Faculty of Science and Technology
Department of Engineering and Safety

Risk of Ship Collision in the Barents Sea in 2030

Academic Year 2014

—  
Jan Børre Hansen Balto
TEK-3901 Master’s thesis in Technology and Safety in the High North
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

  II  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

PREFACE AND AKNOWLEDGEMENT

This master thesis is the culmination of a two year master program in Technology and Safety in the
High North at UiT, The Arctic University of Norway. The thesis is an individual project and is
equivalent to 30 ECTS. The goal is for the student to gain in-depth knowledge and competence
within a selected area in the field of technology and safety, relevant for the high north. Learning
outcome through the project is to improve the student’s ability to independent engineering and
research work, and provide training in planning of projects, systematic processing of information
and report writing.
The topic came during a meeting with Aker Solutions Tromsø in December 2013. Without any
special background in maritime transport, Aker Solutions made me realize and inspire me to this
theme as it is interesting and relevant in the coming years with respect to the development of the
petroleum activity that’s most likely are going to take place in the Barents Sea.
During the course of this thesis and preliminary work a few people have made significant
contributions, and a few acknowledgements are in order. First of all a great acknowledgement to my
thesis advisor, Professor Javad Barabady. All the people at Aker Solutions in Tromsø, and
especially Safety and Environment Engineer Sigve Daae Rasmussen, for giving me a very
inspirational stay at Aker Solutions office, and with the help to navigate my way through a large
amount of documents in the initial phase of the thesis. A big gratitude goes to Trond Langemyr at
Kystverket for good help and giving me aceess to the extended version of the simulation tool used
in this thesis. I will also thank PhD student Jaap Van Rijckevorsel for his infectious commitment
and help with the structure of the task.
A big thank you goes to my class for all the good experiences, friendships, and hardships prevailed
together, you know who you are. I am grateful to my girlfriend Ragna for always supporting and
encouraging me along the thesis process and to my lovely daughter Alma for constantly reminding
me about what is actually important in life.

Tromsø 1st of June 2014

Jan Børre Hansen Balto

  III  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

  IV  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Master thesis – TEK-3901 Master Thesis in Technology and Safety in the High North

UiT - The Arctic University of Norway ABSTRACT OF THE MASTER‘S THESIS

Departement of Engineering and Safety Availability:
Faculty of Science and Technology Open

Author: Jan Børre Hansen Balto

Title: Risk of Ship Collisions in The Barents Sea in 2030

Degree Program: Technology and Safety In The High North

Subject: TEK-3901 Master Thesis 30 ECTS

Supervisors:
Professor Javad Barabady – UiT – The Arctic University of Norway
C&T Manager Northern Norway Øystein Mikelborg – Aker Solutions

Abstract:
This thesis evaluates the risk of ship collisions in the Barents Sea in 2030 between three future
scenarios; Minimum, Basis and Maximum Scenario. IWRAP Mk2 program is utilized to make the
calculations. Automatic Identification System (AIS) data of 2013 is used to parameterize current
traffic density, while the increased traffic in the different scenarios is derived from an analysis of
multiple sources, including Rystad Petro Foresight, government documents and reports from DNV.

The petroleum production in the North Sea is expected to decline, while exploration and production in
the Northern part Norway is expected to increase. This will lead to that the Barents Sea will be a major
contributor to oil and gas production, instead of the North Sea and the southern Norwegian Sea
towards the end of 2030s.

The petroleum industry is on its way north to an area that earlier mainly has been associated with high
fishing activity, but may now be more dominated by larger supply vessels. This change will cause an
increase in ship traffic in the area, and the probability of ship collisions may therefore be elevated. The
issues discussed in this report are important for the industry, and necessary for predicting the future
risk picture in the Barents Sea. It is vital to idenfity the future risk of ship collision with regards to the
increase in ship traffic due to the potential consequences with respect to the harsh and vulnerable
environment and lack of infrastructure in the northern part of Norway.

This thesis will investigate how the probability of ship collision change, and also identify the risk of
ship collisions in the Barents Sea within the different scenarios of petroleum development. There are
five types of collision between ships which are taken into account in this thesis, these are; Head on
collsion, crossing collsion, overtaking collsion, bend collision and merge collsion.

The thesis will answer the research problems regarding to how the environmental conditions in the
Barents Sea are, how the increased offshore-related traffic increase the probability of ship collision in
the Barents Sea in 2030, and how the risk of ship collisions change between the three scenarios.

The results show that there will be significantly differens in the likelihood for ship collisions in the
three scenarios. The total likelihood for minimum scenario is 5,80E-04 incidents/year, while the
likelihood in basis and maximum scenario is calculated to 1,8E-03 and 1,75E-03. The final leg into
`Polarbase` (Hammerfest) is the leg in all scenarios that will have the greatest likelihood for ship
collisions, and will also be the most critical leg with respect to the high density of ships in it, despite
its short length. The ship type that will be the biggest contributer to ship collisions is both support
ships and crude oil tankers, these collisions will be by type; head on collision and overtaking collision.

  V  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

A critical situation will occur in the Barents Sea if a ship collision takes place, and especially
collisions with crude oil tankers with its chemicals. This will put great demand on the oil spill
management in the region.
Date: Language: Number of pages:
June 1st 2014 English 60

Keywords: Ship, Collision, Marine Accident Modelling, The Barents Sea, IWRAP Mk2

  VI  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

PREFACE  AND  AKNOWLEDGEMENT  ..................................................................................................  III  

ABBREVIATIONS  AND  NOMENCLATURE  ..........................................................................................  IX  
FIGURES  AND  TABLE  LIST  .....................................................................................................................  XI  
1   INTRODUCTION  ...................................................................................................................................  1  
1.1   BACKGROUND  AND  RESEARCH  PROBLEM  ........................................................................................  1  
1.2   AIM  OF  THE  THESIS  ...................................................................................................................................  1  
1.3   RESEARCH  QUESTIONS  ............................................................................................................................  2  
1.4   SCOPE  .............................................................................................................................................................  2  
1.4.1   Minimum  Scenario  ................................................................................................................................................  2  
1.4.2   Basis  Scenario  .........................................................................................................................................................  2  
1.4.3   Maximum  Scenario  ...............................................................................................................................................  3  
1.5   LIMITATION  AND  ASSUMPTIONS  .........................................................................................................  4  
1.6   THESIS  OUTLINE  ........................................................................................................................................  5  
2   LITERATURE  REVIEW  ........................................................................................................................  7  
2.1   DESCRIPTION  OF  ANALYSE  AREA  .........................................................................................................  7  
2.1.1   Description  of  the  Barents  Sea  .........................................................................................................................  7  
2.1.2   Physical  Conditions  ...............................................................................................................................................  7  
2.1.3   Existing  and  future  fields  in  the  Barents  Sea  ..........................................................................................  11  
2.2   MARITIME  TRAFFIC  ................................................................................................................................  12  
2.2.1   Russian  Traffic  .....................................................................................................................................................  13  
2.2.2   Description  of  Activity  ......................................................................................................................................  13  
2.2.3   Traffic  Development  2005-­‐2009  .................................................................................................................  15  
2.2.4   Ship  Traffic  In  The  Northeast  Passage  .......................................................................................................  15  
2.3   SHIP  COLLISION  ........................................................................................................................................  16  
2.3.1   Types  of  Accidents  .............................................................................................................................................  16  
2.3.2   Main  Causes  of  Ship  Collision  ........................................................................................................................  17  
2.3.3   Concepts  of  Risk  and  Risk  Analysis  .............................................................................................................  19  
2.3.4   Collision  Avoidance  ...........................................................................................................................................  20  
3   RESEARCH,  METHODS  AND  MATERIAL  ....................................................................................  23  
3.1   METHOD  .....................................................................................................................................................  23  
3.2   DATA  COLLECTION  ..................................................................................................................................  23  
3.3   DATA  ANALYSIS  ........................................................................................................................................  24  
3.3.1   IWRAP  Mk2  ...........................................................................................................................................................  24  
3.3.2   Causation  Factor,  Pc  ...........................................................................................................................................  26  
3.3.3   Operating  Vessels  and  Size  .............................................................................................................................  27  
4   RESULTS  OF  ANALYSIS  AND  DISCUSSION  ................................................................................  29  
4.1   MINIMUM  SCENARIO  ..............................................................................................................................  29  
4.2   BASIS  SCENARIO  .......................................................................................................................................  32  
4.3   MAXIMUM  SCENARIO  .............................................................................................................................  35  
4.4   COMPARISON  OF  RESULTS  ...................................................................................................................  38  
4.5   SENSITIVITY  ANALYSIS  ..........................................................................................................................  40  
4.6   FURTHER  DISCUSSION  ...........................................................................................................................  41  
5   CONCLUSION  AND  RECOMMENDATIONS  .................................................................................  43  
REFERENCES  .............................................................................................................................................  45  

  VII  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

  VIII  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

ABBREVIATIONS AND NOMENCLATURE

Abbreviations

AIS Automatic Identification System

ARPA Aumatic Radar Plotting Aid
CO2 Carbon Dioxide
DNV Det Norske Veritas
DWT Deadweight Tonnes
FPSO Floating, Production, Storage and Offloading
GT Gross Tonn
IMO International Maritime Organization
IMR Inspection, Maintenance and Repair
LNG Liquefied Natural Gas
MACHINE Model of Accident Causation Using Hierarchical Influence Network
NCA Norwegian Coastal Administration
NEZ Norwegian Economical Zone
NG Natural Gas
NGL Natural Gas Liquids
PSV Platform Supply Vessel
RNNP Petroleumstilsynet/Petroleum Authority
SOLAS Safety of Life At Sea
TSS Traffic Separation System
VVTS Vardø Vessel Traffic Service

Nomenclature
⁰C Temperature – degrees centigrade
3
Sm Standard cubic meters is a device for volumetric measurement of oil, NG and
gas condensate at standard conditions defined in ISO standard 13443.
P Probability
N Number of accident candidates
Pc Causation factor
Lw Length of the waterway
PGi,j Probability that two ships of this classes i and j collide in a head-on meeting
situation if no evasive manoeuvres are made
Qi n Number of passages per time unit for ship class i moving in direction n
Φ Standard normal distribution function
Average vessel breadth
Ɵ Angle
Fflag Multiplication factor for flag state
Fage Multiplication factor for age of the ship
Fwind Multiplication factor for wind
Fvis Multiplication factor for visibility
Fnav Multiplication factor for the navigation status
EXP(i) Exposure for certain accident type
CASRAT(I, Casualty rate for a certain accident type (i), ship type and ship size
type, size)

  IX  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

  X  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

FIGURES AND TABLE LIST

Figures

Figure 1.1 Relative daily production in Norway from 2013 to 2050…………………………... 1

Figure 1.2 Minimum Scenario…………………………...…………………………...………… 2
Figure 1.3 Basis Scenario…………………………...…………………………...……………... 3
Figure 1.4 Maximum Scenario…………………………...…………………………...………... 4

Figure 2.1 Overview over the Barents Sea…………………………...………………………… 6

Figure 2.2 Lowest air temperature with an annual probability greater than 10^-2……………... 7
Figure 2.3 Occurence (%) of temperature below -1.8 and wind above 10m/s January 1961-
2010…………………………...…………………………...………………………... 8
Figure 2.4 Ice extent in the Barents Sea from 1979 – 2013………………………….................. 9
Figure 2.5 Monthly distribution of polar lows in the Norwegian Sea and the Barents Sea,
which is registered in the Norwegian Meteorological Institute from 2000 to 2012… 9
Figure 2.6 Overview over existing and future fields in the Barents Sea……………………….. 11
Figure 2.7 Traffic  density  between  Vardø  and  Røst  in  the  second  half  of  2010………………… 11
Figure 2.8 Number of passing ships versus cargo volume development……………………….. 14
Figure 2.9 The   MACHINE   model   reflect   the   relationship   between   human   errors,  
hardware  failures  and  environmental  elements…………………………............. 17
Figure 2.10 The  AIS-­‐Sat  I  in  orbit…………………………...………………………….............. 21

Figure 3.1 Methodology used in the thesis…………………………...………………………… 23

Figure 3.2 Fault three for calculating the causastion probability PC for collision……………… 27

