565
1 INTRODUCTION
Safetyofnavigationisoneofthekeyissuesaffecting
thetransportprocesses,whichareusedformaritime
transport. Among the many threats that could
undermine the proper course of the voyage,
accidentsrelatedtothemanfallingoverboardbelong
to a small group of events that can ta
ke place
regardless of the hydro‐meteorological conditions,
specificsoftheoperatedwaterarea,trafficintensity,
orotherexternalcircumstances.
Despite the tendency to constantly improve the
levelofsafetyforoverseavessels,correctresponseof
the ship navigator, resulting in the correct
implementation of the requiredʺman overboardʺ
manoeuvreisnotalwaysobserved.TheissueofPOB
(PersonOverboard)accidentsisthereforest
illvalid,
particularly that for a decade you can see steady
growthoftheworld merchant fleet, whichoverthe
last eight years has increased by 20 357 vessels
[EuropeanMaritimeSafetyAgency2005‐20013].
Figure1.Thenumberoftheworldmerchantfleetinyears
2005‐2013[EuropeanMaritimeSafetyAgency2005‐20013].
2 SIMULATIONSTUDIES
Becauseofthedirectimpactoftheʺmanoverboardʺ
manoeuvreontotherescueofhumanlifeatsea,the
membersoftheScientificSocietyʺWatchersʺactiveat
the Faculty of Navigation of Gdynia Maritime
Comparison of the Efficiency of Williamson and
Anderson Turn Manoeuvre
K.Formela,M.Gil&H.Śniegocki
GdyniaMaritimeUniversity,Gdynia,Poland
ABSTRACT:Thepaperpresentscomparisonofthe‘Personoverboard’manoeuvres.Thearticlewasbasedon
researchcon‐ductedonagroupofstudentsoftheFacultyofNavigationofGdyniaMaritimeUniversity.People
involved in studies previously conducted on board training on school training ships:ʺDar Młodzieżyʺ and
ʺHoryzont IIʺ, but did not have experience as an officers. Research was carried on the Polaris Ships Bridge
Simulator,locat
edatthefacultyofNavigation,ofGdyniaMaritimeUniversity.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 9
Number 4
December 2015
DOI:10.12716/1001.09.04.14
566
University,decidedtocarryouttherelevantresearch
inordertodeterminetheeffectivenessofindividual
manoeuvres.
Themeasurementpartconsisted,successively,of
theʺWilliamson Turnʺ and theʺAnderson Turnʺ
performed by two groups of 17 students of the
Navigation‐MaritimeTransportFaculty.Persons
involved in the studies previously underwent the
seagoing service onboard of training vessels‐
s/v
ʺDarMłodzieżyʺorʺHoryzontIIʺ.
Selectedfortestingwasnavigationand
manoeuvring simulator, POLARIS type,
manufactured by the Norwegian company named
Kongsberg. All devices operate on the basis of
complex mathematicalmodels that provide realistic
reactions of an individual and his environment. In
addition, they have DNV (Det Norske Verit
as)
certificate confirming the possibility of their use
during specialized courses for crews of merchant
ships).
Figure2. Simulators used for researches [Gil,Śniegocki
2015].
2.1 Objectivesandcourseofthestudies
According to the assumptions, each of the tested
individuals performed the exercise three times
manoeuvringtothestarboard.Allparticipantswere
operatinginthesamewaterareaandwiththesame
hydro‐meteorologicalconditions,whichareshownin
Table1.
Table1. Hydro‐atmospheric conditions simulated in the
researchedwaterarea[Gil,Śniegocki,2015].
_______________________________________________
Depth 100montheentireworkingarea
Windspeedand320°‐5kn(2°B)
direction
SeaState 1–Calm(rippled),0‐0,5m
Currentspeedand none
direction
Airtemperature21°C
Visibility 8–verygood–10Mm
CloudsNoclouds,clearvisiblesky
_______________________________________________
Toaccomplishtheta
sktheBULKC06Lmodelwas
selected–itisafullyloaded,215‐meterbulkcarrier
with a displacement of more than 60 000 tonnes
[KongsbergMaritime, 2015]. In ordertoshortenthe
exercise, it was beginning with a course 000 ° with
the settings of an engine order tel
egraph (E.O.T.)
ʺFULLSPEEDAHEADʺandthemaximumspeedof
theship‐15.9knots.
