93
1 INTRODUCTION
Theincreasingtrafficintensityalongwithvesselsizes
and higher speeds continue to draw researchers
interestandefforttotheproblemsofmarineshipping
safety andeffectiveness.Thisrefers,inparticular,to
areaswherevesseltrafficisdense.Oneofthewaysto
reducethenumberofaccidentsintheseareas is the
i
ntroduction of additional principles of traffic
management. An example of these are traffic
separation schemes (TSSs), regulating vessel traffic.
TSSsarecomposedoftrafficlanes,eachindicatingthe
directionsoftrafficflowswithintheTSS‐coveredarea.
The increased, organized vessel traffic makes it
difficulttoa
pplytheclosestpointofapproach(CPA),
one of the fundamental criteriaof navigational
safety. The shipʹs domain may be an alternative for
theCPAcriterion.Theshipdomainisdefined asan
areathatthenavigatorshouldmaintainfreeofother
vessels (Fuji & Tanaka 1971). The most frequent
ly
reporteddomainsaretwo‐dimensionaldomainswith
defined shape and size. It is possible to include the
third dimension‐domain depth. Its introduction
permits to take account of safe distances to
underwater dangers. A number of factors affecting
thesizeandshapeofthedomaincausesdifficultiesin
itsdetermination.Thesefact
orscomprisethetypeof
area (open, restricted), types and sizes of vessels,
hydro‐meteorological conditions. The human factor
should also be taken into account. The relevant
research found in the literature deals with both
restrictedandopenareas.TSSsarespecificareasdue
to intensive vessel traffic. For thi
s reason they are
classified as restricted areas, even if they often lack
physicalboundariesofthemanoeuvringarea relating
todepth,lengthorwidth.Ananalysisofthecriteria
for assessing navigational situations in these areas
seemstobe essentialdue tothat specificity.Forthis
reasonTSSsmayrequirecrit
eriasomewhat different
from those used in both open and restricted areas
wherenavigation is restricted by physically
available manoeuvring area. This concerns traffic
lanesaswellasprecautionaryareasinTSS.
Ship Domain as a Safety Criterion in a Precautionary
Area of Traffic Separation Scheme
Z.Pietrzykowski&J.Magaj
M
aritimeUniversityofSzczecin,Szczecin,Poland
ABSTRACT: The ship domain is one of the criteria for navigational safety assessment. It is particularly
importantinrestrictedareaswithhighintensitytraffic,wherethecriteriaofclosestpointofapproach(CPA)
andtimetoCPAaredifficulttoapply.ThisresearchcontinuestoexamineshipdomainsinTrafficSeparation
Schemes (TSS). We have analyzed p
recautionary areas established within TSSs in connection with changed
arrangementsofvesseltraffic.Besides,wehavedefinedshipdomainsinaprecautionaryareaofa specificTSS,
andcomparedthemtodomainsofvesselsproceedingalongtrafficlanes.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 11
Number 1
March 2017
DOI:10.12716/1001.11.01.10
94
2 TRAFFICSEPARATIONSCHEMES
2.1 TSScharacteristics
Trafficseparationschemesareestablishedtoenhance
the safety of navigation by introducing specific
principles of vessel traffic organization in order to
keepitundercontrol.InaccordancewiththeSOLAS
Convention SOLAS (SOLAS 1974),, the IMO is the
only international institution that
issues guidelines
andrulesforTSSsystemsattheinternationallevel.
Principles for the establishment and organisation
of TSS have been laid down in the IMO Resolution
ʺGeneral provisions on ships routingʹ(IMO 1985)..
These systems, by separating the flows of vessel
traffic,vitallyincreasethelevelofsafetyof
lifeatsea,
navigational safety and effectiveness, and
environment protection. The designated traffic lanes
encompassone‐wayvesseltrafficwithinthescheme.
TSSareasaregovernedbyIMOregulations.Rule10
of the International Regulations for Preventing
Collisions at Sea (COLREGs 1972) defines the
behavior of ships navigating in these
systems. This,
however,doesnotrelieveshipsfromcomplyingwith
the other rules of the Collision Regulations. An
example of a traffic separation scheme is shown in
Figure 1a. Traffic lanes are often defined by
conventionalboundaries.Thismeansthatabreachof
the lane boundary does not lead to a
direct risk of
grounding or collision with a shore structure. In
accordancewith the definition, the traffic laneisʹan
areawithindefinedlimitsinwhichone‐waytrafficis
established(IMO1985).
a)
b)
Figure1.Trafficseparation Scheme: a)separationoftraffic
at a crossing; b) precautionary area at a junction, with
recommendeddirectionsoftrafficflow(IMO1985)
Some TSS areas comprise precautionary areas
(Figure1b).Precautionaryareaisdefinedasʹarouting
measure comprising an area within defined limits
where ships must navigate with particular caution
andwithinwhichthedirectionoftrafficflowmaybe
recommendedʹ(IMO1985).