Figure 4.1 Minimum scenario in IWRAP…………………………...………………………….. 29

Figure 4.2 Incidents per year for each facility in the minimum scenario………………………. 30
Figure 4.3 Incidents per year, per ship type in minimum scenario…………………………....... 31
Figure 4.4 An Overview on how the different ship collides in Minimum Scenario……………. 31
Figure 4.5 Basis scenario in IWRAP…………………………...…………………………......... 32
Figure 4.6 Incidents per year for each facility in the basis scenario. …………………………... 33
Figure 4.7 Incidents per year, per ship type in basis scenario………………………….............. 34
Figure 4.8 An Overview on how the different ship collides in basis Scenario…………………. 34
Figure 4.9 Maximum Scenario IWRAP…………………………...………………………….... 35
Figure 4.10 Incidents per year for each facility in the maximum Scenario……………………… 36
Figure 4.11 Incidents per year, per ship type in maximum scenario…………………………...... 37
Figure 4.12 An Overview on how the different ships collide in Maximum Scenario…………… 37
Figure 4.13 Comparison of Incidents per year for all scenarios…………………………............. 38
Figure 4.14 Comparisson of all scenarios of likelihood for ship collision for the different
offshore fields…………………………...…………………………...……………… 39
Figure 4.15 Comparison of the different scenarios of the likelihood of ship collision for each
ship type…………………………...…………………………...…………………… 39
Figure 4.16 Comparison of collision type for each scenario…………………………...………... 40
Figure 4.17 Comparison of the sensitivity analysis for each scenario………………………….... 41

  XI  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Tables

Table 1.1 Type of offshore facilities in Minimum Scenario………………………………………... 2

Table 1.2 Type of offshore facilities in basis scenario…………………………...………………… 3
Table 1.3 Type of offshore facilities in maximum scenario…………………………...…………… 4

Table 2.1 Visibility distribution…………………………...…………………………...…………… 10

Table 2.2 Sun activity in different locations…………………………...………………………….... 10
Table 2.3 Maritime Traffic for existing and future fields…………………………...……………… 10
Table 2.4 Ship Traffic along the Norwegian coast linked to exports from Russia…………………. 12
Table 2.5 Traffic pattern for different ship types…………………………...………………………. 12
Table 2.6 Description of the different collisions type…………………………...…………………. 15
Table 2.7 Causation factor, personal…………………………...…………………………...……… 19
Table 2.8 Causational factor, organization……………...…………………………………...………………………… 19

Table 3.1 Causation Factors used in the analysis…………………………...……………………… 27

Table 3.2 Operating vessels and size…………………………...…………………………………... 28

Table 4.1 Collision type in minimum scenario…………………………...………………………… 30

Table 4.2 Collision type in basis scenario…………………………...……………………………... 33
Table 4.3 Collision type in maximum scenario…………………………...………………………... 36
Table 4.4 Comparison of incidents per year for all scenarios, values…………………………...…. 38
Table 4.5 Sensitivity Analysis of minimum scenario…………………………...………………….. 40
Table 4.6 Sensitivity Analysis for basis scenario…………………………...……………………… 40
Table 4.7 Sensitivitiy Analysis for maximum scenario…………………………………………… 40

  XII  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

1 INTRODUCTION
The Master thesis is the final assignment for the two year Master of Science program, Technology
and Safety in the High North at the UiT The Arctic Unversity of Norway. The thesis is independent
and equivalent to 30 ECTS. In the Master thesis, the student should demonstrate knowledge about
the research methodology presented in the program, as well as skills in scientific reflection and
analysis.
In this chapter will the background and research problem be presented along with the aim of the
thesis, research questions, scope, limitation and assumptions, and the thesis outline.

1.1 BACKGROUND AND RESEARCH PROBLEM

The petroleum production in the North Sea is expected to decrease, while exploration and
production in the Northern part Norway is expected to increase (Figure 1.1). This will lead to that
the Barents Sea will be a major contributor to oil and gas production, instead of the North Sea and
the southern Norwegian Sea towards the end of 2030s.
The petroleum industry is on its way north to an area that earlier mainly has been associated with
high fishing activity, but may now be more dominated by larger supply vessels. This change will
cause an increase in ship traffic in the area, and the probability of ship collisions may therefore be
evaluated. The issues discussed in this report are important for the industry, and necessary for
predicting the future risk picture in the Barents Sea. It is vital to idenfity the future risk of ship
collision with regards to the increase in ship traffic due to the potential consequences caused by
harsh and vulnerable environment and lack of infrastructure.

Figure 1.1: Relative daily production in Norway from 2013 to 2050 (Rystad Petro Foreseight 2030, 2012)

1.2 AIM OF THE THESIS

The main objective of the thesis will be to analyze the risk of ship collisions in the Barents Sea.
This thesis will study three future scenarios for development of petroleum activity, developed by
Rystad Petro Arctic. The enivornmental conditions in the Barents Sea will be identified and
discussed. Secondly this thesis will discuss how the risk of ship collisions will increase with respect
to the three scenarios, minimum, basis and maximum scenario, and analyse wheter or not there are
any areas that will experience a higher risk than other.

  1  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

1.3 RESEARCH QUESTIONS

The following research questions are posed on the basis of the research problem.
• How are the environmental conditions in the Barents Sea?
• How does the increased offshore-related traffic increase the probability of ship collision in
the Barents Sea in 2030?
• How will the risk of ship collision change between the three scenarios in 2030?

1.4 SCOPE
The thesis will be based on three possible scenarios for future the field development in the Barents
Sea in 2030. These scenarios (chapter 1.4.1 – 1.4.3) have been prepared by Rystad Energy for Petro
Arctic in the report Rystad Petro Foresight 2030 (Petroarctic, 2014). The thesis will only consider
the increment of future traffic associated to the petroleum industry, and only focus on the
probability part of the risk.

1.4.1 Minimum Scenario

Minimum scenario (Figure 1.2), consisting of proven resources: Goliat, Snøhvit and Johan Castberg.
An overview over the different facilties is listed up in Table 1.1.

Figure 1.2: Minimum Scenario (Rystad Petro Foreseight 2030, 2012)

 
Table 1.1: Type of offshore facilities in Minimum Scenario
Facility FPSO Pipeline to shore
Johan Castberg X
Goliat X
Snøhvit X

1.4.2 Basis Scenario

Basis scenario (Figure 1.3), includes proven resources and fields with relatively high probability of
discovery: Goliat, Snøhvit, Gohta, Johan Castberg, Hoop, Lopparyggen øst and Barentshavet sydøst.
An overview over the different facilties is listed up in table 1.2.

  2  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Figure 1.3: Basis scenario (Rystad Petro Foreseight 2030, 2012)

 
Table 1.2: Type of offshore facilities in Basis Scenario.
Facility FPSO Pipeline to shore
Johan Castberg X
Gohta X
Goliat X
Hoop X
Barentshavet Sydøst X
Snøhvit X
Lopparyggen X

1.4.3 Maximum Scenario

Maximum scenario (Figure 1.4), also includes the fields from basis scenario and fields in areas
which today are considered to have a low probability of discovery, as well as fields in non-opened
areas. The different facilities will vary between an Floating, Production, Storage and Offloading
(FPSO) platforms and semi-submersible platforms with pipeline to shore and to Barents Pipe which
is a pipeline (marked with red line) which is going from the east of the Barents Sea down to
southern Norway, see table 1.3. The purpose of this scenario is to cover the entire geograpichal area
of analysis.

  3  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Figure 1.4: Maximum Scenario (Rystad Petro Foreseight 2030, 2012)

 
Table 1.3: Type of offshore facilities in maximum scenario

Facility FPSO Pipeline Pipeline

to shore Export
Goliat X
Hoop X
Barentshavet sydøst I X
Fingerdjupet X
Johan Castberg X
Gohta X
Barentshavet sydøst II X
Snøhvit X
Lopparyggen øst X

1.5 LIMITATION AND ASSUMPTIONS

The thesis is governd by the following limitation:
• There are great uncertainties about how the Barents Sea will look like in 2030. Information
regarding sizes and associated shipping traffic to each field in the future scnearios is based
on the information from different companies, experts and government reports.
• Consequences of ship collisions are not considered in the risk anlysis.
• The focus of the thesis is on the Norwegian Economic Zone of the Barents Sea.
• The calculation and simulation is based on Automatical Information Source (AIS) data from
2013

The following assumptions have been considered for the thesis:

• The export traffic for each field must take the shortest route until the separation zone is
reached.

  4  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

• The maritime traffic, except the added offshore fields and Russian transit traffic is assumed
to be the same in 2013 as in 2030
• The same amount of Inspection, Maintenance & Repair (IMR) vessels is chosen for each
field, this assumption is made in collaboration with ENI Norge through a telephone meeting.
• The incoming traffic for transporting equipment to Polarbase from the marked is assumed to
be by land.
• Assumes that the capacity at Polarbase is acceptable for each scenario.
• There is no WOW (Waiting on weather) in the simulations.
• Assumes that all vessels have successful operations.
• The export tankers from Hammerfest will use the same sailing route as Arctic Princess.
• The helicopter technology and capasity is improved, with respect to crew transportation. No
crew boats are included in the model.

1.6 THESIS OUTLINE

This thesis consists of the 5 chapters. Chapter 1 contains background information relevant for the
thesis and a description of the research problem, an overview over the three scenarios, aim, research
questions, the scope, limitations and the thesis outline. Chapter 2 follows with a literature review.
This chapter starts with an environmental description of the Barents Sea, with physical conditions
and an overview of the existing and future fields in the Barents Sea. The two next sub chapters
contains of the maritime traffic and a description of the theory of ship collisions. Chapter 3
describes the research, method and materials, how the method is conducted and the challenges
connected to data collection and data analysis. Chapter 4 contains of discussion of the findings,
literature and the results. This chapter starts with results of the data analysis, comparison of results,
a sensitivity analysis and ends with a literature discussion. Chapter 5 presents the conclusion to the
thesis and suggestions for further research.

  5  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

  6  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

2 LITERATURE REVIEW
This chapter presents the basic theories and knowledge that are used to answer the research
questions.

2.1 DESCRIPTION OF ANALYSE AREA

2.1.1 Description of the Barents Sea
The Barents Sea (Figure 2.1) is a subarctic shallow continental shelf of approximately 1 400 000
km2. The Barents Sea is delineated in the Norwegian Sea in the west by a line from the North Cape
of Bear Island to the South Cape of Spitsbergen, the rest of Zemlja Frantsa losifa (Frans Josef Land),
Novaja Zemlja in the east, the Arctic Ocean to the north and the Russian and Norwegian coast in
the south. The depth varies between 200 m and 500 m, but the ocean is shallower than 50 m in the
Spitsbergen Bank (SNL, 2014).

Figure 0.1: Overview over the Barents Sea (World Atlas, 2014)

There have been drilled more than 100 wells since 1980 in the Barents Sea. However it was only at
the beginning of the new millennium that the Barents Sea could be termed as the third oil and gas
province (SNL, 2014).

2.1.2 Physical Conditions

In this sub-chapter the physical conditions in the Barents Sea are described. All the mentioned
physical conditions may have an impact on the sailing conditions, and may therefore be a
contribution for increasing the risk for ship collisions.

Air Temperature
The average minimum air temperature in the Barents Sea is -7.7 ˚C with an annual range between -
6.0 ˚C to -9.0 ˚C. The minimum air temperatures that can be expected in the southwest are in the
range of -15˚C to -20˚C. Towards the north and east, the temperatures decrease to the range of -
20˚C to -30˚C. The minimum air temperatures are shown in Figure 2.2 (Jacobsen, 2012).

  7  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Figure 0.2: Lowest air temperature with an annual probability greater than 10^-2 (Jacobsen, 2012)

Wind
A comparison of wind conditions in the Barents Sea and the North Sea shows no major differences
in wind speed. The highest wind speed have been measured at Bear Island, and then found to
decrease towards the east and north. Metrological conditions in the Barents Sea are dominated by
storms that forms in the North Atlantic and the wind direction during winter is typically from the
southwest, except near the coast where the wind direction normally is northeast (Thelma, 2010).

Icing
Icing is a well-known hazard to traditional operations in the northern waters, where about 80
vessels capsized due to icing in the period 1955-1970 (Løset, et. al., 2006). In combination with low
temperatures, icing is caused by:
• sea spray
• undercooled rain
• rain
• fog
Sea spray is the most frequent cause of icing and the factor with the biggest contribution of ice on
ships, and a combination of spray and atmospheric icing can cause extreme ice loads. Icing due to
sea spray is a phenomenon which occures at low temperatures combined with strong winds from the
south and southeast bringing cold air masses from the east. According to the Meteorological
Institute icing from sea spray will occur with temperatures below -2˚C and with wind speed in
excess of 11 m/s, however the data for sea spray icing is limited. Observations of air temperatures
at the Norwegian coastal stations indicate that icing will be a problem in the part of the Barents Sea
which is opened for petroleum activities (Figure 2.3). The icing problem in the North Barents Sea
can be extreme, and spray and mist can cause build up reaching four centimeters of ice per hour on
the surface of a device (Thelma, 2010). Figure 2.3 shows the occurence (percentage) of temperature
below -1.8 C and wind above 10 m/s in January from 1961 to 2010.
Ice accretion on ships and structures is a concern for operations in cold climates and can lead to a
variety of problems, as even light ice accretions can lead to many operational difficulties, e.g.
slippery decks, ladders and handrails. Ice accretion can be a safety hazard; if equipment such as
winches, derrick, valves, life-saving and fire fighting equipment are rendered inoperable, causing
delayes in operation or potentially necessitates an evacuation of the platform. For vessels, the effect
are more serious, in that ice accretion increase the draught, reduces the freeboard, and moves the
centre of gravity of the vessel, thereby compromising stability (Løset, et. al., 2006).

