447

1 INTRODUCTION

TheArcticmaybecomeintegralpartoftheshipping

industry on a global scale if current climate trends

continue. If that does happen it will involve a

transitional period, where many lessons will be

learned as the boundaries of normal shipping

operations are broadened. Experienced shipping in

the

 Arctic is limited, information is scarce, and not

widelyshared.Inordertobecomepreparedforsuch

an increase in shipping traffic in the Arctic (and

Antarctic), information we do have should be

examinedasthoroughlyaspossible.Thismayhelpus

betterunderstandtheconditionsandhowtooperate

inthem.

The present work uses the Functional Resonance

AnalysisMethod(FRAM)tobuildanunderstanding

ofArcticshipnavigationandusestheExxonValdez

grounding as a case to examine the model’s utility.

Thisworkisintendedtoinitiatediscussionacrossthe

maritime domain about FRAM and understanding

Arctic operations.

 We can use the FRAM to help

understand different elements of ship navigation,

including the so called “soft factors,” which are

difficult to assess with traditional techniques. This

willbecomeeven more important whenconsidering

Arcticshippingbecausetheinformationisbothvague

andscarce(ArcticCouncil,2009).TheFRAM

provides

a structured framework to consider anecdotal

experience from successful shipping operations,

which can help formalize lessons learned and share

themacrossthedomain.Byconsolidatinginformation

acrossthedomainitwillimproveourunderstanding

ofshippingsafety. Byimprovingour understanding

this way, we can then improve ship operations (the



way they function) and safety in the maritime

domain.

Using the FRAM to Understand Arctic Ship Navigation:

Assessing Work Processes During the Exxon Valdez

Grounding

D.Smith,B.Veitch,F.Khan&R.Taylor

emorialUniversityofNewfoundland,St.John’s,NL,Canada



ABSTRACT:Arcticshippinginvolvesacomplexcombinationofinter‐relatedfactorsthatneedtobemanaged

correctlyforoperationstosucceed.Inthispaper,theFunctionalResonanceAnalysisMethod(FRAM)isusedto

assessthecombinationofhuman,technical,andorganizationalfactors

thatconstituteashippingoperation.A

methodology is presented on how to apply the FRAM to a domain, with a focus on ship navigation. The

methoddrawsonshipnavigatorstoinformthebuildingofthemodelandtolearnaboutpracticalvariations

thatmustbemanagedtoeffectivelynavigatea

ship.TheExxonValdez caseisusedtoillustratethemodel’s

utilityandprovidesomecontexttotheinformationgatheredbythisinvestigation.Thefunctionalsignatureof

theworkprocessesoftheExxonValdezonthenightofthegroundingispresented.Thisshowsthefunctional

dynamicsofthatparticular

shipnavigationcase,andservestoillustratehowtheFRAMapproachcanprovide

anotherperspectiveonthesafetyofcomplexoperations.



http://www.transnav.eu

the International Journal

on Marine Navigation

and Safety of Sea Transportation

Volume 12

Number 3

September 2018

DOI:10.12716/1001.12.03.03

448

2 BACKGROUND

Ashippingoperationisasocio‐technicalsystemthat

requires many combinations of social and technical

factors to be managed to succeed. There has been a

movementtowardsadaptiveapproachestosafety to

help manage such systems (Borys et al., 2009). This

approachreliesonnotonlymodeling

theelementsin

the system, but the relationships in the system, eg.

how elements interact together (Vicente, 2004).

Becauseofthisshiftinthinking,othertechniques are

being adopted from resilience engineering to help

managecomplexsystemsaswell(Ayyub,2015,2014;

Hollnageletal.,2006).

Additionally,thereisacceptance

thatmanyofthe

conditionsthat operationsare beingsubjectedto are

sodynamicthatitisverydifficulttoprescribeasingle

safetyprotocolto managethem.The Society ofRisk

Analyst’srecentreviewstatesthatinthesecasesitis

bettertohaveadynamicsetofsolutionsto

adaptto

thesedynamicconditions(Avenetal.,2015).Safetyis

then approached by understanding how to best

monitorareasofthesystemandhowtocontrolthem:

inotherwords, by designing systemsthatadapt(or

maintain control) when subjected to dynamic

conditions.

Thereare anumber ofmethods

thatarefounded

on adaptive safety methodologies: the Functional

Resonance Analysis Method (FRAM), Systems‐

Technical Accident Model and Processes (STAMP),

and Human‐Tech approach (Hollnagel, 2012;

Leveson, 2004; Vicente, 2004, respectively). Each

method has the potential to improve safety by

incorporatingsystemsthinkingintothe approach.In

this paper the FRAM

 is used to perform an

investigation of ship navigation in the Arctic. The

FRAM was chosen for two reasons: 1) it focuses on

functionality, and 2) it promotes communication

between assessors and workers. To understand

functionality, you must understand the conditions

thatcanbeoperatedin,andtheconditionsthat

cause

problems.Thismeansthataccidenteventsshouldnot

be isolatedfrom thetypicaloperationaloutcomes to

develop understanding of accident mechanisms. By

isolating the accidents, biases may enter the

interpretations of events. Safety solutions should

show consideration of both the event(s) one would

like to prevent and promotion of the

 event(s) one

would like to achieve. When understanding

functionality, it is best to obtain an understanding

from the operational perspective. This concept

promotesunderstandingtheworkasitisdone,rather

than as it is imagined by assessors. This can help

reduce the communication gap that exists between

assessors and

operators, thereby, promoting safety

solutionsthataregroundedinreality.

