Marine Safety Investigation Unit

MARINE SAFETY INVESTIGATION REPORT

Safety investigation into the partial flooding of the

engine-room on the Maltese registered bulk carrier

CAPRI

whilst alongside at Dampier, Australia

on 26 December 2017

201712/034

MARINE SAFETY INVESTIGATION REPORT NO. 24/2018

FINAL

ii

Investigations into marine casualties are conducted under the provisions of the Merchant

Shipping (Accident and Incident Safety Investigation) Regulations, 2011 and therefore in

accordance with Regulation XI-I/6 of the International Convention for the Safety of Life at

Sea (SOLAS), and Directive 2009/18/EC of the European Parliament and of the Council of 23

April 2009, establishing the fundamental principles governing the investigation of accidents

in the maritime transport sector and amending Council Directive 1999/35/EC and Directive

2002/59/EC of the European Parliament and of the Council.

This safety investigation report is not written, in terms of content and style, with litigation in

mind and pursuant to Regulation 13(7) of the Merchant Shipping (Accident and Incident

Safety Investigation) Regulations, 2011, shall be inadmissible in any judicial proceedings

whose purpose or one of whose purposes is to attribute or apportion liability or blame, unless,

under prescribed conditions, a Court determines otherwise.

The objective of this safety investigation report is precautionary and seeks to avoid a repeat

occurrence through an understanding of the events of 26 December 2017. Its sole purpose is

confined to the promulgation of safety lessons and therefore may be misleading if used for

other purposes.

The findings of the safety investigation are not binding on any party and the conclusions

reached and recommendations made shall in no case create a presumption of liability

(criminal and/or civil) or blame. It should be therefore noted that the content of this safety

investigation report does not constitute legal advice in any way and should not be construed

as such.

© Copyright TM, 2019.

This document/publication (excluding the logos) may be re-used free of charge in any format

or medium for education purposes. It may be only re-used accurately and not in a misleading

context. The material must be acknowledged as TM copyright.

The document/publication shall be cited and properly referenced. Where the MSIU would

have identified any third party copyright, permission must be obtained from the copyright

holders concerned.

MARINE SAFETY INVESTIGATION UNIT

Maritime House

Lascaris Wharf

Valletta VLT 1921

Malta

iii

CONTENTS

LIST OF REFERENCES AND SOURCES OF INFORMATION .......................................... iv

GLOSSARY OF TERMS AND ABBREVIATIONS ................................................................v

SUMMARY ............................................................................................................................. vi

1 FACTUAL INFORMATION .............................................................................................1 1.1 Vessel, Voyage and Marine Casualty Particulars ..........................................................1 1.2 Description of Vessel .....................................................................................................2

1.2.1 MV Capri ...............................................................................................................2 1.2.2 Ballast system .........................................................................................................3

1.2.2.1 Ballasting procedures .....................................................................................5 1.3 Crew ..............................................................................................................................5 1.4 Environment ..................................................................................................................5 1.5 Narrative ........................................................................................................................6

1.5.1 Events leading up to the accident ...........................................................................6 1.5.2 Post accident events ..............................................................................................10

1.6 Damages to the Engine-room ......................................................................................11

2 ANALYSIS .......................................................................................................................14 2.1 Purpose ........................................................................................................................14 2.2 Fatigue and Alcohol .....................................................................................................14 2.3 Immediate Cause of the Flooding ................................................................................14 2.4 Water Hammer Effect ..................................................................................................16 2.5 Caution and Procedures to Eliminate Water Hammer Effects ....................................16 2.6 Missed Procedural Step ...............................................................................................18 2.7 Mode Errors .................................................................................................................19

3 CONCLUSIONS ...............................................................................................................22 3.1 Immediate Safety Factor ..............................................................................................22 3.2 Latent Conditions and other Safety Factors .................................................................22 3.3 Other Findings .............................................................................................................23

4 ACTIONS TAKEN ...........................................................................................................23 4.1 Safety Actions Taken During the Course of the Safety Investigation .........................23

5 RECOMMENDATIONS ..................................................................................................24

ANNEXES ...............................................................................................................................25

iv

LIST OF REFERENCES AND SOURCES OF INFORMATION

Gergant, A., Simpson, A. R., & Sijamhodžić, E. (1991). Water hammer analysis of

pumping systems for control of water in underground mines. Paper presented at the 4th

International Mine Water Congress.