  8  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Figure 0.3: Occurence (%) of temperature below -1.8 and wind above 10m/s January 1961-2010
(Kystverket,2014)

Atmospheric icing may occure in the Barents Sea throughout the year as low air temperatures are
possible at any time. For moving ships the rate of icing depends on wind speed, air temperature, sea
temperature, characteristic speed and heading of the ship. Atmospheric icing occurs through fresh
water precipitation, like snow, rain and super-cooled droplets. Atmospheric icing normally form
when the air temperature is between 0 ˚C and – 20 ˚C and the wind speed is less than 10 m/s. As a
result of atmospheric icing, the higher parts of the ship can get covered with 1-2 cm (rarely up to 6
cm) thick ice (Løset, et. al., 2006). Atmospheric icing may produce a uniform layer of ice on all
exposed surfaces. This may pose various problems in operations, communication and navigation, as
it also adheres to antennas and other technical aids on the deck, e.g. cranes, winches and valves.
Black frost can cause a critical reduction of stability, especially for smaller vessels, if atmospheric
icing occurs simultaneously as sea spray icing. Sea spray can only cause ice accumulation up to a
certain height above the waterline, but atmospheric icing can occur at all heights (Løset, et. al.,
2006).

Status for sea ice in the Barents Sea

The ice extent in the Barents Sea is as its greatest in April. From 1979 to 2013 there has been a
negative trend in sea ice extent in April, although the yearly variations are large. The ice extent is
lowest in September, and there has also been a negative trend from 1979 to 2013.
The last eight years have had yearly variations that have been more moderate compared to previous
years. The lowest ice exent in April was in 2006, and in September in 1979, 2001, 2004, 2011, 2012
and 2013 the area have been nearly ice-free (Figure 2.4).

  9  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Figure 0.4: Ice extent in the Barents Sea from 1979 - 2013 (Meterologisk institutt, 2012)

Polar Lows
Polar lows are small but intense low pressure formed in the Arctic waters during the winter season
from October to April. A potentially damaging aspect of polar lows are the rapid changes, as the
wind can increase from breeze to storm in just a few minutes, and the wave height is observed to
increase by up to 5 meters in under an hour. Generally, polar lows are difficult to forecast, since
they occur in areas with few points of observations, and they are of a comparatively small scale in
relation to the observation coverage (Meteorlogisk institutt, 2012).
Figure 2.5 presents the monthly distribution of polar lows in the Norwegian Sea and the Barents Sea,
which is registered at the Norwegian Metrological Institute from 2000 to 2012.

Figure 0.5: Monthly distribution of polar lows in the Norwegian Sea and the Barents Sea from 2000 to 2012
(Meterologisk institutt, 2012).

Visibility
The sight parameter is based on the assessment of an observer, and therefore only manned stations
have visibility data. The nearest weather stations catering for the area are stations on the coast of
Finnmark and on Bear Island. The conditions, as observed at these stations are shown in table 2.1
(Meteorologisk institutt, 2012).

  10  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Table 0.1: Visibility distribution (Meterologisk institutt, 2012)

Visibility Sight Bjørnøya and Hopen Vardø Radio
Good > 10 km 50 % (July) 80-90 %
60% (Rest of the year)
Moderate 4 - 10 km 12 – 19 % 5–8%
Low 1 – 4 km 10 – 19 % 9 – 12% (Dec&Jan)
3 – 7 % (Rest of the year)

The frequency of fog at Bjørnøya and Hopen is significantly higher than in Vardø and the
percentage is highest from June to September where it varies in the range 11-27 %. The frequency
of fog is for the rest of the year in the range of 4-8 % at these stations. Vardø radio has the greatest
frequency of fog in July-August and in February when there is fog 4-7 % of the time. The rest of the
year is in the range 1 % (Meteorologisk institutt, 2012).
Darkness
The sun is below the horizon for a given period during winter. This results in total darkness, called
polar night, in the middle of the winter with only limited periods of twilight during the day. The
length of the daylight period decreases rapidly from the autumn equinox until the sun falls below
the horizon. Similarly the daylight period increases rapidly from the return of the sun until the
spring equinox. Table 2.2 shows the dates when the sun falls below the horizon and when it returns
in different locations (Jacobsen, 2012).

Table 2.2: Sun activity in different locations (Jacobsen, 2012)

Location Sun disappears Sun returns
Vardø 23. November 19. January
Hammerfest 22. November 20. January
Nordkapp 20. November 22. January

2.1.3 Existing and future fields in the Barents Sea

The Rystad report includes planned, possible and probable offshore fields through their scenarios.
Considering the uncertainties associated to the size of the future fields, some assumptions with
respect to the ship traffic have been done. Table 2.3 shows a total overview of the maritime traffic
for the existing and future fields in the Barents Sea, which are used in the simulations.
Table 2.3: Maritime Traffic for existing and future fields
Field Size Supply Vessel IMR Vessel Export
mill. Sm3 [75-100m] [100-125m] Tankers
[275-300m]
Goliat1 10 70 4 36
Johan Castberg2 100 130 4 100
Gohta3 10 70 4 36
Fingerdjupet 10 70 4 36
Hoop 10 70 4 36
Lopparyggen Øst 10 70 4 36
Barentshavet sydøst 10 70 4 36
Barentshavet sydøst II 50 100 4 50
Snøhvit AIS AIS AIS
                                                                                                               
1
Parameters from telephone meeting with ENI Norge  
2
Parameters are taken from the government’s document for scenarios for year-round petroleum activity in the Lofoten
and the Barents Sea 2005-2020
3
The parameters for the rest of the offshore fields are reached in consultation with Project Manager in Petro Arctic,
Kjell Giæver.  

  11  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

A location overview of the different offshore fields is shown in figure 2.6.

Figure 0.6: Overview over existing and future fields in the Barents Sea

2.2 MARITIME TRAFFIC

With the forseen increase in petroleum
activity in the Norwegian and the Barents
Sea an increase in the shipping traffic in the
area is expected. In this context, the main
goal should be that the risk of environmental
damage caused by ship collisions and spills
should be kept at a minimal level, while
continuously striving to further reduce the
risk.
Det Norske Veritas (DNV) prepared a report
commissioned by the Coastal Administration
where the probability of acute pollution from
shipping along the Norwegian coast is
analysed. The analysis is based on traffic data
from 2008 and forecast for 2025. It is
concluded in the DNV report that the
predicted increase in Russian traffic, in
combination with increased Norwegian
exports of petroleum from the Barents Sea
will cause the likelihood of emission to
increases significantly by 2025 along most of
the coast of Nordland, Troms and Finnmark
(Figure 2.7) Without the introduction of
further maritime safety measures, an increase
in tanker traffic will result in a greater
probability of a major accident in the area.
Figure 2.7: Traffic density between Vardø and
Today’s emission probability is low due to
Røst in the second half of 2010 (Det Kongelige
the relatively low level of activity and the Miljøverndepartement, 2011)
introduction of effective maritime safety measures.

  12  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

As of today, the same effective maritime safety measures is about to be introduced south of Lofoten
(Det Kongelige Miljøverndepartement, 2011).

2.2.1 Russian Traffic

There are several reviews of the extent of the transit traffic to/from Russia. Kystverket (2003) have
discussed the uncertainties of the development based on meetings with different Russians groups. It
is assumed that there will be established a pipeline to Murmansk before 2015, and the total export
will be approximately 80 milion tonnes crude oil. In addition, any transport of gas/condensate with
ship is assumed to be approximately 6 milion tonnes of Liquefied Natural Gas (LNG) and 1 milions
tonn Liquefied Petroleum Gas (LPG) (DNV, 2003).
The traffic is assumed to be of 656 ships from Russland every year with different vessel size, the
total Russian traffic is presented in Table 2.4 by the length of the ship and the volume of
Deadweight tonnes (DWT).
Table 2.4: Ship Traffic along the Norwegian coast linked to exports from Russia (DNV, 2013)
Ship Traffic in the analyze area Total
100 000 DWT (225m - 250m) 320
180 000 DWT (300m - 325m) 150
280 000 DWT (325m - 350m) 86
100 000 DWT LNG (275m - 300 m) 60
25 000 DWT condensate (150m - 175m) 40
Total vessels 656

2.2.2 Description of Activity

There is expected to be an increase in all type of ships, but especially gas and oil tankers, except
fishing boats that are expected to decline (Det Kongelige Miljøverndepartement, 2011). The
reduction of fishing boats due to implementation of improved technology, better resource
management and continued restructuring of the fishing fleet. The reduction of fishing boats does not
necessarily mean less tons caught fish, but the fishing boats sails less to catch allocated quotas
(DNV & Kystverket, 2012). Table 2.5 shows the traffic pattern for all ships except fishing ships.
For 2011 represents this traffic around 61 % of ‘all ship traffic, and 30-40% of these are over 5,000
DWT.

Table 2.5: Trafitic pattern for different ship types (DNV & Kystverket, 2012)

Description
Oil Tankers

In 2011 oil tankers constitutes about 3,1 % of

total travelled distance.

More than 86% of the sailed distance, is made

by vessels with a deplacement of 5000 DWT
or more. These vessels are required to follow
the Traffic Separation System (TSS) when
traveling along the coast.

  13  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Gas tankers and chemical-/product

tankers

In 2011 gas tankers and chemical-/product

tankers constitutes about 3,4 % of total
travelled distance.

More than 99% of the sailed distance is made

by vessels that are over 5000 DWT. These
vessels are required to follow the TSS when
traveling along the coast.

All vessels except fishing vessels

In 2011 these vessels constitute about 61 %

of total travelled distance.

Approximately 30-40 % of sailed distance is

made by vessels that are over 5000 DWT.
These vessels are required to follow the TSS
when traveling along the coast.

All vessels

A density plot of total travelled distance for

all vessels in 2011.

  14  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

2.2.3 Traffic Development 2005-2009

The traffic of seismic vessels, offshore supply vessels and tankers has increased significantly more
than for other groups of vessels, altough from relatively low levels. Fishing vessels accounted for
most ship movements in 2008, about 58 % of the total distance travelled in the Barents Sea. Over
80 % of all distance travelled for ships with 10,000 gross tons in the planning area is now taking
place in the separation system Vardø-Røst, including close to 100 % of all traffic. The remaining
traffic in the area is dominated by cargo ship on 1,000-5,000 gross tons, but there are also activities
with other cargo, offshore vessels and other support vessels (Det kongelige miljøverndepartement,
2011).
Transit traffic consists of large tankers and bulk carriers to/from Russian ports. Until 2008, the
traffic volume was stable in terms of both cargo volume and the number of passing ships. The total
cargo volume is in the range of 10 million to 12 million tonnes per year, and is taken by 200 to 240
fully laden tankers. In 2009 the volume rose significantly (Figure 2.8). There are indications that the
volume transported from the transit operations will continue to increase in the coming years, and
the average size of tankers carrying oil is expected to increase (Det kongelige miljøverndepartement,
2011).

Figure 2.8: Number of passing ships versus cargo volume development (Det Kongelige
Miljøverndepartement, 2011)  

2.2.4 Ship Traffic In The Northeast Passage

Interest in ship traffic in the Arctic Ocean, including the Northeast Passage has increased with the
rapid retreat of ice sheet in recent years. Summer ice has retreated sufficiently to create time
windows in which all or part of the shipping lanes north of Russia and Canada / USA is open before
freezing starts again. Today's traffic in the Arctic Ocean is low, and it is expected that the ship
traffic in the Arctic Ocean over the next few years will still dominated by ships that have
destinations in the area (Det kongelige miljøverndepartement, 2011).
Ships in transit through the Northeast Passage are currently at a very low number. Ship traffic
through the Passage is likely to increase as global warming increases. The Northeast Passage will
shorten the distance between Rotterdam and Yokohama from 11,200 nautical miles to 6,500
nautical miles, which can provide a significant cost reduction (DNV & Kystverket, 2012).
For shipping, the predictability in relation to absence of ice, especially multi-year ice, is a very
important factor when the route through the Northeast Passage is chosen. Due to the annual
variations, with some cold winters, multi-year ice must still be expected in the coming years. Other
factors, such as the difficulty of identifying and quantifying risk factors associated with sailing in
the Arctic, compared to traditional routes brings negative impact on the attractiveness of the

  15  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Northeast Passage. There are also considerable uncertainties with regards to Russian policies and
particularly to the development of infrastructure. If transit through the Northeast Passage is to
increase to a commercial scale, it would be more than a pure economics and logistics question
(DNV & Kystverket, 2012). For liners, such as container ships, the reliability and predictability of
the passage plan is the most important factor, one must know the exact date when the goods will
arrive. This is challenging through the Northeast Passage due to uncertainties regarding ice, weather
and politics (DNV & Kystverket, 2012).  