2.1 FRAM

The FRAM is built on identifying functional

resonance. Functional resonance is an analogy to

stochastic resonance, where multiple signals of low

amplitude noise are inputted to a  system and, if

resonanceoccurs,theoverallsystemsignalcanhavea



muchgreateramplitude.Infunctionalresonance,the

outputofthesystemfunctionsarevariableandslight

variations between the many functions in a system

have the potential to combine  in such a way that

resonance occurs. The resonance will be some

variationoftheoverallsystemperformancethatgoes

beyond

what is typical or expected, regardless of

whether the outcome is viewed as good or bad. By

modeling the system functions and variability in

sufficient detail, safety solutions will emerge that

focusonmonitoringandcontrollingthesystem.

TheFRAMisbasedonfourunderlyingprinciples

(Hollnagel,2012):

 Failures and

 successes are equivalent in the way

that they happen for the same reason.

Alternatively,itcanbesaidthatthingsgowrong

forthesamereasonsthattheygoright.

 Daily performance of socio‐technical systems,

includinghumansindividuallyandcollectively,is

alwaysadjustedtomatchthesystemconditions.

 Many of the outcomes of the system that we

notice, and also the ones we don’t notice, are

emergentratherthanresultant.

 Relationsanddependenciesmustbedescribedas

they develop in a particular situation and not as

cause‐effectlinks.Thisisdonethroughfunctional

resonance.

The first

 step of the FRAM is to describe the

functions of the system and the aspects of the

functions that occur when work happens. Each

functioncanhave6aspectsthatshouldbeconsidered,

asseeninFigure1.

Output:Eachfunctionshouldhaveanoutput(s).If

work is being done

there should be something

producedbythe work. Theoutputsare thenpassed

throughoutthesystemand have theability to affect

otherworkinthesystemin5possibleways.

1 Input:Theinputstartsthefunctions.Iftheinputis

anoutputthatarriveslatefromanotherfunction,it



will affect the functionality of the downstream

function.

2 Preconditions: Preconditions must be available

prior to the function starting, but they do not

initiatethefunction.Theycanlaydormantinthe

systemuntilthefunctionbegins.

3 Resources: These are things that are processed

during the function. To limit

the resources that

considered, focus should be placed on resources

that are consumed and subsequently need to be

resupplied by another function in the system.

Resources such as computers, which are not

consumed, should not considered here. They

would be considered as execution conditions,

which can be assessed when understanding the



functionitself.

4 Time:Otherfunctionaloutputshavethepotential

toaffecttheavailabletimetocarryoutafunction.

5 Control: Other functions may interact with

downstream functions in a way that acts as a

control.

After the system functions and aspects are

described at some level of detail.

The variability

should be considered. Step 2 considers the internal

variabilityofthefunctionandthevarietyofwaysan

output can be produced under dynamic conditions.

Step3assessesthecoupledsystemvariability,which

isthewaythevariationsfromupstreamfunctionscan

affect the downstream functions, and in turn

 the

449

entire system performance. The final step is to

identifyappropriatewaystomonitorthesystemand

control the variability in it. In practice, it is very

difficult to obtain all the necessary information at

once,sothisprocessmayneedtoberepeatedasnew

informationisobtained.



Figure1.FRAMfunctiondiagram(Hollnagel,2012)

3 METHODOLOGY

In order to build a FRAM model for Arctic ship

navigation the following methodology was used.

First, the scope was defined. Then the system

functions and connections were imagined by the

assessor(s). The conceptualized model was then

checked with operators to verify that the model

reflects the way the

 work is actually done. At this

point,themodelrepresentedthepotentialfunctional

paths thatcouldbe taken forthesystem to produce

someoutcome.Then thevariabilityof thefunctional

outcomescanbeunderstood.Itisbesttolearnabout

the variability of the functions by either monitoring

the

functionaloutputdirectlyorcommunicatingwith

the workers who carry out each function. Once the

functional model was built and some variability

documented, the model was applied to cases. By

examining cases through the lens of the FRAM,

different findings may emerge that pertain to

functional execution and system variability. These



findingscanthenbeusedtoeitherupdatethemodel,

or manage the operation. This methodology is

mappedoutinFigure2.

3.1 Definingthescope

Thefirststepistodefinethescopeoftheassessment.

This assessment focuses on (Arctic) ship navigation.

From a systemic perspective, there are

many

functionsthatinfluencetheperformanceofashipping

operation and trying to model all of them at once

could be overwhelming.As there is so much

informationtolearnabouttheworkthatiscarriedout

inashippingoperation,theinitialassessmentfocuses

on navigating the vessel. This is

 the most basic

objective for a ship and all other work is

complementary to it. This allows the initial

understanding to reflect the most immediate

functionsrequiredfornavigation,andthenthescope

can be gradually broadened in the future. Also, the

focuswillbeontransitshipping; stationaryoffshore

installationsareoutofscope.



