DNV RESEARCH INNOVATIONresearch.dnv.com/skj/FSA-ENC/ENC.pdf · adequate for fast generation of computer images on the screen. In order to get data structures that facilitate rapid

TECHNICAL REPORT

DET NORSKE VERITAS

DNV RESEARCH & INNOVATION

EFFECT OF ENC COVERAGE ON ECDIS RISK

REDUCTION

REPORT NO. 2007-0304 REVISION NO. 01

DET NORSKE VERITAS

TECHNICAL REPORT

ECDIS - ENC report FINAL 2007-04-13.doc

Tel: Fax: http://www.dnv.com

Date of first issue: Project No.:22.03.2007 91001117 Approved by: Organisational unit:Rolf Skjong BRINO911 Client: Client ref.:Finnish Maritime Administration, Norwegian Mapping Authority - Hydrographic Service, Swedish Maritime Organisation, National Survey and Cadastre (Denmark)

Carolus Ramsay, Frode Klepsvik, Johan Fransson, Jens Petter Hartmann

Summary:

In this report, the results from a study on the effect of Electronic Navigational Chart (ENC) coverage on Electronic Chart Display and Information System (ECDIS) risk reduction is reported. Global traffic data for cargo ships has been evaluated in relation to the present and future global coverage of ENC to provide an overview. Also, eleven specific ship routes, representative for global merchant shipping, has been analysed to assess the ECDIS risk reducing potential in light of actual ENC coverage along these routes. Currently, the global coverage of suitable ENC lie between 82% and 94%. The coverage along selected routes varied from 28% to 100%, with four of the eleven routes having 100% coverage. Cost effectiveness assessments were carried out, and it was found that there were big differences between oil tankers and other cargo ships. Hence, it was deemed appropriate to formulate different recommendations for oil tankers and other cargo ships. Based on the analyses carried out in this study, and the current cost efficiency criteria used at IMO, the following recommendations on mandatory carriage of ECDIS may be proposed, assuming ship lifetime 25 years:

Oil tankers Other cargo ships

new ships > 500 GT. new ships > 3,000 GT. existing ships> 3,000 GT if not older than 20 years. existing ships> 10,000 GT if not older than 20 years. existing ships > 10,000 GT irrespective of age. existing ships > 50,000 GT irrespective of age.

Report No.: Subject Group:2007-0304 Indexing terms

Report title: Key words Service Area

Market Sector

Effect of ENC Coverage on ECDIS Risk Reduction

• ECDIS • ENC • Grounding • FSA

Work carried out by:

Erik Vanem, Gjermund Gravir and Magnus Strandmyr Eide

Work verified by: Arve Lepsøe

Date of this revision: Rev. No.: Number of pages:10.04.2007 01 65

No distribution without permission from the client or responsible organisational unit (however, free distribution for internal use within DNV after 3 years)

No distribution without permission from the client or responsible organisational unit.

Strictly confidential Unrestricted distribution

© 2007 Det Norske Veritas AS All rights reserved. This publication or parts thereof may not be reproduced or transmitted in any form or by any means, including photocopying or recording, without the prior written consent of Det Norske Veritas AS.

DET NORSKE VERITAS

Report No: 2007-0304, rev. 01

TECHNICAL REPORT

Page i Reference to part of this report which may lead to misinterpretation is not permissible.


Table of Content Page

1 CONCLUSIVE SUMMARY....................................................................................... 1

2 ABBREVIATIONS ..................................................................................................... 3

3 INTRODUCTION AND BACKGROUND................................................................. 4 3.1 ENC and ECDIS 4 3.2 Definition of scope 6 3.3 Grounding scenarios and risk model 7 3.4 Description of data sources 9

4 ECDIS COST ESTIMATES...................................................................................... 17 4.1 NPV estimates 18

5 WORLDWIDE ENC COVERAGE........................................................................... 20 5.1 Officially approved vs. unofficial electronic navigational charts 20 5.2 Suitable ENC coverage 20

6 ENC COVERAGE FOR SOLAS SHIPS................................................................... 22 6.1 2006 coverage estimates 24 6.2 Anticipated 2010 coverage estimates 25

7 REPRESENTATIVE SHIPPING ROUTES.............................................................. 27 7.1 Method for selecting representative shipping routes 27 7.2 Selected representative shipping routes 27

8 EFFECTS OF ENC HOLES ALONG SELECTED ROUTES ................................. 32 8.1 Effect of holes in ENC coverage 32 8.2 ENC coverage along selected routes 32 8.3 Estimated grounding risk reduction on selected routes 54 8.4 Estimated cost effectiveness of ECIDS on selected routes 57

9 COST-EFFECTIVENESS OF ECDIS IN LIGHT OF ENC COVERAGE .............. 59 9.1 Cost-effectiveness for new ships 59 9.2 Cost-effectiveness for existing ships 61

10 CONCLUSIONS AND RECOMMENDATIONS .................................................... 64

REFERENCES......................................................................................................................... 65

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT

Page 1 Reference to part of this report which may lead to misinterpretation is not permissible.


1 CONCLUSIVE SUMMARY This report outlines a recent study on the risk reducing effect of Electronic Chart Display and Information System (ECDIS) to improve the navigational safety on cargo ships. Previous studies have shown that ECDIS is a cost effective risk control option for large passenger ships and for certain cargo ships sailing particular trades. The aim of this study has been twofold: 1) to take the actual coverage of Electronic Navigational Charts (ENC) into account and adjust the risk reduction effect of ECDIS accordingly and 2) to extend the scope to include all cargo ships of different size (above 500 GT) and age. This has been achieved using three complementary approaches. First, actual historic traffic data has been overlaid actual ENC coverage to evaluate the extent of global coverage for suitable ENC. Secondly, eleven specific routes has been analysed to evaluate the grounding risk reducing potential of ECDIS, when considering actual ENC coverage. Finally, the average grounding risk reduction has been used to evaluate cost-effectiveness for ECDIS for varying ship types, sizes and ages.

The study on actual current and near future ENC coverage showed that: a. The global coverage of suitable ENC in coastal areas currently lies between 84% and

96% and is expected to increase to 87 – 98% within a few years. b. The coverage of suitable ENC along selected routes varies between a minimum of

28% to a maximum of 100% for selected representative routes. c. 36% of the selected routes (4 of 11 routes) already have 100% ENC coverage in

coastal areas. d. For 5 of the 7 routes currently without 100% ENC coverage, an increase of coverage

is planned for the near future. e. The grounding frequency reductions achievable from implementing ECDIS vary

between 11% and 38% for the selected routes. This variation is due to variations in ENC coverage.

f. It is expected that ECDIS may result in 9.1 x 10-3 groundings averted per shipyear on average, and that this will increase to 10.0 x 10-3 groundings per shipyear in the near future, when planned extended ENC coverage become available.

The assessment of cost effectiveness revealed that Net Cost of Averting a Fatality (NCAF) is less than USD 3 million for various combinations of ship types, sizes and ages. For most of these combinations, NCAF < 0 Based on the analysis reported herein, and the current cost efficiency criteria used at IMO, the following recommendations may be proposed:

i. ECDIS should be made mandatory for all new oil tankers of 500 gross tonnage and upwards.

ii. ECDIS should be made mandatory for all new cargo ships, other than oil tankers, of 3,000 gross tonnage and upwards.

iii. ECDIS should be made mandatory for all existing oil tankers of 3,000 gross tonnage and upwards.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



iv. ECDIS should be made mandatory for all existing cargo ships, other than oil tankers, 10,000 gross tonnage and upwards.

v. Exemptions may be given to existing oil tankers of less than 10,000 gross tonnage and existing cargo ships, other than oil tankers, less than 50,000 gross tonnage when such ships will be taken permanently out of service within 5 years after the implementation dates given for iii) and iv) above.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



2 ABBREVIATIONS

AMVER Automated Mutual-assistance Vessel Rescue

COADS Comprehensive Ocean-Atmosphere Data Set

DWT Deadweight tonnes

ECDIS Electronic Chart Display and Information System

ECS Electronic Chart System

ENC Electronic Navigational Charts

FSA Formal Safety Assessment

GCAF Gross Cost of Averting a Fatality

GPS Global Positioning System

GT Gross Ton

IHB International Hydrographic Bureau

IHO International Hydrographic Organization

IMO International Maritime Organization

NCAF Net Cost of Averting a Fatality

NM Nautical mile (1 nm = 1.852 km)

NPV Net Present Value

RCDS Raster Chart Display System

RENC Regional Electronic Navigational Chart Coordinating Centre

RNC Raster Navigational Charts

SENC System Electronic Navigational Chart

SOLAS International Convention for the Safety of Life at Sea

TEU Twenty-foot equivalent unit

VOS Voluntary Observing Ships

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



3 INTRODUCTION AND BACKGROUND

3.1 ENC and ECDIS 3.1.1 ENC Only up to date official charts may be used to fulfil carriage requirements of ships. Other nautical charts are often referred to as private charts, and these are not accepted as the basis for navigation under the SOLAS convention. There are two kinds of official digital charts available, Electronic Navigational Charts (ENC) and Raster Navigational Charts (RNC).

RNC stands for Raster Navigational Charts and official RNCs are digital raster copies of official paper charts. These can only be issued by, or on the authority of, a national Hydrographic Office. According to the IMO performance standard, ECDIS operated in the Raster Chart Display System (RCDS) - mode may be used to meet carriage requirements for areas where ENCs are not available. However, in these areas ECDIS must be used together with an appropriate portfolio of up to date paper charts

ECDIS operation in RCDS - mode is acknowledged to have limitations compared to using ENCs. Hence, in order to fully exploit the risk reducing effect of ECDIS, ENCs need to be available and for the remainder of this study, the availability of RNCs will not be considered.

ENC stands for Electronic Navigational Charts. ENCs are produced by or on the authority of a government authorised Hydrographic Office or other relevant government institution. ENCs should be the responsibility of the responsible Hydrographic Office and be based on their source data or official charts. They should be compiled and coded according to international standards and regularly updated with official update information distributed digitally. All ENCs should be referred to World Geodetic System 1984 Datum (WGS84), the world-wide datum used by Global Positioning System (GPS). For the purpose of this study, only ENCs will be considered.

ENCs are vector charts compiled from a database of individual geo-referenced objects from Hydrographic Offices’ archives. IMO offer the following definition for ENC [1]: ENC means the database, standardized as to content, structure and format, issued for use with ECDIS on the authority of government-authorized hydrographic offices. The ENC contains all the chart information necessary for safe navigation, and may contain supplementary information in addition to that contained in the paper chart (e.g. sailing directions) which may be considered necessary for safe navigation. Being a database, ENC content may be continuously retrieved by special operational functions in ECDIS to give warnings of impending danger related to the vessel’s position and its movements.

ENCs are optimized to absorb the Hydrographic object information and this structure is not adequate for fast generation of computer images on the screen. In order to get data structures that facilitate rapid display of ENC data, ECDIS first converts each ENC into an internal format called System Electronic Navigational Charts (SENC) which is optimized for creating chart images. In contrast to the ENC format that is common and uniform, SENC formats are proprietary for each ECDIS manufacturer. Presentation rules for the display of the abstract geographic entities of ENCs are contained in the presentation library as a separate ECDIS software module.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



3.1.2 ECDIS ECDIS (Electronic Chart Display and Information System) is a type of navigational electronic chart system that may be installed on the bridge of a vessel. The IMO ECDIS Performance Standards [1] defines ECDIS equipment as follows: Electronic chart display and information system (ECDIS) means a navigation information system which, with adequate back-up arrangements, can be accepted as complying with the up-to-date chart required by regulation V/20 of the 1974 SOLAS Convention, by displaying selected information from a system electronic navigational chart (SENC) with positional information from navigation sensors to assist the mariner in route planning and route monitoring, and by displaying additional navigation-related information if required.

Another class of navigational electronic chart systems exist, simply referred to as Electronic Chart System (ECS). Such systems do not meet the SOLAS chart carriage requirements. Hence, the use of ENCs in a tested, approved and certified ECDIS (with appropriate back-up arrangements) is the only alternative option to paper charts for vessel navigation. Appropriate back-up systems may either be in the form of paper charts or an independent, separate ECDIS. For the purpose of this study, dual ECDIS are assumed, i.e. with a complete, independent ECDIS as the back-up arrangements.