Crew members – MV Capri

TMS Dry Ltd.

v

GLOSSARY OF TERMS AND ABBREVIATIONS

AMSA Australian Maritime Safety Agency

bar Unit of pressure

BV Bureau Veritas

cSt Centistokes

ECR Engine control room

DG Diesel generator

GT Gross tonnage

IMO International Maritime Organization

ISM The International Management Code for the Safe Operation of Ships

and Pollution

kgcm-2

Kilogram per centimetre squared

kW Kilowatt

m Metre(s)

mm millimetres

NM Nautical mile(s)

SMS Safety Management System

STCW The International Convention on Standards of Training, Certification

and Watchkeeping for Seafarers, 1978 as amended

UTC Universal Co-ordinated Time

VHF Very High Frequency

vi

SUMMARY

Capri, a Maltese registered capsize bulk carrier arrived at Dampier anchorage,

Australia on 22 December 2017. The vessel was scheduled to load 164,000 tonnes of

iron ore for China.

On 24 December, while at anchor, the vessel deballasted ballast tanks nos. 1 and 4

(port and starboard) in order to reduce the number of ballast tanks which she would

have to deballast once alongside when loading operations commence. The vessel then

berthed in the evening of 25 December and since loading had not started, the chief

mate opted to strip the ballast tanks that had been emptied during the day before. On

completion at 2330, he advised the engine-room to line up the ballast system so that

deballasting could take place when loading started.

Loading started at 0229 on 26 December and the chief mate instructed the engine-

room to prepare for deballasting operations; this required the starting of an additional

generator. The chief mate then opened the remote suction valves to ballast tanks

nos. 2 port and starboard. Soon afterwards, a loud bang was heard and the vessel

blacked out.

Power was eventually restored and the crew discovered a spray of water coming from

the ballast pump suction strainer cover that was reaching the electrical distribution

panel located close to the strainer. Water was steadily rising in the engine-room

bilges. Despite isolating the ballast system, about 1,100 tonnes of sea water flooded

into the engine-room.

The ballast pump strainer cover and a ballast valve appeared to have been damaged

when valves to ballast tanks port & starboard nos. 2 were opened. A number of

electrical motors, switchboards, cabling, as well as the disabling of the main engine

was directly attributed to the ingress of seawater into the engine-room. The vessel

was subsequently towed to Singapore where the vessel’s equipment was reinstated.

The safety investigation concluded that the ballast system valves had been incorrectly

set up and the procedures when initiating deballasting were not followed. This led to

the ballast lines remaining empty and a hammer effect was created by the head of

water.

1

1 FACTUAL INFORMATION

1.1 Vessel, Voyage and Marine Casualty Particulars

Name Capri

Flag Malta

Classification Society Bureau Veritas

IMO Number 9248526

Type Bulk Carrier

Registered Owner Aurelia Owning Co. Ltd.

Managers TMS Dry Ltd.

Construction Steel (Double bottom)

Length overall 289.00 m

Registered Length 280.20 m

Gross Tonnage 87,390

Minimum Safe Manning 15

Authorised Cargo Cargo in bulk

Port of Departure Rio Tinto, Australia

Port of Arrival Dampier, Australia

Type of Voyage Short International

Cargo Information Iron Ore (164,400 mt)

Manning 22

Date and Time 26 December 2017 at 0232 (LT)

Type of Marine Casualty Serious Marine Casualty

Place on Board Engine-room

Injuries/Fatalities None reported

Damage/Environmental Impact Vessel was rendered unfit to proceed due to water

damage to electrical system and physical damage

to the ballast system

Ship Operation Alongside / moored – Deballasting

Voyage Segment Arrival

External & Internal Environment Daylight, Good visibility, wind West Northwest

19 knots and moderate sea. Air temperature was

29 ⁰C and sea temperature was 28 °C.

Persons on Board 22

2

1.2 Description of Vessel

1.2.1 MV Capri

Capri (Figure 1) was built by Nippon Kokan KK (NKK Corp) - Tsu ME, Japan in

2001. She is a capesize bulk carrier and classed by Bureau Veritas for unrestricted

navigation.

Figure 1: MV Capri

Capri has nine cargo holds with holds nos. 1, 3, 5, 7 and 9 strengthened for alternate

hold loading. Her gross tonnage (GT) is 87,390 and net tonnage is 57,416. The

vessel has a length overall of 289.0 m and a beam of 45.0 m. Her depth is 24.10 m

and the maximum deadweight is 172,579 tonnes at a summer draught of 17.81 m.

