Oil Spill Risk Assessment

REPORT

New Zealand Marine Oil Spill Risk Assessment 2004 Prepared for

Level 8, gen-i Tower 109 Featherston Street PO Box 27006, Wellington

December 2004

Contents

Executive Summary-------------------------------------------------------------------------------------------- ES-1 1

Introduction ------------------------------------------------------------------------------------------------- 1-1 1.1 1.2 1.3 1.4 1.5

2

2.2

Information Gathering 2.1.1 Consultation Plan 2.1.2 Information Sources for Assessment of Spill Likelihood 2.1.3 Information Sources for Assessment of Spill Consequences Categorisation Of Oil Types

2-1 2-1 2-1 2-3 2-3

Risk Creator Groups ------------------------------------------------------------------------------------- 3-1 3.1 3.2

3.3

4

1-1 1-2 1-3 1-3 1-4

Background Data ----------------------------------------------------------------------------------------- 2-1 2.1

3

Overview Background Risk Assessment Quantitative Risk Analysis – Approach and Limitations Methodology

Introduction Vessel Activity 3.2.1 Tankers 3.2.2 International Cargo and Passenger Vessels 3.2.3 Coastal Cargo and Passenger Vessels 3.2.4 Fishing Vessels 3.2.5 Small Craft Non-Vessel Activity 3.3.1 Offshore Oil and Gas Production 3.3.2 Exploration Activity 3.3.3 Bulk Oil Transfer Operations 3.3.4 Bulk Storage Terminals 3.3.5 Product Pipelines and Bunker Lines 3.3.6 Bunkering Operations

3-1 3-1 3-2 3-3 3-4 3-4 3-5 3-6 3-6 3-6 3-7 3-7 3-7 3-8

Changes Since 1997 and Future Trends ---------------------------------------------------------- 4-1 4.1

4.2

4.3

Oil Industry Changes Since 1997 4.1.1 Refinery Receipts and Exports 4.1.2 FPSO / New Plymouth Liftings 4.1.3 Product Imports 4.1.4 Product Tanker Movements 4.1.5 Patterns of Coastal Voyages 4.1.6 Tankers 4.1.7 Port and Related Activities Future Trends in the Oil Industry 4.2.1 Marsden Point Refinery 4.2.2 Taranaki/New Plymouth 4.2.3 Coastal Deliveries and Product Imports 4.2.4 Port Infrastructure Other Changes Since 1997

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

i

4-1 4-1 4-1 4-2 4-2 4-3 4-3 4-4 4-5 4-5 4-5 4-6 4-6 4-6

Contents

4.4 5

5.2

5.3

Consequence Assessment 6.1.1 Approach 6.1.2 Methodology 6.1.3 Conclusions

6-1 6-1 6-1 6-4

Oil Spill Scenarios Coastal Spill Scenarios Port Spill Scenarios 7.3.1 Maui Field and FPSO

7-1 7-1 7-1 7-2

Spill Likelihood Model----------------------------------------------------------------------------------- 8-1 8.1 8.2 8.3

8.4 8.5 8.6 8.7 8.8 8.9 9

5-1 5-1 5-1 5-2 5-2 5-2 5-2 5-3 5-3 5-3 5-3 5-3 5-5

Oil spill Scenarios ---------------------------------------------------------------------------------------- 7-1 7.1 7.2 7.3

8

New Zealand Oil Spills 5.1.1 Sizes of Spills 5.1.2 Where Spills Occur 5.1.3 Oil Spill Causes 5.1.4 Sources of Oil Spills International Oil Spill Data 5.2.1 Size and Frequency of Oil Spills Globally 5.2.2 Causes of Spills 5.2.3 Australia 5.2.4 United States Accident/Incident Data 5.3.1 New Zealand Data 5.3.2 International Data

Oil Spill Consequence Assessment ---------------------------------------------------------------- 6-1 6.1

7

4-7

Oill Spills and Accidents ------------------------------------------------------------------------------- 5-1 5.1

6

Recent Developments and Future Trends

Generalised Structure Common Features Of Both Models Risk Measures 8.3.1 Return Period 8.3.2 Spill Rate 8.3.3 Coastal Exposure Coastal Spill Model Inputs Coastal Spill Model Outputs Port Spill Model Inputs Port Spill Model Outputs Limitations of Spill Likelihood Model Sensitivity

8-1 8-1 8-2 8-2 8-3 8-3 8-3 8-4 8-7 8-9 8-9 8-10

Results of Spill Likelihood Analysis---------------------------------------------------------------- 9-1 9.1

Coastal Spill Risk 9.1.1 Spill Size vs. Return Period 9.1.2 Risk Weighted Average Spill rate 9.1.3 Spill Exposure

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

ii

9-1 9-1 9-1 9-2

Contents

9.2 9.3 10

Port Spill Risk Total Spill Risk

9-8 9-8

Discussion and Conclusions of Spill Likelihood Modelling-------------------------------10-1 10.1 General Comments on Results 10.1.1 Benefits of Approach 10.1.2 Spill Return Period 10.1.3 Spill Rate 10.2 Comparison with 1998 10.2.1 Return Period 10.2.2 Spill Rate 10.3 Conclusions and Recommendations

11

10-1 10-1 10-1 10-1 10-2 10-5 10-5 10-5

Sensitivity Analysis and Data Limitations-------------------------------------------------------11-1 11.1 Sensitivity Analysis 11.1.1 Average oil carried - 2004 model 11.1.2 Spill Rate Analysis 11.2 Quality and Limitations of Input Data

11-1 11-1 11-3 11-4

12

References -------------------------------------------------------------------------------------------------12-1

13

Glossary of Terms ---------------------------------------------------------------------------------------13-1

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

iii

List of Tables, Figures, Plates & Appendices

Tables Table 2-1: Categorisation of Oil Types .................................................................................................... 2-4 Table 4-1: Hull Types of Tanker Visiting Marsden Point 2002/2003...................................................... 4-4 Table 4-2: Recent Developments and Future Trends ............................................................................... 4-8 Table 6-1: Example of a Consequence Assessment Summary at Cell Level ........................................... 6-2 Table 8-1: Vessel Activity by Coastal Area ............................................................................................. 8-6 Table 8-2: Vessel Activity and Bunker Fuels by Port.............................................................................. 8-8 Table 9-1: Coastal Spill Summary ........................................................................................................... 9-3 Table 9-2: Port Spill Risk Summary ........................................................................................................ 9-9 Table 9-3: Spill Size Exceedance by Port and Coastal Area.................................................................. 9-10 Table 10-1: Effect of Improved Data and Model Changes .................................................................... 10-3

Figures Figure 6-1: New Zealand Oil Spill Consequences Map ........................................................................... 6-3 Figure 8-1: Coastal Areas Used in Model ................................................................................................ 8-5 Figure 9-1: Return period for 1000-tinne Spill by Coastal Area.............................................................. 9-4 Figure 9-2: Coastal Spill Rate in Tonnes/Year......................................................................................... 9-5 Figure 9-3: Coastal “Exposure” in Tonne-Years...................................................................................... 9-6 Figure 9-4: Tonne-year and Spill Rate Summary by Coastal Area and Source ....................................... 9-7 Figure 9-5: Spill Return Period (Combined Port, Coastal and FPSO) ................................................... 9-11 Figure 10-1: Average Spill Rate by Vessel Category ............................................................................ 10-4

Appendices Appendix A

Terms of Reference for the 2004 Risk Assessment

Appendix B

Information on Non-Vessel oil activity (cargo transfers and bunkering, oil storage terminals and transfer pipelines)

Appendix C

Information on Vessel activity with the potential for oil spills

Appendix D

Analysis of historical New Zealand and international data for oil spills and shipping accidents

Appendix E

Consequences Assessment

Appendix F

Description of the Quantitative Risk Model for spill likelihood

Appendix G

Results and Summary of Quantitative Modelling

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

iv

Executive Summary

Introduction The 2004 New Zealand Marine Oil Spill Risk Assessment has been prepared in accordance with the Maritime Safety Authority’s New Zealand Marine Oil Spill Response Strategy. The Strategy requires a comprehensive national risk assessment to be carried out and for this to be reviewed at least every six years. A risk assessment was first prepared in 1992, followed by a second risk assessment in 1998. The risk assessment comprises three phases: Phase 1 – Preparation Phase 2 – Risk Assessment Consultancy Phase 3 – Analysis and Review This study comprises Phase 2 of the risk assessment and updates the information on oil industry and maritime activity from 1998 and the corresponding quantitative assessment of spill likelihood. The study uses the same model, with some minor changes. A new framework for the assessment of oil spill consequences has also been developed, building on previous work on coastal sensitivity by the Maritime Safety Authority (MSA).

Background Data A consultation plan was developed for the purposes of information gathering. Organisations contacted for the likelihood assessment included the main oil industry players, port companies, shipping operators and harbourmasters. Data on oil related activity was collected for the period July 2002 – June 2003 for comparison with calendar year 1997 used in the previous study. For the consequences framework, information sources included central and regional government agencies. The framework was developed co-operatively through a series of workshops.

Risk Creator Groups The study adopted the same risk creator groups used in the 1998 work, covering: •

vessel activity (four categories of tanker and six other vessel categories) and

•

non-vessel activity (cargo transfer, bunkering and offshore exploration and production).

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

ES-1

Executive Summary

Vessel activity for the 2002/2003 period is summarised below: Vessel category

Number of port calls

Equivalent vesseldays in NZ coastal waters (estimated)

Tankers

539 (includes 47 large crude carriers)

580

Foreign passenger and cargo

2,800 container vessels 3,600 others

8,500

Coastal passenger and cargo

9,400

2,700

Fishing (> 6m)

10,000 (estimated)

83,000

Small commercial craft (>6m)

46,000 (estimated)

69,000

In terms of non-vessel activity for the period: •

Around 5 million tonnes of crude oil were imported into New Zealand at the Marsden Point refinery;

•

Around 13 million tonnes of indigenous crude oil, condensates and petroleum products were transferred to and from tankers at New Zealand ports, comprising 590 cargo transfer operations; and

•

Around 570,000 tonnes of bunker fuels and lube oils were loaded in over 3,000 bunkering operations at the main ports.

Changes in Activity Since 1997 Refinery intake volumes were up by about 9% on 1997, but local crude and condensate was down by about 70%, reflecting declining production from the Maui field. Fuel import volumes increased by around 35%, but volumes distributed by the coastal tanker fleet were down about 2% and there was a significant reduction (around 15%) in port visits due to improved fleet utilisation. A significant proportion of foreign tankers visiting (around 80%) were either double-hulled, double-sided or doublebottomed and this is expected to increase. The last six years has seen the introduction of larger container vessels to New Zealand – US – Europe services, significant growth in traffic in regional ports such as Tauranga and Napier, new deepwater berths at Marsden Point and Picton and changes to coastal vessel fleets. In addition there has been an increase in cruise vessel activity with several new large vessels now visiting regularly in the cruise season.

Oil Spills and Accidents – New Zealand For the period July 1998 – April 2004, there were 778 oil spills reported to the MSA. However, only about 60% of the spill records contained sufficient detail for useful analysis. Of these, there were

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

ES-2

Executive Summary

20 spills exceeding 1 tonne for the period, four of 20 tonnes or greater and one of around 300 tonnes (in 1998). For the same period there were three major accidents in New Zealand involving spills of 50 tonnes or greater, and at least a further six where there was the potential for a major spill, but where fortunately this did not occur. In addition there was a much larger number of other maritime accidents and incidents reported involving vessels over 5000 GRT, where serious damage resulting in spills could have occurred, but was averted.

Oil Spills and Accidents – International International figures for large oil spills (over 700 tonnes) show a significant decrease from the 1970’s (around 24 spills per year of this size on average) to approximately 9 per year in the 1980’s and 8 per year in the 1990’s. The average since the beginning of 2000 has been 4 per year. Statistics for serious incidents involving tankers show a significant downward trend over the last 15 years with over 900 incidents per year reported in 1990 and on average less than 200 per year since 2001. Comparable figures for vessels other than tankers were not available.