Figure2:MethodologyforbuildingFRAMmodel



Figure3:GeneralshipnavigationFRAMmodel(scope)

First,buildaFRAMmodeltohelpdefinethescope

(Figure 3). We define a function, “Navigate ship,”

which describes the function that is carried out to

physicallymovetheshipfromporttoport.Thenwe

can define the aspects of the “Navigate ship”

function.Theoutputcanbethat

theshipisnownear

the destination, and other functions involved in,

“Arrive at port,” can begin bringing the ship to the

destination.Theinputisthefunction“Decidetoleave

port.”Whilethisdecisiontoleaveisinfluencedbythe

shippingschedule,theshipdoesnotnecessarilyleave

exactly

whenscheduled.Manyfactorscouldaffectthe

time at which the ship actuallyleaves port, but this

decisionis controlledbytheschedule.Thetime that

thisdecisionwillbemadewillberoughlyaroundthe

scheduled time, but could be ahead or behind

schedule, due to inspections, cargo or consumable



loading,etc.Theshippingschedulecanalsoinfluence

theshipnavigationfunctionwithrespecttotime.The

ship navigator may make decisions to speed up or

changeroutetostayonschedule.Amajorcontrolling

aspectforshipnavigationisto“Consideroperational

regulations.” By considering these operational

450

regulations, best practices, and guidance can be

transferred to the ship navigator, helping to control

thefunctionality.Apreconditionisthat a shipmust

eitherbedesignedand/orprocuredandcrewmustbe

hired in order to navigate this ship. This is a

preconditionbecauseitmusthappenpriorto

theship

navigation,butitdoesnotinitiatetheshipnavigation

as theinputdoes.The ship and crew can remain at

portuntilthe decisiontoleave port has beenmade,

then“Navigateship”canbegin.Lastly, let’sconsider

the resources necessary to navigate a ship. In the

FRAM,

resourcesshouldbefocusedonitemsthatare

consumed during or need to be resupplied after a

function is executed. While, we could think of the

shipasbeingaresource,itwillnotbeconsumed(at

least not over a single voyage), and is more

appropriately considered as a precondition

 aspect.

Resources such as cargo and consumables (fuels,

stores, ballast, maintenance materials) will be

consumedduringavoyageandshouldberesupplied

beforeanothervoyageistobegin.

This generalized model (Figure 3) has helped us

definescopeandstartthinkingaboutshipnavigation 

intermsoftheFRAM.However,

themodelisnotyet

detailedenoughtoprovidemuchusefulinsight.Now

thatthescopeis betterunderstood, thefocus canbe

shifted to understanding how ship navigation is

carriedout.

3.2 BuildingaconceptualizedFRAMmodel

In the FRAM, it is best to have your assessment

informed by the

 workers who carry out or interact

closely with the system functions. However, it is

usefultofirstbuildaconceptualizedmodelfromthe

perspective of the assessors to help illustrate the

FRAM to the worker(s) in the context of their

operation.Thisconceptualizedmodelcanbeseenin

Figure4.

In Figure 4, the ship navigation process is

describedasacontinualassessmentoftheconditions

that result in a decision to maintain a course or to

change course.Thiscan be done many times overa

single voyage. The decision then leads to the

navigatorfollowingthe chosen courseand

 notifying

thecrewofanyadjustments,ifnecessary. Inorderto

reasonably make an assessment, the ship navigator

must consider many conditions comprehensively to

makethemostinformeddecisions.Theoutputsfrom

these functions may be produced at different rates

andassessmentsbythenavigatorwillbemadewith

varying

levelsofinformation.Someoftheinputsthat

we can imagine are important to a navigator’s

assessmentare:

 Observingthecurrentweatherconditions

 Obtainweatherforecasts

 Observetheiceconditions

 ObtainIceforecasts

 Considertheintendedorpredictedroute

 Monitortheconditionofthevessel



 Be aware of the surrounding location and

geography.



Figure4:ConceptualizedFRAMmodelforshipnavigation

451

3.3 Verifyingwithworkers

To inform our assessment, we spoke with 3 ship

captains. The discussions were focused on

understanding how ship navigators navigate ships,

and making note of any unusual variations or

conditions that they shared. The representation of

shipnavigation(Figure4)wascritiquedbythe3 ship



navigatorsanditcontainedmanyofthefunctionsthat

thenavigatorsused butit was incomplete.Consider

thefunctionaldescriptionsandtheinitialdescription

oftheaspectsforshipnavigationinTable1.Theonly

times that an output will by omitted is when it has

been left out to

define the scope of the analysis.

Similarly,when“notinitiallydescribed”islisted,this

doesnotmeanthatthataspectisnotpresent.Itmeans

thatthescopehasinitiallybeenlimitedtodescribing

the coupling of the immediate functions that have

been described. This will help prevent becoming

overwhelmed with

 complexity initially. Additional

aspectscanbefurtherdescribedlater,ifneeded.

Itcanbeseenthatadditionalfunctionshavebeen

identified through conversations with ship

navigators. The visual representation of the FRAM

modelwithinputfromshipnavigatorscanbeseenin

Figure 5. It can be seen that this

more detailed

descriptionofshipnavigationshowsamorecomplex

representationthantheoneinFigure4.Itisimportant

understandthecomplexities that arepresentin ship

navigation beca use  these complexities must be

managed in the operation, whether we decide to

modelthemornot.

