International Journal

on Marine Navigation

and Safety of Sea Transportation

Volume 3

Number 3

September 2009

275

1 REGULATION AND BEST PRACTICE IN

MARINE SAFETY INVESTIGATION

1.1 Casualty Investigation Code

The global harmonisation of marine casualty

investigation was taken a step further last year with

the approval by the International Maritime Organi-

zation of the new Code of the International Stand-

ards and Recommended Practices for a Safety Inves-

tigation into a Marine Casualty or Marine Incident

(Casualty Investigation Code). The Maritime Safety

Committee adopted the Casualty Investigation Code

by resolution MSC.255(84). And a new regulation 6

in chapter XI-1 of the SOLAS Convention was also

adopted (resolution MSC.257(84)), giving mandato-

ry status to the Code, which takes effect on 1 Janu-

ary 2010. However, the IMO has invited Govern-

ments to start implementing the new Code on a

voluntary basis prior to the effective date of the

Code [1].

1.2 Common approach of 1997

This most recent Code incorporates and builds on

the best practices in marine casualty investigation

that were established by the IMO’s Code for the In-

vestigation of Marine Casualties and Incidents,

adopted in November 1997. That Code sought to

promote co-operation and a common approach to

marine casualty and marine incident investigations

between States. While the new Code specifies some

mandatory requirements, it does recognize the varia-

tions in national laws in relation to the investigation

of marine casualties and incidents. But the broad aim

is to facilitate and promote objective marine safety

investigations for the benefit of flag States, coastal

States, and the shipping industry in general.

1.3 Objectives and purpose

The objectives and purpose are well stated in the

Code’s opening chapter:

“1.1 The objective of this Code is to provide a

common approach for States to adopt in the con-

duct of marine safety investigations into marine

casualties and marine incidents. Marine safety in-

vestigations do not seek to apportion blame or de-

termine liability. Instead a marine safety investi-

gation, as defined in this Code, is an investigation

Reconstructing a Marine Casualty: The

Effectiveness of the Full-Mission Simulator as a

Casualty Analysis Tool

E. Doyle

Cork, Ireland

ABSTRACT: The primary purpose of a marine casualty investigation is to seek to establish the causal factors

of the casualty with a view to learning the hard lessons and avoiding a repetition.

The broad questions of an investigation: "who?, what?, when?, where?, why?, and how?", all help to uncover

the facts. The investigation sequence will cover a diverse range of fact-finding activities, amongst which, as

often the case, may be a requirement for “conducting specialised studies”. Following the fact-finding stage

the typical investigation progresses to analysis of the facts, reaches conclusions and makes recommendations.

Keeping an open mind, to avoid premature conclusions, requires the separation of the fact-finding and

analysis phases. But the analysis may well help to identify missing pieces of evidence, or different lines of

enquiry that may otherwise have gone undetected. As an effective reconstruction tool, a full-mission bridge

simulator offers an opportunity to examine a broad spectrum of environmental condition

s and vessel

characteristics, as well as equipment failures, human factors and operating procedures. A casualty incident

can be reconstructed in a real-time simulated environment, to aid more detailed analysis.

Within the usual confines of the legal process, comprehension of nautical ‘black magic’ is greatly simplified

for non-mariners, by seeing the simulated casualty incident unfold, in real-time or in selected short-time seg-

ments.

276

conducted with the objective of preventing ma-

rine casualties and marine incidents in the future.

The Code envisages that this aim will be achieved

through States:

.1 applying consistent methodology and ap-

proach, to enable and encourage a broad rang-

ing investigation, where necessary, in the in-

terests of uncovering the causal factors and

other safety risks; and

.2 providing reports to the Organization to

enable a wide dissemination of information to

assist the international marine industry to ad-

dress safety issues.”

1.4 Causal factors rather than blame or fault

As we can see, the primary purpose of a casualty

investigation is to seek to establish the causal factors

of the casualty with a view to learning the hard

lessons and avoiding a repetition. And while it is

not, and never should be, the role of a marine safety

investigation team to attribute blame or fault, that is

not to say the investigating authority should refrain

from fully reporting the causes because fault or lia-

bility might be inferred from its findings.