Capri’s propulsive power is provided by a two-stroke, single acting, 6-cylinder Mitsui

MAN-B&W medium speed engine, producing 14,711 kW at 80 RPM. The vessel has

one fixed pitch propeller and her service speed is 15 knots.

The vessel is registered in Malta and owned by Aurelia Owning Co Ltd. It is

managed and operated by TMS Dry Ltd. (Company) which is based in Athens,

Greece. The vessel is traded on the bulk market on a worldwide basis usually either

between Brazil and the Far East or between Australia and the Far East.

3

1.2.2 Ballast system

The ballast water system comprised of two main ballast pumps capable of operating

either independently or in parallel. The normal operating pressure of each pumps was

3.0 kgcm-2

and the maximum operating pressure was 4.5 kgcm-2

. A suction strainer

was fitted between the sea chests and the ballast pump suction valves.

The ballast pumps could be started either locally or remotely from the ballast control

mimic board in the ship’s office (Figure 2). The ballast pump discharge valves were

also operated from the ballast control mimic board.

The ballast water system was also connected to the fire, general service and the cargo

hold bilge systems.

Figure 2: Ballast Control Mimic Board

The ballast pumps were primarily used for ballasting / de-ballasting the vessel’s

ballast tanks and cargo hold no. 6, which was the ballast hold, as well as to drive the

ballast eductor for stripping the ballast tanks. The fire and general service pumps

were used for pumping out the cargo hold bilges and driving the cargo hold bilge

eductor when necessary. A schematic drawing of the ballast piping is provided below

at Figure 3.

4

Figure 3: Bilge and ballast system

Most of the valves operating the ballast system were remotely operated from the

ballast control mimic board located in the cargo control room. However, a number of

valves had to be locally operated from inside the engine-room.

5

1.2.2.1 Ballasting procedures

Ballasting and deballasting instructions were posted on the ballast control mimic

board for the attention of deck officers. The instructions detailed the sequence of

operation of valves and pumps as can be found in Annex A.

Ballasting and deballasting procedures (Annex B) with a simplified ballast piping

diagram were similarly posted in the engine-room, in the vicinity of the ballast pumps,

for the attention of engine-room personnel.

1.3 Crew

The Minimum Safe Manning Certificate issued by the flag State Administration, stipulated a

minimum crew of 15. The vessel was manned by 20 crew members. All crew members were

from the Philippines, bar for the master, who was from Ukraine and the Romanian electrician.

The engine-room was manned on a 24 hour basis. The engine-room officer of the watch was

normally assisted by an oiler. The watches kept on board were the conventional 4-on, 8-off.

The routine day-to-day planned maintenance was supervised by the second engineer,

supported by a third engineer, the fourth engineer, the electrician and three oilers.

At the time of the flooding, the following crew members were present in the engine room:

chief engineer;

third engineer (watchkeeper);

electrician; and the

oiler.

The ship’s office was attended by the chief mate.

1.4 Environment

The accident happened during daylight hours. Visibility was good, with a West

Northwesterly fresh breeze and moderate sea. The air temperature was 29 °C and the

sea temperature was 28 °C.

6

1.5 Narrative

1.5.1 Events leading up to the accident1

Capri arrived at Dampier anchorage, Australia, in ballast at 2200 on 22 December

2017. She was scheduled to load a cargo of iron ore of about 164,000 tonnes, bound

for China.

At 0945 on 24 December, the vessel started deballasting ballast tanks nos. 1 and 4

port and starboard in order to reduce the number of tanks she would have to deballast

when loading would start, once she arrived alongside. This was completed at about

1800.

At 1936, on 25 December, the vessel came alongside berth no. 5 at Parker Point. Her

drafts were noted to be forward 7.10 m and aft 8.90 m. Soon afterwards, at 2000,

ballast tanks nos. 1 and 4 port and starboard were stripped using ballast pump no. 2

and the ballast eductor (Figure 4).

At 2044, the auxiliary boiler alarm (emergency shut off) activated, which was

acknowledged in the engine control room.