Table1.InitialdescriptionofFRAMfunctionsandaspectsforshipnavigation

__________________________________________________________________________________________________

Nameoffunction Obtainweatherforecast  Setnew/maintaincourseObserveIceconditions

__________________________________________________________________________________________________

Description  ObtainweatherforecastfromAdecisionismadetoeither  Observethecurrenticeconditions.

meteorologicalorganization maintainthecurrentcourseortoThiscanbedonefromthebridgeoron

ordepartmentmakeadjustmentstocourse.  deck,butalsotheconditionsahead

canbeobservedvia

helicopteror

aircraft

__________________________________________________________________________________________________

AspectDescriptionofAspectDescriptionofAspectDescriptionofAspect

__________________________________________________________________________________________________

InputNotinitiallydescribedCompleteorpartialassessment Notinitiallydescribed

made

OutputWeatherforecastobtained RoutingdecisionmadeIceconditionshavebeenvisually

observedonboard

 Uprouteiceconditionsassess.with

Helicopter

Precondition  NotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

ResourceNotinitiallydescribedNot

initiallydescribedNotinitiallydescribed

ControlExperiencebasedweather NotinitiallydescribedExperiencedvisualassessmentofice

judgementRadarimageobserved

TimeNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

__________________________________________________________________________________________________

Nameoffunction ForecastIceconditionsAssesslocationandInformcrewofcourse

surroundinggeography

__________________________________________________________________________________________________

Description  Obtaintheforecastedice  Locatethevesselwithrespect Informcrewofanychangeofcourseif

conditions.Thismaybedonetointendedroute,shipping  necessary.

byhistoricaltrendsinarea lanesandregionalgeographic

and/ortacticalicedriftmodelsfeatures.

__________________________________________________________________________________________________

AspectDescriptionofAspectDescriptionofAspectDescriptionofAspect

__________________________________________________________________________________________________

InputNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

OutputObtainedforecastedice  Geographicalassessmentmade Responsiblecrewmembernotified

conditions

Dailyicechartobserved

Precondition  NotinitiallydescribedAwareofthepresentroute  Routingdecisionmade

ResourceIcechartdownloadedNotinitiallydescribedNotinitiallydescribed

ControlExperience

basediceforecast HaveshippinglanemapsNotinitiallydescribed

 Improvedknowledgeofregional

specificconditions

TimeNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

__________________________________________________________________________________________________

Nameoffunction AssesslocationandMakesituationalassessmentPerformcrewwork

surroundinggeography

__________________________________________________________________________________________________

Description  LocatethevesselwithrespectThecaptainandbridgeteam  Thecrewwillperformtheirnecessary

tointendedroute,shipping makeasituationalassessment worktomaintaincourseoradjusttheir

lanesandregionalgeographicbasedontheavailableworktoaccommodateanychanges.

features.informationata

giventime.

__________________________________________________________________________________________________

AspectDescriptionofAspectDescriptionofAspectDescriptionofAspect

__________________________________________________________________________________________________

InputRoutingdecisionmade Weatherforecastobtained  Responsiblecrewmembernotified

 Uprouteiceconditionsassess.

withHelicopter

 Obtainedforecastediceconditions

 Geographicalassessmentmade

 Weatherhasbeenobserved

 Awareofapparentvesselcondition

 Iceconditionshavebeenvisually

452

observedonboard

 Proximatetrafficcommunicatedwith

OutputNotinitiallydescribedCompleteorpartialassessment Notinitiallydescribed

made

Precondition  NotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

ResourceNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

ControlNotinitiallydescribedIceNumeralcomputedNotinitiallydescribed

Time

NotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

__________________________________________________________________________________________________

Nameoffunction ObserveweatherConsiderpredicted/updatedComputeIceNumeral

route

__________________________________________________________________________________________________

Description  Thecurrentlocal(ship)  Considerthecurrentrouteyou Computetheicenumeralasper

weatherconditionsare   aretransiting.Thismaybe  Canadianregulatoryrequirements.

observed.Thiscanbefrom suggestedbyoperational

thebridgeorondeck.plannersoradjustedbythe

navigator.

__________________________________________________________________________________________________

AspectDescriptionofAspectDescriptionofAspectDescriptionofAspect

__________________________________________________________________________________________________

InputNotinitiallydescribedNotinitiallydescribedDailyicechartobserved

OutputWeatherhasbeenobserved Awareofthepresentroute  IceNumeralcomputed

Precondition  NotinitiallydescribedNotinitiallydescribedShipclassificationassigned

ResourceNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

ControlNotinitiallydescribedNotinitiallydescribed

Notinitiallydescribed

TimeNotinitiallydescribedShippingschedulemadeNotinitiallydescribed

__________________________________________________________________________________________________

Nameoffunction Monitorvesselcondition  AssignshipclassificationDownloaddailyicecharts

__________________________________________________________________________________________________

Description  Thevesselʹsconditionis  TheshipisassignedaDownloadthedailyicechart(s)that

monitoredtounderstandtheclassification.Inparticular,this areapplicabletoyourregion.These

vesselʹscurrentcapabilities. classificationherepertainstothe chartsareproducedbyCanadianIce

categorythatwill

beusedto  Services (CIS)inCanada.

computetheicenumeral.

