deck

(P)(S)

Forecastledeck

205

Figure 27: Photo mosaic of bridge front

206

Figure 28: Control console (front, looking to port)

Figure 29: Control console (starboard forward corner)

Autopilot mssing

Radar

Elac echosounder

Echosounder

Rottingpanellingaroundconsole

Note: See Figure 4 for photograph of control console

207

Figure 30: Steering position at bridge front

Tiller to port

Gyrocompassrepeater

Autopilot/Tillerchangeoverswitch

208

Figure 31: Radar set below bridge windows

Note: (1) Hood is missing(2) Cathode Ray Tube has imploded(3) See Figure 2 for position and, appearance of original radar

Hatch cover Access opening

Figure 32: Open access hatch on aft face of port funnel

209

Figure 33: Looking aft onto the open fish loading hatches

Note: 1) The dark border to the hatch covers is the rubber seal which ensures that they are weathertight when closed.2) The hatches are undamaged.

Looking down onto open fish loading hatches

Starboardhatch cover

Porthatchcover

210

Figure 34: View down outboard edge of port fish loading hatch

Note: 1) Locking pin should have been visible in this view, it is missing.2) Securing clip is in closed position, but is undamaged.

Portbobbin rail

Port fish loading hatch

Securingclip

211

Figure 35: View onto port bobbin rail showing port stern ramp gate fully open and in its recess

Aft

Port bobbinrail

Port sternramp gate

Port fishloading hatch

Forward

212

‘C’ Class boat davit, stowed

Liferaft rack, empty

Forward

Nets rising from net arenaAft door to net store, open

Spare trawl door stowage rack, empty

Aft

Figure 36: Photo mosaic of port side trawl deck

Liferaft rack empty

Inflatable boat missing

Forward

Door to factory, open

Nets fallen out of arena tostarboard

Aft

Spare trawl door rack, empty

213

Figure 37: Photo mosaic of starboard side trawl deck

214

Figure 38: Bobbins lying inside net arena

Starboard side forecastledeck (aft)

Starboard coamingof net arena

Bobbin (1)

Forward Aft

Aft

Forward

Port coaming of net arena Bobbin (1)Bobbin (2)

Danlenobobbin

Note: 1) Bobbins (1) and (2) are connected.2) The danleno bobbin is a spare attached to the port bobbin rail.

215

Figure 39: Photo mosaic of ‘A’ frame mast

The net appears to be hungover a wire / rope which isthrough this block

Port side ‘A’ Frame stay, broken at lower end. Currently through ‘A’ frame, net tangled /wrapped around broken end.

Rope/Cable through top block on ‘A’ Frame ends here, broken end.

216

Figure 40: Partition bulkhead burst forward

Note: 1) This bulkhead was the forward bulkhead of the No. 2 four crew man cabin.

Upper partof bulkhead

Lower partof bulkhead

217

Figure 41: Desk drawers in mate’s cabin

Note: 1) Desk is facing to port, drawers have fallen out to starboard.

218

Line ofthinner wire

Line

of

thic

k ca

ble

240˚

295˚

Figure 42: Seabed cables in vicinity of wreck

Photograph of wires crossingApproximate position

of Gaul (length 66m)

245

219

Figure 43: Bow damage – view from ahead

Line of stem

Anchorbox

Fishingnet drapedover bow

Platingdished-in

Anchorbox

Platingdished-in

220

Figure 44: Section through net arena looking forward

BOBBINS

UPRIGHT

HEELED < 90º

HEELED > 90º

COAMING TO NET ARENA

BOBBINS

TRAWL DK

NETNET

TRAW

L DK

BOBBINS UNMOVED

TRAWL DK

TRAWL DK

NET

221

Figure 45: Gaul – Factors tree for the open fish loading hatches

Fish loading hatchesopen and unsecured

Opened as GAUL sankto the seabed

Opened without the crew’sinvolvement before the

GAUL sank

Opened by the crew

Pressure of trapped air in thefactory lifted hatches

The hatches fell openStoring a damaged

net in fish chuteEmptying a cod-end

in fish chute

GAUL rolledbeyond 90°

GAUL trimmedpast the vertical

Hatches began to fallopen when GAUL rolledpast 90° in knockdown

Hatches “jumped” opendue to GAUL ’s pitching

motion in wavesHatches fell open as shesank steeply by the stern

or or or

or oror or

or or

No No NoYes

No No No

COMMENT

(1) Calculationsshowed that there wouldbe insufficient pressure.