At about 2300, the chief mate called the engine-room which, at the time, was manned

by the fourth engineer and oiler to inform them that the stripping operation of ballast

tanks nos. 1 and 4 port and starboard could be stopped because the ballast tanks were

empty. He then instructed them to prepare the valves for deballasting ballast tanks

nos. 2 port and starboard. The watchkeeping oiler then operated the ballast valves at

about 2310 and left them as indicated in Figure 5.

1 Unless otherwise stated, all times are local time.

7

Figure 4: Ballast valves status as at 2000

Source: TMS Dry Ltd.

8

Figure 5: Ballast valves status as at 2310

Source: TMS Dry Ltd.

9

At about 2330, generator no. 1 was stopped as no cargo or deballasting operations

were anticipated. At 0000 (26 December), the engine-room watch was taken over by

the third engineer and duty oiler.

By 0130 on 26 December, the chief engineer, third engineer, and electrician were all

busy working on the auxiliary boiler that had previously alarmed and not operating in

the automatic mode.

At 0229, the chief mate was informed that cargo loading operations had commenced

and instructed the engine-room personnel to prepare for the de-ballasting of ballast

tanks nos. 2 port and starboard in accordance with the vessel’s loading plan. The

chief mate then also opened the remotely operated tank valves for ballast tanks nos. 2

port and starboard from the ballast control mimic board.

At about 0231, generator no. 1 was started to be put on line and synchronised with

generator no. 2 to provide power for operation of the ballast pumps. Soon afterwards,

at around 0232, a loud bang was heard, and a blackout occurred. The emergency

generator automatically started supplying power to the emergency circuits including

emergency lighting2.

Electrical power was initially restored using generator no. 2 and an inspection of the

engine-room was carried out. The crew noticed a spray of water coming from the

ballast pump suction strainer cover that was reaching the electrical distribution panel

fitted close to the strainer. Moreover, the water level was steadily rising in the

engine-room bilges.

Numerous breakers were tripped in the engine control room to isolate the affected

distribution panel and this resulted in a loss of cooling water to the generators since

the sea water cooling pumps were supplied from this panel. In addition, ballast valves

‘BW508V’, ‘BW502V’ and ‘HV509V’ were closed to isolate the strainer and stop the

ingress of water.

At 0250, emergency power was again supplied from the emergency generator.

2 At the time of the blackout, the chief mate was reportedly opening valve ‘BW506V’ (ballast pump

no. 2 discharge valve) from the ballast control mimic board.

10

A draft survey carried out after the flooding indicated that about 1,123 tonnes of sea

water had flooded the engine-room.

1.5.2 Post accident events

At 0310, the master completed the ‘Emergency check list for flooding’. An

assessment indicated that the water level in the engine room had reached a level of

approximately 2.3 m above the tank top, and numerous electric motors were

submerged with switchboard and cabling affected by seawater impingement, as well

as seawater ingress in the main engine sump tank.

At 0320, the master informed the Company of the situation and then stopped cargo

operations at 0330.

Power to the vessel was restored through generator no. 3. The cooling water system

was connected to the vessel’s emergency fire pump using flexible hoses. The crew

then started transferring the water from the engine-room bilges to the aft peak tank,

using portable pumps.

At about 0625, the harbour master boarded the vessel to assesses the situation and at

0925, two surveyors from the Australian Maritime Safety Agency (AMSA) attended

the vessel to carry out an investigation. By 1200, AMSA had issued a detention

notice to the vessel and requested that the transferring of water to the aft peak tank is

suspended.

Capri remained alongside but on 27 December at 1030, AMSA granted permission to

the crew to re-commence the transferring of water from the engine-room bilges to the

aft peak tank.

Prior to the vessel being given permission to be towed out to Dampier anchorage, both

AMSA and Bureau Veritas (BV) required the following works to be carried out:

a) all water to be transferred from the engine-room bilges to the aft peak tank;

b) engine-room bilges to be cleaned of oil;

c) ballast pump suction strainer to be repaired and re-fitted; and

11

d) operation of the three pumps to supply cooling water to generators, air

conditioning and domestic refrigeration plants to be restored and that one pump

was also made readily available to supply water for the fire mains.

On 28 December, an engine manufacturer’s representative attended the vessel and

confirmed that the main engine could not be operated. The same day, Dampier

Harbour Master informed the vessel that permission to remain at anchorage would be

withdrawn as there was the potential of a cyclone to form in the area. He also advised

that the vessel would require towage to a convenient repair facility.