Precautionaryareasmayalsobeintroducedforthe
termination
ofanysingleroute.
Traffic separation schemes are established
throughouttheworld.OntheBalticSeatheseinclude
TSSAdlergrund,TSSInBornholmsgate,TSSNorthof
Rügen, TSS Słupska Bank, TSS Gdansk Bay
(Pietrzykowskietal.2015).
2.2 TSSBornholmsgat
TSS Bornholmsgat is one of the traffic separation
schemes
ontheBalticSea(IMO2005).Densetrafficof
different vessel types and sizes is observed, moving
from the Danish Straits and the Kiel Canal to the
eastern Baltic Sea and in opposite directions (Figure
2a).Withinthatareatherearesixtrafficlanesandone
precautionaryarea.
The number of
vessels that transited the area in
2011 was over 64000(HELCOM 2011). Figure 2b
illustrates trails of vessels registered in the AIS sys‐
temduringfourdaysofJune2011.
a)
b)
Figure2. TSS Bornholmgate: a) traffic lanes and
precautionary area; b) vessel tracks recorded in the AIS
systemwithinfourdaysofJune2011
95
Table1liststhenumbersofshipsregisteredinthe
sameperiodoffourdaysineachtrafficlaneandthe
precautionaryarea.
Table1. Numbers of ships in the examined TSS
Bornholmsgatareas.
______________________________________________
AreaShips
______________________________
Allbulkcarrier
______________________________________________
trafficlane1298215
trafficlane2313235
trafficlane310784
trafficlane412489
trafficlane5213145
trafficlane6175124
precautionaryarea 618443
_____________________________________________
Additionally, the most numerous group of bulk
carriers was considered. The main flows of vessel
trafficwereobservedalongtrafficlane2–trafficlane
4trafficlane3–1.
3 THERESEARCH
3.1 Methodofshipdomaindetermination
Thereareanumberofmethodsfordomaindetermi‐
nation (Fuji
& Tanaka 1971, Zhao et al. 1993,
Rutkowski1998,Smierzchalski&Weintrit1999,Zhu
et al. 2001, Pietrzykowski 2008, Pietrzykowski &
Uriasz 2009, Wang et al. 2009, Wielgosz & Pie‐
trzykowski 2012, Hansen et al. 2013, Wang 2013).
These are mainly analytical, statistical and artificial
intelligence methods. By statistical and
artificial in‐
telligencemethods,theshipdomainisdeterminedon
the basis of ship trails obtained from simulations
performed on navigational simulators, or real data
studies, in which shipsʹ positions in the examined
area are recorded. Currently such data are acquired
fromtheAIS.Themethodusedforshipdomainde
‐
termination in this study includes the following
stages(Pietrzykowski&Magaj2016a,b):
1 Acquisitionof data on ship positions in the
analyzedarea(AIS)
2 Transformationofshippositionsfromtruemotion
to relative motion display, with the coordinate
system centre associated with the shipʹs AIS
antennaposition.
3 Determinationofvesseltracksdensity.
4 Determination of the ship domain boundary for
shipʹssectors(inthisstudy72sectorsaredefined).
5 Approximation of the domain boundary to a
geometricform(2Ddomain).
6 Stages3,4and5aredepictedinFigure3.
This method has been
used for the ship domain
determination in the precautionary area of TSS
Bornholmsgat.
3.2 Theshipdomainintheprecautionaryarea
The determination of ship domains in the
precautionaryareaTSSBornholmsgatewasbasedon
AISdatarecordedwithinfourdays(cf.Table1).All
vesselsthatwerefoundwithin
theanalyzedarea and
period were included. Bulk carriers,the most
numerousshipgroup,wereconsideredseparately.