Consequence Assessment Framework A framework for assessing the consequences of oil spills on coastlines was developed in conjunction with the MSA and central and regional government agencies. This built on the earlier work in the 1998 study and subsequent work by the MSA in developing the Voluntary Vessel Routeing Code in 2001. The resulting framework is robust and intuitive and will provide the MSA and regional authorities with a consistent way of assessing and presenting the impacts of oil spills on people and the environment at a sufficiently detailed level for regional spill response planning.

Modelling of Oil Spill Likelihood The quantitative model developed in 1998 for estimating oil spill likelihood in ports and coastal areas was updated with the 2002/2003 data. Some refinements to the model were made to improve its reliability and cater for improved input data. The model calculates: •

Return period for spills of different sizes

•

An average risk weighted spill rate (equivalent tonnes per year of oil spilled

•

A measure of exposure to oil spills (expressed in tonne-years) which reflect oil volumes and transit times for each coastal area.

Given the inherent level of uncertainty associated with any quantitative modelling (both in the input data and the assumptions), the model output is far more reliable when used as a comparative tool (relative spill

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

ES-3

Executive Summary

rate or exposure across ports, risk creators or oil types, for example) than when used in absolute sense, and this should be borne in mind when interpreting the results.

Results of Modelling The estimated 1% annual PEL (Probability of Exceedance Level) spill size - the size of spill that is expected to be equalled or exceeded once every 100 years for all of New Zealand - is around 3,500 tonnes. The corresponding figure for 1998 is marginally higher. A more detailed comparison cannot readily be made as the 1% PEL spill size is derived from interpolation of plotted data rather than estimated directly. This comparison is also affected slightly by some changes to modelling inputs (as opposed to the level of activity), however this has limited effect on the return period values. (It is noted that as part of this study, the 1998 figures have been checked and errors subsequently corrected, hence the difference from the figure quoted above and the 7,000 tonnes previously derived). Within the model, spills of the order of 3,000 tonnes derive solely from tankers, and as tanker data for both 1997 and 2002/2003 years was quite comprehensive, the comparison is considered valid. As expected, the difference is not significant between the two years, with some changes in patterns of tanker activity offsetting others. For the ports, Auckland has the largest spill at the 1% PEL, followed by Wellington and Lyttelton. For the coastal areas, East Auckland (encompassing the area between Whangarei and the Coromandel Peninsula). This is a reflection of the dominance of international and coastal cargo vessel activity in overall port and coastal movements. For the coastal areas, the area with the highest estimated average annual spill rate and spill exposure is East Northland (North Cape to Whangarei) which reflects the activity of large crude carriers. However, of the ports, Auckland has the highest spill rate reflecting the level of vessel activity of all sizes. The combined average risk-weighted spill rate for ports is around 3 – 3.5 times higher than for the combined coastal areas, reflecting the greater risks associated with vessel movements in and out of harbours. The contribution to the overall spill rates from bunkering and cargo transfer are not significant. The contribution to the estimated spill rate by vessel group (including both ports and coastal areas) is as follows: Tankers

28%

International cargo and passenger vessels

27%

Coastal cargo and passenger vessels

5%

Maui platforms, pipelines and FPSO

5%

Fishing vessels

13%

Small craft

21%

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

ES-4

Executive Summary

Non-persistent oils (including petrol, condensates and gas oils) contribute around 62% of the total spill rate, but the split between persistent and non-persistent oils varies significantly across the vessel categories. The overall estimated average spill rate for 2002/2003 is about 5% lower than in 1998, when the changes in the model are taken into account.

Conclusions The 2004 risk assessment gives an updated (and we believe) more reliable picture of the likelihood of an oil spill in New Zealand waters than the previous study. It also includes better information on fishing vessels and smaller vessel activity and their contribution to the overall spill risk. Overall, it should provide greater insight into the patterns of shipping activity and the relative contribution to oil spill risk from the different risk creators, as well as giving an improved picture of the geographical spread of spill risk. It must be emphasised that the oil spill risk assessment is an ongoing process, with the aim over time of improving the characterisation of the risk so as to better understand it, while at the same time actively working to reduce that risk. There are two remaining areas of significant uncertainty that should be addressed prior to the next risk assessment being carried out: •

The number of vessel movements through each coastal area.

•

The quantity of oil (fuel) carried on average by each category of vessel.

These items are not really an issue for tankers, where the overall numbers of coastal movements in a year is small (and can be quite accurately estimated) and where the quantity of oil being carried is relatively easily ascertained. However, for other foreign vessels, which account for over 8,500 vessel-days in coastal waters per annum, the lack of hard information is significant. The uncertainties are also less significant for coastal cargo and passenger vessels, which contribute 50% more port movements than foreign vessels, but only one third the number of vessel-days and carry on average only a fraction of the fuel. In order to improve the accuracy of the risk assessment, the MSA should consider focussing efforts on improved reporting and data collection from these foreign vessels.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

ES-5

Introduction

1

Introduction

1.1

Overview

SECTION 1

This report describes the 2004 Marine Oil Spill Risk Assessment prepared for the Maritime Safety Authority (MSA). The report is set out as follows: Section 1

Sets out the background to the risk assessment and the terms of reference for the work.

Section 2

Provides background information on data sources and oil types.

Section 3

Identifies and characterises potential oil spill creators (vessel and non-vessel activity).

Section 4

Discusses changes in risk creating activity in New Zealand waters since the last risk assessment in 1998.

Section 5

Presents a summary of historical oils spill and maritime accident data.

Section 6

Describes the approach to assessing the consequences of oil spills.

Section 7

Describes the oil spill scenarios used in the quantitative model for spill likelihood.

Section 8

Describes the methodology used in the quantitative model.

Section 9

Presents the results of the analysis.

Section 10

Discusses the results and provides conclusions and recommendations.

Section 11

Includes a sensitivity analysis for the model and discusses the quality and limitations of the data collected.

Section 12

Includes the main references used in the study.

Section 13

Provides a glossary of terms used in the text.

The appendices provide more detail to supplement specific sections of the main report. Appendix A

Contains the Terms of Reference for the 2004 Risk Assessment

Appendix B

Provides information on Non-Vessel oil activity (cargo transfers and bunkering, oil storage terminals and transfer pipelines).

Appendix C

Provides information on Vessel activity with the potential for oil spills.

Appendix D

Presents an analysis of historical New Zealand and international data for oil spills and shipping accidents.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

1-1

Introduction

SECTION 1

Appendix E

Contains the report on the Consequences Assessment.

Appendix F

Describes the Quantitative Risk Model for spill likelihood in detail.

Appendix G

Contains more detailed summaries of the results of the quantitative model.

1.2

Background

Under the Maritime Transport Act 1994, the Maritime Safety Authority (MSA) has a mandate to promote a safe marine environment and provide effective marine pollution prevention. As part of its obligations, the MSA is responsible for the New Zealand Marine Oil Spill Response Strategy. The strategy provides: “Unless other factors dictate an earlier timetable, a comprehensive national risk assessment will be carried out every six years. This risk assessment should focus on existing contributions to risk from the various maritime sectors, as well as those new or potential activities which are reasonably foreseeable. The first national risk assessment for New Zealand was completed in 1992. This established the basis for the first Marine Oil Spill Response Strategy. In 1998, the MSA completed its second national marine oil spill risk assessment. The 1998 risk assessment measured and presented marine oil spill risk in a manner similar to that used for other forms of emergency response, and also addressed the possibility of an unpredictable, catastrophic spill which the 1992 study did not. The probability of a spill event of a particular size occurring in any given year (Probability of Exceedance Level or PEL) was estimated and assigned a value. New Zealand has chosen to plan for a spill event with a PEL of 1% to ensure a cost-efficient domestic level of response capability, which meets international standards for marine oil spill preparedness. From the 1998 study, a 1% PEL spill size of approximately 7,000 tonnes was derived. Beyond this level, New Zealand is able to augment its domestic capability with international assistance agreements. In accordance with the Strategy, the 1998 risk assessment is now being reviewed. The 2004 New Zealand Marine Oil Spill Risk Assessment comprises three phases: Phase 1 – Preparation: Defines the data models and determined the Terms of Reference for the Risk Assessment Consultancy. Phase 2 – Risk Assessment Consultancy: Involves the collection and collation of local and international data on activities which create risk, the investigation and reporting of best practice, definition of appropriate return times for spill incidents, and analysis of locations which have threatened resources and their relative sensitivity. Phase 3 – Analysis and Review: Involves the policy analysis of the data as analysed and provided to MSA from Phase 2, the review of existing strategies and programmes and the development of future policy options.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

1-2

Introduction

SECTION 1

Phases 1 and 3 are being carried out by the MSA. URS New Zealand was commissioned in December 2003 to carry out Phase 2. This phase has three components; •

to establish a methodology to review the likelihood of a range of oil spill scenarios;

•

to identify those location specific external environmental factors that may contribute to the overall risk; and

•

to assess the magnitude of consequences of a range of oil spill scenarios on coastal resources.

The terms of reference for the risk assessment are presented in Appendix A.

1.3

Risk Assessment

The objective of risk management is to avoid or reduce the impact of (usually) undesirable events. These may include loss of life, illness or injury; damage to property and consequential loss; or environmental impact. By carrying out a formal and structured assessment of risks, the nature of the risks can be better understood, allowing informed decisions to be made about how best to deal with them. Formal risk management involves the following steps: 1.

Establishing the context;

2.

Identification of the risks;

3.

Analysis of risks: estimation of the consequences and the likelihood of occurrence where risk = likelihood x consequence;

4.

Evaluation of risks - ranking of risks from different sources and assessment against some appropriate criteria for the purposes of prioritisation;

5.

Treatment of risk.

The approach taken is consistent with that outlined in the joint Australian/New Zealand Risk Management standard AS/NZS4360:1999.

1.4

Quantitative Risk Analysis – Approach and Limitations

As part of this assessment, a quantitative analysis of the likelihood of marine oil spills was carried out using the model developed for the 1998 study. The sources of risk and types of events that could lead to oil spills were determined and the frequency of these events or accident scenarios estimated using relevant historic data, predictive techniques including

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

1-3

Introduction

SECTION 1

event trees, and expert judgement. The model predicts the size of the spill and the oil type for each spill source (vessel type, or non-vessel as applicable) and for each location (coastal area or port). There are uncertainties associated with the data, methods and models used to identify and estimate risk. Sensitivity analysis can be used to determine the effect of changes in the input parameters to the overall outcome and determine which of these parameters has the most influence on the estimation of risk. Those parameters identified as being most significant can then be refined to a greater degree of certainty. Generally, conservative figures are used in the analysis, so that the risk will tend to be over-estimated rather than under-estimated. The greater value of risk quantification lies in its use in comparing the risk in different areas or for different scenarios, as the inherent uncertainties in the methods are reduced when the results are used on a comparative basis. The main value derived from the process of systematic analysis is a better understanding and appreciation of the oil spill risk and the identification of important issues for improving management mechanisms that might otherwise have been overlooked.