As the water level in the engine-room bilges was gradually reduced, it was found that

ballast valve ‘BW507V’ was also damaged and leaking water. AMSA required that

the leakage was isolated prior to the vessel’s towage operation.

On 29 December, AHTS Pacific Centurion was fixed by the managers to tow the

vessel from Dampier to Singapore. In the meantime, restoration of the cooling sea

water pump capability continued and on 31 December, five pumps were available for

use with two of these also being capable of being used for fire fighting.

On 01 January 2018, permission was granted to tow the vessel to Dampier anchorage

and await AHTS Pacific Centurion’s arrival and subsequent onward towage to

Singapore for permanent repairs. On 02 January 2018, the vessel was shifted by local

tugs to Dampier anchorage where temporary repairs continued as per requirements of

BV and AMSA, in preparation for the towage operation to Singapore.

On 08 January 2018, the vessel tow to Singapore commenced. Capri arrived in

Singapore on 20 January after which, she proceeded to Keppel Shipyard for

permanent repairs.

1.6 Damages to the Engine-room

The ballast pump’s suction strainer cover (Figure 6) had an outside diameter of

1,500 mm. The cover was found deformed and detached from the casing due to the

failure of the retaining bolts, leaving a gap of approximately 20 mm around half of its

circumference.

12

Figure 6: Ballast pump suction strainer cover

The seawater ingress occurred along this gap (from between the detached and

deformed cover and the strainer housing). The safety investigation found that 9

retaining bolts had elongated and seven bolts had parted.

The body of valve ‘BW507V’ was cracked and required renewal (Figures 7a and 7b).

Figure 7a: Leaking valve ‘BW507V’

13

Figure 7b: Valve ‘BW507V’ as seen in the repair yard

A large number of pumps, including the ballast pumps, main sea water cooling and

main engine lubricating oil pumps and electric motors were submerged in sea water.

Electrical distribution panels were sprayed with sea water and cabling submerged.

The main engine crankcase and lubricating oil sump tank had also been subjected to

water ingress, rendering the main engine inoperable.

14

2 ANALYSIS

2.1 Purpose

The purpose of a marine safety investigation is to determine the circumstances and

safety factors of the accident as a basis for making recommendations, to prevent

further marine casualties or incidents from occurring in the future.

2.2 Fatigue and Alcohol

The master carried out an alcohol test on all crew members at 0800 on 26 December,

i.e., a few hours after the accident. The results were negative for all crew members.

Similarly, the crew members’ hours of rest records, as required by STCW and the

Maritime Labor Convention, were analysed. The safety investigation concluded that

neither fatigue nor alcohol contributed to the events leading to the accident.

2.3 Immediate Cause of the Flooding

Although the damaged strainer was designed to withstand a hydrostatic pressure of

2 kgcm-2

(i.e., about 20 m water head), it was apparent that for it to fail, it was

subjected to a force in excess of its designed load. The physical evidence indicated

that the flooding was caused soon after ballast tanks nos. 2 port and starboard valves

were opened.

It may be stated that since ballast tanks nos. 1 port and starboard had just been

educted, a section of the ballast pipe lines would have been empty (Figure 8). It was

hypothesised that the opening of the valves for ballast tanks nos. 2 port and starboard

led to the rapid flow of water and air down the ballast pipes, driven by the head of

water of the ballast tanks (nos. 2 port and starboard). As the water and air

encountered closed valves ‘BW507V’, ‘BW514V’, ‘HV508V’, ‘BW503V’ and

‘BW506V’, a hydraulic shock was initiated when the rapidly moving column of liquid

and air was forced to an abrupt stop. As the weakest part of the pipe system was the

bolted suction strainer cover, the hydraulic shock deformed the cover and elongated and broke

the retaining bolts. The body of valve ‘BW507V’ was also damaged and cracked, resulting in

the uncontrollable flooding of the engine-room.

15

Figure 8: Ballast line showing section with entrapped air

Source: TMS Dry Ltd.

Pipe section with

possible entrapped

air following

stripping with

eductor

16

2.4 Water Hammer Effect

Water hammer is also known as hydraulic transient and it happens when the velocity

of the flow of the liquid in a pipe changes. Water hammer is the transmission of

pressure waves along pipelines, which result from a change in flow velocity. One of

the possible outcomes of water hammer is dangerously high pressures created in the

pipelines. High pressures can lead to either pipe ruptures or damages to other parts of

the system. The safety investigation believes that water hammer was a direct cause of

the pipe failure.