2.3 SHIP COLLISION

There can be many reasons for a ship collision to occur. Rule 7 in Farewells rules of the nautical
road says this about the risk of collision: “Every vessel shall use all available means appropriate to
the prevailing circumstances and conditions to determine if risk of collision exists. If there is any
doubt such risk shall be deemed to exist.” (Allen, 2005).

2.3.1 Types of Accidents

Since other collision types, e.g., collision with a floating object, are not considered in this thesis,
ship-ship collisions are referred as collision hereafter. A ship-ship collision occurs if a ship strikes
another ship (Kristiansen, 2005). Collisions can be divided into head-on, overtaking, merging,
crossing and bend collision (Table 2.6).
Table 2.6: Description of the different collisions type

Collision type Illustration

Head-on collision

Overtaking collision

Merging collision

Crossing collision

Bend collision

  16  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

2.3.1.1 The Head-On Situation

“When two power-driven vessels are meeting on reciprocal or nearly reciprocal courses so as to
involve risk of collision each shall alter their course to starboard so that each shall pass on the port
side of the other.” (Allen, 2005).
Vessels approach each other in a head-on situation at a rate equal to the sum of their individual
speeds, whereas in the overtaking situation the rate of approach is, off course, the difference
between the velocities of the ships involved. The closing speeds in crossing collision is somewhere
in the middle. Thirty-knot closing speed are common in head-on encounters, fifty-knot relative
speed are an unremarkable occurrence for container ships and naval vessels, and closing speeds for
high-speed craft can approach terrestrial highway magnitudes. When vessels collide full on, even at
slow speed, the result can be extremely destructive. The general rule for calculating the vessels’
respective kinetic energies is (Allen, 2005):

Kinetic Energy = ½ Mass x Velocity2

2.3.1.2 The Overtaking Situation

Of the five types of ship collisions – overtaking, head-on, merging bend and crossing – many
consider the overtaking situation the least risky due to the low relative speed and reduce force of
impact in the event of collision. The risks of overtaking collisions are greater in narrow channels,
where increased traffic density, limited manoeuvring room, and the risk of interaction expose the
vessels to the danger of grounding or collision.
Overtaking situations develop slowly, often placing the two vessels in close proximity – and
exposed to crossing or meeting traffic – for considerable periods of time (Allen, 2005).
2.3.1.3 Crossing Situations
Rule 15 in “Farewells Rules of the nautical road” describes crossing situations like this:
“When two power-driven vessels are crossing so as to involve risk of collision, the vessel which has
the other on her own starboard side shall keep out of the way and shall, if the circumstances of the
case admit, avoid crossing ahead of the other vessel.” (Allen, 2005).

2.3.1.4 Merging Situations

The merging situation occur when two vessels are sailing in different directions, and meets in a
waypoint connection of three legs. This in combination with ships who fails to make evasive
actions in order to avoid the collision (IALA Web, 2014).

2.3.1.5 Bend Situations

The bend situtation occur when two ships are sailing in oposite directions meet in a bend (waypoint
connection with two legs). One of the ships fails to change the course at the waypoint, resulting in
the ships ending up on collision course. This in combination with ships who fails to make evasive
actions in order to avoid the collision (IALA Web, 2014).

2.3.2 Main Causes of Ship Collision

In 2002 Liu and Wu of Dalian Maritime University studied 100 written collision reports from the
maritime authorities in the UK, USA, Australia, Canada, New Zeeland and Sweden. And listed the
following main causes: (Lee & Parker, 2007)

  17  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

1. Improper use of radar and ARPA are still a common problem on board ship. The most
frequently observed errors were misinterpreting the information showing on the radar screen,
improper setting of the range scale of the radar and no radar plotting. The underlying human
elements included lack of knowledge, experience, training and fatigue. The organisational
factor also had an influence to some extent.
2. Poor lokout has been involved in most collisions. Factors included the lack of competent
lookouts, improper looking methods, and improper use or no use of all available means.
Poor lookout was usually caused by lack of experience, knowledge and training, manning
problems, lack of safety culture, high workloads and inattention.
3. Error of judgement was another factor commonly found in the cause of collisions. The most
probable underlying human elements were lack of knowledge, training and information
processing ability. Fatigue and workload also played important roles in the navigator`s
ability to make the correct judgements.
4. Communication problems ranked high in the list of unsafe acts. The most frequently made
mistakes were lack of communication and misinterpreting received information. The major
underlying human elements found were the reluctance of navigators to exchange
information.
5. Failure to take early actions frequently appeared in collision cases. The causes of this
unsafe act were poor lookout and the torpor induced by the sheer monotony of keeping
watch.
6. Apparently improper ship manoeuvring failure to comply with good seamanship and
failure to display signals had a close relationship with knowledge, skill, training and
experience.
7. Visibility was an important influencing factor in ship collisions. Failure to sound signals,
failure to arrange appropriate lookout, failure to reduce speed and failure to communicate
with others frequently appeared in this situation.
8. Collisions studied were caused by a combination of several factors in general.
As we see, the different causes influence each other, and there is no single reason that caused a
collision, but often a combination of several factors. In figure 2.9 we can se an illustration of the
connection between different elements that leads to an accident. This generic model is called Model
of Accident Causation Using Hierarchical Influence Network (MACHINE), and shows how the
direct causes of all accidents are combinations of human errors, hardware failures and external
events.

Figure 2.9: The MACHINE model reflects the relationship between human errors, hardware failures and
environmental elements (Embrey, 1992)

  18  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

2.3.3 Concepts of Risk and Risk Analysis

A measure of potential loss is called risk. It is defined as the product of the probability or frequency
of the unwanted event and its consequences if it occurs:
Risk = Probability x Consequence (1)
This thesis will only focus on defining the probability part and identify the probabilities of ship
collisions in the Barents Sea. If we can find measures which reduces the probability, the risk will
also be smaller.
Typically, marine accident probabilities are modelled based on the work of Fujii et al. (1974) and
Madcuff (1974). Following their first ideas, the frequency of marine accidents is generally
estimated as:
𝑃 = 𝑁  ×𝑃! (2)
Where
N Number of accident candidates
PC Causation factor

Accident candidates are the ships that are on an accident course in the vicinity of another vessel. In
other words, the number of accidents would be N if no evasive maneuvers were made to avoid the
accident. Causation factor is the probability of failing to avoid the accident while being on an
accident course. It quantifies the fraction of accident candidates that are actually colliding with
another vessel (Ylitalo, 2010).

2.3.3.1 Probability of a collision

It is many different aspects in the cause of a collision, like human or organizational factor. The
probability is calculated as followed: (Geijerstam & Svensson, 2008)

Fcp = F x Fd x P1 x P2 x P3 (3)

Fcp – Frequency of powered passing vessel collision

F – Total traffic in the lane
Fd – Proportion of vessels that are in the part of the lane directed towards another ship
P1 – Probability that the passage planning stage is not carried out correctly
P2 – Probability that the vessel suffers a watch keeping failure
P3 – Probability that a platform or stand-by vessel fails to alert the ship in time to prevent a collision

2.3.3.2 Causation factor PC

The causation factor specifies the probability that the officer of the watch will fail to react, e.g. in
case the vessel in on collision course with another vessel. 80 % of the PC is estimated to come from
Human Error (Kystverket, 2014).
The causation factors are important for the results since they act as reduction factors on the
calculated number of blind navigation collisions. In the specification on the causaution factor it
should be considered if navigators exhibit extraordinary awereness; possible beacuse of two
navigators being present on the bridge. For ferry routes it is typically the case that the causation
factor is lower than the average due to the navigators increased situation awareness (IALA, 2012)

Table 2.7 and Table 2.8 presents the different factors which contribute to personal and
organizational failures (Kystverket, 2014).

  19  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Table 2.7: Causation factor, personal

Personal:
Physical limitations Wishful thinking Laziness
Inadequate communication Ignorance Greed
Bad judgement Negligence Alcohol
Fatigue Folly Mischief
Boredom Panic Violations
Inadequate training Carelessness Ego

Table 2.8: Causational factor, organization

Organization:
Ineffective regulatory Production orientation Inequitable promotion /
recognition
Poor planning / training Cost-profit incentives Ineffective monitoring
Poor communication Time pressures Ego
Low quality culture Rejection of information Negative incentives
Low worker morale Complex structure Vioaltions

2.3.3.3 How to Perform a Maritime Risk Assessment

It is described by the International Maritime Organization (IMO) that a generic model of collision
risk shall not be viewed in isolation, but rather as a collection of systems, including organisational,
management, operational, human, electronic and hardware aspects. The systems and functions
should be broken down to an appropriate level and aspects of interaction of functions and systems.
The extent of their variability should also be adressed. The human element is regarded as one of the
most contributory aspects to the causation of accidents and must be incorporated in an assessment.
Expert judgement is an important part of an assessment that provides proactive thoughts and ideas
and is necessary where limited data exists (IMO, 2007).
During an identification of possible hazards, it is necessary to combine both creative and analytical
techniques with the aim to identify all relevant hazards. Structured group reviews with experts in
the various appropriate aspects such as ship design; operations and management should be
undertaken followed by a ranking of hazards and scenarios with regards to their contribution to an
accident (IMO, 2007).

2.3.4 Collision Avoidance

The environmental, human and economical consequences of a ship collision in the Barents Sea are
large. It is therefore important to have good and reliable systems on ship collision avoidance. From
a theoretical point of view ship avoidance can be describes as easy as: “Collisions avoidance
involves two or more seagoing vessels that have to cooperate and coordinate their individual
operations to avoid ending up in the same place at the same time.” (Nielsen & Petersen (2004).
The central problem facing the mariner in selecting the appropriate collision avoidance action is the
absence of mutual cognition – understanding not only the conduct required of the mariner´s own
vessel, but of the other vessel as well. In short, what is that other vessel going to do? Regimes
aimed at fostering coordinated action by approaching vessel seek to ameliorate the problem. A
coordinated system for collision avoidance requires three elements. Each approaching vessel must
mutually perceive: (Allen, 2005)
1. The risk of collision
2. The strategy to be applied in avoiding collision

  20  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

3. The point in distance and/or time at which manoeuvres are to be made.

Since 2002 new ships and later all larger sea-going vessels (>300 Gross Tonn (GT)) and all
passenger vessels are required to carry AIS on board. Through dedicated VHF frequencies, AIS
information is transmitted between vessels, from vessels to shore, or vice versa. In simple terms
AIS is a technology to make ships “visible” to each other. As an aid to collision avoidance, it
records the information of ship behaviour, including the effects of human action and ship
manoeuvrability. The information includes the vessel's name, its particulars, ship type, registration
numbers, and destination as well as the vessel's position, speed, and heading (Mou, Tak &
Ligteringen, 2010).
“Even with this technology on ship collision avoidance, ship collisions still occures. It has, in fact,
become commonplace to hear that human factors are involved to a considerable degree (some say
as much as 80 %) in most collisions. But such obvious conclusions should not distract us from
searching out the true root causes of human failures. The errors may be in risk detection,
communication, assessment, or management” (Allen, 2005).
After approval by the IMO, a sailing led-system was established in the Norwegian economic zone
from Vardø to Røst on July 1st 2007. The system requires all tankers and cargo ships over 5000
gross tons in transit to stay at least 30 nautical miles from land. By moving the risk further away
from the coast the likelihood of accidents and spills are reduced. The increased distance gives the
authorities more time to intervene if the ships are having problems and requires assistance, and
opportunities to prevent accident is improved. Vardø Vessel Traffic Service (VTS) was established
in 2007 and monitors all tankers and other risks along the Norwegian coast. Vardø VTS also
monitors the vessels compliance with rules of the sail lead system from Vardø to Røst. If a vessel
departs form the lead, the VTS call up and guide vessels on the wrong course and requisition
assistance when needed (Det Kongelige Miljøverndepartement, 2011).
Measures to improve safety at sea conducted after 2005 (Det kongelige miljøverndepartement,
2011).
o Automatic Identification System o Norwegian Coastal Administration
(AIS) for ships (information, tracking (NCA) has developed a procedure for
and collision prevention) have been the government´s overall handling of
introduced. This is estimated to situations where it is necessary for a
reduce the risk of collsion by 20% vessel to seek a port of refuge.
o The satellite-based monitoring of sea o To limit the potential for damage by
areas has been developed. oil spills, a requirement was
o Vardø VTS was established in 2007 introduced in 2007 that ships calling
and monitors the risk of traffic along at nature reserves in East Svalbard
the Norwegian coast, including shall not bring or use other fuel than
Svalbard light marine diesel. From 1st January
o Tow preparedness has been improved 2010, a similar claim made applicable
and three new tugs have been to the three major national parks on
chartered on short-term contracts. the west coast of Svalbard. It made
o New harbour act came into force in temporary exemption for the approach
2010. The harbour act became to Ny-Ålesund and Magdalenefjorden
effective on Svalbard in 2008 through until 2015
a separate regulation