1.5

Methodology

Phase 2 of the Oil Spill Risk Assessment was initiated in December 2003, and the data in this report has been collected and collated following an extensive data gathering exercise. This has involved contact with a wide range of organisations to obtain information on port activities, vessels moving around the coasts and entering New Zealand ports, oil movements, coastal and port hazards and sites sensitive to oil spills. A framework for consequences assessment was developed using existing information, case studies and workshops. It built on previous work in this area - particularly the MSA’s Vessel Routeing Code Review. New Zealand’s coastline was divided into cells, which were then ranked in terms of their likely sensitivity to oil impacts on a 5 point scale. Sensitive areas were identified taking into account environmental, social, cultural, economic and scientific factors. This information was visually represented on a map of the New Zealand coastline. The methodology was developed by progressive review and workshops. This report and the accompanying appendices discuss the various aspects of the project, including the data gathering and collation, identification and description of risk creator groups, development of a framework for assessing spill consequences and the quantitative analysis of likelihood (scenarios and model), including the many assumptions made. Some of the information used in the study has been provided on a confidential basis and/or is commercially sensitive and in most cases this information has been aggregated within the report so that it is not readily identified to a source. However, to protect URS’ interests we have assumed that the whole report will remain confidential. It should be noted that although the data gathering was extensive, there remain some gaps and inconsistencies in the information provided, resulting, in many cases, from lack of data held or poor

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

1-4

Introduction

SECTION 1

responses from a small number of organisations contacted. The report comments on the significance of this in Section 10.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

1-5

Background Data

2

Background Data

2.1

Information Gathering

2.1.1

Consultation Plan

SECTION 2

For the purposes of obtaining the required data for the risk assessment, a Consultation Plan was prepared. The purpose of the plan was to: •

Set out the purpose of the data collection;

•

Establish key messages to be communicated to stakeholders during the project;

•

Identify known data sources and develop contact lists;

•

Agree an approach and programme for data collection and consultation;

•

Identify any communication issues prior to starting collection of data, including lessons learned from the 1998 risk assessment.

As information requirements for the quantitative risk modelling (likelihood) and the consequence assessment were somewhat different, these were detailed separately in the consultation plan.

2.1.2

Information Sources for Assessment of Spill Likelihood

Organisations contacted URS engaged a sub-consultant, Hale & Twomey Limited, to gather and analyse data from the oil industry. This included bulk oil transport and handling (both crude oils and refined products) as well as bunkering, throughout New Zealand. The data collected by Hale & Twomey covered operations by: •

the New Zealand Refining Company (NZRC);

•

Shell Todd Oil Services (STOS);

•

the major oil companies (BP, Caltex, Gull, Mobil and Shell) and;

•

the coastal tankers - Silver Fern Shipping (SFSL)

Data collection from other organisations was carried out by URS directly. Groups contacted included: •

all port companies

•

all Regional Councils (harbourmasters and Regional On-scene Commanders (ROSCs)

•

shipping operators

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

2-1

Background Data

•

SECTION 2

Royal New Zealand Navy

Data was also collected from government agencies, including: •

Maritime Safety Authority (New Zealand historical oil spill data and maritime accident data, vessel movements and fleets from the MSA’s MSC1 and SSM2 databases)

•

Ministry of Fisheries (catch effort data)

Reference Period The 1998 risk assessment was started in March 1998 and collected data for calendar year (January – December) 1997. For the 2004 risk assessment, data was requested for the 12-month period 1 July 2002 – 30 June 2003. The rationale for this was that it matched the financial reporting year of many organisations. Calendar year 2002 was considered to be too long ago, and data for the year ending December 2003 was not all available at the time the data collection was started.

Outcome and comment on data received Information was received back from the most of the organisations contacted. Given the use of a subconsultant for the oil industry data, and the existence of established relationships with many of data providers from the 1998 risk assessment, the majority of contact was conducted by phone and email. As for the 1998 work, the information received has been varied in detail, mainly because the collection and reporting of this data within the numerous organisations contacted is largely focussed on their own commercial interests. However, we consider that the quality of the data compiled for this study has improved significantly. Specifically: •

The use of an oil industry sub-consultant has enabled a more complete and accurate picture of oil industry activity to be generated. In particular, we believe that data on tanker movements through coastal areas and on bunkering operations is more complete and reliable.

•

Vessel movement data from port companies is now available electronically. While reporting measures (for example, ship descriptions) still vary between ports, the availability of raw electronic data allows for improved data manipulation and analysis.

•

The provision of catch effort data by MFish has allowed a far more accurate picture of commercial fishing vessel activity in New Zealand coastal waters to be developed.

1

MSC – Maritime Safety Charge

2

SSM – Safe Ship Management System

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

2-2

Background Data

•

SECTION 2

The establishment by MSA of the Safe Ship Management (SSM) system and Safe Operational Plans (SOPs) since the 1998 study has provided more complete data on the number, type and distribution of smaller commercial vessels operating in New Zealand. However, there remains a significant level of uncertainty about movements (i.e. vessel activity) for small vessels in particular and the average quantity of oil carried for all vessels (vessel profiles). In the quantitative risk model, the likelihood of an oil spill is primarily a function of these two measures.

2.1.3

Information Sources for Assessment of Spill Consequences

Information sources used for the development of a consequences framework included: •

Work on coastal sensitivity in the 1998 risk assessment and previous MSA work including the Vessel Routeing Code

•

Regional and central government – including Regional Councils, Ministry of Fisheries (Marine Biosecurity), Oceans Policy, Ministry for the Environment and the National Rural Fire Authority (e.g. Wildfire Threat Analysis Workbook)

•

International sources – a number of initiatives have been taken internationally that provided useful background including work done by:

2.2

-

The National Oceanic and Atmospheric Administration of the United States (NOAA);

-

The Pacific States/British Columbia Oil Spill Task Force and the US Coast Guard;

-

The Department of the Environment, Transport and Regions of the United Kingdom; and

-

The National Oceans Office of Australia.

Categorisation Of Oil Types

Following the methodology used in the 1998 study, the types of oil handled in New Zealand were split into five categories that broadly represent the characteristics of the oil when spilled on water. This ranking was based on the work of Allen (1997), which in turn was largely derived from classification systems used by NOAA and ITOPF (International Tanker Owners Pollution Federation). The categories, which are defined by specific gravity (SG) and a description, are presented in Table 2-1.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

2-3

Background Data

SECTION 2

Table 2-1: Categorisation of Oil Types Group

Specific Gravity (SG)

Description and Examples

I

< 0.8

II

0.80 - 0.85

Middle distillates and light-medium crudes AGO, MGO3, NGO (Navy gas oil) Light-medium crudes

III

0.85 - 0.95

Medium – heavy crudes, light fuel oils LFO4 LCO – Light Cycle Oil (gas oil blendstock) Lube oils and lube blendstocks Medium – heavy crudes

IV

0.95 - 1.00

Fuel oils and residues HFO, HBFO5 Residues

V

> 1.0

Light distillates Maui and Kapuni condensate Gasoline blendstocks Motor spirit (RMS/PMS), Avgas Jet A1, kerosene

Very heavy fuel and bunker oils Bitumen (B18, B45)

In order to more accurately categorise the oils carried in New Zealand waters, some variations have been made against the SG scale. These include: •

Light crude oils have been included as Group II although their SG is less than 0.8 in some cases (there are some light crudes of this category). It is noted that even crude oils of SG less than 0.8 qualify as persistent crude by the ITOPF definition (see Section below).

•

All residues have been included in Group IV even if their SG is less than 0.95. Residue is ‘topped’ crude oil similar to fuel oil so is best represented by this grouping.

•

While some crude oils (both local and imported) have high pour points due to their wax content, which can make spill characteristics worse, this has not been used to adjust the categorisation. The pour points of all crude oils were identified in the NZRC intake data supplied. In previous studies the local crude oils (McKee Blend and Maui F) were classified as Group IV. While their pour points

3

AGO, MGO – automotive and marine gas oil (diesel)

4

LFO – light fuel oil (40 cSt)

5

HFO – heavy fuel oil (180 cSt), HBFO (BFO) – heavy bunker fuel oil (380 cSt)

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

2-4

Background Data

SECTION 2

are relatively high (10-200C) they are lower than some imported crude oils, so were not deemed a special case. Both of the indigenous crude oils have a very low residue percentage, despite the higher pour points. Gas oil specifications (mainly in terms of cloud point) vary with time of year and delivery location, but this should not significantly effect the fate and behaviour of this product in the event of a marine spill. It is noted that marine gas oil (MGO) was not available in New Zealand during the period surveyed although it has since been reintroduced at a limited number of locations. Other than sulphur and cold properties, MGO is similar to AGO and therefore should behave in similar fashion in the event of a marine spill. In this report, the term “finished products” is used to describe refined oils such as motor spirits, gas oils, fuel oils and bitumens - in order to distinguish them from refinery feedstocks such as crude oils and condensate.

Persistent and Non-persistent Oils For the purposes of assessing environmental impacts, the persistence of the oil in the marine environment is the main concern. The above five groups have been split into two – non-persistent oils (Groups I and some Group II) and persistent oils (Groups III, IV and V, and the remainder of group II). The spill risk has therefore been assessed in terms of the non-persistent and persistent oils according to this split. In very general terms, non-persistent oils are those which are more volatile and will evaporate and/or disperse relatively quickly, but are also generally more toxic to marine life. Persistent oils are denser, more likely to form oil-water emulsions and are slower to disperse, but generally have a smothering effect rather than being particularly toxic. There is no universally accepted definition of non-persistent/persistent oils although ITOPF has developed the following definition for non-persistent oils. A non-persistent oil is one where: •

At least 50% by volume distils at 3400C;

•

At least 95% by volume distils at 3700C.

This definition works well with refined products (up to AGO) being classified as non-persistent along with condensates, while all crude oil and heavier products are persistent. This split is similar to that used in shipping contracts where feedstocks and products are classified as white (non-persistent) or black (persistent). With this definition, all Group II oils have been classified as persistent except for the gas oil grades (AGO/MGO/NGO). All crude oils have been classified as persistent.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

2-5

Risk Creator Groups

3

Risk Creator Groups

3.1

Introduction

SECTION 3

The main risk creator groups have been identified and are discussed below. These groups have generally been characterised by activity and by vessel type, as follows:

Vessel •

tankers (four sub-categories)

•

international cargo/passenger vessels

•

coastal cargo/passenger vessels

•

fishing vessels

•

small craft

Non-vessel •

offshore oil exploration and production (including the Maui platforms and pipelines and the FPSO6)

•

bulk oil transfers and bunkering operations

•

bulk storage terminals

•

pipelines and bunker lines

For each of the risk creator groups, the risk has been assessed on the basis of the level of activity in the 2002/2003 year. Changes which have occurred since 1997, and current trends and likely changes over the next five years are discussed in Section 4.

3.2

Vessel Activity

For the purposes of this study, coastal and international vessel types were divided into several categories, based on size, and cargo and fuel types. These categories are the same as those used in the 1998 study, except that two additional groups have been included this time. A brief description of the categories used is presented below. Further descriptions of the categories used, and statistics on movements is presented in Appendix C.

6

FPSO – Floating Production, Storage and Offtake Facility

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

3-1

Risk Creator Groups

3.2.1

SECTION 3

Tankers

Tankers have been divided into four groups - large crude carriers, indigenous crude/condensate/naphtha tankers, foreign product tankers and coastal product tankers. While the latter three categories have ships that are interchangeable, the categories provide a useful distinction between cargo types. (1) Large Crude Carriers These vessels have between 80,000 and 153,000 tonnes capacity and bring in crude oils to the Marsden Point refinery, primarily from Australia, South East Asia and the Middle East. During the period July 2002 through June 2003 there were 47 such vessel arrivals. It is noted that larger crude carriers have restricted cargo capacities when coming to Marsden Point because of draft limitations; the largest cargo received during the period was 131,000 tonnes. (2) Indigenous Crude/Condensate/Naphtha Tankers These are foreign tankers of up to 40,000 tonnes capacity. They are often the same tankers used for product imports (Category 3) and in the case of movements to Marsden Point, for coastal product deliveries (Category 4). They operate in a number of different services when in New Zealand waters: •

shipment of Maui and Kapuni condensates and onshore crude from New Plymouth to Marsden Point (note: there were no shipments of onshore crudes to Marsden Point in the period covered)

•

export of Maui condensate and onshore crude from New Plymouth direct to refineries overseas (all shipments went to Australia in the period covered)

•

export of naphtha7 from New Plymouth to Asia

•

shipment of Maui F sands from the FPSO direct to Marsden Point (has not happened for 4 to 5 years)

•

export of Maui F sands from the FPSO direct to overseas refineries

Typically, a tanker will either load a combination of condensate and crude oil from New Plymouth for Marsden Point or export, or load at both New Plymouth and the FPSO for export. There was only one FPSO only loading in the period covered, a significant change from 1997. Naphtha loadings are full, single grade shipments from New Plymouth. (3) Foreign Product Tankers These are similar to the crude/condensate tankers (sometimes the same vessels) with cargo capacities of up to 40,000 tonnes. They are primarily used for the following: •

delivery of imported semi-refined blendstocks and residues to the Marsden Point refinery;

7

Naphtha is a light cut from condensate that is used as a petrochemical feedstock and also as an unrefined gasoline blendstock. The naphtha stream is cut from the Maui condensate. This operation commenced after 1997.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

3-2

Risk Creator Groups

•

SECTION 3

delivery of imported finished products (motor spirit, gas oil and jet fuel) direct to coastal ports (generally, but not exclusively from Australia and Asia).