Inthedesignedtaskamanwasinitiallyheldata
distanceof380mfromthebowofthevesselsothat
thevessel,whenfloatingpastthemanʹsposition,was
positionedparalleltoit
.Theappliedsolutionmadeit
possible to simulate a situation in which the victim
has just fallen overboard. Information about the
initiationofthePOBalarmwassenttosteeringunits
with radio communications at a time when the
survivorhaspassedtherighttraverseoftheship.In
pra
ctice, this meant the circumstances in which the
distancefromtheinitialpositionofthevesseltothe
POB position was about 500 m. Simulations were
constructedinsuchawaysothatthemanplacedin
the water was not affected by the drift of wind,
making his position constant during the entire
manoeuvre.
Since none of the tested persons had previous
experience with the applied model of the ship, the
firsta
ttemptwascarriedoutinaccordancewiththe
description of the manoeuvre contained in the
ʺInternational aviation and maritime search and
rescueguideʺ(IAMSAR).
Figure3.„AndersonTurn”(a)and„WilliamsonTurn(b)as
perIAMSAR[InternationalMaritimeOrganization,2008].
IfasshowninFigure3,theguidecontainingthe
guidelines of the International Maritime
Organisation(IMO)suggeststhefollowingsequence
of action in case of an immediate manoeuvre – the
ʺAnderson Turnʺ [International Maritime
Organization,2008]:
Move the rudder on the side from which a man
fell;
Afterreachingachangeinthecoursebytheva
lue
of250°,settherudderintheʺzeroʺposition;
In the case of theʺTurn Williamsonʺ manoeuvre
IAMSAR recommends the following procedure
[InternationalMaritimeOrganization,2008]:
Move the rudder on the side from which a man
fell;
Afterreachingachangeinthecoursebytheva
lue
of60°,settherudderontheopposite side;
Whenthecoursereachesthevaluelowerby20°
thanthecountercourseascomparedtothe
initialsettingsofthevessel,settherudderinthe
ʺzeroʺposition.
Both at the first and the second manoeuvre, the
init
iationandintensityoftheprocessofstoppingof
the vessel was carried out at the discretion of an
individual performing the exercise. During the
567
researchfurthertestswerecarriedoutonthebasisof
previous experiences of participants. At the time of
theexercisesexecution,thoseinvolvedintheexercise
modified the course of the manoeuvre‐in
accordance with their convictions based on the
analysis of the chart showing the route of the first
test. Each a
pproach was considered as completed
when the navigated vessel had reached a speed of
lessthantwoknots.Thatcriterionwasadopteddue
totherecognitionthatsuchspeedlevelwassufficient
forsafedeploymentofarescueboat.
During the exercise, at the position of an
instructor(atechnicaloperatorofthesimulator)they
recordedthefollowing real‐ti
medynamicparameters
ofthevessel,intwosecondintervals:
referencenumberofameasurement;
durationoftheexercise[hh:mm:ss];
course[°];
directionofthewaterroute[°];
speed[knots];
speedchange[knot
s/min];
latitude;
longitude;
heading(HDG)[°];
actualpositionoftherudder[°];
definedpositionoftherudderblade[°];
rateofturn[°/min].
Figure4. Recorded tracks and data. Preview on the
instructorstation.
3 DEVELOPMENTOFTHECOLLECTEDDATA
After collecting all the measurements, it was
necessarytoextractfromthemthoseparametersthat
ultimately were to be used to investigate the
effectiveness of POB manoeuvres. The described
attributes included: execution time of the
ʺWilliamson Turnʺ and theʺAnderson Turnʺ along
withthestoppingofthevessel(t
m),aswellaslatitude
andlongitude(φ,λ)attheendofthemanoeuvre(V
K
<2knots).
3.1 Coordinatestransformation
For the implementation of the charts being the
projectionof obtained final positions of the ship on
the plane, it was necessary to transform the
geographicalcoordinates to Cartesian X, Y. For this
purpose they applied the relationship between
rectangularcoordinatesandtheircounterpartsonthe
ellipsoidofrevolution(spheroid).Inordertoincrease
the accuracy of the obtained calculat
ions they used
theGauss‐Krügercoordinatesystemreproductionfor
the axial meridian 006°, based on the following
relationships[Spechtetal.2013]:
 






2
46
3224
5242224
sin cos
2
sin cos 5 9 4
24 720
sin( ) cos 61 58 270 330 445
SB L
xkR B B
R
LL
BBt
BB tt t





 






3
3224
5
524222
cos( ) cos 1
6
cos 5 18 14 58 13
120
L
yL B B t
L
Btt t




After obt
aining the rectangular coordinates they
calculatedtheaveragepositionofallmeasurements
and the distance separating the ship from the
positionofmanattheendofthemanoeuvre.Forthis
purposetheyappliedthepropertiesofa loxodromic
triangle‐andsimilarcalculationswereperformedfor
thedataobtained.