2 THE FACT-FINDING/ANALYSIS

CONUNDRUM

2.1 Uncovering the facts

The investigation must attempt to uncover all the

facts, by seeking answers to such fundamental

questions as: "who?, what?, when?, where?, why?,

and how?" In this regard, the fact-finding sequence

of the investigation is likely to include such

activities as:

− inspecting the location;

− gathering or recording physical evidence;

− interviewing witnesses;

− reviewing of documents, procedure and records;

− conducting specialised studies (as required);

− identifying conflicts in evidence;

− identifying missing information; and

− recording additional factors and possible

underlying causes.

2.2 Progression to analysis

Following the fact-finding stage a typical marine

casualty or incident investigation includes: analyses

of the facts; conclusions; and recommendations.

2.3 Fact-finding and analysis

Investigators need to keep an open mind and avoid

reaching conclusions too early. It may appear self-

evident that the fact-finding stage of the process

should be separate from the analysis. But it must

always be borne in mind that the analysis may well

help to identify missing pieces of evidence, or

different lines of enquiry that may otherwise have

gone undetected.

2.4 Simulator as effective reconstruction tool

In the course of very many marine safety

investigations, the availability of a full-mission

bridge simulator is likely to offer a powerful and

productive analytical tool. Such a tool affords the

opportunity to examine a broad spectrum of

environmental conditions and vessel characteristics,

as well as equipment failures, human factors and

operating procedures. A marine casualty may be

reconstructed in a real-time simulated environment,

to allow detailed analysis of the incident. Mariners

who have had the benefit of full-mission simulator

training will readily appreciate the merits of the

debriefing/playback feature, allowing detailed

examination of the exercise or simulated incident, as

the replay unfolds in real-time or short-time

segments.

2.5 Investigation and legal proceedings enhanced

Such simulation can be replayed at will, with very

obvious benefits for expediting the work of the ma-

rine safety investigation team. In another forum,

such as the civil judicial process, it has the added

benefit for non-mariners of aiding the comprehen-

sion of nautical terminology with the consequent po-

tential to expedite settlement.

3 COLLISION CASE STUDY

3.1 Investigation and litigation

A practical example of the potential beneficial anal-

ysis that a simulated examination might generate is

given from the following marine casualty case study.

It centers on a collision off the southeast coast of

Ireland, in June 2000. The collision was investigated

by the newly established Marine Casualty Investiga-

tion Board (MCIB) [2], who did not have access to

adequate simulation facilities at that time. The case

also generated High Court proceedings which, in the

event, were settled shortly before the scheduled

hearing.

3.2 Summary of the incident

On the morning of 13 June 2000, the beam trawler

mfv STELIMAR (LOA 19m, 200t) was on passage

from her home port of Dunmore East, heading to-

wards her usual fishing grounds. She was steering a

277

course of about 145˚, and making about 8.5 knots.

The weather was fair: Wind SW’ly F 3/4, with good

visibility.

At the same time, the tanker mv ALMANAMA

(LOA 249m, 97,000 dwt) was making a course of

256˚, speed 13.8 knots, bound for Cork Harbour.

The vessel had cleared the traffic separation scheme

at Tuskar Rock and was now on a course that would

take her across the path of STELIMAR. In fact, the

two vessels were on converging courses, in circum-

stances where the bearing between them was not

changing significantly — a collision seemed inevi-

table unless avoiding action was taken by one or

both vessels.

This was a classic "crossing situation" for which

there is clear provision in the COLREGS. Rule 15

obliged ALMANAMA, as the give-way vessel, to

keep out of the way and thus avoid collision, while

Rule 17 required STELIMAR, the stand-on vessel,

to maintain her course and speed — in the early

stage of the encounter, at any rate.

In the event, a collision did occur, at a position

about 14 miles SSE of Hook Head. STELIMAR sus-

tained substantial damage, which necessitated her

being towed back to Dunmore East. Given the

enormous disparity in the size and tonnage of the

two vessels it was nothing less than incredible good

fortune that STELIMAR did not capsize and found-

er.

In addition to her Rule 15 obligation, ALMA-

NAMA was also required by Rule 16, to "…take

early and substantial action to keep well clear." In

discharging her obligations, ALMANAMA could

have made a large alteration of course to starboard

so as to make her intentions very clear to the stand-

on vessel, or she could have made a substantial re-

duction to her speed but this action would not have

been so readily apparent to the stand-on vessel. Ac-

cepting that speed reductions are rarely used by

give-way vessels when taking avoiding action in

open sea situations, ALMANAMA could reasonably

have been expected to make a substantial alteration

of course to starboard. Further, she should have

done so at an early stage in the encounter so as to

avoid putting STELIMAR in the unnecessary and

difficult position of having to take avoiding action

under Rule 17(a)(ii).