From an engineering and design perspective, there are a number of factors which can

be adopted in order to mitigate water hammer problems. Water hammer control

devices come in different forms, such as those which generate inertia, controlled valve

closure, surge tanks, air chambers and pressure relief valves3. It did not transpire that

any of these water hammer control devices were fitted to the ballast system at the

design / construction phase of the ship.

2.5 Caution and Procedures to Eliminate Water Hammer Effects

Although at the design stage, the system was not inherently protected against water hammer

effects4, the vessel’s builder had identified the possibility of water hammer occurring in the

ballast pipe work and a caution notice had been posted on the ballast control mimic board

(Figure 9). In addition, the vessel had its own procedures and precautions on how to avoid

water hammer effect (Annex A).

Figure 9: Caution notice on water hammer

3 The type of water hammer control devices depends on various factors, not least the characteristics

of the system (e.g., length of pipeline, height of head and expected pressures).

4 In addition to the devices listed in Footnote 3, the redesigning of the pipeline layout, thicker

pipelines or higher strengths and alteration of operational parameters were possible measures to

mitigate water hammer effects.

17

It would appear that these procedures were not followed. After completion of the

stripping operation of ballast tanks nos. 1 port and starboard and nos. 4 port and

starboard at 2300, the chief mate instructed the engine-room personnel to stop the

pumps and line up the valves in preparation for the deballasting of ballast tanks nos. 2

port and starboard. In doing so, they would have followed the following sequence, as

indicated in Figure 10:

Figure 10: Ballast valves referred to in the deballasting procedure

Source: TMS Dry Ltd.

18

1. Open fully overboard discharge valve ‘HV509V’;

2. Open fully delivery valve ‘BW508V’ to overboard and valve ‘BW507V’

to ballast pipe to refill main pipelines by gravity for several minutes to

avoid pressure build-up impact / water hammer effect;

3. Open fully suction valve ‘BW502V’ on the suction side and close valve

‘BW507V’ at the engine-room side;

4. Open fully discharge valves ‘BW505V’ / ‘BW506V’, depending on the

ballast pump to be used;

5. Open fully suction valves ‘BW503V’ / ‘BW504V’, depending on the

ballast pump to be used; and

6. Opening fully the valves of the ballast tanks to be pumped out.

During the course of the safety investigation, it became evident that step 2 of the

procedures was not executed correctly at the time since valve ‘BW507V’ was not

opened to allow the main pipelines to refill by gravity. The closed valve ‘BW507V’

left the main pipelines full of air as they had not been refilled with water.

It was also established that only valves ‘HV509V’, ‘BW508V’ and ‘BW502V’ were

opened by the oiler and the closed valve ‘BW507V’ was not noticed by the duty

engineer as he was occupied trouble shooting the auxiliary boiler faults. At 0000 on

26 December, when the next watch took over, they were advised that the valves had

been lined up as per the instructions of the chief mate, and assumed that this was the

case. It would appear that the duty engineers of the 8-12 and the 12-04 watches did

not check the ballast line system status, very likely because they were occupied by the

fault that had developed on the auxiliary boiler.

2.6 Missed Procedural Step

It is also evident that the ballast procedures posted in the ship’s office were not

followed. Soon after informing the engine-room to prepare for deballasting, the

valves for no. 2 port and starboard were opened, without any cross confirmation of the

actual status of the other ballast valves.

19

The fact that the ballasting operation has surely been carried out on several occasions

by the crew members, is suggestive that they were familiar with the steps required and

therefore leaving one of the ballast valves closed and not filling part of the ballast

system with water to avoid water hammer was an omission which had no roots in the

implementation of the vessel’s safety management system.

The safety investigation was not aware whether the crew members operating the

valves were actually consulting a schematic drawing or a checklist. However, even

then, it would not have been impossible for a step (and one valve) to be accidently

missed. In a multi-step procedure, the vulnerability of missing a functionally

(isolated) task always exists. From a cognitive process point of view, it would appear

that there was an issue with the task planning (an overlooked item) and the monitoring

stage (the omission which was not detected).

Moreover, as for any symbolic barrier systems (the caution notice in Figure 9) and the

corporeal barrier system (the deballasting procedures as part of the vessel’s SMS),

these type of barrier systems are inherently weak because they do not carry an

interface signals mode which would trigger an alarm if not followed for whatever

reason.