Some voyages have also delivered bitumen and fuel oils. There are also a number of smaller tankers in regular use (4,000 – 9,000 GRT) which predominantly carry chemicals, solvents, bitumen and lube oils, but which sometimes bring in motor spirits and Avgas. (4) Coastal Product Tankers There are two coastal tankers in the fleet operated by Silver Fern Shipping (SFSL), the Kakariki, a double-hulled product tanker and the Taiko, an older double-bottomed (single sided) vessel. They carry up to 38,000 tonnes of oil. While they are used primarily for distribution of finished products from Marsden Point to New Zealand ports, they also are sometimes used for condensate shipments from New Plymouth and product imports. Data on coastal tanker activity and oil movements is presented in Appendix C.

3.2.2

International Cargo and Passenger Vessels

International cargo and passenger vessels visiting New Zealand waters have been divided into two broad categories for the current study. These are: •

Container vessels, and;

•

Other shipping, including: –

Bulk carriers - logs, fertiliser, wood chips, ore etc.

–

Refrigerated carriers (reefers) - kiwifruit, apples, meat, dairy products, fish

–

Passenger/cruise vessels/large pleasure vessels

–

Chemical and product (non-petroleum) tankers - methanol, LPG, tallow etc.

–

International roll-on/roll-off (RoRo) vessels – Trans-Tasman, car carriers

–

General cargo vessels

Container vessels make up a significant proportion of foreign vessel visits each year and the number of containers moving across New Zealand wharves has increased steadily year on year. The majority of these vessels operate to fixed schedules. In most cases they travel faster than conventional cargo shipping (21 – 23 kts vs. 14 – 15kts) and carry larger quantities of fuel oils. Since 1997, new vessels have been introduced to New Zealand routes with the largest ones carrying up to 4100 TEU8 and having bunker

8

TEU – twenty-foot equivalent unit; a measure of container carrying capacity

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

3-3

Risk Creator Groups

SECTION 3

capacities of up to 6,500 tonnes. Hence a separate container vessel category has been added to the risk model this time. Figures for international vessel visits to New Zealand ports were drawn from data supplied by port companies. A summary of estimated movements for international vessels at all ports is presented in Appendix C. Information on individual vessels (for the purposes of developing vessel profiles) was provided by a number of shipping operators. Specific information was also obtained on cruise ship activity in New Zealand. This is discussed in more detail in Appendix C.

3.2.3

Coastal Cargo and Passenger Vessels

This category covers New Zealand-registered vessels and foreign-registered vessels operating continuously in New Zealand waters, over 500 GRT. It includes: •

interisland rail ferries

•

other coastal RoRo shipping, operated by Pacifica and Strait Shipping/Bluebridge as well as the Lynx fast ferry

•

LPG carriers

•

cement carriers operated by Golden Bay Cement and Holcim

•

offshore support vessels, barges, dredges and naval vessels

There are 14 vessels registered with MSA under the ISMNZ category, but this number includes the two coastal tankers. Estimated movements for coastal cargo and passenger vessel movements in 2002/2003 are presented in Appendix C.

3.2.4

Fishing Vessels

There are two main categories of vessel operating in New Zealand waters: •

Large deep-sea freezer/trawlers and freezer factory vessels. Many are foreign-owned vessels and engaged on charter or in joint ventures.

•

Smaller in-shore vessels.

The larger fishing vessel fleets generally operate out of the three main ports of Nelson, Lyttelton and Timaru although smaller numbers of larger vessels and small vessel fleets discharge catches and refuel in Auckland, Bluff, Tauranga, Westport and Dunedin.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

3-4

Risk Creator Groups

SECTION 3

The main fishing grounds within the New Zealand Exclusive economic Zone (EEZ) are off the West, South and East coasts of the South Island, with a number stretching up the North Islands East and West coasts. Fishing in most of these areas tends to be seasonal. Cook Strait and Chatham Rise are fished year round. The main species of fish include hoki, orange roughy, oreo dory, squid, southern blue whiting, and jack mackerel. However a number of other species are also targeted. The level of fishing vessel activity around New Zealand coastlines was estimated using catch effort data for the year 2002/2003 supplied by MFish. This data summarises fishing days by vessel type for each reporting area, extending out to the limit of the EEZ, and so reflects the general pattern of activity described above. However, for the purposes of the risk modelling, only those reporting areas immediately adjacent to the mainland coast (typically 150 – 250 km out from the shore) were included. For the purpose of the risk model, fishing vessels have been split into two groups: •

24m LOA (length overall) and greater

•

less than 24m LOA

It was assumed that the larger vessels are bunkered on gas oil or LFO and the smaller vessels on gas oil only. More detail on fishing vessel activity is given in Appendix C.

3.2.5

Small Craft

This category includes commercial vessels generally below 24m or 500GRT and includes harbour ferries (such as operated in Auckland and the Hauraki Gulf), small passenger vessels, tugs and dredges. These vessels typically operate within harbours and close inshore in coastal areas and are bunkered on gas oil. Small craft characteristics and activity were derived from a number of sources including the MSA’s SSM and SOP9 databases, port and shipping operators’ data and schedules for regular ferry and tourist operations. The MSA’s databases report numbers of vessels registered by area, rather than activity (movements), so data from the MSA’s regular operating hours surveys of commercial shipping were used to estimate the number of port and coastal movements. A more extensive discussion is included in Appendix C.

9

SOP – Safe Operational Plan – a safety management system introduced by MSA in 2000 to provide a practical and affordable set of safety requirements for specialised small commercial craft and boats not previously covered by the Safe Ship Management System (SSM).

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

3-5

Risk Creator Groups

SECTION 3

3.3

Non-Vessel Activity

3.3.1

Offshore Oil and Gas Production

At present, offshore oil and gas production in New Zealand is limited to the Maui Field, off the Taranaki coast, where two fixed platforms, Maui A and B, produce natural gas, condensate and F sand crude. The gas and condensate is pumped to Oaonui by pipeline from the field, with the F Sand crude flowing by pipeline to the floating product storage and offtake facility (FPSO), which processes, stores and offloads the crude onto tankers. Total production from July 2002 to June 2003 totalled around 108,000 tonnes with eight shipments loaded out. A number of offshore developments are currently in the planning stages. The partners in the Pohokura gas/condensate field have recently confirmed their intention to proceed with development, based on a fixed platform and pipelines to production facilities in North Taranaki. Preliminary design work is proceeding for the Kupe oil field off South Taranaki, with a fixed platform and pipelines to shore understood to be the most likely option. Planning work is continuing for the development of the Maari oilfield, approximately 80 km off the Taranaki coast, which if it proceeds, is likely to be based on an FPSO. At this stage, neither of the latter two developments is confirmed.

3.3.2

Exploration Activity

At as June 2004, there is one exploration drilling rig, the semi-submersible Ocean Bounty, operating in New Zealand waters. The rig has been engaged in drilling activity in offshore Taranaki since January 2004 and has drilled three wells so far with a fourth and possibly a fifth planned. Last year, the same rig drilled three exploration wells during a five-month programme in Taranaki. There is likely to be an increasing level of exploration activity over the next few years in a number of offshore licence areas, in the light of recent changes to the royalties regime for gas and dwindling Maui gas reserves. Exploration programmes are currently being planned for east coast basins off the Wairarapa and Canterbury coasts for later this year. Generally there is no significant storage of oil associated with exploration rigs other than fuel. A typical semi-submersible rig may carry up to 1,000 tonnes of gas oil. Potential sources of spills include well fluids from blowouts, rainout of liquids from flaring, drilling muds and bunkering/supply operations. There will also be some level of support vessel activity and port bunkering associated with any drilling programme. Rig support and anchor handling duties for semi-submersibles are typically provided by two anchor handling/tow/supply (AHTS) vessels with one on station covering emergency standby and support duties and a second ferrying supplies and equipment from shore. For a one-month drilling programme, there would typically be around ten supply visits, though not all of these would be transferring fuel.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

3-6

Risk Creator Groups

3.3.3

SECTION 3

Bulk Oil Transfer Operations

Oil is loaded and discharged to coastal and foreign tankers at most major ports in New Zealand. At New Plymouth and Marsden Point, tankers both discharge and load routinely, whereas at the other ports, usually only finished products are discharged and backloading is uncommon. Since 1997, tanker deliveries of bulk oil to Whangarei and Gisborne have ceased. A summary of all bulk oil transfer operations (excluding bunkering) and the quantities of finished products discharged from both coastal and foreign tankers at New Zealand ports for 2002/2003 is presented in Appendix B.

3.3.4

Bulk Storage Terminals

Finished products are stored in bulk at terminals at all major ports. In New Plymouth, there is also bulk storage of Taranaki crude oils and condensates. At Marsden Point there is extensive storage of crude feedstocks, blendstocks and finished products. There is also significant storage of automotive fuels and jet fuel at the Wiri terminal close to the Manukau harbour. A summary of the current bulk storage capacity for each product at each port is presented in Appendix B. Terminals are normally located close to port facilities. In general, spills of oil into the marine environment from bulk storage facilities will only occur as a result of failure of secondary containment, as all bulk tanks are bunded. Normal practice is that all stormwater collected in bunded tank compounds must be held and verified as clean before being released through an interceptor to the stormwater systems, which in turn usually discharge to rivers or harbours directly.

3.3.5

Product Pipelines and Bunker Lines

Product pipelines and bunker lines are used for transfer of products between the storage terminals and the ports. In most cases at least some part of these lines run along wharves (above or slung underneath) or adjacent to coastlines, so there is the potential for a spill into the marine environment. Most pipelines range in diameter from 150 mm to 300 mm, except for Seaview (350 mm) and lines at New Plymouth and Marsden Point (up to 450 mm and 650 mm respectively). Most product pipelines (those used for discharging cargoes of white oils – that is, petrol, jet fuel and gas oil) are now “rested” on product (usually gas oil) when not in use. There has been a progressive move away from resting wharflines on water over the last five years. While this would appear to increase the chance of an oil spill because there is product in the line all the time, rather than just during cargo discharge, with no water present there is less corrosion (hence a lower risk of failure) and fewer process issues associated with handling and treating oil-contaminated water. Bunker lines used for gas oil and LFO are normally rested on product whereas HFO lines are either blown clear with air after use, or if left full, must be kept heated. The same generally applies to bitumen pipelines. A summary of wharf pipelines is given in Appendix B.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

3-7

Risk Creator Groups

3.3.6

SECTION 3

Bunkering Operations

Vessels are bunkered at most ports around New Zealand. Gas oil is available at all ports but fuel oils (LFO and HFO) are only available at the major ports. HBFO (heavy bunker fuel oil) is only available at Marsden Point where it is used for bunkering large crude tankers. Fuels available, quantities bunkered and the estimated number of bunkering operations for each port for 2002/2003 are presented in Appendix B. Fuel oils are generally bunkered using dedicated pipelines from the oil storage terminals. In the case of the Waitemata harbour in Auckland, fuel oils are also bunkered to vessels at berth by barge. At most ports, gas oil is bunkered by pipeline, barge and road tanker. Gas oil, and in some cases motor spirit, is also available from marine stops (wharfside dispensers) at numerous harbours and marinas all over the country, supplied by above ground or buried tanks. Road tankers are also used at other minor port locations around the country as required. For international vessels, bunkering in New Zealand can be expensive and so operators tend to avoid it if possible or only top up as needed. Bunkering operations are therefore largely confined to New Zealand coastal operators and fishing fleets.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

3-8

Changes Since 1997 and Future Trends

4

SECTION 4

Changes Since 1997 and Future Trends

This section summarises the major changes in oil-related marine activity since the previous risk assessment and looks at current developments and further changes likely over the next five years. In some cases, where data is more accurate and complete this time around, differences may be more apparent than real, and this is briefly discussed.