In order to be an ECDIS, equipment must be shown to meet a number of requirements laid down by the performance standards. I.e. it must support the whole range of navigational functions that make use of the characteristics of the chart data and their specific presentation. The performance standards contain requirements related to i.a.:

• Display of SENC information • Display of other navigational information • Display requirements for route planning and monitoring • Provision and updating of chart information • Scale indication • Colours and symbols • Route planning, monitoring and voyage recording • Accuracy • Performance tests, malfunction alarms and indications • Back-up arrangements • Power supply

Within the ECDIS, a database of electronic nautical charts (ENC) store chart information in the form of geographic objects represented by point, line and area shapes carrying individual attributes that make each object unique. Mechanisms are built into the ECDIS system so that the data can be inquired and used to perform certain navigational tasks such as anti-grounding surveillance. The ECDIS performance standards also state that the use of ECDIS should reduce the navigational workload related to route planning, route monitoring and positioning compared to the use of paper chart. This means that navigational risks could be reduced when using ECDIS compared to traditional paper charts.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



3.1.3 Previous FSA studies on ECDIS Previously, two studies on navigational safety have been reported to IMO where the effects of ECDIS have been evaluated in particular. These studies focused on large passenger ships [2, 3] and on other shiptypes such as oil tankers, product tankers and bulk carriers [4, 5]. The conclusions arrived at in these previous studies were:

a. ECDIS is a cost effective risk control option for large passenger ships, with a significant potential to save lives by reducing the frequency of collision and grounding

b. ECDIS is a cost effective risk control option for all other vessel types engaged in international trade, with the exception of the smallest vessels.

c. The potential for saving lives is small for cargo ships, but ECDIS represents a net economic benefit in itself.

However, the previous studies did not consider the coverage of ENCs in detail or the effect of this coverage on the ECDIS performance, and the simplifications and assumptions in relation to this introduces uncertainty in the above conclusions. The current study was initiated in order to investigate this assumption in more detail, and to evaluate the actual effect of ECDIS given the actual coverage of ENC. This will be done both for the current situation (2006) and for the anticipated ENC coverage in 2010.

3.2 Definition of scope The scope of this study is to extend the previous studies on ECDIS with considerations of the actual ENC coverage worldwide. Thus, the assumption and simplifications regarding ENC coverage in previous studies will be challenged, and the effect of this will be investigated.

It is expected that ECDIS will reduce the risks related to several generic accident scenarios such as grounding, collision, contact etc. However, for the purpose of this study, only reductions in grounding risks will be analysed, and the effect on other scenarios is considered out of scope.

The study will utilize existing knowledge from the previous studies whenever such knowledge is available and relevant. For example, the risk models developed using Bayesian Networks in the previous FSAs will also be exploited for the purpose of this study. It is deemed out of scope to develop new risk models or to modify the existing ones. However, minor adjustments to the accident cost models will be made in order to accommodate other shiptypes and trades.

It is expected that the costs associated with acquisition and installation of ECDIS equipment may have changed since the original study. Hence, the cost models will be updated accordingly. New cost information will be collected from ECDIS providers in order to adjust cost estimates to the present situation.

The main focus of this study is on the global coverage of ENC. Thus, world ship traffic density distributions will be compared to ENC availability in order to estimate the ratio of world shipping traffic having suitable ENC coverage for safe navigation. This will be done for the present situation (2006) as well as for the anticipated situation in 2010. For both 2006 and 2010 traffic data from 2000/01 will be used. Furthermore, more detailed studies on a number of selected representative routes will be carried out in order to assess the effects of ENC holes along these routes.

It is within the scope of this study to provide a best estimate of the cost effectiveness associated with carriage requirements for ECDIS. Expected risk reduction related to fatalities, property loss

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



and environmental damage will all be within scope. The cost effectiveness will be assessed for both newbuildings and existing ships, and it will be investigated to what extent the cost effectiveness will be dependent on the years left in operation for existing ships, i.e. on the age of the ship. This should form the basis for discussions regarding possible ECDIS requirements for existing ships.

3.3 Grounding scenarios and risk model The grounding scenarios defined in previous IMO submissions on navigational safety will be adopted also in this study. For a full description of these, reference is made to these earlier IMO submissions [2, 5]. A summary of the grounding scenarios will be briefly outlined in the following.

It is normally distinguished between two types of ship grounding events, i.e. powered grounding and drift grounding. Drift groundings are events that occur because the vessel is unable to follow a safe track due to e.g. mechanical failure (loss of propulsion, loss of manoeuvrability) or adverse environmental conditions, and these cannot be prevented by navigational measures. Thus, only the powered grounding scenario will be relevant in this study. Henceforth, grounding is taken to mean powered grounding, if not specified otherwise.

The following five grounding scenarios have been considered, as illustrated in Figure 1: 1. Course towards shore, supposed to change course - does not turn 2. Course along shore, not supposed to change course - turns towards shore 3. Course along shore, drift-off, should correct course - does not correct course 4. Wrong position, should steer away from object - does not steer away 5. Meeting/crossing traffic, supposed to give way - gives way, steers towards shore

Figure 1: The five grounding scenarios

In the previous FSA studies covering ECDIS, grounding risk models incorporating these five scenarios were developed, utilizing Bayesian network models combined with Excel models. These will be exploited in this study as well, and there will not be developed any new risk models for the purpose of this study. However, the Excel models will be slightly modified according to the characteristics of the specific routes that are selected for more detailed investigation. Figure 2 illustrates the grounding risk model adopted for this study.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



GROUNDING

Course towards shore Vessel loose control Technical

Human performance

Safety Culture

Management factorsPersonnel factorsWork conditions

No/insufficient action

Vigilance

Vessel damage

Fatalities

1

23 4

5

Figure 2: The grounding risk model

Due to the complexity of the risk model, it will not be described in full detail herein. The complete model may be found in Appendix A to Annex II of the full FSA report on cruise navigation where also the probability input to the grounding network is included. The nodes from the grounding network are described in Appendix C and the Excel model describing the exposure is included in Appendix D of Annex II of the full FSA report on cruise navigation [6]. However, one particular feature of the risk model is noted: the risk of grounding is assumed negligible for waters where there are no obstacles within 5 nautical miles in any direction.

3.3.1 Consequence model ECDIS is a measure for reducing the probability or the frequencies of groundings, and the possible consequences will not be influenced by whether ECDIS are implemented or not, or by the ENC coverage. Therefore, a simplistic consequence model will suffice.

For the purpose of this study, the consequence model utilized in the previous FSA study will be adopted [5]. These modelled results are deemed appropriate for the purpose of this study. It should be noted that other potential benefits form ECDIS have not been quantified in the current report. ECDIS may, all other things equal, lead to fewer collisions due to the fact that the officer on watch will have more time for lookout.

Fatality rates for generic tankers and bulk carriers were estimated in the previous FSA study [5] based on the risk models developed as Bayesian Networks. According to these models, the probabilities of fatality were 2 x 10-4 for tankers and 6 x 10-4 for bulk carriers per crewmember

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



and per grounding event. These estimates will be adopted for the current study as well. Furthermore, the fatality estimate for generic tankers will be assumed applicable to oil tankers, chemical tankers, LNG carriers and container vessels. The fatality estimate for bulk carriers will be assumed adequate for bulk carriers and general cargo carriers for the purpose of this study. These assumptions agree reasonably well with generic FSA studies carried out at DNV [7, 8, 9, 10] with the exception of general cargo vessels, for which the statistics indicates that the fatality risk associated with grounding may be 3 times as high as for bulk carriers [11], and LNG carriers which have not experienced any fatalities due to grounding to date [12]. However, for the purpose of this study, the assumptions will be adequate, and it is noted that because of this assumption, the results for general cargo in particular will tend to be conservative. The average fatality probabilities (per crewmember and per grounding) for the different shiptypes used in this study are presented in Table 1.

Table 1: Fatality probabilities from risk model for different shiptypes

Oil tanker Bulk carrier Chemical tanker LNG carrier General cargo Container

2 x 10-4 6 x 10-4 2 x 10-4 2 x 10-4 6 x 10-4 2 x 10-4

In addition to a number of fatalities, other consequences of a grounding incident might be environmental damage (e.g. due to oil spill), property damage (e.g. to ship and cargo) and business interruptions. These consequences will be estimated in terms of monetary costs.

The cost models proposed in [13] will be exploited for the purpose of this study for the various shiptypes that will be investigated. This cost models contains direct costs of an accident (i.e. repair cost, cargo loss, salvage and search and rescue), opportunity costs due to unavailability of the ship and the costs associated with environmental damage. However, for the environmental damage category, the associated cost used in this study will be based on the CATS approach, i.e. a cost of 60,000 USD per tonne of oil spilt [14, 15].

The cost model distinguishes between total losses, serious damages and non-serious damages. In order to estimate the total average cost of a grounding accident, it is assumed that there will be ten serious accidents for every total loss, and three non-serious accidents for every serious accident (including total losses). This is in general agreement with previous generic ship type FSA studies carried out at DNV [13, 7, 8, 11]. Hence, the breakdown of accident severity categories will be as follows:

• 75% non-serious accidents • 22.7% serious accidents • 2.27% total losses

3.4 Description of data sources In order to carry out this study, two types of data are essentially needed, i.e. to provide estimates of the distribution of the global ship traffic and to provide information about the global coverage of ENCs. In the following, the sources used to obtain this information will be briefly described.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



3.4.1 AMVER – Data source for ship traffic distribution AMVER1 (Automated Mutual-assistance VEssel Rescue) is a system used for tracking the positions of participating merchant vessels at sea, and data from this system is chosen to represent the global distributions of cargo ships in international trade. Participation in AMVER is free, voluntary and generally limited to merchant ships above 1,000 GT on voyages longer than 24 hours. However, recently enrolment has been expanded to accommodate other types of vessels such as cruise ships, research vessels and fish processors. Ships of all flags may participate.

Ships that participate in AMVER submit information including their position, ship type and size to the AMVER database. The main motivation behind participation is that, by regular reporting, someone knows the position of a ship at all times during voyage in case of an emergency, reducing the time between notification and rescue response. The system is used worldwide by search and rescue authorities to arrange for assistance to ships and persons in distress at sea. Currently, more than 12,000 ships from over 140 nations participate in AMVER. More than 100,000 voyages are tracked annually with around 3,000 daily plots on average.

AMVER data is used to produce monthly density plots illustrating the distribution of AMVER participating vessels. Plots from 2003 and onwards are available at the AMVER website and Figure 3 shows an example density plot for January 2007. AMVER data are generally not stored and aggregated for extended periods of time. However, AMVER data corresponding to a whole year for 2000/2001 has been collected [16], and this aggregated data set will be used in the current study. It is hence assumed that this set of AMVER data covering a whole year will be representative for the current distribution of ship traffic densities. This dataset is illustrated in Figure 4. Comparing the two plots, it appears that the differences between the distribution for 2000/2001 and January 2007 are insignificant.

Figure 3: AMVER density plot for January 2007 (from AMVER website)

1 AMVER website: http://www.amver.com

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Figure 4: AMVER data for one year 2000/2001

3.4.2 COADS – Supplementary data source for ship traffic distribution In order to supplement the ship traffic data from AMVER, an additional set of data has been exploited, i.e. the Comprehensive Ocean-Atmosphere Data Set (COADS). This is an extensive collection of ship location data based on reporting of routine meteorological observations from the Voluntary Observing Ships (VOS) fleet. The VOS fleet contains about 7,000 vessels, of which about 1,800 cargo and passenger vessels.