  21  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

2.3.4.1 VVTS – Vardø Vessel Traffic Service

The main task of the VTS is monitoring risk traffic and enables alert actions if necessary, and to
coordinate vessels in Norwegian tugboat prepardeness. The traffic control center is a part of of the
coastal administrations line of preparedness against acute pollution (Kystverket, 2012).
The responsibility of Vardø VTS was expanded on July 1st 2008 to cover the entire Norwegian
Economic Zone, from the Swedish border in the south to the border between Norwegian and
Russian economic zone in the north, Svalbard and Jan Mayen included. Within the scope of
monitoring, the VTS monitor tankers and other risk traffic in the coverage area (Kystverket, 2012).
Vardø VTS has recently helped to avert a series of incidents that could have resulted in a major
accident. Their ability to quickly get an overwiev of potential hazardous and risk situations has
undoubtedly helped to increase maritime safety throughout the Norwegian economic zone
(Kystverket, 2012).

2.3.4.2 Automatic Identification System

Automatic Identification System (AIS) in an international aid to avert ship collisions and to identify
and monitoring the ships (Figure 2.10). AIS is made applicable to vessels over 300 GT in
international traffic and 500 GT engaged in domestic voyages, and all tankers and passenger ships
irrespective of size. Excempt from the requirement to be equipped with AIS are special categories
such as warships, naval auxiliaries and state-owned or state-operated vessels and small craft yacht.
Since the AIS system has a limited range from the coast there will be traffic that is not captured by
the system, such as fishing vessels, Svalbard traffic (including coal transport and cruise/passenger
accounts for the largest magnitude) and parts of the transatlantic move (including to/from Russia).
The coverage area is still deemed to be sufficient to capture all matarial aspects of shipping in
Barents Sea (Kystverket, 2010).

Figure  2.10:  The  AIS-­‐Sat  I  in  orbit  (SNL,  2014)  

  22  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

3 RESEARCH, METHODS AND MATERIAL

The  aim  of  this  chapter  is  to  present  and  discuss  the  applied  research  approach  and  
methodologies  used  in  this  research.    

3.1 METHOD
Figure 3.1 presents an overview of the used methods from start to the conclusion. It started with a
literature review about the analysis area before the basis for the thesis was ready. From the
simulation tool IWRAP, the results from the different scenarios achieved.

Figure 3.1: Methodology used in the thesis

3.2 DATA COLLECTION

To acquire knowledge regarding the subject in this thesis, an extensive data collection was required.
In order to substantiate the theory chapters it has been performed research in various government
documents developed by DNV and Kystverket. This is considered as very reliable source for this
thesis. These documents were handed to me personally after a meeting with DNV here in Tromsø.
Øyvind Persson in DNV was also a contributing part in the definition part of the thesis. The library
at UiT – The Arctic University of Norway was invaluable to attain basic knowledge about ship
collisions, collision avoidance and the reasons for why ship collisions occur. The amount of
literature concering these issues were huge, but there were little specific literature on ship collisions
in the Barents Sea. Therefore was it a challenge to separate information that would be relevant for
this thesis.
When a basic knowledge of the analysis area and theory behind ship collsions was achieved,
contact was established with senior advisor Trond Langemyr in Kystverket. He was very engaged in
the topic of this thesis and gave access to even more government documents unavailable on the
Internet. He proposed using IWRAP Mk2 to simulate the frequency of ship collisions, and gave
access to the extended version for free. To make the simulations realistic, he sent a complete sample
of AIS data for 2013. These data are not for distribution.
Contact with professionals within the industry has been leading the work. As mentioned there
where early-established contact with specialists in DNV and Kystverket, but also specialists at Aker
Solutions, ENI Norge, Petro Arctic and UiT – The Arctic University of Norway. By participating in

  23  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

a workshop at Hurtigruta by OPLOG (OPerational Logistics and business process management high
north Oil & Gas operations), I had the opportunity to discuss and brainstorm about this subject with
specialist in the industry further valuable data and insight was achieved. It has been very helpful to
discus this subject with specialists through the whole process in order to be able to make the
required assumptions. The exact locations of the various fields are defined in an ongoing project of
Aker Solutions. The sailing routes for each facility are assumption made in consultation with
supervisor Øystein Mikelborg, where the ships sail the shortest route from A to B.

3.3 DATA ANALYSIS

IWRAP has the function to load AIS data for a particular area and create a density plot for this
traffic. After the density plot is made, your own sailing routes should be defined with desired traffic.
This was done for each field in the scenarios, for supply vessels, IMR vessels, exports vessels and
the Russian transit traffic. IWRAP calculate the frequency of accidents in each predefined leg and
connects the results with the chosen AIS data.
The simulation program was not able to process the all the AIS data at once, and it was therefore
decided to split the AIS data into quarters. In practice, each scenario had four simulations, with its
associated result. The relevant results from each quarter were added together to get the average
result that the figures in the result chapter are based on.
It is possible to get vast amount information about the traffic, accident types and locations through
this program. It was therefore a challenge to find the most relevant information and effective
methods of presentate it. It was decided to get statistics for all legs for each offshore field, type of
collisions and an overview over which ship types that involved in the collisions.
IWRAP does not include tools for estimating the uncertainties of the results, so this was performed
manually. All scenarios have a sensitivity analysis with a difference in traffic with plus minus ten
percent. The sensitivity analysis is included to predict the outcome if the basis predictions turns out
to be different.

3.3.1 IWRAP Mk2

The objective of IWRAP is to provide the user with a tool that assists to quantifying the risks
involved with vessel traffic in specific geograpichal areas. On the basis of a specified traffic
intensity and composition the tool allows the user to efficiently evaluate and estimate the annual
number of collisions in the specified navigational area (IWRAP, 2014).
IWRAP gives results as the frequency of head-on, overtaking, merging, crossing and bend
collisions. The relative risk of each waterway and waypoint is marked on the map. It is also possible
to evaluate collision frequencies at certain waterway or waypoint. In addition, overall collision
frequencies, frequencies at certain location or of certain collision type are presented by ship type
(Ylitalo, 2010). The following section discuss different equations which are used to calculate the
risk of the different collision types in IWRAP 8 (IALA Web, 2014)

i) Head-on Collisions
The relative speed of two ships approaching each other is expressed as:

(!) (!)
𝑉!" = 𝑉! + 𝑉! (4)

Where:

  24  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Vi (1) is the speed of the ship of the ship class i moving in the direction 1
Vj (2) is the speed of the ship of the ship class j moving in the direction 2
The number of collision candidates for head-on collisions on a waterway is evaluated as

!
𝑁!!!"#!!"   = 𝐿!   !!"#!!"
!,! 𝑃!!,!   (!)!" (!)  (  𝑄! ! 𝑄!!  ) (5)
!! !!

Where:
LW is the length of the segment
Qi,j(n), (1,2) is the number of passages per time unit for each ship type and size, in each
direction, (1) and (2),
fi (1)(y) and fj(2)(y) is the geometrical probability distribution of the lateral traffic spread on the
route. The traffic spread is typically defined by a Normal Distribution but
may in principle be of any type.
𝑃!!!"#!!"
!,!
depends on traffic distributions across the waterway, fi (1)(y) and fj(2)(y). Typically, traffic
spread across a waterway is defined by a normal distribution but any distribution may be used.
Distributions have to be assumed to be independent. When traffic is normally distributed with
parameters (µi(1), σi(1)) and (µj(2), σj(2)), the mean sailing distance between vessels headon to opposite
direction is

𝜇!" = 𝜇! (!) + 𝜇! (!) (6)

The standard deviation of the joint distribution is

𝜎!" = (𝜎 ! )! + (𝜎 ! )! (7)

In the case of normal distribution, PG i,j can be calculated as

!!" !  !!" !!" !  !!"

𝑃!!!"#!!"
!  ,!
= Φ   !!"
−  Φ − !!"
(8)
Where
Φ is the standard normal distribution function and
𝐵𝑖𝑗 is the average vessel breadth:
(!) (!)
!! !!!
𝐵!" = !
(9)

Where 𝐵! is the average breadth of vessel of ship class i.

ii) Overtaking Collisions

When estimating the number of overtaking collisions, the relative speed in equation (4) is replaced
by

(!) (!)
𝑉!" = 𝑉! − 𝑉! (10)

Where Vij > 0 otherwise no overtaking will occur. The geometric probability of meeting (5) is
replaced by

  25  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

(!) (!) (!) (!)

!"#$%&'()* (!) (!) !! !!! (!) (!) !! !!!
𝑃!!,! = 𝑃 𝑦! − 𝑦! < !
− 𝑃 𝑦! − 𝑦! <− !
(11)

For normally distributed traffic, µ in equation (6) is now

𝜇!" = 𝜇! (!) − 𝜇! (!) (12)

Thus, the number of overtaking collision candidates is calculated as in the case of head-on collision
(equation (6)). (Friis-Hansen, 2008)

iii) Crossing Collisions

The frequency of crossing collisions depends on the angle between two lanes. Figure 2.6 shows two
crossing waterways for which the ship traffic also is given. The geometric number of crossing
collision for crossing waterway can similarly to equation (5) be expressed as,

(!) (!)
!"#$$%&'
𝑄! 𝑄! 1
𝑁! = 𝐷 𝑉
(!) !" !"
                                   𝑓𝑜𝑟  10° < 𝜃 < 170°
𝑉!
(!)
𝑉! sin 𝜃
!,!

Where
(!) ! (!) ! (!) (!)
𝑉!" = (𝑉! ) + (𝑉! ) − 2𝑉! 𝑉! cos 𝜃

gives the relative speed between the vessels.

iv) Merging and Bend Collisions

Merging collision is considered as crossing collision. A bend collision may occur if a ship does not
turn at a bend of a waterway and as a result is on a collision course with another vessel (Friis-
Hansen, 2008).

v) Crossing Collisions with Small Vessels

In IWRAP, it is also possible to include the small vessels that do not carry AIS equipment by
inserting “area traffic”. However, area traffic is assumed to be uniformly distributed to the analysis
area around the year.

3.3.2 Causation Factor, Pc

“The causation factor Pc is the reduction factor with which the number of accident candidates has
to be multiplied to get the estimated frequency of marine accidents. Causation factor quantifies the
probability of failing to avoid the accident while being on an accident course.” (Ylitalo, 2010)
The most traditional and most exact way to estime the value of the causation factor is to use
historical data from the wanted area. The traditional approach for calculation of PC is to formulate a
fault tree or an event tree analysis, see figure 3.2.

  26  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Figure 3.2: Fault three for calculating the causastion probability PC for collision (IALA Web, 2014)

The default causation factor (Table 3.1) for collisions is adopted from IWRAP, these numbers are
the “IALA defintions” which are predefined in the program. No work has been published about
adjusting the causation factor for the Barents Sea, so the default value is used to get an estimate of
collision frequency. To get exact values for the causation factor, more research in the Barent Sea
region must be done.
Table3.1: Causation Factors used in the analysis
Type of collision Value
Merging 1,300E-04
Crossing 1,300E-04
Bend 1,300E-04
Headon 0,500E-04
Overtaking 1,100E-04

3.3.3 Operating Vessels and Size

For the analysis of vessel traffic in the planning area and close to the coast area are the identified
vessels divided into 12 ship types and 7 size categories. Table 3.2 presents the ship types and size
categories, which are used in the analysis.