(2) If GAUL sank bythe stern and factoryflooded completely, therewould be no trapped air.

COMMENT

(1) Research hasshown that sinkingvessels return closer tothe upright and do notroll over.

COMMENT

(1) Positions of floormats in bridge indicatethis did not occur.

COMMENT

(1) Consistent withevidence and results ofmodel experiments.

COMMENT

(1) There is noevidence that this had everoccurred on GAUL

(2) It could only occurwith GAUL dodging head tosea in which case it wouldnot have led to dangerousflooding.

COMMENT

The fish chute wasfound to be empty.

COMMENT

(1) All theevidenceindicates thatGAUL was notfishing.

222

Figure 46: Gaul – accident factors tree

GAUL listing 35° tostarboard landed on

the seabed

GAUL lies on the seabedlisted 35° to starboard in

280m of water

Sides of fore peak waterballast tank crushed by

water pressure

Fish loading hatchesopened fully

Sides of forward oil fueldeep tank crushed by

water pressure

After GAUL left the surfaceshe levelled off and took up

a list to starboard

Partition bulkheadsburst forward

Water flooded throughsuperstructure from aft

to forward

GAUL sank steeplyby the stern

Water flooded her interiorthrough open doors and

hatches on the trawl deck

Doors and hatches immersed

GAUL listed at least40° to starboard

Cargo and loosegear shifted to

starboard

Trawl deckswamped with

water

GAUL heavilytrimmed by

stern

Factory flooded

Forward fueltank empty

Fuel usedfrom forward

tank first

Factory door open

Fish loading hatchesfell fully open

Tanks had not filledwhen GAUL reached

crushing depth

Tanks were filling throughtheir air pipes as GAUL

sank rapidly

Forward tanks emptybefore GAUL sank

Air pipes kept abovewater until GAUL sank

below surface

Fuel used fromforward tank first

Forward water ballasttank empty to give

stern trim for towing

Fish loading hatchespartially open

Hatches started tofall open

Non-return valve inhydraulics prevented

hatches closingEngine roomescape open

After GAUL had made apartial recovery from the

knock down, someone triedto escape from the engine

room

GAUL Rolledbeyond 90°

Hatches not securedagainst opening

Machinery stopped,engine room flooded

through ventilator

GAUL was“knocked down”

Impact dented the outboardface of the port funnel

KEY

As seen on GAUL’s wreck

Taken from evidence to the FI in 1974

Observed in MAIB’s model experiments

Taken from the vessel’s constructiondrawings

HOW TO USE THE “TREE”

(1) The tree represents the linkagebetween the many factors which theMAIB believes contributed to GAULsinking.

(2) For any particular factor, thosefeeding into it (i.e. those physicallylower in the tree) should give theimmediate reasons for its status.

(3) For example: the fish loadinghatches were found on the wreck to befully open, why was this? It wasbecause they had fallen fully open.Why had they fallen fully open? Therewere two necessary conditions for thatto have occurred: First, they werealready partially open and second,GAUL sank steeply by the stern. Whywere the fish hatches partially open?Again for two reasons:..........

..........and so on.

Hit by a group of verylarge breaking waves

GAUL beam on to sea

Some combination ofwind and waves push

bow around to starboard

Wind and waves 4 to 5points on starboard bow

GAUL making 170°in the direction of

Honningsvaag

Winds from ESEforce 8 - 10

GAUL heading forMalangan Bank whereother trawlers are still

fishing

GAUL heading forHonningsvaag forrepair or supplies

Rudder at 10° to 15°to port

or or

GAUL heading towardsHonningsvaag to gain

shelter of land

or