The domains were determined by the method
described in section 3.1, approximating them to
ellipses with semi‐axes A and B (minor and major,
respectively). Thus, domain length and width were,
respectively,2Aand
2B(Fig.4).
a)
b)
0 10 20 30 40 50 60 70 80 90
1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
distance x37 [m]
density
c)
-2000 -1000 0 1000 2000
-2500
-2000
-1500
-1000
-500
0
500
1000
1500
2000
2500
x [m]
y [m]
without rotation
with rotation
boundary points
ship domain
Figure3.Themethodforshipdomaindeterminationbased
onshiptracks:a)methodoffirstmaximumforthedomain
boundary determination for a given sector; (c)
approximationofdomainboundary(ellipse)(Pietrzykowski
&Magaj2016a,b).
96
a)
b)
Figure4.Theshipdomainintheprecautionaryarea: a)ship
trackdensity(allships),b)shipdomains (all vessels, bulk
carriers)
The domain parameters are given in Table 2.
Relative dimensions of the domains refer to mean
shiplengths,136mand127mforallshipsandbulk
carriers,respectively.
Table2. Parameters of ship domain in the precautionary
areaofTSSBornholmsgat.
______________________________________________
ParameterShips
______________________________
Allbulkcarrier
______________________________________________
Length(2B)[m]27022880
Breadth(2A)[m]11301136
Length(2B)[L]*19.822.6
Breadth(2A)[L]*8.28.9
_____________________________________________
*relativelength/breadthinshipslengthL.
Theresultsshowavisiblechangeinthelengthto
breadth ratio: increased length, reduced breadth. At
thesametime,comparedtotestresultsfordomainsof
vesselsinareasrestrictedphysicallybytheavailable
manoeuvring area (Hansen et al. 2013), the
determined lengths and breadths of domains are
larger.
The ratiosof lengths to breadths are also
different.Inonecase,theratioapproximatelyequals
2,intheother(precautionaryarea)itis2.4to2.5.
Further,wecomparedthesizesofdomainsinthe
precautionary area to those in the adjacent traffic
lanes.
4 COMPARATIVEANALYSISOFDOMAINSIN
TSS
4.1 Domainsonthetrafficlanesandintheprecautionary
area(PA)
Ship domains on all traffic lanes were determined,
then compared to the domains in the precautionary
area.TheresultsareshowninFigure5.
a)
b)
Figure5. Ship domains in the distinct TSS Bornholmsgat
areas:a)allvessels);b)bulkcarriers
A large diversity of domain lengths can be seen,
while the breadths are comparable. Similar domain
breadths can be explained by equal widths of all
trafficlanes(2.7Nm). Supposedly, asimilardomain
breadthintheprecautionaryareastemsfromthefact
97
thatforthetrafficlanesgoinginandoutofthePAthe
domainbreadthsarealsosimilar.
Thesmallestvaluesofdomainlengthswerefound
in the PA and traffic lanes 1, 2 and 5. These waters
weretransitedbythelargestnumberofvessels(cf.
Table1).Traffic
lanes2and5arepartoftheshipping
route from the Eastern Baltic towards the Danish
StraitsandtheKielcanal.
In this connection the ship domains for traffic in
bothdirectionsweredeterminedandcompared.
4.2 Shipdomainsforselectedtrafficflows
We analyzed vessel traffic flows from
the Eastern
Baltic towards the Danish Straits and the Kiel canal
andintheoppositedirection.Tothisendthedomains
were determined for vessels passing the route
runningthrough trafficlane 2,PA andtraffic lane5
(stream 1) and, separately, traffic lane 6, PA and
trafficlane1(stream
2).Thedefinedshipdomainsare
showninFigure6.Forcomparison,domainsforeach
areaareadditionallymarked.
There is a clear difference of domain lengths for
both streams. The domain length for stream 1 is
approximately about half shorter than for stream 2.
Thismaybecausedbe
thevesseltrafficflowfromthe
OresundtowardstheEasternBaltic(trafficlane4‐PA
‐trafficlane1)intersectingtheanalysedstream1.
5 CONCLUSIONS
Thisarticlepresents theresults of researchinto ship
domains in the precautionary area of TSS
Bornholmsgat.Domainsoftwogroupsofshipswere
taken
into consideration. The domains had similar
lengthsandbreadths.Theresultsshowaclearchange
in thedomain length to breadth ratio for both open
areas and areas restricted by the available
manoeuvringarea.
The determined domains were then compared to
the domains defined for all traffic lanes of the
analyzed
TSS.Thedomainbreadthswerefoundtobe
similar, while domain lengths differed considerably.