4.1

Oil Industry Changes Since 1997

4.1.1

Refinery Receipts and Exports

Total intake volumes at the Marsden Point refinery have increased over the last six years, although local condensate volumes processed have fallen and have been replaced by light crudes. The main changes since 1997 have been: •

Total volumes have increased by about 9%.

•

Imported crude and residue volumes have increased by about 20%.

•

Local condensate volumes (Maui and Kapuni) are lower by nearly 70%.

•

Residues are tending to be rather lighter than previously.

•

The split of the total intake by oil type (on average) has changed slightly.

There continue to be small numbers of product exports from Marsden Point; most of these are fuel oil. This is not a new occurrence, but these volumes have been captured more completely this time.

4.1.2

FPSO / New Plymouth Liftings

Total crude and condensate liftings from Taranaki have fallen significantly in line with the changing production patterns of the various fields. Since 1997, extraction and exports to Asia of naphtha (a Group I oil) from Maui condensate have begun, but the liftings in Groups II to IV have fallen considerably. Main changes have been: •

Aggregate liftings for the 2002/2003 were much lower – down about 70% on 1997. Specifically:

•

o FPSO liftings were down by around 90% o Maui condensate volumes were down by around 67% o McKee (Fletcher Blend) volumes were down by around 60% o Kapuni condensate volume was down by around 40% No local crude was shipped to NZRC in 2002-03.

•

Naphtha exports in 2002/2003 totalled 300,000 tonnes (10 voyages).

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

4-1

Changes Since 1997 and Future Trends

SECTION 4

It is noted that there was a significant reduction in liftings of oil from the FPSO between 1997 (26) and 1998 (10). For 2002/2003 there were eight. However, for the previous risk assessment, the 1998 (lower figure) was used in the modelling as the previous year’s activity did not reflect the situation at the time of the study or the likely pattern over the subsequent period. Hence the risk profile presented by the FPSO has not changed significantly since the last study. As noted earlier, all except one of the tankers lifting from the FPSO also uplifted crude and condensate from New Plymouth on the same voyage. There was only one voyage which did not enter a New Zealand port, a significant change from 1997.

4.1.3

Product Imports

Total product imports have continued to increase since 1997, driven by increasing demand that outstrips the refinery’s capacity increases from modifications and process improvements. Small parcels of bitumen are now being imported regularly (there were four in 2002-03). The previous study did not record any such imports, although they did occur occasionally. Overall, the main changes are: •

Total product import volumes have increased by 35%

•

AGO volume was up significantly by 33%

•

Gasoline volume was also up markedly by 21%

•

Jet volume was up slightly by 4%

•

Bitumen imports totalled 38,000 tonnes in 2003/03 (none were recorded previously)

4.1.4

Product Tanker Movements

There have been some significant changes in the movements of product tankers (both local and import vessels) on the New Zealand coast since 1997. These reflect the fact that the volumes available for vessel loadout from Marsden Point are declining and that there are more foreign tankers importing products directly to New Zealand ports. In addition, better vessel management and voyage planning have resulted in operational efficiency gains. Major changes that have occurred are: •

Total “coastal loadings” volumes from NZRC have dropped 10% (partly resulting from increased throughput on the Refinery-Auckland pipeline (RAP))

•

Whangarei is now supplied by road, as is Gisborne.

•

BP has decommissioned its storage terminal at Freeman’s Bay in Auckland.

•

Gull Petroleum established a new fuel storage terminal at Mt. Maunganui in 1998 and now imports fuel directly to that site. Challenge Petroleum (now Caltex) has also commissioned new storage facilities in Timaru and New Plymouth since 1997. Additional tankage has been added in other ports such as Nelson and Lyttelton, though some of this replaces existing storage, and so is not all new capacity.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

4-2

Changes Since 1997 and Future Trends

SECTION 4

•

Gas oil storage for reserve electricity generation has been re-established at Napier and storage for fuel oil has been recommissioned at the New Plymouth Power Station.

•

About 15% of the total volume was delivered by foreign vessels in 2002/2003 (though the Kakariki was out of service for some of the period)

•

There are now only two local product tankers (three previously) – the Kakariki and the Taiko, both of which carry white and black oils (except only the Taiko carries bitumen). Previously, the Taiko carried predominantly black oils and the other two vessels white products only.

•

Foreign product tankers now lift regularly from NZRC for coastal distribution with a total of 16 such voyages during 2002/2003. Liftings from Marsden Point range in size from 1,500 tonnes to 35,000 tonnes.

•

Total discharge volumes from product tankers fell by 2% but the number of port visits fell by 15% and the number of discharges fell by 16%.

4.1.5

Patterns of Coastal Voyages

There have been some changes in the pattern and frequency of product tanker movements through the various coastal areas. These changes reflect elements such as: •

The local coastal fleet is now two vessels, which tend to travel up and down the east coast of both islands, with some visits to Wellington and relatively infrequent voyages on the west coast of the North Island.

•

The increase in foreign vessels doing imports or in-charters, with many coming from Australia or Southeast Asia and (if not calling at Marsden Point to discharge and/or load) often transiting Cook Strait at least once on their voyage

The most obvious change in coastal voyage patterns is the increase in the number of passages through Cook Strait (126 tanker movements in 2002/2003 vs 77 in 1997). Some of that increase may be due to the way in which voyages into and out of Wellington have been counted this time as the available data on import tanker movements is more complete.

4.1.6

Tankers

While there are no comparable figures available from the tanker activity data for 1997, there has been a clear trend towards more double hulled vessels. Table 4-1 summarises data on the configuration of tankers visiting in New Zealand for 2002/2003.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

4-3

Changes Since 1997 and Future Trends

SECTION 4

Table 4-1: Hull Types of Tanker Visiting Marsden Point 2002/2003 Tanker hull

Crude and residue tankers

Foreign product tankers

type

(47 visits, average cargo

(36 visits, average cargo 40,800

122,800 tonnes)

tonnes)

% by volume

% by visits

% by volume

% by visits

Double hull

63

60

67

53

Double side

6

9

6

6

Double bottom

4

4

17

28

Single hull

28

28

10

14

The future trends are discussed in Section 4.2.1 below.

4.1.7

Port and Related Activities

There have been a number of changes involving specific ports: •

Whangarei port has closed for tanker discharges and the region is now supplied by road vehicle from Marsden Point

•

LFO is no longer stored at Mt Maunganui

•

Gisborne is no longer an oil port

•

Wellington’s two lubricants blending plants have been shut down, so there are no longer discharges of bulk lubricants at Seaview (see also below).

•

Bitumen is now being imported to and stored at Timaru

More generally, there have been various efficiency gains and other operating improvements at ports: •

The numbers of port visits and discharges have fallen (as noted above)

•

The average sizes of parcels discharged have increased

The numbers of berth shifts (between discharges) at certain ports (such as Nelson and Bluff) have been captured more accurately than in 1997. The installed wharflines and the operation of them have changed relatively little over the period but the recorded data has improved significantly: •

Ten new wharflines have been recorded compared to the previous study but only three of those are genuinely new (the others were not captured previously). Two lines have been taken out of service.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

4-4

Changes Since 1997 and Future Trends

•

SECTION 4

There is now no sea-water in wharflines and many white oil lines now rest on product (usually AGO).

4.2

Future Trends in the Oil Industry

4.2.1

Marsden Point Refinery

The trend from the last six years of more crude carriers carrying on average larger parcels of crude (100,000 tonnes vs 90,000 tonnes) is expected to stabilise with the constraints of draft and jetty size at Marsden Point limiting any significant increase. There are no current plans to relieve either of these constraints. New operating procedures established following the Eastern Honour and Capella Voyager incidents last year will also limit maximum cargo sizes. However an increasing proportion of the tankers will be double-hulled (currently 60%) with the oil companies using the refinery having a policy of now using double-hulled ships if they are available. IMO regulations mean ships carrying residue (common) or crude with an SG greater than 0.9 (unusual for Marsden Point) have to be double-hulled from 2005 although double-bottomed or double-sided ships can be used if they have the appropriate certification (CAS). From the current statistics, it appears the oil companies chartering the tankers for delivery of oil to Marsden Point are moving ahead of the worldwide trends and would be expected to continue to do so. By the time of the next risk assessment (2010) crude deliveries are likely to be almost exclusively in doublehulled ships, although regulations will not rule out deliveries of lighter crude in single-hulled ships under 25 years of age. Total delivery volumes are not expected to change significantly although there will be less residue being delivered by product tankers, replaced by residue deliveries on large crude carriers (because of changing Australian refinery dynamic). Blendstock deliveries are expected to continue to increase with a peak in deliveries in 2004/2005 before dropping to a level similar to that in the 2002/2003 period. Some export, primarily of fuel oil, is expected to continue over the next six years.

4.2.2

Taranaki/New Plymouth

Volumes are forecast to continue to reduce with the depleting of the Maui gas field. Over the next six years a number of new fields may be developed including Pohokura (resulting in export condensate from New Plymouth), Kupe (resulting in export crude/condensate from New Plymouth) and Maari (expected to be an offshore FPSO). STOS has recently signalled the likely decommissioning of the Maui FPSO but no timeframe has been given at this stage. Of these projects, only Pohokura has been confirmed at this time but if they all proceed, volumes should begin to rise again from 2006. However levels are only likely to be similar to the 2002/2003 period rather than the high levels seen in 1997.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

4-5

Changes Since 1997 and Future Trends

SECTION 4

The ships used for export to Australia (most of the crude oil and a large proportion of the condensate) are expected to be double-hulled from 2004 with an Australian requirement (not legislated, we understand) that ships are to be less than 15 years old. In 2003/2004 the double-bottomed tanker Onyx is being replaced by the double hulled Resolve for the regular export run. The makeup of the exports in the future may become more slanted to crude as opposed to condensate at present.

4.2.3

Coastal Deliveries and Product Imports

There are no further coastal port closures (for product deliveries) currently planned although that does not rule out changes in the next six years. Volumes carried by coastal tanker are expected to reduce slowly as a greater proportion of the refinery’s production is carried by pipeline. Imported volumes will increase to balance the increasing demand. Much of the current growth is in diesel. Silver Fern Shipping (the coastal fleet operator) is expecting to continue with a two-ship fleet in the foreseeable future. The trend of the coastal tankers doing some offshore and import work while import tankers move local product from the refinery should continue. Under the latest IMO regulations, SFSL should be able to continue using the single sided/doubled bottom Taiko until it reaches 25 years old in 2009. From that point forward, all ships used on the coast will be fully double-hulled. Over the last six years, two of the four lubricant blending plants in New Zealand have closed down. The Caltex plant is due for closure this year (2004) leaving the BP plant in Auckland as the only blend plant in the country. This will further reduce bulk lubricant discharges from tankers at Auckland’s Wynyard Wharf, replaced with increased use of iso-tanks on container ships. A small number of bitumen imports are expected to continue. From 2005 these will need to be in double hulled ships.