3.3 ‘Factual’ conflict

The MCIB investigation report noted the “Factual

Report of the Collision…” from STELIMAR’s per-

spective, and a similar “Factual Report…” from

ALMANAMA. There should be no surprise that the-

se ‘factual reports’ were in conflict. The real surprise

was that the MCIB analysis failed to resolve the con-

flict adequately.

3.4 STELIMAR’s perspective

STELIMAR'S skipper first noticed a large merchant

vessel visually, broad on his port bow at a distance

of 6 or 7 miles, on a general W'ly heading, shaping

to cross his path — she would need closer attention

as the range closed.

When the radar image of this large ship, soon to

be identified as ALMANAMA, first appeared at the

extremity of his 3-mile radar display, the skipper be-

gan to pay continuous attention to her progress. He

believed that she was making 14 or 15 knots, and his

concern was heightened by the developing situation,

as presented in Fig. 1: he was in a crossing situation

with a large vessel, whose bearing appeared to re-

main the same or nearly so.

Figure 1. A reconstruction of the crossing encounter

3.5 Imminent collision

When the vessels were about 1.5 miles apart, and

ALMANAMA had still not altered course,

STELIMAR came to the conclusion that he would

have to take avoiding action.

He could have altered course to starboard but the

skipper felt this would have prolonged the period of

uncertainty. In the event, he chose to de-clutch the

main engine and allow STELIMAR'S speed to

quickly run down. He estimated he did this when

the vessels were about 0.75 to 0.5 miles apart — or

about 2 to 2.5 minutes before impact.

In taking the speed off his vessel the skipper an-

ticipated that ALMANAMA would pass safely

ahead of him. However, very shortly afterwards

(perhaps when 0.5 to 0.25 miles apart) he was

alarmed to see a man on ALMANAMA'S starboard

bridge wing running into the bridge in an agitated

state. He was deeply concerned at this and, believing

that he now had a full emergency on hand, he put his

engine to "Full astern". He estimates the two vessels

were about 0.25 miles apart at this point and that he

was at "Full astern" for 30 to 40 seconds until the

collision. He believed that ALMANAMA was turn-

278

ing slowly to starboard, towards STELIMAR as she

gained stern-way.

3.6 ALMANAMA’s perspective

Meanwhile, the “Factual Report…” from the other

vessel has the OOW on the bridge of ALMANAMA,

plotting a fix for 1115 and altering course to 256°.

The vessel’s speed was about 13.8 knots.

At 1120 he observed a small target (STELIMAR)

some 40˚ to 50˚ on the starboard bow at a distance of

5 or 6 miles. He claims he acquired and plotted this

target on the ARPA, which predicted a CPA of 1 to

1.5 miles with the target crossing ahead. He also

took a series of visual bearings, which indicated that

the vessel was passing ahead, but did acknowledge

that the bearings were changing very slowly. He es-

timated that the fishing vessel was heading on a

course of about 150° at about 10 or 11 Knots.

When the fishing vessel was between 2.5 and 3.5

miles off and about 1.5 to 2.5 points on the starboard

bow, the OOW tried to call it on VHF Channel 16,

but there was no reply from STELIMAR.

3.7 Belated course alteration

He now altered course to starboard, to 268°, though

the fishing vessel was still fine to starboard and

about 1 mile off. He claimed that STELIMAR al-

tered course to port to about 120° and possibly re-

duced speed also. ALMANAMA then applied hard-

to-port helm in a final, and ultimately vain, attempt

at trying to avoid collision.

4 CLOSE-QUARTERS ANALYSIS

4.1 Course recorder trace

ALMANAMA’s course recorder trace confirmed her

course alteration from 230˚ to 256˚at 1115. It also

showed that her next course alteration, to 268˚, was

made just about two minutes prior to the collision,

and that the hard-to-port manoeuvre had practically

no effect before impact.

4.2 Hard lesson on failure to “take early and

substantial action”

Deconstruction of the final phase of this collision

encounter was clear to all; STELIMAR took emer-

gency "full astern" action when it seemed clear to

her that collision could not otherwise be avoided, but

the action was unsuccessful because the beneficial

effect of her stern-way motion was nullified by the

very belated turn to starboard by ALMANAMA,

culminating in the collision.