Thesmallestdifferencesconcernedthosetrafficlanes
withthelargestnumberofvesselsrecorded.
The authors also examined ship domains for
identified vessel traffic streams. Large differences in
domainsizeswereobserved.Itwaspointedout
that
one of the reasons may be the interaction of two
trafficflows.
The determined domain data can be used for
analysisandassessmentofanavigationalsituationby
vesseltrafficservices.
It seems purposeful from the perspective of ship
controltodefineasingledomain,e.g.foraparticular
type
ofshipinaspecifictrafficflow.Furtherstudies
are planned in this field, also for other traffic
separationschemes
a)
b)
Figure6.Shipdomainsforselectedtrafficflows intheTSS
Bornholmsgat: a) stream 1 (all vessels); b) stream 2 (all
ships)
REFERENCES
COLREGs 1972. Convention on the International
Regulations for Preventing Collisions at Sea [with
amendments2009],IMO
Fuji Y., Tanaka K. 1971.Traffic Capacity. Journal of
Navigation.24,(543‐552).
Hansen M., Jensen T., Lehn‐Schiøler T., Melchild K.,
Rassmussen F. & Ennemark F. 2013. Empirical ship
domainbasedonAISdata,Journal
ofNavigation,no66.
HELCOM 2011. Helsinki Commission Baltic Marine
Environment Protection Commission, Report on
shippingaccidentsintheBalticSeaareaduring2011
http://www.helcom.fi/Lists/Publications/Annual%20rep
ort%20on%20shipping%20accidents%20in%20the%20Bal
tic%20Sea%20area%20during%202011.pdf.
IMO 1985. Resolution A.572(14): General Provisions on
ShipsʹRouteing
IMO2005.ResolutionA.977(24):ShipsRouteing
98
Nazeri Z., Donohue G., Sherry L. 2008. Analyzing
RelationshipsBetweenAircraftAccidentsandIncidents:
A Data Mining Approach, Proceedings International
ConferenceonResearchinAirTransportation (ICRAT‐
2008),Fairfax,VA,
PietrzykowskiZ.2008.Ship’sfuzzydomain–acriterionof
navigational safety in Narrow Fairways, Journal of
Navigation,no61.
Pietrzykowski
Z., Uriasz J. 2009. The ship domain – a
criterion of navigational safety assessment in an open
seaarea,JournalofNavigation,no62,2009.
Pietrzykowski Z., Wołejsza P., Magaj J. 2015. Navigators’
Behavior in Traffic Separation Schemes, TransNav, the
International Journal on Marine Navigation and Safety
ofSea
Transportation,Vol.9,No.1inMarch2015,pp.
123‐128.
Pietrzykowski Z., Magaj J. 2016. Ship Domains in Traffic
Separation Schemes, Scientific Papers Maritime
UniversityofSzczecin,no45(117),pp.143‐149.
PietrzykowskiZ.,MagajJ.2016.AnalysisofShipDomains
in Traffic Separation Schemes, Scientific Journals
MaritimeUniversityof
Szczecin2016,48(120),pp.88‐95.
Rutkowski G. 1998. Ship domain in navigation on water
areas difficult for navigation, Scientific Papers of the
Faculty Navigation, Maritime Academy in Gdynia,
Volume6
SOLAS1974.InternationalConventionfortheSafetyofLife
atSea,IMO.
Smierzchalski,R,Weintrit,A.1999.Domainsof
navigational
objectsasanaidtorouteplaningincollisionsituationat
sea. In: Proc. of 3rd Navigational Symposium,Gdynia,
265–279,(inPolish).
Wang N., Meng X., Xu Q., Wang Z. 2009. A unified
analytical framework for ship domains, Journal of
Navigation,no62,.
WangN.,2013.Anovelanalyticalframework
fordynamic
quaternionshipdomains,JournalofNavigation,no66.
Wielgosz M., Pietrzykowski Z. 2012. Ship domain in the
restrictedarea–analysisoftheinfluenceofshipspeed
on the shape and size of the domain, Scientific Papers
MaritimeUniversityofSzczecin,no30.
ZhaoJ., WuZ.,Wang
F.1993.Commentsofshipdomains,
JournalofNavigation,no46,.
ZhuX.,XuH.,LinJ.2001.Domainanditsmodelbasedon
neuralnetworks,JournalofNavigation,no54.