Figure5.Spreadofreceivedfinalpositionswithrespectto
the POB position for the Williamson Turn and the
AndersonTurn.
3.2 Breakdownoftheresultswithrespecttoimplemented
trials
Followingthealignmentofgeographicalcoordinates
they analysed the results as divided into three
attempts. They sorted and counted the average
values of the distance (from the POB position) and
the duration for individual approaches. The
summaryofresultsispresentedinTable2.
568
Table2.Averagevaluesofmanoeuvrestimeanddistances
toamanforindividualtests.
_______________________________________________
WilliamsonTurn AndersonTurn
______________________________________
Attempt DPOB[m]tm[min] DPOB[m]tm[min]
_______________________________________________
1.608.84 18.12  344.09 11.43
2.542.45 15.56  267.65 10.79
3.606.50 14.73  263.45 11.24
_______________________________________________
According to the above a table can be seen a
gradualshorteningofthedurationofmanoeuvreof
each successive trial‐especially in the case of the
ʺWilliamsonTurn.ʺThisfactresultsfromapractical
familiarization of a tested individual with the
handling/manoeuvring characteristics of the vessel
and the modification
of the generalized scheme of
actionbasedontheanalysisoftheearliermanoeuvre.
Figure6.Graphwithaveragedurationofthemanoeuvres
inparticularattempts
Basedontheresultsconcerningthecourseofthe
ʺWilliamson Turnʺ it can be concluded that the
smallestdistancesfromtheaveragepositionandthe
positionofmanwereobtainedatthesecondattempt.
In the case of theʺAnderson Turnʺ the approach
resultsinboththe2ndand3rd
trialwereveryclose,
but definitely different from the results obtained in
the1stapproach.Thismayresultfrom theanalyses
of the previous manoeuvre carried out by students,
gettingpracticallyacquaintedwiththepropertiesofa
marine vessel and the implementation of
amendments to the generalized schemeof
manoeuvreproposed
intheIAMSAR.
The values obtained atthe third time (especially
in the case of theʺWilliamson Turnʺ) being smaller
than at the first approach, while higher than at the
second, may result from probably too much
confidenceemerginginthehelmsmanwhonoticeda
significant improvement as compared
to the
manoeuvre implemented as the first one. This may
resultinareductionofconcentration,beingatease,
andultimatelygettingworseresults.
Visible differences between the results obtained
withbothmanoeuvres areslightlydifferent‐in the
case of theʺ270° turn” data are similar. This is
probably due to
the simplicity of executing of the
saidmanoeuvreandthelimitedabilitytomodifythe
approachitself.InthecaseoftheʺWilliamsonTurn,ʺ
which is assumed to be a longer manoeuvre, the
differences in the results obtained in subsequent
attemptsaremuchmoreapparent.
4 CALCULATIONOFEFFICIENCYOF
POB
MANEUVERS
Inordertotesttheeffectivenessofthemanoeuvre,it
was necessary to allow the indicator to quickly
determinewhetheramanoeuvrehadbeenperformed
successfully. While attempting to define a model
describingtheindicator, the following requirements
wereaccepted:
zero‐dimensionality;
obtainedvalues≤1;
taking
intoaccountthetimeofamanoeuvre(tm);
takingintoaccountthedistanceoftheshipfroma
POBposition(DPOB);
beingshotmanoeuvreʺmanoverboardʺ;
grasping the essence of the POB manoeuvre
performance.
Given the above assumptions, it was decided to
translateallthevalues
obtainedtooneform.Because
tm is not the total time that separates the
commencementofaʺPOBʺtypemanoeuvrebyaship
untilrescuingaman,itwas necessary to determine
the total duration of the manoeuvre (t
c). For this
reasonallitscomponentsweresummed:
cmlr
tttt

where
t
m=durationof“POBmanoeuvre”[min],
t
l=durationofrescueboatlaunching[min],
t
r=timerequiredtopassDPOBdistancebytherescue
boat[min].
1852
60
P
OB
r
r
D
t
V
 
where
V
r=rescueboatspeed[kn].
Due to the assumption regarding the occurrence
of the worst conditions‐in accordance with the
chartspresentedintheIAMSAR‐theyassumedthe
averagesmallestamountoftimeamancan survive
inthewaterwiththelowesttesttemperature,namely
60minutes.
Figure7. Effects of the hypothermia in different water
temperatures[11].
569
AccordingtotheacceptedmodeltheGindicator
wascalculatedforallmeasurementsmade.Themost
effectivemanoeuvre made by the student using the
ʺAndersonTurnʺreachedavalueof0.263;whilethe
least efficient 0.341. For theʺWilliamson Turnʺ the
bestvalueobtainedwas0.380;andtheworst
0.571.