__________________________________________________________________________________________________

AspectDescriptionofAspectDescriptionofAspectDescriptionofAspect

__________________________________________________________________________________________________

InputNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

OutputAwareofapparentvessel  Shipclassificationassigned  Icechartdownloaded

condition

Precondition  Engineroommaintenance/ NotinitiallydescribedNotinitiallydescribed

issuesinformed

Awareofvesselʹstypical

capability

ResourceNotinitiallydescribedNotinitiallydescribedNotinitially

described

ControlNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

TimeNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

__________________________________________________________________________________________________

Nameoffunction IcenavigatormakesObtainmapofshippinglanes Observeradarimage

assessments

__________________________________________________________________________________________________

Description  IcenavigatormakesPriortoshippingthroughan  Theradarimageisobservedandthen

assessmentsoftheconditionsareaitisgoodpracticetoobtain shouldbevisuallyinspectedto

andupcomingtasksand  mapsoftheshippinglanes.The determinewhatcausedtheradar

sharesexperience

withships shippinglanestypicallyhasmoreimagetobeproduced

bridgeteam.reliablesoundingsandhavebeen

practicedovertheyears.

__________________________________________________________________________________________________

AspectDescriptionofAspectDescriptionofAspectDescriptionofAspect

__________________________________________________________________________________________________

InputNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

OutputExperiencedvisualHaveshippinglanemapsRadarimageobserved

assessmentofice

Experiencebasediceforecast

Improvedknowledgeof

regionalspecificconditions

Experiencebasedweather

judgement

Precondition  IcenavigatorhasbeenNotinitiallydescribedAradarsignal

hasbeendetected

assignedbyshipsradar

ResourceNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

ControlNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

TimeNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

__________________________________________________________________________________________________

Nameoffunction ObserveothertrafficCommunicatewithproximate Communicatewithengineroom

traffic

__________________________________________________________________________________________________

Description  Observeanyothershipping Communicatewithproximate Thereiscommunicationbetweenthe

trafficthatmaybeinthearea traffic.Thiscanbedonevia  engineroomandthebridgetodiscuss

lights,hornsorradio.anyissuesorneededmaintenance.

__________________________________________________________________________________________________

AspectDescriptionofAspectDescriptionofAspectDescriptionofAspect

__________________________________________________________________________________________________

InputNotinitiallydescribedOthertrafficobservedNotinitiallydescribed

OutputOthertrafficobservedProximatetrafficcommunicated Engineroommaintenance/issues

453

withinformed

Precondition  NotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

ResourceNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

ControlRadarimageobservedNotinitiallydescribedNotinitiallydescribed

TimeNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

__________________________________________________________________________________________________

Nameoffunction Assigncertifiedicenavigator DetectradarimageBecomeawareofvesselʹscapability

__________________________________________________________________________________________________

Description  Toassignanicenavigatorto Radarsignalhasbeensent fromThenavigatorbecomesawareofthe

assistwithnavigationofthe shipsradarandisreadyto  vesselʹscapabilities.Thenavigational,

vessel.Thisisrequiredfor receiveanysignalsthatbounce structuralandoperationalcapabilities.



NavigationintheCanadian backfromobjects

Arctic.

__________________________________________________________________________________________________

AspectDescriptionofAspectDescriptionofAspectDescriptionofAspect

__________________________________________________________________________________________________

InputNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

OutputIcenavigatorhasbeenAradarsignalhasbeendetected Awareofvesselʹstypicalcapability

assignedbyshipsradar

Precondition  NotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

ResourceNotinitiallydescribedNotinitiallydescribedNotinitially

described

ControlNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

TimeNotinitiallydescribedNotinitiallydescribedNotinitiallydescribed

__________________________________________________________________________________________________

Nameoffunction Makeshippingschedule

__________________________________________________________________________________________________

Description  Expecteddepartureandarrivaltimesaredetermined. 

__________________________________________________________________________________________________

AspectDescriptionofAspect

__________________________________________________________________________________________________

InputNotinitiallydescribed

OutputShippingschedulemade

Precondition  Notinitiallydescribed

ResourceNotinitiallydescribed

ControlNotinitiallydescribed

TimeNotinitiallydescribed

__________________________________________________________________________________________________



Figure5.FRAMmodelforshipnavigationwithinputfromshipnavigators



3.4 LearningVariations

Figure5showsamapofthepotentialwaysthata

shipcouldbenavigated.Buttherearemanywaysthe

ship could be navigated, including combinations of

thepotentialfunctionalpathsshowninFigure5.This

variabilitymustbeunderstood,ifitistobeproperly

managed. Also, there will be more Arctic specific

knowledge here, because Arctic shipnavigationisa

variationofshipnavigation.SeeTable2forsourcesof

variability and additional notes along with some

ways this variability has been managed in the past.

This model can help to better understand some

shipping

scenarios.

454

Table2.Variability,notesandmanagementstrategieswithfocusonArcticshipping

__________________________________________________________________________________________________

AssociatedFunction  SourcesofpotentialvariabilityNotesandManagementtechniques

__________________________________________________________________________________________________

Setnew/maintaincourse MorethanonepossiblecourseSlowdown‐allowtimetoreceivemore

information–makemoreinformeddecision

Schedulingandexpectedprofitscan

influencedecisionmaking

Theamountofconsumableonboardalso

affectdecisionmaking(routeselection)

Assesslocationand   GPSmaynotbe

accurateathighlatitude

surroundinggeography Coastlineandunderwatermappingmay

bepoorinareasofArctic

Soundingcouldbeinaccurateoutsideof

shippinglanes

Considerpredicted/Possiblemultipleroutes‐NWPhas3Dynamicsetofsolutions

updatedrouteIceconditionsmaytakeyououtsideof

shippinglanes



Searchandrescueoperationcantakeyou

outsideofshippinglanes

ComputeIceNumeralThisiscomputedoncedaily‐whenanew

icechartispublished.