For the purpose of this study, it was assumed that a joint AMVER and COADS data set would provide the best overall description of world shipping traffic data. Hence, global traffic distributions are based on a dataset containing both COADS and AMVER data corresponding to a complete year. However, when distributions for particular shiptypes are needed, only AMVER data will be used. The joint ship traffic density distribution from AMVER and COADS data sets are illustrated in Figure 5.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Figure 5: Combined AMVER and COADS data for one year 2000/2001

3.4.3 Data sources for ENC coverage A catalogue with an overview of worldwide coverage of ENC is available from the IHO website2. The coverage can be investigated for the different usage bands (i.e. “overview”, “general”, “coastal”, “approach”, “harbour” and “berthing”) and the catalogue distinguishes between ENCs in three different maturity levels:

• ENCs that are commercially available • ENCs that have been completed (issues), but are yet not made commercially available • ENCs that are in production or are planned for production

It should be noted the IHO ENC catalogue will not be completely up to date, and some areas might have better coverage than the catalogue indicates. In particular, actual ENC coverage in the following areas is known to exist even though they are not included in the catalogue3:

• Straights of Malacca, Malaysia • Tunisia • Canada • North-west Russia • India (charts labelled as completed but not commercially available or in

production/planned for production in the catalogue should be commercially available by now)

Notwithstanding the known gaps in the IHO DNC catalogue, this will still be used as the primary source of information for the purpose of this study. The advantages of this approach are the transparency of using a single, publicly available source and that it is known in which direction the results will be biased, i.e. the results will tend to be conservative.

2 http://www.iho.shom.fr 3 According to personal communications with Tony Pharaoh at IHB on March 6. 2007

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



The global availability of ENCs with detail levels equal to or above “coastal” level (i.e. “coastal”, “approach”, “harbour” and “berthing”) according to the IHO ENC catalogue is shown in Figure 6 and Figure 7 below. Figure 6 displays charts that are available and Figure 7 displays charts that are either in production/planned for production or completed but not commercially available. The figures are taken from the IHO ENC catalogue as of February 2007. It can be seen that extended coverage of ENCs is expected in the time to come, most notably in areas of South-America, the Indian subcontinent, North-East Asia and parts of the Mediterranean.

One of the aims of the current study is to investigate the current ENC coverage as well as the anticipated coverage in 2010. Hence, it will be assumed that all currently completed ENCs and all ENCs currently in production or planned for production will be available in 2010. Thus, the information from the IHO ENC catalogue will be used to establish the global coverage of ENC for the current situation as well as for 2010. IHO has provided further information regarding the available ENCs.

It is noted that this is a rather conservative approach of two reasons: 1) the IHO ENC catalogue is presumably not completely up to date, and the coverage according to this catalogue will be less than the actual coverage. 2) The timeline from planning until an ENC is commercially available may be as little as a few months in certain cases. Therefore, the additional coverage in 2010 will presumably be considerably larger than what is currently planned or in production. At any rate, the IHO catalogue will be used as the prime source of information for estimating the ENC coverage, but it should be kept in mind that this will introduce a conservative bias into the analysis and the results arrived at herein. Figure 8 shows the current availability as well as the anticipated future availability of ENCs in the categories “Overview” and “General”. Such maps are needed for route planning and navigation in areas not regarded as critical for grounding risk. Figure 8 illustrates that the coverage of such maps are good worldwide, and that areas not covered by maps of resolution “Coastal” or higher (Figure 6 and Figure 7) are often covered by lower resolution maps.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Figure 6: Global coverage of ENCs of detail level “Coastal” or better, commercially

available

Figure 7: Global coverage of ENCs of detail level “Coastal” or better, that are not yet

commercially available (blue), and in production or planned for production (red)

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Figure 8: Global ENC coverage for map categories “Overview” and “General”

ENCs are made commercially available through resellers as well as two Regional Electronic Navigational Chart Coordinating Centre’s (RENCs), i.e. PRIMAR Stavanger and IC-ENC. Both RENCs maintain their own ENC coverage catalogues and these catalogues may serve as additional information on global ENC coverage. For some areas these catalogues may be expected to be kept better up-to-date than the IHO catalogue. However, there are some countries that do not have any agreement with the RENCs, and the focus of the RENC catalogues is on the ENCs that they provide. Hence, these catalogues could not be expected to be complete either. According to estimates received from IHB4, there are about 6000 ENC available globally, and 4000 of these are distributed by PRIMAR Stavanger and IC-ENC, i.e. about 67% of globally available ENCs are available through the RENCs.

The global PRIMAR Stavanger ENC catalogue for coastal, approach, harbour and berthing charts is illustrated in Figure 9. The corresponding IC-ENC catalogue is shown in Figure 10. For illustrative purposes, the ENC coverage for the Straights of Malacca and parts of Malaysia is compared in Figure 11 for the three sources outlined above, and it can easily be seen that the catalogues are not identical for this particular area. Other areas may have similar discrepancies.

Nevertheless, for the purpose of this study, a single source (the IHO global ENC catalogue) has been used for estimating the global ENC coverage, and no attempts have been made to merge the information available from the various sources into one single dataset. The gaps in this source are acknowledged, and it is noted that this will lead to conservative results and conclusions.

4 Personal communications with Tony Pharaoh at IHB on March 6. 2007

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Figure 9: PRIMAR chart catalogue for ENCs of detail level “Coastal” or better

Figure 10: IC-ENC catalogue for ENCs of detail level “Coastal” or better

Figure 11: Catalogue comparison for Straights of Malacca. Left: IHO catalogue (online),

Middle: PRIMAR Stavanger catalogue, Right: IC-ENC catalogue

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



4 ECDIS COST ESTIMATES The implementation of ECDIS will be associated with a cost related to acquisition and installation, training and maintenance. The previous study [4, 5] estimated the Net Present Value (NPV) of such cost to be USD 75,000. However, these estimates are not considered to be up to date, and new cost data has been collected in this study. The cost model will be similar to what has been used previously with one exception: Related to training costs, also costs related to travel, board and lodging and overtime pay will be included in the updated cost model. The following cost elements are included in the ECDIS cost model:

• Initial acquisition and installation costs of dual ECDIS (including ECDIS as backup arrangements)

• Initial training cost • Annual maintenance cost

It should be noted that no extra cost is associated with keeping an updated ENC portfolio, as this cost is assumed to be no greater than maintaining an updated portfolio of paper charts. This is in accordance with the previously used cost model for ECIDS [4, 5]. In order to estimate the initial installation and acquisition costs, four ECDIS providers in three different European countries were contacted, and their cost estimates agreed surprisingly well. For newbuildings, the highest cost estimate was USD 43,000 and the lowest was USD 39,000. For retrofit on existing vessels, the highest cost estimate was USD 44,000 and the lowest was USD 39,000. Hence, the initial cost of installing a dual ECDIS is assumed to be USD 40,000 for both newbuildings and existing ships.

There are no specific references to ECDIS in the STCW convention but it is stated that mariners should be competent to carry out the duties they are expected to perform. For ships using ECDIS as the primary means of navigation, the watchkeeping officers and master should therefore be properly trained in the operation and use of such systems before using them operationally at sea. This may be achieved by e.g. attending IMO model course 1.27 or similar courses. Even though there are room for interpretations related to the extent of ECDIS training that is required by current regulations, for the purpose of this study, it is assumed that an ECDIS course is required for bridge personnel for ships where ECDIS is the primary means of navigation. Hence, these additional costs are taken into consideration in the following. The training cost should also include costs related to travel, lodging, board and overtime pay in connection with the course. Several ECDIS courses are currently offered, and the fees for classroom courses vary from USD 550 to almost USD 1500 for courses of 3 to 5 days duration5. Computer based on-board training on ECDIS is also offered which might be a cheaper option. However, for the purpose of this cost effectiveness assessment, it is assumed that necessary training will be achieved at four-day courses and an average course fee of USD 1000 is assumed. It is further assumed that a total of 6 officers per vessel need to attend this course and that there will be no requirement for repetition courses. The estimates for travel costs, board and lodging and overtime pay for attendance at a four-day course are presented in the summary below. 5 Based on information about courses offered at SMSC in Norway (http://www.smsc.no/), Seagull AS in Norway

(http://www.seagull.no), IDESS in the Philippines (http://www.idess.com/), Lairdside Maritime Centre in UK (http://www.lairdside-maritime.com/) and Warsash Maritime Centre in UK (http://www.warsashcentre.co.uk/).

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



ECDIS training is currently a part of most curricula in the education of new navigational officers, and it is assumed that future newly qualified mariners will not need this additional training courses. Thus, training costs are assumed to be limited to sending 6 officers on a four-day training course when ECDIS is implemented on the vessel, i.e. it is an initial cost and no future costs are assumed in relation to training of officers.

The annual maintenance cost is assumed unchanged since the original study on navigational safety, and this estimate will be kept for the current study, i.e. USD 500 per year on average. The costs associated with implementing ECDIS assumed in this study are:

Initial acquisition and installation cost, dual ECDIS USD 40,000 Training cost, per officer

Course fee USD 1,000 + Travel cost USD 1,000 + Board and lodging USD 700 + Overtime pay USD 800 х Number of officers to attend course 6 USD 21,000 Maintenance costs - annual USD 500

Hence, implementation of ECDIS for paperless navigation will be assumed to be associated with an initial cost of 61,000 USD and an annual cost of USD 500 per year throughout the lifetime of the ship. The net present value will be calculated based on a depreciation rate of 5%.

4.1 NPV estimates For newbuildings, the net present value of the costs presented above will depend on the expected lifetime of the ship. Typical lifetimes might be different for different shiptypes. Table 2 indicates assumed lifetimes for the six shiptypes that are included in this study as well as the corresponding NPV of ECDIS costs. All shiptypes’ lifetimes are assumed equal to the average historic scrapping age according to statistics from Lloyd’s Register – Fairplay World Fleet Database (February 2007) except for LNG carriers. For this shiptype, too little statistics are available, but it is recognized that many LNG vessels are still in operation after more than 40 years.

Table 2: NPV of ECDIS cost for various shiptypes

Shiptype Expected lifetime (years)

NPV

(USD) Oil tanker 22 67,400 Container ship 26 68,000 Bulk carrier 25 67,900 General cargo ship 27 68,200 Chemical tanker 23 67,600 LNG carrier 40 69,500

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



It is noted that the expected lifetime is quite similar for the various shiptypes, with the exception of LNG carriers, and that the NPV is not very sensitive to this. Hence, for the purpose of this study, an average expected lifetime of 25 years will be assumed for all shiptypes. This corresponds to a NPV of the costs associated with ECDIS of USD 68,000.

The net present value for retrofitting ECDIS on existing ships will be dependent on the expected remaining years in service. This will simply be the ship age subtracted from the expected lifetime of that shiptype. For the purpose of this study, NPV will be estimated for ships of ages 5, 10, 15 and 20 years for the various shiptypes. These estimates are presented in the tables below, rounded to the nearest 1,000 USD.

Table 3: NPV of ECDIS retrofit cost for existing ships

Ship age Expected remaining years of service

NPV

(USD) Newbuilding 25 68,000

5 years 20 67,000 10 years 15 66,000 15 years 10 65,000 20 years 5 63,000

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



5 WORLDWIDE ENC COVERAGE

5.1 Officially approved vs. unofficial electronic navigational charts An ENC is a digital chart produced by National Hydrographic Offices which complies with the IHO's (International Hydrographic Organisation's) S-57 Edition 3 product specification. In short, this is the specification for the electronic charts to be used in the approved ECDIS. However, the availability of an ECDIS system onboard a vessel enables use of not only ENCs, but also unofficial electronic charts (i.e. charts not approved by the IHO) and Raster Navigational Charts (RNCs) when ENCs are not available. However, for navigational purposes, IMO requires that paper charts are carried for areas not covered by ENCs.

5.2 Suitable ENC coverage The evaluation of global ENC coverage is not a straightforward matter, considering that the context is the effect on grounding risk. This context demands that not ENC coverage as such is considered, but coverage of suitable ENC. A suitable ENC is an ENC with sufficient level of detail for safe navigation in a specific area. For instance, navigating the mid-Atlantic safely requires an ENC with very few details, while navigating the coastal shipping lanes of Norway requires ENCs of great detail.

A very central element to consider when analysing the coverage of suitable ENCs is that there are no official requirements defining what level of detail is required in any given area. SOLAS Chapter V Regulation 9 states that coastal states are obligated to make available the charts needed for safely navigating their waters. This applies to paper charts as well as ENCs. But SOLAS also states that it is the prerogative of the ship master to select the charts suitable for navigation.

The scale of a chart is determined by the type of navigation for which it is intended, the nature of the area to be covered and the quantity of information to be included. Various scale terms are used in the Specifications, such as medium-scale, large-scale, continuous coastal series. These are intended to indicate the type of chart rather than actual scale, which may vary from area to area; the specific scale of charts cannot be defined by universally prescriptive rules. In [17], the scale ranges in Table 4 were suggested to be assigned to the 6 navigational purpose bands.