  27  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Table 3.2: Operating vessels and size

Ship type Size category (Gross tonn)
Oil tankers
Chemical-/Prod tanker4
Gas tanker < 1000 GT
Bulk carrier 1000 – 5000 GT
General cargo 5000 – 10 000 GT
10 000 – 25 000 GT
Container ships
25 000 – 50 000 GT
RoRo
50 000 – 100 000 GT
Reefer > 100 000 GT
Passenger
Offshore supply vessel
Other offshore service vessel
Other activities
Fishing vessel All sizes

                                                                                                               
4  Only  a  small  proportion  of  the  vessels  in  the  group  is  registered  as  pure  chemical-­‐/product  tankers,  the  rest  are  

combined.  Although  the  distribution  of  load  types  on  these  ships  are  not  known,  from  experience  different  types  
of  refined  oil  products  may  be  a  typical  load.  
 

  28  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

4 RESULTS OF ANALYSIS AND DISCUSSION

This chapter will presents the results of the data analysis of the three scenarios, and study the
uncertainties in the results through a sensitivity analysis. The available data were analysed using
IWRAP Mk2 software. Due to the capacity of the program/computer the simulations containing
AIS data for 2013 is separated into four quarters. The average of the results for each quarter is
presented in this chapter. It presents an overview picture of the routes in the scenario, note that the
density plot for the historical data is not shown, but included in the calculations. The different color
at the legs indicates the likelihood of a collision, where deeper color means higher likelihood of
collisions.

4.1 MINIMUM SCENARIO

The minimum scenario consists the Russian transit traffic and two offshore installations; Goliat and
Johan Castberg. Figure 4.1 presents an overview of where the fields are located and how the ship
traffic in the area is. The presented likelihood of ship collisions includes traffic to/from Polarbase
and its associated export tankers.

Figure 4.1: Minimum scenario in IWRAP

By studying Figure 4.1 and the color of the legs, we can identify where the likelihood of ship
collision is greatest, which in this scenario will be the leg into Polarbase. With regards to Table 2.3,
by adding the supply and IMR vessels for Johan Castberg and Goliat, 200 supply vessels and 8 IMR
vessels will sail in that waterway a year, in addition to the existing traffic in the area.
The presented results for example Johan Castberg will include the legs for export to the separation
zone, while the supply vessels which sails the legs to Polarbase. The leg to Polarbase is taken
separately, and will be in addition for all the fields connected to Polarbase.
Figure 4.2 presents which fields and waterways that generate most incidents per year in the
minimum scenario. In total, the Russian transit traffic will be the biggest contribution for the
likelihood of incidents per year. This waterway will include all the traffic in the separation zone,
included the export tankers in this zone. The waterway in to Polarbase has higher level of
probability for incident compare to Goliat and Johan Castberg which can be considered as most
critical because of the high amount of traffic in the short length of the leg.

Through calculations in IWRAP, the result shows that the waterway into Polarbase will have 1.6E-
04 incidents per year. The Russian transit traffic will have likelihood on 3.1E-04 incidents per year,
this is roughly twis as likely compared with the likelihood of a collision on the leg into Polarbase.

  29  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

1,0E-­‐03  
1,0E-­‐04  
1,0E-­‐05  
1,0E-­‐06  

Figure 4.2: Incidents per year for each facility in the minimum scenario.

Table 4.1 presents how often each collision type will occur per year for minimum scenario, the
results is given in incidents per year.
The result indicates that the probability for head on collision is the most common collision type.
Head on collisions will occure 4.1E-04 per year, while overtaking collisions will have 1.7E-04
incidents per year. The total number incident per year in this scenario is 5.80E-04.
Table 4.1: Collision type in minimum scenario
Minimum Total
Scenario [Incidents/Year]
HeadOn 4.1E-04
Overtaking 1.7E-04
Crossing 0
Merging 2.0E-09
Bend 8.0E-09
Sum 5.80E-04

  30  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Figure 4.3 presents an overview of how many incidents per year the different ship types will be
involved in. The results indicate that the crude oil tankers and support ship will have the largest
contribution to the frequency of collisions. The support ships will in this scenario only go back and
forth between the offshore installation and Polarbase, while the crude oil tankers will export the oil
from the fields and be a big part of the Russian transit. Crude oil tankers will have 3.9E-04
incidents per year, while support ships have 1.5E-04 incidents per year.

Sum  
Other  Ship  
Pleasure  boat  
Fishing  Ship  
Support  Ship  
Fast  Ferry  
Passenger  Ship  
Ro-­‐Ro  Cargo  Ship  
Sum  
Bulk  Carrier  
General  Cargo  Ship   Ship  type  
Conatiner  Ship  
Gas  tanker  
Chemical  tanker  
Oil  Product  tanker  
Crude  Oil  Tanker  

Figure 4.3: Incidents per year, per ship type in minimum scenario

 
Figure 4.4 presents the frequency of the different types of collisions decribed in chapter 2.3.1 Types
of Accidents. The results indicates that the probability of head on and overtaking collsion by crude
oil tankers will be the most frequent collision type, while head on and overtaking collisions by
support ship will be the second largest contributer in this scenario.

1,0E-­‐03  
1,0E-­‐04  
1,0E-­‐05  
HeadOn  
1,0E-­‐06  
1,0E-­‐07   Overtaking  
1,0E-­‐08  
1,0E-­‐09   Crossing  
1,0E-­‐10   Merging  
1,0E-­‐11  
1,0E-­‐12   Bend  
Crude  Oil  Tanker  

Conatiner  Ship  

Fishing  Ship  
General  Cargo  Ship  

Ro-­‐Ro  Cargo  Ship  

Pleasure  boat  
Passenger  Ship  

Support  Ship  
Oil  Product  tanker  

Bulk  Carrier  

Other  Ship  
Fast  Ferry  
Gas  tanker  
Chemical  tanker  

Figure 4.4: An Overview on how the different ship collides in Minimum Scenario

  31  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Figure 4.4 shows that the crude oil tankers and the bigger ship doesnt have any contribution from
the collision type; crossing, merging and bend collision. This may caused by that these ships are
following the separations zone, and they sail back and forth to their destination, which in this case is
the offshore facility.

A summary of the minimum scenario:

• The largest likelihood of collision in one leg is the traffic in the separation zone with the
Russian traffic.
• The most critical leg is Polarbase because the high amount of traffic in its short length.
• Head on and overtaking collision is the most common collision type
• Crude oil tankers will have the biggest contribution to the frequency of collisions

4.2 BASIS SCENARIO

The basis scenario consists of the Russian transit traffic and six offshore installations; Goliat, Gohta,
Johan Castberg, Hoop, Lopparyggen and Barentshavet Sydøst. Figure 4.5 presents an overview of
where the fields are located and how the ship traffic in the area is distributed. The presented
likelihood of ship collisions includes traffic to/from Polarbase andassociated export tankers. The
supply vessels connected to Barentshavet Sydøst, will go back and forth to the intended onshore
facility in Vardø, while the export tankers will go the shortest way to the separations zone.

Figure 4.5: Basis scenario in IWRAP

By studying figure 4.5 and the color of the legs, we can identify where the likelihood is greatest of
ship collision, which in this scenario will be the leg into Polarbase. With regards to table 2.3, by
adding the supply and IMR vessels for which is sailing into Polarbase, it will be 410 supply vessel
trips and 20 IMR vessels in that waterway a year, in addition to the existing traffic in the area. Note
that the leg for Polarbase is taken separately, and will be in addition for all fields.
Figure 4.6 present which field and waterway that will have the highest likelihood of incidents per
year in basis scenario. By looking at the color of the legs, it shows that the number of incidents will
be highest into Polarbase. Through calculations in IWRAP, the waterway in to Polarbase will have

  32  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

1,1E-04 incidents per year. The second largest contributer to the total probability is the russian
transit traffic were we have likelihood of 2,4E-04 incidents per year.

1,0E-­‐02  

1,0E-­‐03  

1,0E-­‐04  

1,0E-­‐05  

1,0E-­‐06  

Figure 4.6: Incidents per year for each facility in the basis scenario
Table 4.2 present how often each collision type will occur per year for basis scenario, the results is
given in incidents per year.
The result indicates that the probability for head on collision is the most common collision type in
this scenario. Head on collision will occure 1,1E-03 times per year, while overtaking collisions will
have 5,9E-04 incidents per year. The total incident per year in this scenario is 1,8e-03. Table 4.2
presents the frequency of the different collision types in Basis Scenario.
Table 4.2: Collision type in basis scenario
Basis Scenario Total
[Indicent/year]
HeadOn 1,1E-03
Overtaking 5,9E-04
Crossing 2,0E-09
Merging 3,0E-09
Bend 9,0E-09
Sum 1,8E-03

Figure 4.7 presents an overview of how many incidents per year the different ship types will have.
The results indicatess that crude oil tankers and support ship will have the largest contribution to the
frequency of collisions. The support ships will in this scenario only sail back and forth between the
offshore installation and to the associated onshore facility, while the crude oil tankers will export
the oil from the fields and be a big part of the Russian transit, exept from the export tankers from
Lopparyggen which will go from Hammerfest and in to the separation zone. Crude oil tankers will
have 1,0E-04 incidents per year, while the support ships will have 1,8E-04 incidents per year.

  33  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Sum  
Other  Ship  
Pleasure  boat  
Fishing  Ship  
Support  Ship  
Fast  Ferry  
Passenger  Ship  
Ro-­‐Ro  Cargo  Ship   Sum  
Bulk  Carrier  
General  Cargo  Ship   Incidents/Year  
Conatiner  Ship  
Gas  tanker  
Chemical  tanker  
Oil  Product  tanker  
Crude  Oil  Tanker  

Figure 4.7: Incidents per year, per ship type in basis scenario

Figure 4.8 presents the frequency of how the different ships will collide. The results indicates that
the probability of head on and overtaking collision by support ship will be the most frequent
collision type, while head on and overtaking collisions by crude oil tankers will be the second
largest contributer in this scenario.

1,0E-­‐02  
1,0E-­‐03  
1,0E-­‐04  
HeadOn  
1,0E-­‐05  
1,0E-­‐06   Overtaking  
1,0E-­‐07   Crossing  
1,0E-­‐08  
1,0E-­‐09   Merging  
1,0E-­‐10   Bend  

Figure 4.8: An Overview on how the different ship collides in basis Scenario
A summary of the basis scenario:
• The largest likelihood of collision in a leg will be in the waterway into Polarbase
• Head on collision is the most common collision type
• Support ships will have the biggest contribution to the frequency of collisions
• The most common way for the support ships to collide, is by head on and overtaking.

  34  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

4.3 MAXIMUM SCENARIO

The maximum scenario consists of the Russian transit traffic and eight offshore installations; Goliat,
Gohta, Johan Castberg, Hoop, Lopparyggen, Fingerdjupet, Barentshavet Sydøst and Barentshavet
sydøst II. Figure 4.9 presents an overview of where the fields are located and how the ship traffic in
the area is. The presented likelihood of ship collisions includes traffic to/from Polarbase and its
associated export tankers. Barentshavet sydøst and Barentshavet Sydøst II supply vessels will be
located in Vardø. The export from Johan Castberg and Gohta will in this scenario be from Veidnes,
the associated supply vessels will still go to Polarbase. Export from Barentshavet sydøst II and from
Lopparyggen will be transported in pipelines.

Figure 4.9: Maximum Scenario IWRAP

By studying figure 4.9 and the color of the legs, we can identify where the likelihood is greatest of
ship collision, which in this scenario will be the leg into Polarbase. With regards to Table 2.3, by
adding the supply and IMR vessels for which is sailing into Polarbase, it will be 480 supply vessel
trips and 24 IMR vessels in that waterway a year, in addition to the existing traffic in the area. Note
that the leg for Polarbase is taken separately, and will be in addition for all fields.
Figure 4.10 presents which field and waterway have the most incidents per year in Maximum
scenario. The results of the analysis show that the number of incidents will be highest into
Polarbase. The graph of total incidents per year is presented in red. Through calculations in
IWRAP, the result shows that the waterway in to Polarbase will have 1,0E-03 incidents per year.
The second biggest contributer to the probability is the Russian transit traffic were we have
likelihood on 2,9E-04 incidents per year. The highest probability for collision is through the
waterway into Polarbase.
Figure 4.10 shows which field and waterway that generates most incidents per year in maximum
scenario. The graph of total incidents per year is presented in red.

  35  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

The leg for Polarbase is taken separately, and will be in addition to all fields heading for Polarbase.
The result of analysis shows that the number of incident will be highest into Polarbase. The
waterway in to Polarbase will have 1,0E-03 incidents per year.

1,0E-­‐02  

1,0E-­‐03  

1,0E-­‐04  

Incidents/Year  
1,0E-­‐05  
Total  

1,0E-­‐06  
Goliat  

Fingerdjupet  

Gohta  
Hoop  
BS  SØ  I  
BS  SØ  II  

Veidnes  
Loppa  
Polarbase  

Total  
Johan  Castberg  
Russia  

Figure 4.10: Incidents per year for each facility in the maximum Scenario.