Thecomputationisbasedontheiceassessment

fromtheicechart‐Ifthechartcontainserrorsit

willaffecttheappropriateness

ofthecomputation

DetectradarimageSmallicebergs(growlers)canbedifficult Reducespeed‐increasereactiontimeifdetected

todetectinicelate

Smallicebergs(growlers)canbedifficult

todetectinlargeseastates

Domeshapedicebergsmaybeproblematic

todetect

Sleetcanaffectperformance

ofradar

Qualityoftheinstalledradartechnology

ObserveIceconditionsDarknessaffectsabilitytoseeiceconditions Goodsearchlight‐veryvaluableandbackup

searchlights

ExperienceofIcenavigatorandCaptainWithuncertainconditions,reducespeedto

minimizeforceofunexpectedimpacts

Realconditionscanbeworsethan

was  Dealwithitand/orturnaround

forecasted

Icechartsarepublished24hours‐over  Trytouseicechartandradartopredictships 

24hourstheicewillmovepositioninchangingicefield.Alsosend

helicopterforvisualinspectionifavailable.

Importanttorememberthat

icemoveswithwind

andicebergswillmovewithcurrent

ForecastIceconditionsQualityofIcechartQualityusuallyimprovesifaerialassessmentof

theregionhasbeendone

Forecastmodelsmaybepoorforcertain Experiencedicenavigatorcanalsoprovide

regionsexperiencebasedforecasts

Obtainweather

forecast Forecastmaybepoorqualityornon‐existentExperiencedicenavigatorcanalsoprovide

forsomeregionsoftheArcticexperiencebasedweatherforecastsoflocal

weatherpatterns

Howmanyweatherforecastsareavailable

daily?

Communicationsproblemsathighlatitudes

canaffectabilitytoobtainforecast

Observeweather

conditionsCanobservevarietyofconditions‐Wind

andsnowcanaffectvisibility‐Coldraincan

expecticing

Noticedifferencesfromweatherforecasts Icenavigatormaybeabletohelpdeterminehow

weathermightchange

MakesituationalIsfullbridgeteampresent?Otherworkcommitmentsmaytakethemfrom

assessmentbridgewhenassessmentismade

HowmuchtimetomakeassessmentCanslowdowntomakemoretime

Hereisthefunctionthatinfluencedby  Variationsofeveryupst reamfunctionwill

allotheranalysisfunctionsinfluencethequalityoftheassessmenthere

FatiguecanaffectassessmentsandShift

schedulescanaffectfatigue‐Ice‐induced

decisionmakingvibrationscanaffectfatigue

Icepressurecanbeproblematicforship

navigation,eveninlowicethickness

Longerperiodsofdarknesscanaffect

decisionmaking

SlushhasthepotentialtoclogcoolingwaterFinerscreenoverwaterintakes

intakes,

andrisklosingengine‐thishas

beenseeninthepast

Icebreakerassistancemaybecalledforif Whenfollowing/beingtowedbyicebreaker: Keep

conditionsbecomeunmanageableforvessel. propturning,Mayhavetofollowverycloselyin

455

Thiscouldtakesometimeifnotplannedfor highicepressurefield(channelwillclosein).

inadvanceUseicetohelpstopwhenfollowingclosely

(preventcollision)

CommunicatewithCommunicateupcomingmaintenance  Workculturemayinfluencecommunication

engineroomfrequency

Communicateperformanceissues

Monitorvesselcondition Wet

conditionsoropenwatercanpromote Breakingofftheicecanalsobeadangerous

marineicingprocedureandisusuallyavoideduntilabsolutely

necessary

Thereareicingallowancesinstabilitybook Itisverydifficulttomonitortheweightofice

buildupanddistributionoftheweight

Parallelmid‐bodystresswillbehighif  Avoidifpossible

enteringamobileicefieldfromfastice

(shearzone)

Difficulttomonitor(feel)bowimpactsif

bridgeispositionedastern

BackingupiniceKeeprudderstraightwhenmovingastern

PerformcrewworkCrewmaynot

bepreparedforandhave

experienceincoldclimate

__________________________________________________________________________________________________



4 DISCUSSION

Itisimportanttounderstandthatthismodelstillhas

missing elements.It can be expandedtoincorporate

more elements to improve our understanding of

socio‐technical system that is ship navigation. It is

acknowledgedthatthereareregulatoryfunctionsand

organizational functions omitted from this model.

These functions are carried out at lower frequencies

than the onboard functions, but will influence the

onboardwork.Thenextstepistobetterunderstand

how these regulations and organization affect the

functionality of ship navigation. It may be also

appropriate to further define some functions. For

example, it may be

 appropriate to break down the

“Monitor vessel condition” function into separate

functions,asinFigure6.