Table 4: Suggested scale ranges assigned to navigational purpose bands

Navigational purpose Name Scale range 1 Overview < 1 : 1,499,999 2 General 1:350,000 – 1 : 1,499,999 3 Coastal 1 : 90,000 – 1 : 349,000 4 Approach 1 : 22,000 – 1 : 89,999 5 Harbour 1 : 4,000 – 1 : 21,999 6 Berthing > 1 : 4,000

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



ENCs labelled “Coastal” must be assumed appropriate or suitable for navigation in waters within a certain distance from shore. In other words, inside this distance from shore, navigation without charts labelled Coastal can be assumed to be unsafe.

Furthermore, [17] recommends that the Hydrographic Offices set the compilation scales of their ENCs to be consistent with the standard radar range scales shown in Table 5 in the future. This is to facilitate the display of radar video over ENCs.

Table 5: standard radar ranges and scales ([17])

Selectable radar range Standard scale (rounded) 48 NM 1 : 700,000 24 NM 1 : 350,000 12 NM 1 : 180,000 6 NM 1 : 90,000 3 NM 1 : 45,000 1.5 NM 1 : 22,000 0.75 NM 1 : 12,000 0.5 NM 1 : 8,000 0.25 NM 1 : 4,000

Coastal ENCs have a scale ranging from 1:90,000 to 1:350,000 (Table 4), corresponding to the recommended radar range settings between 6 and 24 nautical miles according to Table 5. This means that when sailing with Coastal charts the radar screen will display objects at a distance between 6 and 24 nm from the ship. It is customary to apply a radar range setting which displays the coastline at the edge of the screen. Thus, it can be assumed that coastal charts should be available when sailing closer than 24 nautical miles from the coastline. For practical reasons, 20 nautical miles from the coastline will be used as definition of suitable ENC coverage for the purpose of this study. I.e. suitable ENC coverage will be assumed if, for all parts of a voyage closer to the shore than 20 nautical miles, officially approved ENCs of scale coastal or larger are available.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



6 ENC COVERAGE FOR SOLAS SHIPS In this part of the study, the global coverage of suitable ENC in areas typically operated by SOLAS ships has been investigated. Worldwide coverage of ENC has been mapped to global shipping traffic distributions in order to investigate the percentage of the global traffic that operates in areas with sufficient ENC coverage. For the purpose of this study, sufficient ENC coverage corresponds to ENC with resolution corresponding to coastal or better for all stretches of the trade closer to shore than 20 nautical miles.

World traffic patterns have been collected for one complete year 2000/2001 (see section 3.4 of this report). For the purpose of this study, these traffic distributions have been utilized, and the world traffic pattern is assumed unchanged in 2006 and 2010.

The ship traffic is reported on in a global grid with a 1x1 degree grid cell resolution. Along the equator, a 1x1 degree grid cell corresponds to a 60x60 nm. Moving away from the equator the “height” of the cell stays the same, while the “breadth” of the cell decreases (1 degree shrinks to 0 nm on the poles). The size of the observation grid cells has been a challenge when trying to determine the number of observations within a short distance from shore, such as 10 nm, 20 nm and 25 nm. For example, a 20 nm band extending from land will sometimes leave most of a grid cell outside the band. Sometimes most of the grid cell will be inside the band. Hence, the 1x1 degree grid cell resolution introduces considerable uncertainties into this part of the study.

In order to overcome this problem, two different approaches have been employed in order to estimate which part of the traffic takes part in areas close to shore. The first approach is to count the traffic in all the grid cells intersecting with the 20 nm band. The second approach is to count only the traffic in the grid cells whose centre point intersects with the 20 nm band. These counting approaches will be referred to as intersecting and centre-intersecting for the remainder of the report respectively.

Table 6 presents the percentage of the world maritime trade that passes within different coastal bands according to the two approaches employed (based on merged AMVER and COADS data). Table 7 presents the percentage of ship traffic within the 20 nm coastal band for main shiptypes for the “all touching traffic squares” counting approach (based on AMVER data).

Table 6: Percentage of world total traffic* within different coastal bands (all ship types

included)

Distance to land Counting intersecting traffic grid cells

Counting centre-intersecting traffic grid cells

within 10 nm 31.2 13.7 within 20 nm 35.2 18.9 within 25 nm 36.7 20.9

*Average relative number of reported ships at any given time.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Table 7: Percentage of traffic* within 20 nm off coast for major shiptypes, counting intersecting traffic grid cells

Ship type % Bulk 28.2 Tanker 25.6 Container 18.5 General Cargo 23.5 *Average relative number of reported ships at any given time.

From Table 6 it can be seen that, on average, 35.2% of all ships will be within 20 nm from a coast at any given time, using the most inclusive counting method. Using the alternative counting method, Table 6 shows that, on average, 18.9% of all ships are within 20 nm of a coast at any given time. Similarly Table 7 gives that, on average, 18.5% of Container ships are within 20 nm of a coast at any given time, while this number is 28.2% for Bulk ships. Note that the figures in Table 7 are generally lower than in Table 6 because the COADS data are included in the data used to produce Table 6.

The different counting approaches are illustrated in Figure 12. The different shades of colour for the different grid cells reflect variations in traffic intensity. The cross-hatched grid cells represent those that are counted by the different approaches. Comparing the two pictures it can easily be seen that the intersecting counting approach will regard notable more traffic to be within the coastal bands. This is also reflected in Table 6 above. The blue single-hatched squares in Figure 12 represent ENC coverage (of coastal or better resolution).

Figure 12: Traffic cells included in the counting of traffic within the 20 nm coastal band for

the intersecting approach (left) and the centre-intersecting approach (right).

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Two similar approaches have been employed in order to estimate whether there is ENC coverage within a cell or not, i.e. where there are ENC coverage intersects with any part of a cell and where there are ENC coverage intersects with the centre of the grid cell respectively. These counting approaches will be referred to as intersecting and centre-intersecting for the remainder of this paper.

Thus, for each coastal band (defined by a distance to shore) there will be four estimates of the ENC coverage, combining the two approaches for counting traffic densities and ENC coverage. For the purpose of this general study, the ENC coverage for each coastal band will therefore not be presented as a single best estimate, but rather as a range of values within which the actual coverage will lie. This is deemed sufficiently accurate for the purpose of this general study and in light of the inevitable uncertainties involved. The estimated ENC coverage according to these approaches is presented in the following for the current situation as well as for the anticipated case in 2010.

6.1 2006 coverage estimates The percentages of the world shipping traffic within different distances to shore, and according to the different approaches to count traffic and ENC coverage within the 1x1 degree grid cell resolution are summarized in Table 8 for the situation in 2006.

Table 8: Percentage of world traffic within different coastal bands with sufficient ENC coverage – 2006

Counting intersecting traffic grid cells Counting centre-intersecting traffic

grid cells

Distance to land

Counting traffic grid cells

intersecting ENC

Counting traffic grid cells centre-intersecting ENC


intersecting ENC


10 nm 88,6 84,4 96,4 94,3 20 nm 86,9 82,1 94,4 91,8 25 nm 86,1 80,9 93,6 90,7

Table 9 presents the percentage of traffic closer than 20 nm miles to shore with sufficient ENC coverage, broken down on major shiptypes and using the intersecting approach to estimate the traffic that passes within the coastal band. Results corresponding to both counting approaches for ENC coverage are reported.

Table 9: Percentage of world traffic within 20 nm off coast with sufficient ENC coverage –

2006, broken down on major shiptypes

Ship type Counting traffic grid cells intersecting ENC


Bulk 86,3 82,4 Tanker 89,3 84,9 Container 93,4 90,4 General Cargo 89,9 86,4

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



It can be seen from the results above that the most conservative approach is to use the intersecting traffic grid cell counting approach for traffic estimation coupled with the ENC centre-intersecting grid cell counting approach for ENC coverage estimation. Accordingly, a conservative estimate of the percentage of world shipping traffic within 20 nm to shore having sufficient ENC coverage would be 82% for 2006. The most optimistic estimate would be 94%. However, in light of the uncertainties involved, it may only be concluded that the actual percentage lies between these limits. I.e. the percentage of the world shipping trade having suitable ENC coverage along their voyages are between 82% and 94% for 2006.

6.2 Anticipated 2010 coverage estimates The percentages of the world shipping traffic within different distances to shore, and according to the different approaches to count traffic and ENC coverage within the 1x1 degree grid cell resolution are summarized in Table 10 for the anticipated situation in 2010.

Table 11 presents the percentage of traffic closer than 20 nm miles to shore with sufficient ENC coverage, broken down on major shiptypes and using the intersecting approach to estimate the traffic that passes within the coastal band. Results corresponding to both counting approaches for ENC coverage are reported.

Table 10: Percentage of world traffic within different coastal bands with sufficient ENC

coverage – 2010

Counting intersecting traffic grid cells Counting centre-intersecting traffic grid cells

Distance to land


intersecting ENC



intersecting ENC


10 nm 92,2 86,8 97,6 95,1 20 nm 90,8 84,7 96,3 93,0 25 nm 90,1 83,7 95,7 92,1

Table 11: Percentage of world traffic within 20 nm off coast with sufficient ENC coverage –

2010, broken down on major shiptypes

Ship type Counting traffic grid cells intersecting ENC


Bulk 89,1 84,4 Tanker 92,7 86,8 Container 95,3 91,4 General Cargo 92,7 87,9

Again, it can be seen from the results above that the most conservative approach is to use the intersecting counting approach for traffic estimation coupled with the ENC centre-intersecting grid cell counting approach for ENC coverage estimation. According to this, a conservative

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



estimate of the percentage of world shipping traffic within 20 nm to shore having sufficient ENC coverage would be 85% for 2010. The most optimistic estimate would be 96%. However, in light of the uncertainties involved, it may only be concluded that the actual percentage lies between these limits. I.e. the percentage of the world shipping trade having suitable ENC coverage along their voyages are between 85% and 96% for 2006.

Comparing the results for the anticipated coverage in 2010 to the estimates for 2006, a slight increase in ENC coverage can be observed. This is expected and is attributable to the expected increase in available ENCs in the near future. However, the increase is not significant, and this may be explained by the fact that ENC coverage is already quite extensive along coastal areas carrying a great portion of world shipping traffic.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



7 REPRESENTATIVE SHIPPING ROUTES

7.1 Method for selecting representative shipping routes There are several ways of quantifying the size of shipping routes. One common way is to count the number of ship passages across an area. Another is counting the amount of cargo transported across an area. A third option could be to count the number of ships, and weigh the ships by size. The approach used in this study tries to employ the two first methods to obtain a balanced view on the global shipping routes.

The other problem one face when defining “representative” shipping routes is to decide on representative ship types. The objective is to represent as much of the world fleet as possible with as few ships as possible. The ships selected should in other words belong to large groups of ships, identified by similar characteristics and similar trade patterns. Determining the size of a group of ships is, however, not straightforward. Three approaches are commonly used to determine the size of a group: number of ships, gross weight tonnage and dead weight tonnage. A fourth approach is monetary value.

7.2 Selected representative shipping routes In order to investigate the effects of lack of ENC coverage in more detail, a set of representative routes are selected for further analyses. These should be representative of the global traffic of merchant shipping, and for the most important shiptypes. Figure 5 indicates the world shipping traffic densities in terms of number of ships according to AMVER and COADS data, and the representative shipping routes chosen in this study should correspond to this picture as closely as possible. Table 12 contains the distribution of shiptypes in the world merchant fleet as of 2005 in descending Gross Ton order [18]. It also contains the number of ships larger than 100 GT in the world fleet for each shiptype. It is seen that the most significant shiptypes are dry bulk carriers, crude oil tankers, container vessels and general cargo vessels. This will be reflected in the selection of representative routes, i.e. typical routes for bulk carriers, oil tankers, container vessels and general cargo vessels should be included among the selected routes.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Table 12: World merchant fleet, 2005 Shiptype Total GT (103) # ships > 100 GT Dry bulk 178,300 5,542Crude oil tanker 147,419 1,939Container 96,252 3,531General cargo 51,907 16,086Ro-ro cargo 32,798 2,092Chemical tanker 28,752 3,154Oil products tanker 27,048 5,095LNG tanker 17,199 195Ro-pax 15,372 2,796Cruise vessels 11,928 456Other bulk types 14,912 1,422Other tanker types 10,619 1,402Other dry cargo/passenger types 10,161 4,759

Total 642,667 48,442

For the purpose of this study, eleven different routes will be selected as representative routes for the world seaborne trade, i.e. three typical oil tanker routes, three typical bulk carrier routes, two typical container vessel routes, one typical general cargo route and two routes typical of other vessels. These will be described further below.