4.3 Making relative velocity simple

The most glaring and unresolved conflict between

the two parties was ALMANAMA’s contention that

STELIMAR was expected to cross ahead at a CPA

of 1 to 1.5 miles, this information allegedly predict-

ed from ARPA. Such contention is readily refuted

by means of a standard relative velocity plot, though

not so readily understood by non-mariners. Howev-

er, the use of a bridge simulator easily overcomes

those difficulties.

4.4 Construction of RelVel triangle

The veracity of the relative velocity information is

dependant on the vector accuracy for each ship. In

the case of ALMANAMA, her course and speed

were established from log records and instrumenta-

tion, while STELIMAR’s course and speed were

consistent with her recent departure (about 2 hours)

from her home port. Reversing the vectors from the

collision point allows construction of the basic rela-

tive velocity triangle, as given in Fig. 2. Because of

uncertainty in the precise timing of each vessel’s

movements in the final moments of the encounter

the plot may contain an inherent error, but nothing

of any significant consequence. This was confirmed

by rerunning the incident as a test exercise on the

NMCI bridge simulators.

Figure 2. The Relative velocity triangle, from ALMANAMA’s

perspective

4.5 Critical relative bearings

Given the geometry of this encounter, as outlined at

Fig. 1, it will be seen that STELIMAR was bearing

280˚ from ALMANAMA, or 50˚ on her starboard

bow before she altered course at 1115. The conten-

tion that STELIMAR was observed at 1120, about

40˚/50˚ x 5/6 miles on ALMANAMA’s starboard

bow, conflicts with the Fig. 1 plot which shows that

the vessels were no more than 4.4 miles apart then.

It is certainly the case that STELIMAR could not

have been seen 50˚ on the bow at any time after

ALMANAMA altered course at 1115 — if it were

so, a collision could not possibly have happened.

The only rational conclusion is that STELIMAR was

279

seen broad (50˚) on the starboard bow before AL-

MANAMA altered course to 256˚.

5 ARPA SIMULATION

5.1 ARPA vectors — ‘true’ or ‘relative’?

The change in the relative bearing of STELIMAR

(from 50˚ on the bow, drawing left to 24˚) may well

have misled the OOW on ALMANAMA into believ-

ing that STELIMAR was crossing clear ahead of his

own vessel. It is also possible that he confused the

“true” and “relative” vector information presented

by his ARPA radar. In any event, he chose to disre-

gard (until too late) the warning of his own eyes

when he observed that the compass bearing of

STELIMAR was changing only very slowly.

5.2 ARPA information

The simulated ARPA display in Fig. 3 presents an

early stage of the encounter from the ALMANAMA

perspective; her course is 256˚, and STELIMAR (the

acquired target) is bearing 281˚ (25˚on the starboard

bow), range 5.1NM. The target’s ‘true’ vector is

clearly visible, indicating a crossing condition.

Figure 3. The ARPA simulation displays a ‘true’ vector from

the acquired target on ALMANAMA’s starboard bow

Figure 4. The target’s ‘relative’ vector signals a developing

collision condition

A short time later, as presented in Fig.4, the navi-

gational situation remains the same but the ARPA

vector presentation is now ‘relative’. On any nor-

mally functioning and operated ARPA equipment,

this developing close-quarters situation will trigger

all the usual audible and visible alarms.

6 CONCLUSIONS

As demonstrated in this case study, full-mission

bridge simulation lends itself easily and readily to

collision analysis. The incident was reconstructed in

a real-time simulated environment, aiding the more

detailed analysis than that offered in the MCIB re-

port. The simulator reconstruction exposed possible

equipment failure, human factors and shortcomings

in operating procedures. These are weaknesses that

frequently flag missing evidence, which in turn,

prompt investigators to pursue different lines of en-

quiry. The complexity of nautical technology is

greatly simplified by a simulated reconstruction,

which has clear benefits for all parties within the

strictures of the legal process.

REFERENCES

[1] IMO, 2008. The Maritime Safety Committee, at its eighty-

fourth session (7 to 16 May 2008), adopted the Casualty In-

vestigation Code by resolution MSC.255(84) and a new

regulation 6 in chapter XI-1 of the SOLAS Convention by

resolution MSC.257(84) to make the Code mandatory. The

Committee agreed that the Casualty Investigation Code

should take effect on 1 January 2010, noting that the effec-

tive date should be the same as the date of entry into force

of the new SOLAS regulation XI-1/6.

[2] MCIB, 2005. Investigation Report by the Marine Casualty

Investigation Board into ALMANAMA/STELIMAR colli-

sion.