4.2.4

Port Infrastructure

As noted, new storage tanks have been constructed in a number of ports since 1997, including Mt. Maunganui, Timaru and Lyttelton. The increased storage is focused on white products. No significant investment in pipelines is planned although some may happen in the next six years.

4.3

Other Changes Since 1997

The main changes since 1997 are summarised below. Some of these changes will continue to have impacts in the next five to six years. •

Container shipping, more and larger vessels Overall volumes of containers have increased across the country. Larger, deeper draft vessels of

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

4-6

Changes Since 1997 and Future Trends

SECTION 4

4,100 TEU capacity were introduced in 2001, though at present these only operate on the “Eastabout” service from Europe, with a single weekly call at Auckland, Napier and Port Chalmers. •

Growth of regional ports There has been significant growth at regional ports, in particular Mt. Maunganui and Napier, where container volumes have increased significantly. The development of “inland ports” for container transfer to rail has been a factor in this.

•

Changes to coastal fleets The past five years has seen significant changes to the coastal vessel fleet, with several “new” vessels replacing existing ones on Cook Strait services. However, only one fast ferry is now operating (there were two in 1997).

•

New deepwater berths New deepwater berths at Marsden Point and Picton have been opened since 1997. While these have not seen dramatic increases in vessel movements at these ports as yet, these facilities cater for larger, deeper draft vessels (primarily for the log trade at this stage) than those ports have previously been able to handle and will provide significant capacity for future growth.

•

Cruise ships Cruise ship numbers have increased though the pattern is seasonal and the market is strongly influenced by fluctuations and trends in international tourism. However, a number of very large vessels are now coming to New Zealand each season. Cruise ship activity is discussed in more detail in Appendix C.

•

Navigational Aids The last few years have seen the introduction of wave rider buoys and UKDC systems10, increasing tidal windows for ports where port entry has historically been draft constrained – this is the case in most New Zealand ports.

4.4

Recent Developments and Future Trends

Over the next few years, the New Zealand maritime scene is likely to see a continuation of the current trends discussed in Section 4.3 above. These are summarised in Table 4-2. The effects of these trends in terms of the oil spill risk are not easily quantified. The current patterns for coastal transport are significantly influenced by the availability and cost of alternatives, which can change over a relatively short period, so this makes it somewhat difficult to predict what will happen in the future.

10

UKDC – Under keel dynamic clearance systems

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

4-7

Changes Since 1997 and Future Trends

SECTION 4

Table 4-2: Recent Developments and Future Trends Development

Timing

Comments and Likely Impact on Oil Spill Risk

Deepwater berth at Picton (Shakespeare Bay)

Opened April 2000

Deepest export berth in New Zealand – possibly large coal and log carriers topping up.

Deepwater port at Marsden Point

Opened June 2002

This is servicing log and export timber/ timber products currently.

Channel dredging and port reclamation at Auckland

2005

Increasing operating window for deeper draft container shipping – see below.

Future development of the Western Reclamation Area in Auckland

Next 5 years.

The direction of future development is uncertain but may reduce the extent of bulk liquids trade through Wynyard Wharf in the longer term.

Expanded use of UKDC

Next 5 years.

Being adopted at more ports – see above.

Larger containers ships (4,100 TEU)

Introduced 2001

Significantly larger container capacity than previous vessels operating services to New Zealand. The number of services operated by these vessels is likely to increase, but is likely to remain a small proportion of total ship activity.

Port and Harbour Safety Code

2004 (ongoing)

This initiative is aimed at improving port safety and ensuring consistency in operating practice across all New Zealand ports.

Port Knights’ Exclusion Zone

From Dec 2004

Exclusion of all shipping > 45m overall length inside 5nm from shore between Cape Brett and Bream Head, including the Poor Knights Islands.

Cruise ships

Next 5 years.

Likely to see more of the larger vessels, though numbers are uncertain.

Interisland ferries - Clifford Bay development and rail ferry fleet replacement.

6 – 10 years

Decision whether to proceed with Clifford Bay not yet made. Relocation would affect vessel activity in the Marlborough Sounds. Fleet replacement unlikely to significantly change the pattern of movements or vessel profiles.

LNG and other energy developments mooted (e.g. recommissioning of Marsden B Power Station, coal loading facilities/barging operations)

6 – 10 years

These developments are somewhat speculative at this stage, but could have regional impacts.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

4-8

Oill Spills and Accidents

5

SECTION 5

Oill Spills and Accidents

This section contains a summary of NZ and international oil spill and maritime accident data collected for this study. A more extensive discussion including diagrams and histograms is presented in Appendix D.

5.1

New Zealand Oil Spills

Oil spill records for all regions and the NZ Exclusive Economic Zone (EEZ) were obtained from the MSA database of oil spills. This database is routinely updated with oil spill information from each region. At the time of review (June 2004) the MSA database included 1,390 entries. Of these, 778 were reported during the period since the last risk assessment was prepared (July 1998 – April 2004). The data available at the time was incomplete, as prior to 1995, data collection had historically been poor in many regions. While oil spill reporting and recording has improved since the last review, some data cleaning was still required to delete records with missing data that rendered them unusable. Following this exercise, 472 oil spills form the basis of this review (about 62% of the recorded entries). Appendix D contains detailed analysis and figures illustrating this data.

5.1.1

Sizes of Spills

Spills recorded in New Zealand have been generally very small, with more than 30% being classified as 0.001 tonnes (approx. 1 litre) or smaller. This size of spill is generally indicative of a sheen on the water rather than an accurate estimate of spill volume. During the period of this review 89% of oil spills were 0.1 tonne (approx. 100 litres) or smaller. Spills greater than 1 tonne have been rare with nine on the record up to June 1998. However, between July 1998 and March 2004 there were 20 oil spills 1 tonne or greater in size (see Figure D-1-3). While most of these were less than 4 tonnes in size, there was also one large spill of 300 tonnes of gas oil by the commercial fishing vessel Dong Won 529 off the east coast of Stewart Island in Southland (October 1998). There were also three significant spills of 44 tonnes (Wellington, July 2003), 40 tonnes (Gisborne, February 2002), and 20 tonnes (Taranaki, March 2003) respectively.

5.1.2

Where Spills Occur

There are more spills recorded in ports and harbours than on open coastlines, which is to be expected as there are more vessel movements and oil transfers and there is better surveillance. Canterbury (includes Lyttleton and Timaru) had the highest number of recorded spills during the period of the review (125). Northland had 107, Auckland 81 and Nelson 38 respectively. From a simple analysis of the spill records, it is estimated that around 10% of oil spills occur in coastal waters. However, it is likely that the reporting rate (for the small spills at least) is also much lower in coastal areas, with small bilge discharges remaining unseen and/or unreported. Additionally, there were seven oil spills recorded outside the 12 nautical mile limit.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

5-1

Oill Spills and Accidents

5.1.3

SECTION 5

Oil Spill Causes

In a significant number of cases the cause was unknown (221 out of 472). This is most probably because the spills were too small or slow to be detected so the source could not be traced. Bunkering accounted for the largest number (102) and bilge discharges the second largest number (35). Sinking is also significant. This follows an historic pattern. Land-based discharges, grounding, vandalism, collision and capsize each account for relatively small numbers of spills.

5.1.4

Sources of Oil Spills

The source of over 35% of the 472 oil spills included in the review was unknown. Of the remainder, 25% were from fishing vessels (116). Non-vessels sources accounted for 32 spills, there were 20 from general cargo vessels and 11 from recreational vessels. Naval, tugs, barges and passenger vessels accounted for 36 spills. There were relatively few spills from oil tankers, chemical tankers, exploration activity and bulk carriers (14 in total).

5.2

International Oil Spill Data

The main source of international oil spill data collected for the purposes of the risk assessment was International Tanker Owners Pollution Federation (ITOPF). Information from the Oil Spill Intelligence Report (OSIR) was also referred to. ITOPF and OSIR are the two main organisations collecting international oil spill data. The study also includes Australia Marine Safety Authority (AMSA) and United States Coastguard (USCG) summaries of oil spill records. This data has been used to ensure that the spill scenarios developed are realistic. A more extensive discussion including tables and histograms is presented in Appendix D.

5.2.1

Size and Frequency of Oil Spills Globally

In the previous report there was a general downward trend identified from the 1970’s to the late 1990’s in terms of spill sizes and numbers of events. This trend has continued from the late 1990’s to 2003. The vast majority of spills are small (less than 7 tonnes) with data on numbers and amounts incomplete. The number of large spills (over 700 tonnes) follows the trend with ITOPF reporting an average of 24.2 spills per year from 1970 – 1979, 8.9 spills per year on average from 1980-1989, and 7.3 spills per year on average from 1990 – 1999. An even more marked decrease has occurred in recent years with an average of 3.5 spills per year from 2000 – 2003. It is relevant to note that only 5% of the volume of oil that spills into the ocean comes from major oil spills.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

5-2

Oill Spills and Accidents

5.2.2

SECTION 5

Causes of Spills

ITOPF reports that most incidents are the result of a combination of actions and circumstances, all of which contribute in varying degrees to the final outcome. An analysis was undertaken which explored the incidence of spills of different sizes in terms of the primary event or operation in progress at the time of the spill. The following conclusions were drawn from this analysis: •

most spills from tankers result from routine operations such as loading, discharging and bunkering which normally occur in ports or at oil terminals;

•

accidents involving collisions and groundings generally give rise to much larger spills, with almost a fifth involving quantities in excess of 700 tonnes.

5.2.3

Australia

The Australian Marine Safety Authority (AMSA) recorded 300 oil discharge sightings and oil spills reported during 2002-2003. This was a similar figure to the years preceding with 335 spills recorded over the 1994-1995 period and 351 over the 2000 – 2001 period. At the time of the review, the complete AMSA database of Australian oil spills had 6,000 records. The most common vessel types involved in oil spills from 1994 – 2003 were either unidentified vessels, fishing vessels or barges, ferries or other similar vessels. Australian incident trends are relatively comparable to New Zealand trends as most incidents occur in coastal areas or harbours, with incidents on inland water bodies less frequent.

5.2.4

United States

US data are not as readily comparable to New Zealand data as incidents occur in inland waters with far greater frequency. For example, the USCG reported that from 1991 through 2001, 61% of all spills occurred in lakes, rivers, canals, harbours, bays and sounds. The majority follow the global trend of mostly small spills with the incidence of large spills decreasing from the 1970’s through 2001. In total the USCG has recorded 239,033 spills occurring from 1973 – 2001, with 8,354 spills in 2000 and 7,557 spills in 2001.

5.3

Accident/Incident Data

5.3.1

New Zealand Data

A review was made of maritime accident records from MSA files over the period since 1998, to identify incidents that either resulted in oil spills directly or had the potential to do so. The database was searched for the following accident types: •

groundings

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

5-3

Oill Spills and Accidents

•

collisions

•

fires and explosions

•

capsizes and founderings

SECTION 5

Only a small number of incidents were positively identified as having resulted in oil spills as the primary focus of the data collection is maritime safety. These include: •

The fishing vessel Dong Won 529 sinking in Stewart Island in October 1998, spilling around 300 tonnes of gas oil.

•

The log carrier Jody F Millenium grounding in Gisborne in February 2002, spilling around 25 – 30 tonnes of HFO.

•

The rail ferry Aratere colliding with the fishing vessel San Domenico in Wellington in July 2003, spilling around 50 tonnes of gas oil (from the fishing vessel only).

Other serious incidents with the potential for large oil spills were: •

The tankers Capella Voyager and Eastern Honour which both touched the bottom when entering Marsden Point in April and July 2003 respectively. Both were carrying around 110,000 tonnes of crude oil.