7.2.1 Oil tankers The main region for oil exports is the Middle East Gulf, and the main oil importers are in Asia, North America and Europe [19]. Thus, three representative routes will be selected to represent these trades:

• Middle-East to East Asia • Middle-East to North America, going around South Africa. • Middle-East to Europe via Suez Canal and the Mediterranean Sea.

Saudi Arabia is by far the biggest oil exporter in the Middle East, and they are exporting to North America and Europe from e.g. Yanbu at the Red Sea and to Asia from e.g. Dammam in the Persian Gulf. U.S. main import terminals are located along the Gulf Coast, e.g. the port of Galveston, Texas near Houston. The port of Barcelona, Spain will be used as a proxy for European ports, and Yokohama, Japan will be selected to represent Asian import terminals. Hence, the following specific routes are chosen to represent oil tanker traffic:

1. Dammam, S.A. – Yokohama, Japan 2. Yanbu, S.A. – Galveston, TX, USA 3. Yanbu, S.A. – Barcelona, Spain

7.2.2 Container vessels The main exporters of goods shipped in containers are in Asia, and the main importers are in North America and in Europe. Hence, two selected routes for typical container vessel traffic are between Asia and North America and between Asia and Europe.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Singapore is the world’s biggest container port followed by Hong Kong, Shanghai and Shenzhen. The biggest container port in Europe is Rotterdam, followed by Hamburg and Antwerp. The major North American container port is Long Beach on the Pacific coast [20]. The leading foreign seaports for cargo arriving at Long Beach are Hong Kong, Shenzhen and Pusan. Hence, the following specific routes are selected to represent typical container vessel traffic:

4. Singapore, Singapore – Rotterdam, Holland 5. Hong Kong, China – Long Beach, CA, USA

7.2.3 Bulk carriers The main regions for export of dry bulk commodities (e.g. ore, coal and grain) are Australia and South-America and the main imports are in Asia and Europe [19]. Hence, three selected routes for typical bulk trade will be selected between Australia and Asia, South-America and Europe and North-America and South-America.

Newcastle in Australia is one of the world’s largest export ports for coal. Coal import is increasing in China, and one of the main ports for coal import in China is at Qinhuangdao in Heibei province.

Brazil is the world’s second biggest exporter of iron ore, second only to Australia, with a share of more than 30% of the world iron ore exports [21]. The port of Tubarao at Vitória is the world’s biggest iron ore port. About 15% of the iron ore exported from Brazil is exported to Germany, which makes this the biggest European importer of Brazilian iron ore. Hansaport in Hamburg is Germany’s largest dry-bulk terminal, which also receives iron ore.

Canada is the world’s second largest exporter of wheat, second only to Australia. The biggest markets for Canadian wheat are in Asia and Africa, but Brazil is the biggest Latin-American importer of Canadian wheat. The largest port terminals are in Vancouver on the Canadian west coast. The port of Salvador is the largest port in Brazil’s North and North-eastern regions, and wheat is one of the most important imports in this port. Shipments between Vancouver and Salvador would be expected to go through the Panama Canal.

Hence, for the purpose of this study, the following specific routes will be selected to represent typical bulk carrier traffic:

6. Newcastle, Australia – Qinhuangdao, China 7. Vitoria, Brazil – Hamburg, Germany 8. Vancouver, Canada – Salvador, Brazil

7.2.4 General cargo vessels General cargo vessels exist in great numbers, and serve a variety of purposes, making it hard to analyse and define typical routes. However, in order to include a route covering the heavily trafficked areas of northern and western Europe, a route going from the Baltic Sea to Southern Europe though the Danish straights and English Channel was selected. General cargo vessels carry shipments of every conceivable kind, and the selected route will carry printing and writing paper from the major exporter Finland to Spain. The following route was selected as a representative route for general cargo vessels:

9. Helsinki, Finland – Cadiz, Spain

7.2.5 Other shiptypes Two other shiptypes will be represented, i.e. chemical tankers and LNG carriers.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Chemical tankers carry different types of liquid bulk cargo. EU has the biggest regional share of exports and imports of chemicals and North America is the most important trading partner, in terms of both export and import6. A trans-Atlantic route for chemical tankers will therefore be selected as representative for chemical carriers. Rotterdam is an important European liquid bulk terminal. There are many liquid bulk terminals on the U.S. East Coast, and Savannah, Georgia will be selected as a representative terminal for trade with EU on chemicals. Hence, for the purpose of this study, the following route was selected to represent a typical trade for trans-Atlantic chemical tankers:

10. Rotterdam, Holland – Savannah, GA, USA Although the fleet of LNG carriers is relatively small measured in number of ships, one representative route for LNG shipping will be selected in this study due to the significant size of this fleets gross tonnage. Also, this is a very high-value ship type, with a substantial growth rate. The main importing countries of LNG are Japan and South-Korea; with the most important exporting countries to these are Indonesia and Malaysia. As of 2004, USA was the fourth largest importer of LNG, with Trinidad & Tobago as the most important exporting country. As of 2005, there were five operating LNG importing terminals in USA. Of these, the one in Everett, MA (close to Boston) is the oldest and the one with the highest capacity. LNG export from Trinidad is sent from Point Fortin. Hence, for the purpose of this study, the following route was selected as a representative LNG shipping route:

11. Point Fortin, Trinidad & Tobago – Everett, MA, USA The 11 routes selected above is assumed to be representative for the world merchant shipping, and these will be used for the purpose of evaluating the effect of lack of ENC coverage on the effect of ECDIS in this study. These routes are indicated in a world map in Figure 13. In order to examine the eleven routes in more details, the Netpas Distance software, available from Netpas7, has been used.

6 According to the European Chemical Industry Council, Cerfic. Website: http://www.cefic.org/ 7 Netpas website: https://netpas.net/

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Figure 13: Selected routes representative for world merchant shipping

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



8 EFFECTS OF ENC HOLES ALONG SELECTED ROUTES In order to evaluate the effects of holes of ENC coverage some specific routes will be selected for further investigation.

When estimating the typical duration of a voyage along these routes, a typical average speed of 15 knots will be assumed, although it is recognized that the speed will generally vary significantly between different shiptypes. This average speed will correspond reasonably well to typical speeds for e.g. bulk carriers and tankers, whereas e.g. container vessels and car carriers are normally associated with higher speeds.

The speed may have an influence on the modelled grounding frequency in two ways. 1) Increased speed means that more distance is covered per year, resulting in increased risk exposure and hence increased risk of experiencing critical course. 2) Increased speed will generally increase the risk of grounding conditioned on a critical course since less time will be available for corrective actions. Thus, it can be assumed that the risk will generally increase with increasing speed. At any rate, an average speed of 15 knots is deemed appropriate for the purpose of this study, even though it will give conservative results (low risk estimates) for ships such as containerships and car carriers which typically maintain higher speeds.

8.1 Effect of holes in ENC coverage The effect of holes in the ENC coverage along a route will be to reduce the risk reducing performance of an installed ECDIS, meaning that in areas where no ENCs are available, there will be no benefits from having ECDIS installed. For the purpose of this study, it is assumed that the effect of such holes is proportional to the ratio of the route in coastal and narrow waters where ENC is not available, i.e. the ratio of the route closer to shore than 5 nautical miles where ENC is available. Thus, the net risk reducing effects of ECDIS, ΔRNET will be reduced by this ratio.

ΔRNET = ΔRECDIS x ENCCN/(ENC + No ENC)CN (1)

In equation (1), ΔRNET denotes the net risk reducing effect of ECDIS for the selected route, ΔRECDIS denotes the risk reducing effect of ECDIS for areas where suitable ENCs area available (36% according to previous studies [4, 5]), ENCCN denotes the distance along the route in coastal or narrow waters with suitable ENC coverage and (ENC + No ENC)CN is the total distance along the route in coastal or narrow waters. In the following, these distances will be investigated for the 11 selected routes.

8.2 ENC coverage along selected routes It has been assumed that ENC is needed for safe navigation for all areas where the distance to shore is less than 20 nautical miles. However, in the risk analysis to estimate the frequency of groundings, three types of waters were defined, i.e. open waters, coastal waters and narrow waters. The effect of ECDIS and hence of ENC coverage will be different for these types of waters.

For the purpose of this study, the following defines the types of water: • Open waters: > 5 nautical miles from shore

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



• Coastal waters: between 2 and 5 nautical miles from shore • Narrow waters: < 2 nautical miles from shore

It is acknowledged that these definitions differ slightly from the ones presented in [5], but the definitions above are unambiguous and hence assumed to be more appropriate. In the following, the 11 selected routes will be further examined in order to establish the ratio of the voyage which takes place in each of these types of waters as well as the ratio of ENC coverage for the stretches within coastal and narrow waters.

In order to investigate the ENC coverage along selected routes, the IHO global ENC catalogue has been used to assess the availability of suitable ENCs together with route descriptions obtained using the Netpas Distance tool. The Netpas Distance tool is also used to estimate the time of voyage for each of the selected routes, and an additional three day in port is assumed for each trade.

8.2.1 Dammam, Saudi Arabia – Yokohama, Japan The route between Dammam in Saudi Arabia and Yokohama in Japan covers about 6,500 nautical miles, from the Persian Gulf, through the straight of Malacca, and north of the Philippines to the west coast of Japan. 1160 nautical miles of this route is closer to shore than 20 nautical miles (18%), 730 nautical miles is closer to 5 nautical miles (11%) and 1 nautical miles is closer than 2 nautical miles (1%). The route is illustrated in Figure 15, and the ENC coverage for this route is illustrated in Figure 15. The characteristics of the route together with the ENC coverage today and anticipated for 2010 are presented in Table 13. The voyage excluding time in port is estimated to take about 18 days.

Figure 14: Route from Dammam to Yokohama

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Figure 15: ENC coverage for route Dammam – Yokohama

Table 13: Route characteristics and ENC coverage, Dammam - Yokohama

Total < 20 nm < 5 nm < 2 nm Distance (nm) 6565 1160 730 85 (%) 100 18 11 1

2006 2010 2006 2010 2006 2010 ENC coverage 37 % 56 % 41 % 65 % 40% 81%

The characteristics of this route have been incorporated into the Excel model for critical courses to obtain the probability of critical course per year for this route. The corresponding annual grounding frequencies for ships sailing this trade according to the Bayesian Network model are presented in Table 14, including estimates with and without ECDIS and according to current and anticipated future ENC coverage.

Table 14: Annual grounding frequencies, Dammam - Yokohama Without

ECDIS

With ECDIS – Assume 100%

coverage

With ECDIS – Actual Coverage

2006

With ECDIS – Anticipated

Coverage 2010 Frequency (per shipyear) 4.7 x 10-2 2.9 x 10-2 3.9 x 10-2 3.5 x 10-2

Frequency reduction (%) - 38 % 15% 25%

Groundings averted (per shipyear) - 1.8 x 10-2 7.2 x 10-3 1.1 x 10-2

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



8.2.2 Yanbu, Saudi Arabia – Galveston, TX, USA The route between Yanbu in Saudi Arabia and Galveston in Texas, USA, via Cape Good Hope, South Africa, covers about 12,200 nautical miles, from Yanbu south through the Red Sea, along the East African coast, via Cape Good Hope and across the Atlantic Ocean, along the South American Eastern coast from Brazil to Venezuela, through the Caribbean Sea, into the Gulf of Mexico and finally into Galveston harbour. 2,600 nautical miles of this route is closer to shore than 20 nautical miles (22%), 350 nautical miles is closer to 5 nautical miles (3%) and only 10 nautical miles is closer than 2 nautical miles (< 1%). The route is illustrated in Figure 16, and the ENC coverage for this route is illustrated in Figure 17. The characteristics of the route together with the ENC coverage today and anticipated for 2010 are presented in Table 15. The voyage excluding time in port is estimated to take about 34 days.