It has to be highlighted that crew members had several challenges which they had to

negotiate on board during their day-to-day tasks. Moreover, the loading rate in

Dampier was very high and it was not excluded that this may have distracted the chief

mate during his duties to start the pumps without any delays.

2.7 Mode Errors

Errors of this type are classified as ‘mode errors’, i.e., errors which occur when an

operator (a crew member in this case) executes an action that would be appropriate if

a device were in one configuration rather than another. Very interestingly, these are

neither considered to be human erroneous actions nor equipment failure. Rather, they

are viewed as human –machine system issues. Without a computer-based device to

indicate the status of each and every valve, and with other on-going tasks on board

competing for crew members’ attention, the potential of missing on human-machine

interaction problems increased significantly.

20

Such erroneous actions may also be seen as a breakdown in the situation assessment,

i.e., when track of the system status is lost. As already discussed in sub-sections 2.5

and in this sub-section, it would appear that Capri’s accident was a typical mode

problem issue where:

1. the tolerance of the system for this error mode was limited (from the design

stage); and

2. built-in system mode status indications and feedback was limited.

21

THE FOLLOWING CONCLUSIONS AND SAFETY

ACTIONS SHALL IN NO CASE CREATE A

PRESUMPTION OF BLAME OR LIABILITY.

NEITHER ARE THEY LISTED IN ANY ORDER OF

PRIORITY.

22

3 CONCLUSIONS

Findings and safety factors are not listed in any order of priority.

3.1 Immediate Safety Factor

.1 The immediate cause of the flooding was the failure of a suction strainer and a

valve fitted on the ballast system;

.2 Although the damaged strainer was designed to withstand a hydrostatic

pressure of 2 kgcm-2

, it was apparent that for it to fail, it was subjected to a

force in excess of its designed load;

.3 A hydraulic shock was initiated when the rapidly moving column of liquid and

air was forced to an abrupt stop against a closed valve;

.4 Established procedures against water hammer were not followed;

.5 Step 2 of the procedures was not undertaken correctly at the time as valve

‘BW507V’ was not opened to allow the main pipelines to refill by gravity;

.6 The closed valve was not noticed by the duty engineer as he was occupied

trouble shooting the auxiliary boiler faults;

.7 In a multi-step procedure, the vulnerability of missing a functionally (isolated)

task always exists;

.8 Crew members have several challenges which they have to negotiate on board

during their day-to-day tasks.

3.2 Latent Conditions and other Safety Factors

.1 It did not transpire that any water hammer control devices were fitted to the

ballast system at the design / construction phase of the ship;

.2 Without a computer-based device to indicate the status of each and every

valve, and with other on-going tasks on board competing for crew members’

attention, the potential of missing on human-machine interaction problems

(mode error) increased significantly.

23

3.3 Other Findings

.1 Neither fatigue nor alcohol contributed to the events leading to the accident;

.2 The vessel’s builder had identified the possibility of water hammer occurring

in the ballast pipe work and a caution notice had been posted on the ballast

control mimic board;

4 ACTIONS TAKEN

4.1 Safety Actions Taken During the Course of the Safety Investigation

During the course of the safety investigation, the Company has taken the following actions:

1. initiated a safety campaign, highlighting the hazards of water hammer effect;

2. reminded crew members of the importance to follow the posted Makers' and Shipboard

ballasting/de-ballasting procedures, including the necessity of refilling the ballast lines

with sea water after stripping and before commencing de-ballasting operations;

3. ensured that duty engineers are reminded of the importance to provide information as to

ballast line system set-up during watch handover when ballasting, deballasting and

stripping operations are being performed;

4. revised the Company’s ballast handling procedure to include the risk and hazards of the

water hammer effect;

5. issued a ‘Safety Alert’ that include the lessons learnt to its managed vessels; and

6. installed an engineering modification to enhance monitoring of the ballast

operations by providing a real time indication of ballast piping valves’ open-

close status. The indication is provided by means of indicator panels installed at

the engine-room lower platform near the ballast pumps and in the ballast control

room. The indicator panel receives information on the valve status from the

limit switches fitted on the valves.

24

5 RECOMMENDATIONS

Taking into consideration the safety actions already taken by the Company during the course

of the safety investigation, the MSIU has no recommendations.

25

ANNEXES

Annex A Ballasting and De-ballasting Procedures

26

Annex B Ballasting and De-ballasting Procedures