Figure6.Breakingfunctionintosub‐functions

Thenitmaybeappropriatetoask:1)whenisthe

FRAMmodel“complete”?and2)Howdoweknowif

wehavesufficientgranularity?Themodelwillnever

be complete but each revision should improve the

understanding. There is no guarantee that future

operations will mirror past operations, so there

 are

alwaysnewlessonstolearn.Aslongasthesystemis

operating, there will be new information to add to

your FRAM model. Itwilldepend on what you are

tryingtoexplainandtheexplanationsyouarewilling

toaccept.Thedetailofthefunctionmaybeacceptable

to explain one scenario, but inadequate to explain

another. In this case, it is important to not try to

categorize explanations into two discrete groups,

rightorwrong.Explanationscanrangefrompoorto

acceptable, and further examination will produce

better explanations. As more details are understood

acceptable explanations will emerge.

 The question

then becomes, what is acceptable? Explanations

should be sought that not only describe what

happened, but how it happened and why it

happened. By understanding these 3 parts of a

scenario,bettermanagementstrategieswillbeableto

bedeveloped.

In order to demonstrate the utility of this

information, it should be used to explain certain

scenarios from the shipping domain. The FRAM

modelcanbe usedtoadd totheunderstandingthat

have been obtained from traditional examination

techniques.Insection4.1theExxonValdezcasewill

beconsidered.

4.1 Applyingacase:theExxonValdezgrounding

The

 Exxon Valdez grounded on March 24, 1989 on

Bligh Reef in Prince William Sound while

transporting crude oil from Valdez, Alaska to San

Diego,California.Thisshippingaccidentisoneofthe

most well‐known, which garnered much media

attention and legal intervention because of its

environmental impact and ill‐

defined oil spill

responsepolicy.IntermsofArcticshippingaccidents,

theExxonValdez caseisthe mostwelldocumented

accident that is publicly available.This case may be

themostsuitablecasetoexaminethroughthelensof

the FRAM because of the extent of information

availablecomparedtoother

cases.

All information in this case is taken from the

National Transportation Safety Board’s (NTSB)

marineaccidentreportontheExxonValdezaccident

(NTSB, 1990). The NTSB performed an extensive

investigationandanalysisofthisaccident.Thereport

included 47 findings that were determined to be

relevanttotheaccident,

anaccountofprobablecause,

and recommendations to the

organizations/departments involved. The report has

beenaverysignificantdocumentforshippingsafety

andinfluencedtheadoptionofdoublehulltankships

across the industry. The adoption of double hull

tankships has improved safety of the tankship

456

industry,specificallywithrespecttoitsrelationshipto

theenvironment.

Theaccountofprobablecauseisasfollows(NTSB,

1990): “The National Transportation Safety Board

determinesthattheprobablecauseofthegroundingofthe

EXXON VALDEZ was the failure of the third mate to

properly maneuver the vessel because

 of fatigue and

excessive workload; the failure of the master to provide a

proper navigation watch because of impairment from

alcohol;thefailureofExxonShippingCompanytoprovide

a fit master and a rested and sufficient crew for the

EXXON VALDEZ;thelack of an effective VesselTraffic

Service

because of inadequate equipment and manning

levels, inadequate personnel training, and deficient

management oversight; and the lack of effective pilotage

services.”

This account of probablecausecanbe visualized

bythecausaldependencydiagraminFigure7.



Figure7. Causal dependency diagram produce from the

account of probable cause given in the in the Marine

AccidentReport

Nowconsiderhowthegroundingwouldlookby

applying theinformationinthe groundingreportto

the FRAM model for Arctic ship navigation. The

FRAM model shown in Section 3.3 displays the

potential functional paths to navigating the vessel.

The Exxon Valdez case can be used to illustrate the

functional dynamics

 that contributed to the

grounding. The generalized FRAM model seen in

Figure5representsthepotentialwaysthatanArctic

ship navigator could operate the ship. However,

when a ship navigator operates the vessel, many

combinationsofselectedfunctionsmaybeused.The

marine accident report of the Exxon Valdez

grounding can be used to help understand the

functional dynamics that occurred during that

accident(NTSB,1990).

Figure8 showsthat atabout23h55 onMarch 23,

1989 the Navigator (Third Mate) and his team were

assessingthelocationoftheExxonValdezrelativeto

BusbyIslandLighttodetermineif

itwastimetoturn

backtowardstheshippinglanethatthey hadleft to

avoidglacialice.Atthistime,thenavigatorwasusing

theradartoestimatethevessel’spositionfromBusby

IslandLight,whichheestimatedtobe0.9milesaway.

Also, a fix was plotted on

a chart of the vessel’s

position from visua l  observations, which estimated

BusbyIslandLighttobe1.1milesaway.Therewasa

discrepancyof0.2miles of the navigator’s estimates

of the vessel’s position. Additionally, during this

functional snapshot there was an additional

functional relationship learned that existed between

observingthe

radarimageandassessingthevessel’s

location and surrounding geography. This

relationshipwas notnoticedin previousdiscussions

withshipcaptainsandwasaddedtothemodel(one

of the blue lines in Figure 8) to add to the model’s

comprehensiveness.

In this analysis, the functional signature of the

Exxon

Valdezwaspresented.Thisrepresentsa single

voyage for this vessel. From this data alone, it is

difficulttodeterminewithhighcertaintywhatcaused

this accident. However, if there was data available

aboutother voyagesthatthe ExxonValdezhad and

successfully navigated through Valdez Narrows,

there would be a

 better understanding of the

functional signatures that promoted better

performance of the Exxon Valdez. Presumably, the

vessel successfully navigated the Narrows before

while the captain was away from the bridge, while

workers were fatigued, or while glacial ice entered

into the shipping lanes. By using a method that is

capable

ofalsoanalyzingsuccessfulvoyages,thereis

a better chance of identifying what was different

about the functional signatures that promote such

different outcomes. Additionally, if this information

was available, the value of this analysis could be

increased.