Figure 16: Route from Yanbu to Galveston via Cape Good Hope

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Figure 17: ENC coverage for route Yanbu – Galveston

Table 15: Route characteristics and ENC coverage, Yanbu – Galveston

Total < 20 nm < 5 nm < 2 nm Distance (nm) 12200 2600 350 10 (%) 100 22 3 < 1

2006 2010 2006 2010 2006 2010 ENC coverage 49% 63 % 57% 77% 100% 100%


Table 16: Annual grounding frequencies, Yanbu – Galveston Without

ECDIS


coverage


2006





DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



8.2.3 Yanbu, Saudi Arabia – Barcelona, Spain The route between Yanbu in Saudi Arabia and Barcelona, Spain, covers about 2,100 nautical miles, from the Red Sea, through the Suez Canal and past the south tips of Sicily and Sardinia to the west coast of Spain. 575 nautical miles of this route is closer to shore than 20 nautical miles (27%), 187 nautical miles is closer to 5 nautical miles (9%) and 96 nautical miles is closer than 2 nautical miles (4%). The route is illustrated in Figure 18, and the ENC coverage for this route is illustrated in Figure 19. The characteristics of the route together with the ENC coverage today and anticipated for 2010 are presented in Table 17. The voyage excluding time in port is estimated to take about 6 days.

Figure 18: Route from Yanbu to Barcelona

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Figure 19: ENC coverage for route Yanbu - Barcelona

Table 17: Route characteristics and ENC coverage, Yanbu - Barcelona


2006 2010 2006 2010 2006 2010 ENC coverage 95 % 95 % 94 % 94 % 98% 98%

The characteristics of this route have been incorporated into the Excel model for critical courses to obtain the probability of critical course per year for this route. The corresponding annual grounding frequencies for ships sailing this trade according to the Bayesian Network model are presented in Table 18, including estimates with and without ECDIS and according to current and anticipated future ENC coverage. It is noted that for this particular trade, the anticipated ENC coverage in 2010 is similar to the current situation.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Table 18: Annual grounding frequencies, Yanbu - Barcelona Without

ECDIS


coverage


2006





8.2.4 Singapore, Singapore – Rotterdam, Holland The route between Singapore and Rotterdam covers almost 8,300 nautical miles, going from Singapore, through the Malacca strait, the Suez Canal, Mediterranean and Gibraltar to Rotterdam in the Netherlands. 2,600 nautical miles of this route is closer to shore than 20 nautical miles (32%), 940 nautical miles is closer to 5 nautical miles (11%) and 180 nautical miles is closer than 2 nautical miles (2%). The route is illustrated in Figure 20, and the ENC coverage for this route is illustrated in Figure 21. The characteristics of the route together with the ENC coverage today and anticipated for 2010 are presented in Table 19. The voyage excluding time in port is estimated to take about 23 days.

Figure 20: Route from Singapore to Rotterdam

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Figure 21: ENC coverage for route Singapore - Rotterdam

Table 19: Route characteristics and ENC coverage, Singapore - Rotterdam


2006 2010 2006 2010 2006 2010 ENC coverage 62 % 67 % 63 % 68 % 100% 100%


Table 20: Annual grounding frequencies, Singapore - Rotterdam Without

ECDIS


coverage


2006





DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



8.2.5 Hong Kong, China – Long Beach, CA, USA The route between Hong Kong and Long Beach covers about 6,400 nautical miles, going from Hong Kong, along the Chinese coast, through the Japan Sea and crossing the Pacific Ocean to the US west coast. 820 nautical miles of this route is closer to shore than 20 nautical miles (13%), 100 nautical miles is closer to 5 nautical miles (2%) and 20 nautical miles is closer than 2 nautical miles (0.3%). The route is illustrated in Figure 22, and the ENC coverage for this route is illustrated in Figure 23. The characteristics of the route together with the ENC coverage today and anticipated for 2010 are presented in Table 21. The voyage excluding time in port is estimated to take about 18 days.

Figure 22: Route from Hong Kong to Long Beach

Figure 23: ENC coverage for route Hong Kong – Long Beach

Table 21: Route characteristics and ENC coverage, Hong Kong – Long Beach

Total < 20 nm < 5 nm < 2 nm Distance (nm) 6400 820 100 20 (%) 100 13 2 0.3

2006 2010 2006 2010 2006 2010 ENC coverage 49 % 49 % 100 % 100 % 100% 100%

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



The characteristics of this route have been incorporated into the Excel model for critical courses to obtain the probability of critical course per year for this route. The corresponding annual grounding frequencies for ships sailing this trade according to the Bayesian Network model are presented in Table 22, including estimates with and without ECDIS and according to current and anticipated future ENC coverage. It is noted that for this particular route, the ENC coverage is 100% for coastal and narrow waters, i.e. waters closer than 5 nautical miles to the shore. I.e. the assumption of 100% coverage is valid for this particular trade.

Table 22: Grounding frequencies, Hong Kong – Long Beach Without

ECDIS


coverage


2006



Frequency reduction (%) - 38 % 38 % 38 %


8.2.6 Newcastle, Australia – Qinhuangdao, China The route between Newcastle, Australia and Quihuangdao, China, covers about 4,900 nautical miles, from the west coast of Australia, through the Jomard Entrance, the East Ferguson Isand and the Vitiaz Strait west of Papua New Guinea and past the southern islands of Japan to the North of China. 945 nautical miles of this route is closer to shore than 20 nautical miles (19%), 215 nautical miles is closer to 5 nautical miles (4%) and 6 nautical miles is closer than 2 nautical miles (0.1%). The route is illustrated in Figure 24, and the ENC coverage for this route is illustrated in Figure 25. The characteristics of the route together with the ENC coverage today and anticipated for 2010 are presented in Table 23. The voyage excluding time in port is estimated to take about 14 days.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Figure 24: Route from Newcastle to Qinhuangdao

Figure 25: ENC coverage for route Newcastle – Qinhuangdao

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Table 23: Route characteristics and ENC coverage, Newcastle – Qinhuangdao

Total < 20 nm < 5 nm < 2 nm Distance (nm) 4925 915 215 6 (%) 100 19 4 0.1

2006 2010 2006 2010 2006 2010 ENC coverage 24 % 24 % 28 % 28 % 20 % 20 %

The characteristics of this route have been incorporated into the Excel model for critical courses to obtain the probability of critical course per year for this route. The corresponding annual grounding frequencies for ships sailing this trade according to the Bayesian Network model are presented in Table 24, including estimates with and without ECDIS and according to current and anticipated future ENC coverage. For this particular route, the anticipated ENC coverage in 2010 will be similar to the current ENC coverage for distances up to 5 nautical miles from shore.

Table 24: Grounding frequencies, Newcastle – Qinhuangdao Without

ECDIS


coverage


2006





8.2.7 Vitoria, Brazil – Hamburg, Germany The route between Vitoria and Hamburg covers about 5,200 nautical miles, going from Vitoria, along the Brazilian coast, passing the Cape Verde and Madeira Islands and the Spanish west coast, through the English Channel and into Hamburg. 900 nautical miles of this route is closer to shore than 20 nautical miles (17%), 320 nautical miles is closer to 5 nautical miles (6%) and 80 nautical miles is closer than 2 nautical miles (2%). The route is illustrated in Figure 26, and the ENC coverage for this route is illustrated in Figure 27. The characteristics of the route together with the ENC coverage today and anticipated for 2010 are presented in Table 25. The voyage excluding time in port is estimated to take about 15 days.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Figure 26: Route from Vitoria to Hamburg

Figure 27: ENC coverage for route Vitoria – Hamburg

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Table 25: Route characteristics and ENC coverage, Vitoria - Hamburg


2006 2010 2006 2010 2006 2010 ENC coverage 62 % 84 % 65 % 84 % 79% 83%


Table 26: Annual grounding frequencies, Vitoria - Hamburg Without

ECDIS


coverage


2006





8.2.8 Vancouver, Canada – Salvador, Brazil The route between Vancouver and Salvador covers about 7,600 nautical miles, going from Vancouver, continuing along the US Pacific coast, along the Mexican west coast to Panama, through the Panama Canal and along the South American Atlantic coast down to Salvador. 3,400 nautical miles of this route is closer to shore than 20 nautical miles (45%), 780 nautical miles is closer to 5 nautical miles (10%) and 90 nautical miles is closer than 2 nautical miles (1%). The route is illustrated in Figure 28, and the ENC coverage for this route is illustrated in Figure 29. The characteristics of the route together with the ENC coverage today and anticipated for 2010 are presented in Table 27. The voyage excluding time in port is estimated to take about 21 days.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Figure 28: Route from Vancouver to Salvador

Figure 29: ENC coverage for route Vancouver – Salvador

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Table 27: Route characteristics and ENC coverage, Vancouver – Salvador


2006 2010 2006 2010 2006 2010 ENC coverage 46 % 57 % 49 % 70 % 95% 100%


Table 28: Annual grounding frequencies, Vancouver – Salvador Without

ECDIS


coverage


2006





8.2.9 Helsinki, Finland – Cadiz, Spain The route between Helsinki and Cadiz covers about 2,500 nautical miles, going from The Bay of Finland, passing the Great Belt and the Skaw, the English Channel and the coast of Portugal, to Cadiz on the Atlantic South coast of Spain. 1400 nautical miles of this route is closer to shore than 20 nautical miles (57%), 222 nautical miles is closer to 5 nautical miles (9%) and 29 nautical miles is closer than 2 nautical miles (1%). The route is illustrated in Figure 30, and the ENC coverage for this route is illustrated in Figure 31. The characteristics of the route together with the ENC coverage today and anticipated for 2010 are presented in Table 29. The voyage excluding time in port is estimated to take about 7 days.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Figure 30: Route from Helsinki to Cadiz

Figure 31: ENC coverage for route Helsinki – Cadiz

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Table 29: Route characteristics and ENC coverage, Helsinki – Cadiz Total < 20 nm < 5 nm < 2 nm Distance (nm) 2473 1400 222 29 (%) 100 57 9 1

2006 2010 2006 2010 2006 2010 ENC coverage 100 % 100 % 100 % 100 % 100 % 100 %

The characteristics of this route have been incorporated into the Excel model for critical courses to obtain the probability of critical course per year for this route. The corresponding annual grounding frequencies for ships sailing this trade according to the Bayesian Network model are presented in Table 30, including estimates with and without ECDIS and according to current and anticipated future ENC coverage. It is noted that the current ENC coverage along this route is 100%, thus, the assumption made in the previous studies is valid for this particular route.

Table 30: Annual grounding frequencies, Helsinki - Cadiz Without

ECDIS


coverage


2006





8.2.10 Rotterdam, Holland – Savannah, GA, USA The route between Rotterdam and Savannah covers about 3,900 nautical miles, from Rotterdam, through the English Channel, across the Atlantic Ocean and into Savannah on the U.S. East Coast. 870 nautical miles of this route is closer to shore than 20 nautical miles (22%), 170 nautical miles is closer than 5 nautical miles (4%) and 40 nautical miles is closer than 2 nautical miles (1%). The route is illustrated in Figure 32, and the ENC coverage along this route is illustrated in Figure 33. The characteristics of the route together with the ENC coverage today and anticipated for 2010 are presented in Table 31. The voyage excluding time in port is estimated to take about 11 days.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Figure 32: Route from Rotterdam to Savannah

Figure 33: ENC coverage for route Rotterdam – Savannah

Table 31: Route characteristics and ENC coverage, Rotterdam - Savannah


2006 2010 2006 2010 2006 2010 ENC coverage 100% 100% 100% 100% 100% 100%

The characteristics of this route have been incorporated into the Excel model for critical courses to obtain the probability of critical course per year for this route. The corresponding annual grounding frequencies for ships sailing this trade according to the Bayesian Network model are presented in Table 32, including estimates with and without ECDIS and according to current and

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



anticipated future ENC coverage. It is noted that the current ENC coverage along this route is 100%, thus, the assumption made in the previous studies is valid for this particular route.