•

The bulk carrier Tai Ping which ran aground at Bluff in October 2002.

•

The wood chip carrier Prince of Tokyo which ran aground at Taiaroa Head in April 1999.

•

The RoRo vessel Kent which collided with a barge, puncturing the hull, in Wellington in July 2002.

•

The bulk ore carrier Taharoa Express which almost ran aground off Taharoa in February 2004.

Other incident types with the potential for serious damage resulting in spills include structural damage, machinery failure (involving loss of power or control), fire, flooding and close quarters. A review for the period August 1998 to April 2004 gives the following figures: Vessels > 5000 GRT

169 incidents in total (including the above)

Vessels 500 – 5000 GRT

93 incidents.

The New Zealand data was not used for prediction of future events as the data set is still relatively small, but it was used to confirm possible spill scenarios. A more detailed summary of incidents is included in Appendix D.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

5-4

Oill Spills and Accidents

5.3.2

SECTION 5

International Data

International accident data from INTERTANKO and ITOPF (for tankers) and other overseas studies was also reviewed. A summary and evaluation of this data is included in Appendix D.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

5-5

Oil Spill Consequence Assessment

6

Oil Spill Consequence Assessment

6.1

Consequence Assessment

SECTION 6

A framework for oil spill consequences was developed to establish a methodology for assessing the consequences of oil spill that reaches the New Zealand coast. The framework considers the range of environmental and human factors impacted by an oil spill, and provides a mechanism to semi-quantify their contribution to the regional and national risk profiles. The framework was tested using regional case studies, and existing national data. The development of the consequences assessment framework is detailed in Appendix E.

6.1.1

Approach

The approach adopted was to develop a framework for consequence assessment using existing information, case studies and workshops. Rather than attempting to redo previous work we build on the coastal sensitivity assessment in the 1998 risk assessment study that was further developed by the MSA in their Review of the Voluntary Vessel Routeing Code for Shipping in New Zealand Coastal Waters. This work was based on the concept of coastal Marine Environmental High Risk Areas (MEHRAs) - an approach used by the UK authorities to protect the most sensitive areas of the coastline and surrounding waters without placing blanket restrictions on vessel movements. Specifically the approach adopted in this study involved: •

A review of national and international literature.

•

Review and modification of the consequence framework during workshops with Central Agencies, and regions (i.e. Bay of Plenty and Marlborough District).

•

Re-analysis of data from the Vessel Routeing Code Review has been used to generate the Oil Spill Consequences Map (Figure 6-1). While the approach copes with incomplete information across New Zealand and does not include weighting characteristics. A full complement of information would provide better resolution for each coastal cell.

6.1.2

Methodology

New Zealand’s coastline was divided into cells of 20 km2, which were then ranked in terms of their likely sensitivity to oil impacts on a 5-point scale. Each coastal cell is ranked in terms of its environmental sensitivity based on the following criteria: •

Shoreline character

•

Plants and animals

•

Protected sites

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

6-1

Oil Spill Consequence Assessment

•

Economic value

•

Cultural value

•

Social, Amenity & Recreation values

SECTION 6

The intention is for each cell to be assessed relatively quickly and easily for each element (refer Appendix E for the consequence matrix used to score each resource category). Using the semi-quantitative information generated during the assessment, an index for the sensitivity to oil spills for both environmental resources (Shoreline Character, Plants and Animals, Protected Sites) and human resources (Economic, Cultural, Social Amenity and Recreation) is calculated for each cell. Finally, this summary information for cell sensitivity is both visually represented by different coloured cells (refer Table 6-1), and spatially located in the appropriate coastal position on a map of the New Zealand coastline (refer Figure 6-1) to provide a coastal map of the sensitivity of each resource category (i.e. environmental and human). Table 6-1: Example of a Consequence Assessment Summary at Cell Level Resource Category

Consequence Level Description Zero or

Low

Moderate

Unknown

(score 1)

(score 5)

(score 20)

Negligible

Human

Environment

(score 0)

or High

Shoreline Character

6

Plants & Animals

6

Protected Sites

Extreme (score 50)

6

Economic

6

Cultural

6 6

Social, Amenity & Recreation

In the example above, the cell rates low for environmental resources (i.e. 0 + 1 + 1 = 2), and high for human resources (i.e. 5 + 5 + 20 = 30). This cell would be represented on a map of New Zealand as a square where the bottom half is green (for low environmental resources) and the top half orange (for high human resources). Each cell represents a 20km2 section of coastline. It must be emphasised that the sensitivity rankings used are exclusively related to the sensitivity of the resource to an oil spill – and do not refer to more general environmental sensitivity or conservation value. It is intended that oil spill issues such as oil spill intensity and size, and seasonal effects, will be dealt with on a case-by-case basis. The proposed methodology applies a precautionary approach by giving a high score to criteria where information is uncertain or absent.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

6-2

Oil Spill Consequence Assessment

SECTION 6

Figure 6-1: New Zealand Oil Spill Consequences Map The top triangle of the cell represents the sensitivity of environmental resources, while the lower triangle represents the sensitivity of human resources.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

6-3

Oil Spill Consequence Assessment

6.1.3

SECTION 6

Conclusions

Workshop review and testing of the methodology has demonstrated that it is: •

Relevant at a regional level

•

Reasonably consistent and integrated with existing practises

•

Easy to use and useful

•

Practical to implement, visualise and update

•

Adequately comprehensive and appropriately labour intensive

•

Relatively robust to information gaps and uncertainties at the degree of detail we need

•

Requires minor improvements in the area of cultural information

•

Will be greatly improved with better quality data in coming years

Over the coming six years before the next New Zealand Marine Oil Spill Risk Assessment is undertaken, this work will be used as a guide to information collection. It is anticipated that the information gaps will be at least partially addressed, enabling a more comprehensive assessment of consequences. In the meantime, the consequences framework will provide the MSA and regional authorities with a nationally consistent way of representing their information – augmenting their current Tier 2 Oil Spill Contingency Plans, and the national Tier 3 Oil Spill Contingency Plan.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

6-4

Oil spill Scenarios

7

Oil spill Scenarios

7.1

Oil Spill Scenarios

SECTION 7

The approach to modelling the oil spill risk was based on identifying and quantifying a number of representative spill scenarios for each port and coast. The types of incidents that could reasonably occur were determined by a review of the national and international accident and spill records in relation to the type of activity occurring in each port or coastal area. The approach used this time is the same as that in the 1998 risk assessment.

7.2

Coastal Spill Scenarios

For coastal areas, the main events that can result in oil spills are vessel incidents. These were grouped into three types: •

Collisions

•

Groundings

•

Fire, explosion or structural failure resulting in damage or loss of the vessel

These groupings were selected on the basis of the factors which contribute to them and also on the likely outcome. For instance, the rate of collisions and to a lesser extent groundings, are influenced by traffic density. However, oil will almost always come ashore if a grounding results in a spill, whereas this may not occur in the other cases. Ten vessel categories were identified, and for each the average quantity of oil carried and the type of oil (or properties of the average contents) were determined. The number of movements and the location of these was also assessed. The likelihood and impact of the above incidents (in terms of the potential for an oil spill, the size of the spill and the type of oil) were assessed for each vessel type. A summary of which vessels operate in each coastal area is presented in Appendix F along with a detailed description of each type.

7.3

Port Spill Scenarios

In addition to the above vessel incidents, the category of contact (collision with a fixed object such as a wharf) was added for the port scenarios. The same ten vessel categories were used, and the ports where each type operates identified. The other spill scenarios identified for ports were: •

Spills during bunkering

•

Spills during bulk cargo transfer

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

7-1

Oil spill Scenarios

•

Spills from wharf pipelines

•

Spills from bulk storage terminals

SECTION 7

For each port, the volume and types of oils handled and stored were identified. For bunkering and transfer operations, a maximum spill size was determined based on the response time in the event of a leak, rupture or tank overfill. These scenarios are discussed in more detail in Appendix F.

7.3.1

Maui Field and FPSO

For the purposes of the risk assessment, the FPSO and Maui platforms A and B (MPA and MPB) were treated as a port rather than an activity on a coastline though vessel movements to and from the offshore installations were included in the relevant coastal areas. The types of activity are in many ways similar to those in ports (bunkering, docking, cargo transfer), though the effects of a spill are more similar to that of a spill in coastal waters. The main events identified as leading to oil spills from the platforms, FPSO or associated pipelines are as follows: •

Tanker collision with FPSO, supply/support vessels or platforms

•

Tanker breakaway during cargo transfer

•

Supply/support vessel collision with FPSO

•

Riser and flowline failure (MPB to FPSO)

•

FPSO mooring failure and subsequent collision with platforms

•

FPSO mooring failure and subsequent grounding

•

Fire and explosion on FPSO

•

Fire and explosion on MPB or MPA

•

FPSO structural failure

•

Spill during bunkering (FPSO and platforms)

•

Well blowouts (F sands and condensate on MPB, condensate on MPA)

•

Riser and pipeline failure (MPB to MPA, MPA to Oaonui).

These events were originally identified and assessed in the assessment of environmental effects (AEE) prepared for the FPSO development, which formed the basis for the risk modelling carried out in 1998. As the nature of the operation has not changed significantly, these scenarios are all still valid, although the level activity may have changed. The same model has been used this time, updated to reflect any

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

7-2

Oil spill Scenarios

SECTION 7

changes in activity since that time. Details of the spill risk modelling for the FPSO, Maui platforms and associated pipelines are given in Appendix F.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

7-3

Spill Likelihood Model

8

SECTION 8

Spill Likelihood Model

This section gives a brief description of the models used for quantifying the likelihood of an oil spill. The model is essentially the same as that used in 1998, but with some changes where noted. A more detailed description is provided in Appendix F. General comments are also included here on the quality of input data used, the main assumptions made and the sensitivity of the model to variations in this data.

8.1

Generalised Structure

The oil spill risk model comprises two distinct components: •

a model for spills in ports

•

a model for spills in coastal waters

The general structure of each model is a set of worksheets containing common inputs and modelling parameters. These feed data into individual worksheets for each port or coastal area in which the risk calculations for that area are performed. Results for individual areas are then compiled into an overall summary sheet for the purposes of analysis. The models are structured so that key inputs and outputs are all easily accessible, with a consistent set of logic being used for the risk calculations for each port or coastal area, ensuring a consistency of approach.

8.2

Common Features Of Both Models

For both port and coastal models, ten categories of vessels were defined, according to size, service and oil types carried. All vessel spill calculations are based on these groupings which are as follows: 1.

Large crude carriers

2.

Indigenous crude/condensate tankers

3.

Foreign product tankers

4.

Coastal product tankers

5.

International container vessels

6.

Other international cargo and passenger vessels

7.

Coastal cargo and passenger vessels

8.

Large fishing vessels (24m LOA or greater)

9.

Small fishing vessels (under 24m LOA)

10. Small craft

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

8-1

Spill Likelihood Model

SECTION 8

For each of these vessel types the average quantity of oil carried was estimated. Probability distributions were applied to determine the proportion of oil carried that is lost, given that an incident has occurred which results in a spill. The models take an event analysis approach to estimating the likelihood of a spill, starting with the activity that creates the risk of a spill, as follows:

Vessel movement or transfer of oil (cargo transfer or bunkering) È Incident (grounding, collision, etc.) È Oil spill occurs È Type of oil spilled È Quantity spilled (proportion of average oil carried for vessels or maximum spill size for transfer spills)

8.3

Risk Measures

Two quantitative measures of oil spill risk have been used in the development of the models. These measures complement each other and should be taken together when assessing the spill risk attributable to a given area or risk creator.