Table 4.3 present how often each collision type will occur per year for maximum scenario, the
results is given in incidents per year.
The result indicates that the probability for head on collision is the most common collision type in
this scenario. Head on collision will occure 1,3E-03 times per year, while overtaking collisions will
have the value of 4,5E-04 incidents per year. The total incident per year in this scenario is 1,75E-03.
By comparing figure 4.10 and table 4.3 one can observere that the greatest likelihood for head on
collisions will occure in in the leg to/from Polarbase.
Table 4.3: Collision type in maximum scenario
Maximum Total
Scnerio [Incidents/Year]
HeadOn 1,3E-03
Overtaking 4,5E-04
Crossing 5,0E-09
Merging 6,0E-10
Bend 9,0E-09
Sum 1,75E-03

Figure 4.11 presents an overview of how many incidents per year the different ship types will have.
The results indicate that crude oil tankers and support ship will have the biggest contribution to the
frequency of collisions. The support ships will in this scenario only sail back and forth between the
offshore installation and to the associated onshore facility. Crude oil tankers will have 5,5E-04
incidents per year, while the support ships will have 1,1E-04 incidents per year.

  36  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Sum  
Other  Ship  
Pleasure  boat  
Fishing  Ship  
Support  Ship  
Fast  Ferry  
Passenger  Ship  
Ro-­‐Ro  Cargo  Ship  
Sum  
Bulk  Carrier  
General  Cargo  Ship   Incidents/Year  
Conatiner  Ship  
Gas  tanker  
Chemical  tanker  
Oil  Product  tanker  
Crude  Oil  Tanker  

Figure 4.11: Incidents per year, per ship type in maximum scenario

Figure 4.12 presents the frequency of how the different ships will collide. The results indicates that
the probability of head on and overtaking collision by support ship will be the most frequent type of
collision, while head on and overtaking collisions by crude oil tankers will be the second largest
contributer in this scenario.

1,0E-­‐02  
1,0E-­‐03  
1,0E-­‐04  
1,0E-­‐05   HeadOn  
1,0E-­‐06  
1,0E-­‐07   Overtaking  
1,0E-­‐08  
1,0E-­‐09   Crossing  
1,0E-­‐10  
1,0E-­‐11   Merging  
Bend  

Figure 4.12: An Overview on how the different ships collide in Maximum Scenario.

A summary of the maximum scenario:

• The largest likelihood of collision in a leg will be in the waterway into Polarbase
• Head on collision is the most common collision type
• Support ships will have the biggest contribution to the frequency of collisions
• The most common way for the support ships to collide, is by head on and overtaking.

  37  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

4.4 COMPARISON OF RESULTS

By comparing the results for each scenario we observes a big gap between minimum scenario with
5,8E-04 incidents per year, compared with basis scenario which had 1,8E-03 incidents per year
(Table 4.4). This has a natural explanation since minimum scenario only added two offshore
facilties (Goliat and Johan Castberg) with its associated traffic, while the basis scenario added six
facilities.
Figure 4.13 presents the incidents per year for each scenario. The difference between basis scenario
and maximum scenario (1,75E-03 incidents per year) is a more interesting case. One might have
assumed that it would be significantly a higher frequency of collision in the maximum scenario than
in the basis scenario, since the maximum scenario has added eight offshore facilities compared with
basis´ six. This may be explained in the installed pipeline from east of the Barents Sea which goes
to southern Norway. By inserting this pipeline, the export tankers connected to Barentshavet Sydøst
II and Lopparyggen will be exclucded. This means that it would be 36 less crude oil tankers to
Lopparyggen in the basis scenario. The pipeline from Johan Castberg and Gohta in maximum
scenario redirects their associated export tankers into Veidnes, this is 136 export tankers which will
go for loading at Veidnes.

Incidents/Year  
1,0E-­‐02  
1,0E-­‐03  
1,0E-­‐04  

Figure 4.13: Comparison of Incidents per year for all scenarios, graphical

Table 4.4: Comparison of incidents per year for all scenarios, values
[Incidents/Year] Minimum Basis Maximum
Scenario Scenario Scenario
Total 5,8E-04 1,8E-03 1,75E-03

Figure 4.14 presents the difference in incidents per year for each offshore facility between the three
scenarios. We observe that the highest likelihood of collisions is located in the leg to/from
Polarbase. There could have been made some further measures with respect of this traffic. This
could be:
1) Separate in and outgoing traffic from VTS in this area.
2) Establish an additional base in another location
The first measure will separate the traffic and release the pressure at the narrow waterway into
Polarbase, which would reduce the likelihood of collision on this leg. Introducing traffic lanes the
distances from A to B will be increased, and it still needs to cross other traffic lanes. A measure
from VTS by introducing a “separation zone” here would be an option.
The second measure would have a positive ripple effect both to the infrastructure and increased
population to the districts with more workplaces. This would be an important political issue, as
chosing the coastal city to locate the new base in would cause a high level of interest from the local
communities. This measure will further more create challenges for the existing infrastructure at the
coast of Finnmark, since there are no existing infrastructure today that would be able to receive the

  38  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

amount of supply vessels that is required for the extent described in this scenario. The infrastructure
will not only require a major upgrade at the chosen location, but the road network in Finnmark will
require a major update to implement this measure. The existing roads in Finnmark are bound
together by a larger number of bridges, tunnels and ferry crossings. All these elements must be
considered and taken into account in the planning stage. The many high number of fjords can make
things complicated to get equipments and personnel to where it is needed. The relocation of the
main base for some of the installations would however undoubtedly be a positive measure with
respect to ship collisions.

1,0E-­‐02  

1,0E-­‐03  

1,0E-­‐04  
Minimum  Scenario  
Basis  Scenario  
1,0E-­‐05  
Maximum  Scenario  

1,0E-­‐06  

Figure 4.14: Comparisson of all scenarios of likelihood for ship collision for the different offshore fields

Figure 4.15 presents which kind of ships that is colliding in each scenario. The results are based on
the average value for the quarters in the scenario.

Incidents/Year  

Maximum     Other  Ship  

Pleasure  boat  

Basis   Fishing  Ship  

Support  Ship  
Fast  Ferry  
Minimum  
Passenger  Ship  
Ro-­‐Ro  Cargo  Ship  

Figure 4.15: Comparison of the different scenarios of the likelihood of ship collision for each ship type.

Figure 4.16 presents an comparison of the different collision types in each scenario. Head on and
overtaking collisions will have the largest contribution for all scenarios. By analyzing the figure we
observes that the highest amount of head on collisions is in the maximum scenario, while the
overtaking collisions has the highest amount in the basis scenario.

  39  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Maximum  

Bend  

Basis   Merging  
Crossing  
Overtaking  
Minimum    
HeadOn  

Figure 4.16: Comparison of collision type for each scenario.  

4.5 SENSITIVITY ANALYSIS

The aim of sensitivity analysis is to study the relative influence of the inputs and their uncertainty to
the results. The analysis for each scenario is conducted with an increment and decrement of the
ship traffic with 10%, the results are presented in table 4.5, 4.6 and 4.7.
The sensitivity analysis is only conducted for the 1st quarter in each scenario. This is because of the
large amount of data needed to complete one simulation. It is the change in percentage which is
interesting to identify. The 1st Quarter is chosen since the probability was greatest in this quarter for
all the scenarios, and will therefore be “worst case scenario”.
Table 4.5: Sensitivitiy Analysis of Minimum Scenario
1. Quarter Minimum Scenario Minimum Scenario Minimum Scenario
-10% +10%
Incident/Year Incident/Year Incident/Year
Overtaking 2,5E-04 2,32E-04 [-7,6%] 3,1E-04 [23,4%]
HeadOn 6,65e-04 5,42E-04 [-18,6%] 8,21E-04 [23,5%]
Crossing 0 9,77e-09 9,77E-09
Merging 0 2,79e-09 2,79E-09
Bend 5,50E-09 1,52E-08 [175,7%] 1,52E-08 [175,7%]
Total Collision 9,16E-04 7,74E-04 [-15,6%] 1,13E03 [23,46%]

Table 4.6: Sensitivitiy Analysis of Basis Scenario

1. Quarter Basis Scenario Basis Scenario -10% Basis Scenario +10%
Incident/Year Incident/Year Incident/Year
Overtaking 1,30E-03 6,23E-04 [-51,9%] 7,64E-04 [-41%]
HeadOn 2,42E-03 2,82E-03 [16,8%] 3,70E-03 [53,3%]
Crossing 0 0 0
Merging 2,12E-09 0 0
Bend 5,50E-09 1,52E-08 [175,7%] 1,52E-08 [175,7%]
Total Collision 3,71E-03 3,44E-03 [-7,16%] 4,47E-03 [20,4%]

Table 4.7: Sensitivitiy Analysis of Maximum Scenario

1. Quarter Maximum Scenario Maximum Scenario Maximum Scenario
-10% +10%
Incident/Year Incident/Year Incident/Year
Overtaking 8,29E-04 7,3E-0,4 [-12%] 9,24E-04 [11,5%]
HeadOn 2,81E-03 2,31E-03 [-17,9%] 3,36E-03 [19,5%]
Crossing 0 0 0
Merging 0 0 0
Bend 1,19E-08 1,19E-08 [0%] 1,19E-08 [0%]
Total Collision 3,64E-03 3,04E-03 [-16,6%] 4,29E-03 [17,7%]

  40  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

By studying the results in the minimum scenario, we observe that the probability of ship collisions
will decrease with -15,6% with a reduction of ten percent on the totalt traffic. By increasing the
total traffic, it will then be an increasment of the probability by 24,5%. The minimum scenario will
be the scenario with the greatest difference in the probability of ship collision in percent. The reason
for this may be that the amount of the traffic in this scenario is initially small, and a change in the
number of ships will have a greater effect.
The collision type, bend collision will increase 175,7% in both minimum and basis scenario for
both of the sensitivity analysis. This increment of likelihood of collision will find place in the
waypoint where the support vessels from each field will enter the legs to Polarbase. The biggest
contributor her will be collisions between support ship against support ships which will have a
likelihood of collsions 1,3E-08.
By analyzing the sensitivity results we can conclude that the effect of increasing the total ship
traffic with 10%, the likelihood of collisions will increase more than it will decrease by decreasing
the traffic with 10%. Figure 4.17 presents the results of the sensitivity analysis graphically.

Basis  
Minimum   -­‐10  %  

Maximum     Scenario  
10  %  

Figure 4.17: Comparison of the sensitivity analysis for each scenario

4.6 FURTHER DISCUSSION

In 2002 Liu and WU of Dalian Maritime University studied 100 written collision reports from the
maritime authorities in the UKS, USA, Australia, Canada, New Zealand and Sweden, and listed the
following main causes: Poor lookout, Improper use of radar and ARPA, Error of judgement,
Communication problems, Failure to take early action, apparently improper ship manoeuvring,
visibility and the last was a combination of several factors. All these factors must be considered
when sailing in the Barents Sea. Some factors will have a larger role than others in terms of the cold
climate.
Icing on a vessel will take place by either atmospheric icing or spray icing and may cause that some
components of the boat will lose its functionality partly or completely. As a consequence, it will
have a direct role in several of the points mentioned as causes for incidents. Icing may lead to poor
lookout, radar trouble and therefore improper ship manoeuvring. The icing will be most extreme
when the weather phenomena “Polar Lows” occures. The sudden change in weather, with a drop in
temperature and pressure will lead to icing on vessels. This weather phenomenon is difficult to
predict, so the vessel must be designed for to handle icing and the crew must always be aware of
icing while operating in the Barents Sea.
There are still som uncertainties associated with the future of shipping in the Barents Sea, but the
traffic will most likely increase significantly in the coming years, given the increased interest in oil
and gas development in the Barents Sea. As seen in table 2.5, chemical, product and oil tankers
constitute about 61% of total travelled distance. It is expected an increase of all ship types until
2030, but especially gas and oil tankers, fishing vessels are however expected a reduction (Det
kongelige Miljøverndepartement, 2011). The reduction of fishing boats is due to the
implementintation of improved technology, better resource management and continued

  41  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

restructuring of the fishing vessel. The reduces of fishing vessel does not necessarily mean less tons
caught fish, but the fishing boats sails less to catch allocated quotas (DNV & Kystverket, 2012).
There are also still some uncertainties with regards to the possible opening of traffic in the north
easth passage, but if the passage opens in the coming years, the transit between Europe and Asia
will most likely pass through this passage. This will lead to a big increment of the ship traffic in the
Barents Sea, and there will be a proportional increasement in the Russian transit aswell, it should
then be considered a separation zone even further north than today.
There are some actions with respect to ship collision avoidance made, since 2002. All new ships
and later all large seagoing vessels and all passenger vessels are required to carry an AIS on board.
After approval by the International Maritime Organization (IMO) there was 1st of July 2008
established a sailing led-system in the Norwegian economic zone from Vardø to Røst.
If a ship collision finds place in the Barents Sea in the coming year due to the increased ship
activity in this area, it will have a large impact on the vulnerable environment. There are done
several studies of the effects of oil spill in the Barents Sea, and how it will affect the environment;
this is because the Barents Sea has many vulnerable components, as sea and birdlife. Especially
with respect to the long response time for fields that are far from land. This can be critical in
relation to the large spread of the oil before the collection gets started. A thorough and professional
assessment must be made in relation to the upgrading of the emergency preparedness in the analyse
area to reduce the response time to a minimum.
By looking at figure 3.2, we observers how the causation probability is calculated, and how the
special wind and weather conditions will affect the values of the causation factor used in this thesis.
With greater values at for example the visibility in the faul tree, would have led to greater
likelihood of the probability of ship collision regard to equation 2. This applies for all of the
environmental contributors in the Barents Sea. By having historical data on all of the environmental
contributors, an exact causation factor for the Barents Sea could have been calculated for use in
IWRAP.