Table 32: Annual grounding frequencies, Rotterdam - Savannah Without

ECDIS


coverage


2006





8.2.11 Point Fortin, Trinidad & Tobago – Everett, MA, USA The route between Point Fortin and Everett (Boston) covers about 2,000 nautical miles, going from The Bay of Finland, passing the Great Belt and the Skaw, the English Channel and the coast of Portugal, to Cadiz on the Atlantic South coast of Spain. 340 nautical miles of this route is closer to shore than 20 nautical miles (17%), 122 nautical miles is closer to 5 nautical miles (6%) and 23 nautical miles is closer than 2 nautical miles (1%). The route is illustrated in Figure 34, and the ENC coverage for this route is illustrated in Figure 35. The characteristics of the route together with the ENC coverage today and anticipated for 2010 are presented in Table 33. The voyage excluding time in port is estimated to take about 6 days.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Figure 34: Route from Point Fortin to Everett (Boston)

Figure 35: ENC coverage for route Point Fortin - Everett (Boston)

Table 33: Route characteristics and ENC coverage, Point Fortin - Everett (Boston)


2006 2010 2006 2010 2006 2010 ENC coverage 100 % 100 % 100 % 100 % 100 % 100 %

The characteristics of this route have been incorporated into the Excel model for critical courses to obtain the probability of critical course per year for this route. The corresponding annual grounding frequencies for ships sailing this trade according to the Bayesian Network model are presented in Table 34, including estimates with and without ECDIS and according to current and anticipated future ENC coverage. It is noted that the current ENC coverage along this route is 100%, thus, the assumption made in the previous studies is valid for this particular route.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Table 34: Annual grounding frequencies, Point Fortin - Everett (Boston) Without

ECDIS


coverage


2006





8.3 Estimated grounding risk reduction on selected routes The ENC coverage, grounding frequency reduction and the expected number of averted groundings per year are summarized for all of the selected routes in Table 35, for both the current situation and for the anticipated situation in 2010. The following general observations can be made: • 4 of the 11 selected routes already have 100% ENC coverage in coastal areas • 6 of the 11 selected routes sees no anticipated changes in the ENC coverage between 2006

and 2010 • The grounding frequency reductions due to ECDIS are between 11 - 38% for the selected

routes • The different routes have ENC coverage between 28% and 100%. The global ENC coverage

for ship traffic closer to shore than 10 nm was estimated to be between 84 – 96%.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Table 35: Grounding frequency reduction and averted groundings due to ECDIS for selected routes and for ENC coverage in 2006 and 2010 respectively

ENC coverage

(< 5 nm to shore) Grounding

frequency reduction Groundings averted

(per shipyear)

Route 2006 2010 2006 2010 2006 2010 Dammam – Yokohama 41% 65% 15% 25 % 7.2 x 10-3 1.1 x 10-2

Yanbu – Galveston 57% 77% 22% 29% 1.8 x 10-3 2.4 x 10-3

Yanbu – Barcelona 94% 94% 36% 36% 2.6 x 10-2 2.6 x 10-2

Singapore - Rotterdam 63% 68% 24% 26% 1.5 x 10-2 1.6 x 10-2

Hong Kong – Long Beach 100% 100% 38% 38% 3.1 x 10-3 3.1 x 10-3

Newcastle – Qinhuangdao 28% 28% 11% 11% 1.3 x 10-3 1.3 x 10-3

Vitoria – Hamburg 65% 84% 25% 32% 8.7 x 10-3 1.1 x 10-2

Vancouver – Salvador 49% 70% 19% 26% 7.9 x 10-3 1.1 x 10-2

Helsinki – Cadiz 100% 100% 38% 38% 1.2 x 10-2 1.2 x 10-2

Rotterdam – Savannah 100% 100% 38% 38% 8.9 x 10-3 8.9 x 10-3

Point Fortin – Everett 100% 100% 38% 38% 8.1 x 10-3 8.1 x 10-3

In order to estimate the risk reduction for each of the selected routes, a representative vessel must be assumed for all routes. Based on the ship size, the bunker capacity and the size of the crew, the accident cost and risk models will estimate the average property damage cost, the average environmental damage costs and the average number of fatalities associated with a grounding accident. The characteristics of the selected vessels for each route are presented in Table 36 below. It is noted that the cost models uses the ship deadweight (DWT) as input for estimating costs, and the following generic relationships to gross tonnage (GT) and cargo capacity C (in TEU for container ships and m3 for gas carriers) is assumed [13]:

DWT = 2 GT (for all ships) (2)DWT = 14 C(TEU) (for container ships) (3)

DWT = ½ C(m3) (for gas carriers) (4)

All ship sizes in Table 36 are presented in terms of gross tonnage.

The reduced fatality rates due to ECDIS for each route/shiptype are estimated based on the results from the risk models presented in Table 1.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Table 36: Representative vessels for each selected route

Route Shiptype Ship size Bunker capacity crew

Dammam – Yokohama Oil tanker 40,000 GT 3000 ton 30Yanbu – Galveston Oil tanker 80,000 GT 4000 ton 30Yanbu – Barcelona Oil tanker 20,000 GT 2000 ton 29Singapore - Rotterdam Container ship 10,500 GT 2200 ton 20Hong Kong – Long Beach Container ship 35,000 GT 8000 ton 28Newcastle – Qinhuangdao Bulk carrier 37,500 GT 2500 ton 30Vitoria – Hamburg Bulk carrier 20,000 GT 1500 ton 29Vancouver – Salvador Bulk carrier 10,000 GT 800 ton 25Helsinki – Cadiz General cargo ship 2,500 GT 350 ton 15Rotterdam – Savannah Chemical tanker 8,500 GT 750 ton 23Point Fortin – Everett LNG carrier 34,500 GT Boil-off gas 26

The estimated annual risk reductions with these vessels representing each of the selected routes are presented in Table 37 for the situation in 2006 and Table 38 for the anticipated situation in 2010.

Table 37: Estimated risk reduction for each route with representative ship, 2006 Risk reduction

Route

Groundings averted

(per shipyear)

Average grounding

accident cost (USD million)

Accident cost (USD per shipyear)

# Fatalities (per

shipyear) Dammam – Yokohama 7.2 x 10-3 29 210,000 4.3 x 10-5 Yanbu – Galveston 1.8 x 10-3 57 100,000 1.1 x 10-5 Yanbu – Barcelona 2.6 x 10-2 15 390,000 1.5 x 10-4 Singapore - Rotterdam 1.5 x 10-2 1.8 27,000 6.0 x 10-5 Hong Kong – Long Beach 3.1 x 10-3 5.5 17,000 1.7 x 10-5 Newcastle – Qinhuangdao 1.3 x 10-3 1.9 2,500 2.3 x 10-5 Vitoria – Hamburg 8.7 x 10-3 1.3 11,000 1.5 x 10-4 Vancouver – Salvador 7.9 x 10-3 0.89 7,000 1.2 x 10-4 Helsinki – Cadiz 1.2 x 10-2 0.51 6,100 1.1 x 10-4 Rotterdam – Savannah 8.9 x 10-3 1.2 11,000 4.1 x 10-5 Point Fortin – Everett 8.1 x 10-3 3.2 26,000 4.2 x 10-5

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Table 38: Estimated risk reduction for each route with representative ship, 2010 Risk reduction

Route

Groundings averted

(per shipyear)

Average grounding

accident cost (USD million)

Accident cost (per

shipyear)

# Fatalities (per

shipyear) Dammam – Yokohama 1.1 x 10-2 29 320,000 6.6 x 10-5 Yanbu – Galveston 2.4 x 10-3 57 140,000 1.4 x 10-5 Yanbu – Barcelona 2.6 x 10-2 15 390,000 1.5 x 10-4 Singapore - Rotterdam 1.6 x 10-2 1.8 29,000 6.4 x 10-5 Hong Kong – Long Beach 3.1 x 10-3 5.5 17,000 1.7 x 10-5 Newcastle – Qinhuangdao 1.3 x 10-3 1.9 2,500 2.3 x 10-5 Vitoria – Hamburg 1.1 x 10-2 1.3 14,000 1.9 x 10-4 Vancouver – Salvador 1.1 x 10-2 0.89 9,800 1.7 x 10-4 Helsinki – Cadiz 1.2 x 10-2 0.51 6,100 1.1 x 10-4 Rotterdam – Savannah 8.9 x 10-3 1.2 11,000 4.1 x 10-5 Point Fortin – Everett 8.1 x 10-3 3.2 26,000 4.2 x 10-5

According to these results, the highest cost savings are associated with oil tankers. This is due to the high costs associated with oil spills, which contributes most in the oil tanker case. Grounding is also known to be one of the most relevant accident scenarios for oil spills. Thus, averting oil spills is one of the most important aspects of averting grounding accidents in terms of accident cost savings.

8.4 Estimated cost effectiveness of ECIDS on selected routes Based on the analyses outlined above and on the cost estimates of ECDIS presented in section 4.1 of this report, the cost effectiveness corresponding to implementing ECDIS on a newbuilding expected to sail its entire lifetime on each of the selected routes can be estimated. The corresponding GCAF and NCAF values are reproduced in Table 39.

Whenever future costs and benefits are depreciated to Net Present Value, a depreciation rate of 5% has been used.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Table 39: Cost effectiveness of ECDIS along selected routes GCAF (106 USD) NCAF (106 USD)

Route 2006 2010 2006 2010

Dammam – Yokohama 63 41 < 0 < 0 Yanbu – Galveston 247 194 < 0 < 0 Yanbu – Barcelona 18 18 < 0 < 0 Singapore - Rotterdam 45 43 < 0 < 0 Hong Kong – Long Beach 160 160 < 0 < 0 Newcastle – Qinhuangdao 118 118 54 54 Vitoria – Hamburg 18 14 < 0 < 0 Vancouver – Salvador 23 16 < 0 < 0 Helsinki – Cadiz 25 25 < 0 < 0 Rotterdam – Savannah 66 66 < 0 < 0 Point Fortin – Everett 65 65 < 0 < 0

Based on these results, the following general observations can be made: • The GCAF values associated with each route exceeds USD 3 million. This is due to

the somewhat limited effect of ECDIS in terms of number of lives saved on cargo ships.

• The NCAF values for all routes except one are negative. This indicates that ECDIS is a cost effective risk control options when other benefits than the life-saving potential are taken into account (e.g. environmental and property protection).

• For cargo ships, the most important effect of ECDIS is the prevention of oil spills and the prevention of ship and cargo loss in case of grounding.

• The NCAF value is exceeding USD 3 million for one particular route, and this was the route with the poorest ENC coverage. Hence, only on routes with poor ENC coverage will ECDIS cease to be cost effective.

• The observations listed above are equally true for 2006 as for 2010.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



9 COST-EFFECTIVENESS OF ECDIS IN LIGHT OF ENC COVERAGE In the previous subsection, the associated NCAF and GCAF values indicating cost effectiveness of ECDIS was estimated for particular selected routes. However, in order to formulate recommendations for the world fleet, some globally applicable estimates are required. It is therefore assumed that the 11 selected routes together are representative for the world merchant shipping. Thus, the arithmetic average of the reduction in grounding risk for each of the routes will be assumed to represent the global risk reduction from ECDIS implementation. Hence, the following average risk reducing effect will be assumed attributable to implementation of ECDIS in this study:

2006: 9.1 x 10-3 groundings averted per shipyear

2010: 1.0 x 10-2 groundings averted per shipyear

The different shiptypes that were considered in the analysis of particular routes were also associated with very different accident costs. However, some global average will be needed. It is realized that the accident costs are considerably higher for oil tankers than for the other cargo ships, and that the accident cost generally increase with ship size. Hence, a simple average accident cost per GT will be assumed for oil tankers and other cargo ships respectively. Based on the cost estimates reported in this report, the following average accident costs will be assumed (for grounding accidents):

Oil tankers: 720 USD/GT

Other cargo ships: 120 USD/GT

The expected number of fatalities in a grounding accident is a function of the crew size and it was assumed higher for bulk carriers and general cargo vessels than for tankers and containers. Furthermore, the crew size was assumed proportional to the size of the ship. However, in order to obtain average estimates, an average crew size of 25 will be assumed for all ships. This may not be accurate for all ship sizes, but it is noted that the results are rather insensitive to this assumption. According to the risk models used in this study, this will correspond to an average fatality rate per grounding accident of 0.015 for bulk and general cargo ships and 0.005 for tankers and container vessels. However, for the purpose of obtaining global averages, is will be assumed that the average fatality rates in all groundings will be 0.01 irrespective of shiptype.