By considering  systemic safety solutions and

understandingthe navigationalprocesses, additional

safetyrecommendation

canbemade.Forinstance,in

addition to recommending minimizing fatigue by

analyzing ideal shift schedules, elements could be

introduced into the system that help navigators

performbetterevenwhenfatigued.Itcanbereasoned

that even under ideal sleeping conditions, e.g. a

personworkinga9‐5deskjob,

apersoncanarriveat

worktired orfatigued.Additionalrecommendations

ofupdatingtheautopilotsystemto bemoreevident

astowhenitwasengagedordisengaged,asthiswas

a source of confusion for the crew of the Exxon

Valdez during the grounding. This could help

fatigued workers be

 more aware of the conditionof

theirvessel.Additionally,othertechnologiescouldbe

recommended that help ship navigators  more

accuratelyassesstheir locationinawaterway.Inthe

present,theadditionofGPSonvesselsmayhelpwith

thisalthough,someoftheCaptainsusedinsection3.3

have expressed

concern aboutGPSaccuracyat high

latitudes.

457



Figure8. Functionalrepresentation ofthe ExxonValdez grounding atabout 23h55 with updated functionalrelationship

(bluelines)

5 CONCLUSIONS

In this work, the FRAM has been used to start an

investigation into Arctic shipping by trying to

understandshipnavigationanditsvariations in ice.

TheprocessofbuildingaFRAMmodelwasdiscussed

andanapplicationofthemodelwasillustratedusing

theExxonValdezgrounding.After

speaking withthe

ship navigators, a more detailed FRAM

representationofshipnavigation hasbeendeveloped.

Someofthevariationsandconditionsthatarepresent

in Arctic navigation  are discussed along with the

ways that ship navigators manage these conditions.

The grounding of the Exxon Valdez was examined

and provided

context to the information that was

made available by the Marine accident report. This

case allowed for an alternative perspective and

complementary discussion of the case than could

havebeenhadwithouttheFRAM.

Itisacknowledgedinthisworkthattherearestill

elementsthatfactorintotheshipnavigation

process

thatareomittedfornow,includingmanyregulatory

functions and organizational functions. This work

servesasaninitialstartingpointtousetheFRAMto

helpbetterunderstandthecomplexitiesthatexistfor

ship navigation in the Arctic. This work can be

improved in the future by further defining

the

functionaldescriptions,incorporatingmorevariations

thathavebeenexperienced,andextendingthescope

of the assessment. The framework to do this is

presentedinthis paperandnew informationcan be

usedtoupdatethemodel.

ACKNOWLEDGEMENTS

The financial support of the Lloydʹs Register

Foundation is acknowledged with gratitude.

 Lloydʹs

by supporting engineering‐related education, public

engagementandtheapplicationofresearch.

REFERENCES

Arctic Council, 2009. Arctic Marine Shipping Assessment

2009Report.

Aven, T., Andersen, H.B., Cox, T., Droguett, E.L.,

Greenberg,M.,Guikema,S.,Kröger,W.,Renn,O.,Zio,

E.,2015.RiskAnalysisFoundations.Soc.RiskAnal.

Ayyub,B.M.,2015.PracticalResilienceMetricsforPlanning,

Design, and Decision Making. ASCE‐ASME J. Risk

Uncertain.

 Eng. Syst. Part Civ. Eng. 1.

doi:http://dx.doi.org/10.1061/AJRUA6.0000826

Ayyub, B.M., 2014. Systems resilience for multihazard

environments: definition, metrics, and valuation for

decision making. Risk Anal. Off.  Publ. Soc. Risk Anal.

34,340–355.doi:10.1111/risa.12093

Borys,D.,Else,D.,Leggett,S.,2009.Thefifthageofsafety:

theadaptiveage.J.HealthSaf.

Res.Pract.1.

Hollnagel, E., 2012. FRAM:  The Functional Resonance

AnalysisMethod.AshgatePublishingLtd.

Hollnagel, E., Woods, D.D.,  Leveson, N., 2006. Resilience

Engineering:ConceptsandPrecepts.AshgatePublishing

Ltd.,Hampshire,UK.

Leveson,N.,2004.ANewAccidentModelforEngineering

SaferSystems.Saf.Sci.42,237–270.

NTSB, 1990. Marine Accident

 Report‐Grounding of the

U.S. Tankship EXXON VALDEZ on Bligh Reef, Prince

William Sound, near Valdez, Alaska,  March 24, 1989

(No.NTSB/MAR‐90/04).NationalTransportationSafety

Board,Washington,D.C.

Smith,D.,Veitch,B.,Khan,F.,Taylor,R.,2015.AnAccident

ModelforArcticShipping,in:ProceedingsoftheASME

2015

34th International Conference on Ocean,Offshore

andArcticEngineering.PresentedattheOMAE2015,St.

John’s,NL,Canada.

Vicente, K., 2004. The Human Factor: Revolutionizing the

Way People Live with Technology. Routledge, New

York.