Figure8.GFactorforWilliamsonandAndersonTurn.
Both from the above graph and the values
obtaineditfollows that inthe caseofanimmediate
rescue action, when the moment of a man falling
overboard was observed by a person leading the
vessel or the crew on the bridge, the most effective
method of approach to a man
from among those
describedaboveistheʺAnderson Turn.ʺItisworth
noting that even the most effective test performed
while applying theʺWilliamson Turnʺ was only
slightlybetterthantheleasteffectiveapproachusing
theʺsingleturnʺmethod.
Astheobtainedresults show,thebiggestimpact
ontheeffectiveness
ofthemanoeuvrehasthetimeof
its execution, and in a lesser extent‐the distance
betweentheshipandtheman.
Anexampleofsuchmeasurementcanmanoeuvre
in whichwithʺTurnWilliamsonʺ achieved
D
POB = 517.9 m (which is a value close to the
average).Gtotalled0.344;whichisthesecondofthe
best values obtained. This was possible due to the
favourableoveralltimemanoeuvret
c=20.63min.
An example of such measurement can be a
manoeuvre in which an application of the
ʺWilliamsonTurnʺhelpedtoachieveD
POB=517.9m
(which is a value close to the average). G totalled
0.344;whichisthesecondofthebestvaluesobtained.
Thiswaspossibleduetothefavourableoveralltime
ofthemanoeuvret
c=20.63min.
5 CONCLUSIONS
Basedonthecarriedoutsurveyitwasfoundthatthe
greatest impact on the effectiveness of the
implemented manoeuvre is that of the acquired
experience.Fromthepresentedchartsitfollowsthat
with each new attempt, the duration of the
manoeuvre was reduced. The
variable that most
significantlyaffectsthepresentedefficiencyratioofa
manoeuvreist
m,soonecanconcludethateverynext
POBmanoeuvreperformedbythesamepersonwas
moreeffective.
The research shows that, especially in good
hydro‐atmosphericconditions, the distance between
amanandastoppedvesselislessimportantthanthe
durationofthewholemanoeuvre.Thisisbecauseit
isfasterandmorepreciseto reachtheposition ofa
manwitharescueboatratherthantoapproachhim
byaship.Itwasnottestedhowtheaforementioned
situationaffectedtheeffectivenessofthemanoeuvre
incircumstances in which, due totheoccurrenceof
adverseweatherconditions,
thereisaneedtoshield
themanandtherescueboatfromtheseawaves.
The presented results also demonstrated the
superiority of theʺAnderson Turnʺ over the
ʺWilliamson Turnʺ during the implementation of
aforementionedmanoeuvresintheimmediaterescue
operation. This is due to the simplicity of theʺ
270°
turn,ʺwhichtranslatesdirectlyintoitsdurationand
also efficiency. The values obtained indicate the
saving of time of up to 5 minutes when the
ʺAnderson Turnʺ is selected as a variant of the
approach to the man who fell overboard, with
respecttothemeanvaluesofall
thetimesobtained
bybothmethods.Inasituationwhenthedrowning
man is overboard, this value is a huge gain for the
benefitofallinvolvedintheaction.
BIBLIOGRAPHY
Gil M. &Śniegocki H., „Efektywność manewru «Pętla
Williamsona»”,Logistyka4/2015,ss.3381‐3392,2015.
European Maritime Safety Agency, „The world merchant
fleet in 2005 : Statistics from Equasis”, European
MaritimeSafetyAgency,2005‐2013.
International Maritime Organization i International Civil
Aviation Organization, IAMSAR Manual : In‐
ternational Aeronautical and
Maritime Search and
RescueManual.London/Montreal:IMO/ICAO,2008.
International Maritime Organization, „Resolution
MSC.218(82): Adoption of amendments to the In‐
ternational Life‐Saving Appliance (LSA) Code”. In‐
ternationalMaritimeOrganization,2006.
International Maritime Organization, LIFE‐SAVING
APPLIANCES: 2003 Edition, Combined edition.
London:InternationalMaritimeOrganiza‐tion,2003.
International Maritime Organization, Red.,
SOLAS:
Consolidated Edition, 2004. London: International
MaritimeOrganization,2004.
Kongsberg Maritime, „Description of Ship Model :
BULK06L:Version14”.KongsbergMaritime,2013.
Specht C., Szot T., & Specht M., „Badanie dokładności
personalnych odbiorników GPS w pomiarach
dynamicznych”, Technika Transportu Szynowego, nr
10,ss.2547–2555,2013.