Based on these estimates and the estimates related to the cost of ECDIS, generic cost effectiveness estimates for new and existing ships may be carried out. An average expected lifetime of 25 years is assumed for all vessels. All estimates are assumed to be valid for all SOLAS ships larger than 500 GT. Ships smaller than this is considered out of scope of this study.

9.1 Cost-effectiveness for new ships Based on the estimates presented above, the following GCAF values are associated with implementing ECDIS on newbuildings:

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



GCAF = USD 30 million

This value would be reduced to USD 27 million for the anticipated coverage in 2010.

The NCAF value will be a function of shiptype and size.

For oil tankers, ships of 500 GT are associated with an NCAF of USD 8.2 million (reduces to USD 5.5 million for 2010). It can be shown that NCAF will be less than USD 3 million for all ships greater than 630 GT and negative for ships larger than 700 GT. Hence, ECDIS have been assessed to be cost effective for all oil tankers larger than 630 GT. For the anticipated situation in 2010, this is reduced to approximately 570 GT and 640 GT respectively.

For other cargo ships, ships of 500 GT are associated with an NCAF of USD 26 million (reduces to USD 24 million for 2010). It can be shown that NCAF will be less than USD 3 million for all ships greater than 3,800 GT and negative for ships larger than 4,200 GT. Hence, ECDIS have been assessed to be cost effective for all other cargo ships larger than 3,800 GT. For the anticipated situation in 2010, this is reduced to approximately 3,400 GT and 3,800 GT respectively.

The NCAF as a function of gross tonnage can be illustrated for the cases in 2006 and 2010, for new oil tankers (Figure 36) and other cargo ships (Figure 37).

200 400 600 800 1000GT

-1´107

1´107

2´107

NCAF

200 400 600 800 1000GT

-1´107

1´107

2´107

NCAF

Figure 36: NCAF value as a function of size (GT) for oil tanker newbuildings, 2006 (left)

and 2010 (right). Horizontal line represents USD 3 million.

1000 2000 3000 4000 5000GT

-5´106

5´106

1´107

1.5 ´107

2´107

2.5 ´107

3´107NCAF

1000 2000 3000 4000 5000GT

-5´106

5´106

1´107

1.5 ´107

2´107

2.5 ´107

NCAF

Figure 37: NCAF value as a function of size (GT) for other cargo ships newbuildings, 2006

(left) and 2010 (right). Horizontal line represents USD 3 million.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



9.2 Cost-effectiveness for existing ships For existing ships, the cost effectiveness will be a function of the ship age.

The GCAF value associated with ships of different age is illustrated in Figure 38, with the horizontal line representing USD 3 million. As was already demonstrated in the previous subsection, it can be seen that the GCAF value will never be less than USD 3 million, not even for newbuildings, due to the rather limited potential for saving lives on cargo vessels (because of few fatalities in grounding accidents).

5 10 15 20Age

2´107

4´107

6´107

8´107

1´108

1.2 ´108

1.4 ´108

GCAF

Figure 38: GCAF value as a function of ship age. Horizontal line represents USD 3 million. The NCAF value of implementing ECDIS on existing cargo ships will generally be a function of the type of ship, i.e. oil tanker or other type of cargo ships, the ship size and the age of the ship. This will be further elaborated in the following.

The relationship between NCAF, ship size (GT) and ship age for oil tankers are illustrated in Figure 39 for the case in 2006 and anticipated in 2010.

500010000

15000

20000

GT

0

5

10

15

20

Age

-1´109-5´108

05´108

NCAF

500010000

15000GT

500010000

15000

20000

GT

0

5

10

15

20

Age

-1´109-5´108

05´108

NCAF

500010000

15000GT

Figure 39: NCAF value as a function of ship age and size for oil tankers, for 2006 (left) and

2010 (right).

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



It may be difficult so read from these plots when NCAFs will be less than USD 3 million or negative, so this information is presented in Table 40, where the size of an oil tanker of 5, 10, 15, 20 and 24 years as well as newbuildings that corresponds to these NCAF limits are summarized. Ships larger than this will have better cost effectiveness (lower NCAF). This is based on the 2006 case.

Table 40: Oil tanker sizes corresponding to NCAF < USD 3 million and NCAF < 0

Ship age Size (GT)

(NCAF < USD 3 million)

Size (GT)

( NCAF < 0) Newbuilding 630 700

5 years 720 780 10 years 870 920 15 years 1,200 1,200 20 years 2,000 2,100 24 years 9,300 9,300

The relationship between NCAF, ship size (GT) and ship age for other cargo ships are illustrated in Figure 40 for the case in 2006 and anticipated in 2010.

2000040000

60000

80000

GT

0

5

10

15

20

Age

-5´108

0

5´108

NCAF

2000040000

60000GT

2000040000

60000

80000

GT

0

5

10

15

20

Age

-5´108

0

5´108

NCAF

2000040000

60000GT

Figure 40: NCAF value as a function of ship age and size for other cargo ships, for 2006

(left) and 2010 (right).

It may be difficult so read from these plots when NCAFs will be less than USD 3 million or negative, so this information is presented in Table 41, where the size of other cargo ships of 5, 10, 15, 20 and 24 years as well as newbuildings that corresponds to these NCAF limits are summarized. Ships larger than this will have better cost effectiveness (lower NCAF). Estimates are based on the 2006 case.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



Table 41: Other cargo ship sizes corresponding to NCAF < USD 3 million and NCAF < 0

Ship age Size (GT)

(NCAF < USD 3 million)

Size (GT)

( NCAF < 0) Newbuilding 3,800 4,200

5 years 4,300 4,700 10 years 5,200 5,500 15 years 7,000 7,300 20 years 12,000 13,000 24 years 56,000 56,000

The size/age relationships that correspond to cost effective implementation of ECDIS presented in Table 40 and Table 41 may also be presented graphically, as is shown in Figure 41.

Ship size/age corresponding to cost effecitve ECDISOil Tankers

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

0 5 10 15 20 25

Ship age (years)

Shh

ip s

ize

(GT)

NCAF < USD 3 millionNCAF < 0

Ship size/age corresponding to cost effecitve ECDISOther cargo ships

0

10000

20000

30000

40000

50000

60000

0 5 10 15 20 25

Ship age (years)

Shhi

p si

ze (G

T)

NCAF < USD 3 millionNCAF < 0

Figure 41: Graphical representation of ship size/age combinations for cost effective ECDIS.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



10 ONCLUSIONS AND RECOMMENDATIONS The cost effectiveness of ECDIS as a risk control option for cargo ships has been assessed, in light of the current and expected near future worldwide coverage of ENC.

The cost effectiveness has been assessed in terms of GCAF and NCAF. It was found that GCAF would always be higher than USD 3 million for all shiptypes and sized. Hence, ECDIS were found to have a limited effect in terms of saving lives onboard cargo vessels alone. However, when loss of property and environmental damage costs were taken into account in the calculation of NCAF, ECDIS was found to be a cost effective risk control option for many combinations of ship types, size and age.

Various shiptypes have been investigated, and it was generally found that there are major differences between oil tankers and other types of cargo ships. This is due to the high expected costs ascribed to major oil spills. Hence, it is deemed adequate to formulate separate recommendations for oil tankers and for other cargo ships.

The cost effectiveness in terms of NCAF as a function of ship size and age is presented in Table 40 (oil tankers) and Table 41 (other cargo ships). It is suggested that recommendations regarding carriage requirements for ECIDS are based on these results. Thus, in the following, proposed carriage requirement recommendations for ECDIS are outlined based on the study reported herein.

Based on the analysis reported herein, and the current cost efficiency criteria used at IMO, the following recommendations may be proposed:

i. ECDIS should be made mandatory for all new oil tankers of 500 gross tonnage and upwards.

ii. ECDIS should be made mandatory for all new cargo ships, other than oil tankers, of 3,000 gross tonnage and upwards.

iii. ECDIS should be made mandatory for all existing oil tankers of 3,000 gross tonnage and upwards.

iv. ECDIS should be made mandatory for all existing cargo ships, other than oil tankers, 10,000 gross tonnage and upwards.

v. Exemptions may be given to existing oil tankers of less than 10,000 gross tonnage and existing cargo ships, other than oil tankers, less than 50,000 gross tonnage when such ships will be taken permanently out of service within 5 years after the implementation dates given for iii) and iv) above.

DET NORSKE VERITAS

Report No: 2007-0304 , rev. 01

TECHNICAL REPORT



REFERENCES [1] IMO (1995). Performance standards for electronic chart display and information systems

(ECDIS), IMO Resolution A.817(19). [2] Norway (2005). FSA – Large Passenger Ships – Navigational Safety, Submitted by Norway,

NAV 51/10. [3] Japan (2006). FSA - Consideration on utilization of Bayesian network at step 3 of FSA

Evaluation of the effect of ECDIS, ENC and Track control by using Bayesian network, MSC 81/18/1

[4] Denmark and Norway (2006). FSA Study on ECDIS/ENCs, Submitted by Denmark and Norway, MSC 81/24/5.

[5] Denmark and Norway (2006). FSA Study on ECDIS/ENCs: Details on Risk Assessment and Cost Benefit Assessments, Submitted by Denmark and Norway, MSC 81/INF.9.

[6] Sæther, L.K., Motrøen, S., Listerud, S.H., Lepsøe, A., Georgantzis, A. and Hoffmann, P. (2003). Formal Safety Assessment of Cruise Navigation, DNV Report no. 2003-0277, Det Norske Veritas.

[7] Hoffmann, P., Kjellström, S., Georgantzis, A. (2003), FSA generic vessel risk, Bulk Carrier. DNV Report 2003-1073.

[8] Kjellström, S., Johansen, C.B. (2004). FSA generic vessel risk, Double Hull Tanker for Oil. DNV Report 2003-0425, rev. 03.

[9] Hoffmann, P., Kjellström, S., Georgantzis, A. (2003). FSA generic vessel risk, Tanker for Chemicals. DNV Report 2003-1070.

[10] Kjellström, S., Johansen, C.B. (2004), FSA generic vessel risk, Container Vessel. DNV Report 2003-1077, rev. 02.

[11] Hoffmann, P., Kjellstrom, S., Georgantzis, A. (2003), FSA generic vessel risk, General Cargo Vessel. DNV Report 2003-1074.

[12] Kjellström, S., Johansen, C.B. (2004). FSA generic vessel risk, LNG Tanker. DNV Report 2003-1072, rev. 02.

[13] Spouge, J.R. (2002), A Simple Model of The Costs of Ship Accidents Rev 1. DNV Job No C383184/5

[14] Vanem, E., Endresen, Ø., Skjong, R. (2007). CATS – Cost-effectiveness in Designing for Oil Spill Prevention. To be published in Proc. of PRADS 2007 (October 2007).

[15] Vanem, E., Endresen, Ø., Skjong, R. (2007). Cost Effectiveness Criteria for Marine Oil Spill Preventive Measures. Submitted to Reliability Engineering & Systems Safety.

[16] Endresen, Ø., Sørgård, E., Sundet, J.K., Dalsøren, S.B., Isaksen, I.S.A., Berglen, T.F. and Gravir, G. (2003). Emission from international sea transportation and environmental impact, Journal of Geophysical research, vol 108, 4560.

[17] Primar Stavanger – IC-ENC (2004). Facts about charts and carriage requirements [18] World Fleet Statistics 2005 (2006). Lloyd’s Register – Fairplay Ltd. [19] Fearnley’s review 2005 (2006). Fearnresearch. [20] ISL Market Analysis 2006 – World Port Development (2006). Institute of Shipping

Economics and Logistics. [21] Minerals Yearbook 2004. Vol. I, Metals & Minerals. U.S. Government Printing Office.

Documents

DNV RESEARCH INNOVATIONresearch.dnv.com/skj/FSA-ENC/ENC.pdf · adequate for fast generation of computer images on the screen. In order to get data structures that facilitate rapid