  42  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

5 CONCLUSION AND RECOMMENDATIONS

To ensure a high level of safety, and continously work towards a lower probability of incidents
requirements must be made that shipping crews which operates in the Barents Sea shall have
knowledge and training in operating in arctic and subarctic environment, as well as programs to
report near-incidents and incidents for future reference and added learning.
Main conclusions about the environmental coniditions in the Barents Sea
-­‐ The Barents Sea is a clean and rich sea
-­‐ The area has mainly a low contamination level, and improvement will be difficult and costly.
-­‐ The environmental conditions in the Barents Sea will have a great impact on calculating the
causastion factor values.
If we look at the equation at that presents the likelihood for a ship collision to occur, we can
immediately conclude that the likelihood will increase in the coming year with respect to the
increasing numbers of vessels in the area. Total traffic in the lane, N, will increase and as a
consequence of that, the likelihood will increase. The remaining components in the equation are
more general, and are not specified especially for our area.
Fcp = N x Fd x P1 x P2 x P3 (3)
Conclusion of the results:
• The leg that will be the most critical is the leg to/from Polarbase, this applies for all
scenarios.
• For future measures with respect to the traffic routes, the fjord into Polarbase/Hammerfest
should be considered.
• The scenario that have the highest likelihood for collisions is the basis scenario
• Crude oil tankers and supply vessels will be the ship type with the highest likelihood of ship
collisions in all scenarios
• The most common collision type is head on and overtaking collision in all scenarios.
• By using pipelines to export gas from the Barents Sea, the amount of ship traffic, N,
regarding to equation (3) will be reduced, and therefor reduce the probability of ship
collisions.
• By analyzing the sensitivity results we can conclude that the effect of increasing the total
ship traffic with 10% will have a greater effect then decreasing it.
Based on the research conlusion and issues, I purpose the following points for future research:
• Do the same calculations, with including the consequences of a ship collision in different
locations in the Barents Sea.
• Reviewing data from other areas which have experienced an increase in ship traffic with to
ensure a proper handling of the added numbers of ships passing through Polarbase
especially
• Try to add a TSS into polarbasen; by changing the distribution of the traffic direction.

  43  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

  44  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

REFERENCES
Allen, C. H. (2005). Farewell`s Rules of The Nautical Road, Naval Institute Press, 8th edition.
Borisenkov, Y. P. and Panov, V.V (1972): Basic Results and Prospects of Research on
Hydrometeorological Conditions of Shipboard Icing. Issledovaniye Fizicheskoy Prirody
Obledeneniya Sudov (Leningrad). Cold Regions Research and Engineering Laboratory. Draft
Translation TL411, 1974, pp. 1-30.
Det Kongelige Miljøverndepartement (2011), Oppdatering av fovaltningsplanen for det marine
miljø i Barentshavet og havområdene utenfor Lofoten, Melding til Stortinget, Norway
DNV (2003), Teknisk Rapport, Olje- og energidepartementet , rapport nr. 2003-1411 revisjon nr. 02
DNV (2011), Det Norske Veritas, Miljørisikio ved akutt oljeforurensning fra skipstrafikken langs
kysten av Fastlands-Norge fro 2008 og prognoser for 2025.
DNV (2012), HFO in the Arctic-phase 2.   DNV Document No.: 2013-1542-16G8ZQC-5.
Norwegian Environmental Agency.
DNV & Kystverket (2012), Sannsynlighetsanalyse for skipstrafikk I Barentshavet Øst. Rapport nr.
2012-1174. Rev. 1, 2012-09-20.
DNV (2013), Utredninger for LOS-utvalget for Kystverket, Analyse av skipstrafikk, ulykkesstatistikk,
los relaterte aktiviteter og sjøsikkerhetstiltak. Rapport nr. 2012–1479. Høvik
ENI Norge (2014), http://www.eninorge.com/no/Feltutbygging/Goliat/ (Located: 04.03.14)
Friis-Hansen, P., (2008) Basic Modelling Principles for prediction of Collision and Grounding
Frequencies. Technical University of Denmark.
Geijerstam, K. A. & Svensson, H. (2008), Ships Collision Risk, Lund University, Sweden.
IALA (2012), IWRAP Mk2 v.4.0.0 Manual, Tracking and Monitoring Solutions.
IALA Web (2014), http://www.iala-
aism.org/wiki/iwrap/index.php/Interpretation_of_causation_factors_in_IWRAP. (Located:
21.05.14)
IMO (2007), Formal Safety Assessment. Consolidated text of the Guidelines for Formal Safety
Assessment for use in the IMO rule-making process. Maritime Safety Committee,
International Maritime Organisation
IWRAP (2014), http://www.iala-aism.org/wiki/iwrap/index.php/Introduction_to_IWRAP, (Located:
21.05.14).
Jacobsen, S.R. (2010) Evacuation from Petroleum Facilities Operating in the Barents Sea,
University of Stavanger
Jacobsen, S. R. (2012) Evacuation and Rescue in the Barents Sea, Critical issues for safe petroleum
activity. Master Thesis. University of Stavanger.
Jutta, Y., (2010), Modelling Marine Accident Frequency, Aalto University
Kristiansen, S. 2005, Maritime Transportation. Safety Management and Risk Analysis. ISBN 07506
59998. Publisher: ELSEVIER, Butterworth Heinemann.
Koldenhof, Y., Van der Tak, C., Glansdorp, C.C, (2009) Risk Awareness; a model to calculate the
risk of ship dynamically,” in the XIII Int. Scientific and Technical Conf. On Maritime Traffic
Engineering, MTE09
Kystverket (2010), Aktivitetsbeskrivelse av skipstrafikk i Lofoten og Barentshavet. Avdeling for
transportplanlegging, plan og utredning (TPU),
Kystverket 2012, Polarkoden – definisjon av polare områder for Utenriksdepartementet
Kystverket (2014), E-mail correspondance with Trond Langemyr.
Lee, G. W. U. & Parker, J. (2007), Managing Collision Avoidance at Sea – a practical guide, The
Nautical Institute
Lundin (2014), http://www.lundin-petroleum.com/Documents/ot_co_presentation_e.pdf (Located
10.03.14)

  45  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

Løset, S., Shkinek, K. N., Gudmestad, O. T., Høyland, K. V., (2006) Actions from ice on arctic
offshore coastal structures, ISBN 5-8114-0703-3
Meteorologisk institutt, (2012), Vær, is og andre fysiske utfordringer ved Barentshavet sørøst,
Konsekvensutredning for Barentshavet sørøst. Utarbeidet på oppdrag fra Olje- og
energidepartementet
Ministry of Petroleum and Energy (2014).
http://www.regjeringen.no/upload/kilde/oed/rap/2003/0004/ddd/pdfv/162996-
scenarier_barentshavet_0710.pdf , [located 27.02.14]
Mou, J. M., Tak, C. V. D. & Ligteringen, H. (2010), Study on collision avoidance in busy
waterways by using AIS data.
http://www.sciencedirect.com/science/article/pii/S002980181000020X (Located: 15.02.14)
Mulherin, N. D., Epper, D.T., Proshutinskii, T. O., Proshutinskii, A. I. U, Farmer, L.D. and Smith,
O. P. (1996): Development and Results of a Northern Sea Route Transit Model, U. S. Army
CRREL Report 96-05, May 1996.
Nielsen, M. & Petersen, J. (2004) Collision avoidance at sea – practice and problems.
http://www.iau.dtu.dk/secretary/pdf/EAM01.pdf, (Located: 15.01.14)
Norweigan Maritime Directorate, Arctic Shipping Activities Into the Next Decade
https://serv.ie.dal.ca/biblio/files/nmd03.pdf (Located: 17.02.14)
Petro (2014), http://oljefakta.petro.no/felt/snohvit (Located: 17.02.14)
Petroarctic (2014), http://www.petroarctic.no/library/Files/Petro-arctic/rystad/Petro-foresight-
/PetroForesight_2013_0811_kl13.pdf, (Located: 20.05.14)
SNL (2014), (http://snl.no/Barentshavet).[located : 10.03.14]
Statoil (2014)
http://www.statoil.com/no/environmentsociety/environment/impactassessments/fielddevelop
ments/pages/skrugard.aspx (Located: 17.02.2104)
Thelma (2010), Kalde utfordringer – Helse og arbeidsmiljø på innretning i nordområdene. Rapport
nr. 10-31, ISBN 978-82-8032-024-7
Wang. J. & Zhang, S.M. (2000), Management of human error in shipping operations, Professional
Safety, Section: Human factors, No. 10, pp. 23-28
Ylitalo, Jutta (2010), Modelling Marine Accident Frequency, Master thesis, Aalto University.
Intorduction to IWRAP: http://www.iala-aism.org/wiki/iwrap/index.php/Introduction_to_IWRAP

Figure references
Figure 1.1-1.4 Rystad Petro Foresight 2030, 2012, http://www.gagama.no/documents?id=4264
(Located: 25.01.14)
Figure 2.1 World Atlas, 2014, http://www.worldatlas.com/aatlas/infopage/barentssea.htm
(Located 31.05.14)
Figure 2.2 Jacobsen, S. R. (2012), Evacuation and Rescue in the Barents Sea, Critical issues
for safe petroleum activity. Pp.43. University of Stavanger
Figure 2.3 Kystverket (2014), E-mail correspondance with Trond Langemyr
Figure 2.4 Meterologisk institutt, (2012),   http://www.miljostatus.no/Tema/Hav-og-
kyst/Barentshavet/Indikatorer-barentshavet/Indikator-Isutbredelse-i-Barentshavet/
(Located: 27.05.14)
Figure 2.5 Meterlogisk institutt (2012), Vær, is og andre fysiske utfordringer ved
Barentshavet sørøst. Pp. 76)
Figure 2.7 Det Kongelige Miljøverndepartement (2011), Oppdatering av forvaltningsplanen
for det marine miljø i Barentshavet og havområdene utenfor Lofoten, 2010-2011,
pp.57
Figure 2.8 Det Kongelige Miljøverndepartement (2011), Oppdatering av forvaltningsplanen

  46  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

for det marine miljø I Barentshavet og havområdene utenfor Lofoten, 2010-

2011.pp.58
Figure 2.9 Embrey, D.E. (1992), Incorporating management and organisational factor into
probablity safety assessment, Reliability Engineering and System Safety. Pp. 3
Figure 2.10 The Norwegian Space Agency, AISSAT-1, http://snl.no/AISSat-1
(Located:01.05.14)
Figure 3.2 (http://www.iala-
aism.org/wiki/iwrap/index.php/Causation_Probability_Modelling).

Table references
Table 2.1 Meteorologisk institutt (2012), Vær, is og andre fysiske utfordringer ved
Barentshavet sørøst. Pp. 63
Table 2.2 Jacobsen S.R. (2012), Evacuation and Rescue in the Barents Sea, Critical issues
for safe petroleum activity. Master thesis.
Table 2.4 DNV (2003), Teknisk Rapport, Olje- og energidepartementet, rapport nr. 2003-
1411 revisjon nr. 02
Table 2.5 DNV & Kystverket (2012), Miljø- og risikoanalyse for skipstrafikk I
Barentshavet sørøst. Pp. 9-10
Table 2.7-2.8 Kystverket (2014), Powerpoint from Trond Langemyr.

  47  
 Risk of Ship Collision in The Barents Sea in 2030 / Master Thesis 2014

  a