8.3.1

Return Period

The first is the return period for a spill exceeding a given size. For example, for a given scenario, the return period for a spill of 1000 tonnes or greater might be 800 years. Therefore an oil spill of 1000 tonnes or greater can be expected to occur on average once every 800 years in that location. The return period for a spill of between 1200 tonnes and 800 tonnes will be the difference between the return periods for exceedance of each spill size in that location. The model estimates the frequency of spills exceeding a given size and this frequency is then converted to a return period. The spill size - return period data is then plotted and spill volumes for given probabilities of occurrence (expressed as return periods) can then be estimated from the curve.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

8-2

Spill Likelihood Model

SECTION 8

This information can be used further to derive a probability that a spill within a given size range will occur within a given period e.g. the probability of a spill exceeding 500 tonnes (or perhaps, between 500 tonnes and 1000 tonnes) within the next 50 years. Some caution must be exercised when interpreting these return period – spill size figures (as with any absolute risk values), but they can also be used (more reliably) to compare the relative occurrence of a spill of a given size between individual ports and coastal areas.

8.3.2

Spill Rate

The second measure of risk is the risk-weighted average spill rate, expressed as tonnes of oil spilled per annum. For a given area, risk creator or oil type etc., this figure represents the quantity of oil that can be expected to be spilled over a long period, averaged over all spill sizes, and expressed as an annual spill rate. It is important to note that this is not a real number, and does not predict the quantity of oil likely to be spilled over some short period in the future, but rather it is a measure of the contribution to the overall spill risk from each particular area or source. This figure does not reflect spill size and large spills will tend to distort it. However, when taken in conjunction with the return periods for different spill sizes it provides a useful picture of the spill likelihood.

8.3.3

Coastal Exposure

An additional measure of coastal spill risk included this time is the coastal exposure. This is a measure of the time that vessels spend in each coastal area together with the quantities of oil being carried. This is expressed as tonne-years and in effect is a comparative measure of the potential for an oil spill rather than the actual oil spilled. For each area and vessel type, the exposure is the product of the number of movements, the average transit time and the average quantity of oil carried. Hence the exposure can be determined for each coastal area, oil type or vessel type. As expected, the exposure correlates quite closely with the average spill rate for each coastal area.

8.4

Coastal Spill Model Inputs

The coastline of New Zealand was split into a number of discrete areas for the purposes of the coastal model. These areas were selected on the basis of shipping traffic patterns and locations of ports as shown in Figure 8-1. The coastal areas are not intended to correspond to Regional Council boundaries and in making any comparisons between the areas it must be recognised that they represent widely varying lengths of coastline. In the 1998 risk assessment, 17 areas were used in the model. This time, three of the larger areas have been split to give 21 areas overall. Table 8-1 shows the types of vessels operating in each coastal area.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

8-3

Spill Likelihood Model

SECTION 8

The main inputs to the coastal spill model are: 1.

2.

8.5

For each coastal area: •

Number of vessel movements per year of each type

•

Average distances from the coastline

•

Average transit times through coastal section

•

Average wind strength and direction

•

Coastal hazard factor (same as used in 1998)

For each vessel type: •

Historical incident rates

•

Oil groups and average quantity carried by each vessel type

•

Probability of a spill occurring as the result of a serious incident

•

Spill size probability distribution as a proportion of the oil carried

Coastal Spill Model Outputs

For each coastal area and vessel group, the model estimates: •

The exposure in tonne-years (spill potential)

•

The average frequency of a serious incident (grounding, collision or fire/explosion or structural failure)

•

The average spill frequency (number of spills per year)

•

The risk-weighted average rate of oil spilled into the sea (tonnes per year)

•

The risk-weighted average rate for oil spilled into the sea reaching the shore (tonnes per year)

•

The frequency or return period with which a spill of a given size will be exceeded.

These figures are aggregated over all vessel types to produce coastal area totals and then combined to give national totals. These totals are split into persistent and non-persistent oils. A detailed description of the model along with sample inputs and outputs is given in Appendix F.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

8-4

Fis hin g

X X X X X X X X X X X X X X

X X X X X X X X X X X X X X

X X X X X X X X X X X X X X

X X X X X

X X X X X X X

X X X X X X

X X X X X X X

X X X X X X X

X X X X X X X

Sm all cra f

t

Fis hin g

X X X X X X X X X X X X X X

Sm all

Co as pa tal c ss en argo ge an r d

X X X X X X X X X X X X X X

La rge

Int er pa natio ss en nal c ge r argo

X X X X X X X X X X X X X X

ne rs

X X X X X X X X

lp rod uc t

Co nta i

hip pin g

&

tan ke rs

tan ke rs gn pro du ct Fo rei

Co as ta

Area

Cr ud tan e/co ke nd en rs sa te

La rge

cru de ca rr

ier s

TABLE 8-1: COASTAL MODEL - COASTAL AREAS AND VESSEL TYPES

EAST COAST

1

East Northland

2

East Auckland

3

East Waikato (Great Barrier - Tauranga)

4

BOP (Tauranga - East Cape)

5a

East Cape - Gisborne

5b

Gisborne - Napier

5c

Napier - Cape Palliser

6

Cook Strait

7

Marlborough Sounds

8

East Marlborough/North Canterbury

9

South Canterbury

10a

Timaru - Dunedin

10b

Dunedin - Bluff

11

Southland West/Fiordland South Coast

X

X

X X X X X X X X X X X X X

X X X X

X X X X X

X X X X X

WEST COAST 12a

West Northland - Manukau

12b

Manukau - New Plymouth

13

Taranaki (Maui - New Plymouth)

14

Wanganui/West Manawatu

15

Tasman Bay

16

Westport/Northwest Nelson

17

West Coast/Fiordland

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\5000 - TECHNICAL\Model\Coastal File: Coastal model4.xls Sheet: Coastal Areas

X X X

20/08/2004

Spill Likelihood Model

8.6

SECTION 8

Port Spill Model Inputs

The port spill model is similar to the coastal model except that it contains one additional category of vessel incident (impact with the wharf) as well as covering cargo transfer, bunkering, pipelines and bulk storage. Table 8-2 shows the ports assessed and the types of vessels operating in each port, as well as the bunker fuels available. The main inputs to the port spill model are: 1.

For each port: •

Number of vessel movements per year of each type

•

Number of oil cargo transfer operations per year (by tanker type)

•

Number of bunker operations per year (by oil type)

•

Cargo transfer rates (by tanker type)

•

Bunkering rates (by oil type)

•

Port hazard factor (same as used in 1998)

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

8-7

TABLE 8-2: PORT MODEL - VESSEL TYPES AND FUELS AVAILABLE FUELS AVAILABLE

X X

X X

X X

X X

X

X

X

X X

X X

X X

X X

Fi s hin g

Fi s hin g

AG O

X

LF O

X

O

X

HF

X X X X X X X X X X X

HB FO

X X X X X X X X X X X

uc t ta pro d

Sm all cra ft

X X X X X X X X X X X

Sm a ll

X X X X X X X X X X X

La rge

Co as pa tal c ss en argo ge an r d

g

Int er pa natio ss en nal c ge arg r o

X X X X

X X X X X X X X X X X

Co as tal

pro d gn Fo rei

&

nk ers

nk ers

X X X

uc t ta

sa te

Co nta in e rs h ip pin

Port

Cr ud tan e/co ke nde rs n

La rge cru d

ec a rr ie r s

VESSEL TYPES

EAST COAST

1

Whangarei/Marsden Pt

2

Auckland

3

Tauranga

4

Gisborne

5

Napier

6

Wellington

7

Picton

8

Lyttelton

9

Timaru

10

Otago

11

Southport

X

X

X X X

X X X

X X

X X

X X

X X X X

X X X X

X X

X

X X X

X X

X X X

X X

X X X X X X X X X X X

WEST COAST 12

Manukau

13

Westgate

14

FPSO

15

Nelson

16

Westport

X X

X X X

X X

AGO available through marine stops or one-off bunkers by road tanker only

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\5000 - TECHNICAL\Model\Port File: Port model4.xls Sheet: Ports

20/08/2004

Spill Likelihood Model

2.

8.7

SECTION 8

For each vessel type: •

Historical incident rates

•

Oil groups and average quantity carried by each vessel type

•

Probability of a spill occurring as the result of a serious incident

•

Spill size probability distribution as a proportion of the oil carried

Port Spill Model Outputs

For each port and vessel group, the model estimates: •

The average spill frequency (number of spills per year)

•

The risk-weighted average spill rate (tonnes per year) – it is assumed that all oil spilled within a port area will come ashore.

•

The frequency or return period with which a spill of a given size will be exceeded.

As for the coastal areas, these figures are aggregated over all vessel types and activities to produce port totals and then combined to give national totals. The totals are also presented as persistent and nonpersistent oils. A detailed description of the port model along with sample inputs and outputs is given in Appendix F.

8.8

Limitations of Spill Likelihood Model

The model has been designed primarily to predict the incidence of large oil spills of the size that have not historically occurred in New Zealand. As such it relies on international data as an input. International spill databases for shipping generally have lower limits of spill size which constitute quite large spills in New Zealand terms. The spill model developed in 1998 used a lower limit of 20 tonnes as the basis for analysis (15 tonnes corresponds to approximately 5000 US gallons). However, as small craft generally carry less than 20 tonnes of oil, and spills during cargo transfer and bunkering are likely to be significantly smaller than this as well, the data has been extrapolated to estimate the return period for smaller spills (down to 5 tonnes for vessels and 1 tonne for transfers). It is noted that the estimates in this size range may not be particularly reliable. Some discrepancies can be expected when attempting to apply a methodology for calculation of spill sizes based on fixed proportions of oil carried to a wide range of vessel types. For example 10% of a large crude carrier’s cargo is around 10,000 tonnes, whereas 10% of a fishing vessel’s fuel may be less than 10 tonnes.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

8-9

Spill Likelihood Model

SECTION 8

Therefore the model is likely to be less accurate in predicting the incidence of spills at the lower end of the scale (5 – 50 tonnes) where the frequency is high and the consequences low than at the higher end (1,000 tonnes +). Tankers, which carry significant quantities of oil, dominate the low frequency highconsequence end, but make up only a minor contribution to the high frequency – low consequence end. For bunkering and transfer spills, a lower limit of 1 tonne has been used, as noted above, based on assumed distributions of spill size as a proportion of the likely maximum spill.

8.9

Sensitivity

The model is structured to facilitate manipulation of input data for sensitivity analysis and the effect of varying different inputs is discussed in Section 11.1. In general terms, the spill risk for coastal areas has been modelled as a function of vessel exposure (expressed as vessel-years for each area) which in turn relates directly to vessel movements. Within ports, the vessel spill risk has been modelled as a function of port movements and the non-vessel risk a function of the number of bunkering and transfer operations. These are therefore the key inputs to the risk model. These higher level inputs will have a much greater influence on the output than the lower level ones. For example, increasing the number of tanker movements in a particular coastal area by a factor of two will increase the average spill rate from tankers on that coast by the same amount, whereas changing the probability that a collision results in a spill by the same factor will have far less influence. Another key input is the average quantity of oil carried by each vessel type. Estimates have been made based on the available data to determine a representative figure for each vessel group. However, some of these groups cover a wide range of vessel sizes eg. international cargo vessels, fishing vessels etc. so the estimates have generally erred on the high side. This is one of the significant areas of uncertainty, as specific data on oil volumes carried (for non-tankers) is sparse and it has been difficult to collect sufficient data to have a high level of confidence in the estimates. However, as a consistent calculation method has been used for each port and coastal area, uncertainties in the common inputs are carried through the model and so should not significantly affect the relative risk between different areas.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

8-10

Results of Spill Likelihood Analysis

9

SECTION 9

Results of Spill Likelihood Analysis

This section gives a brief summary of the results of the modelling of spill likelihood. More detail is included in Appendix F. A discussion of these results is presented in Section 10.

9.1

Coastal Spill Risk

A summary of coastal spill risks is presented in Table 9-1. As noted previously, two different measures of coastal risk have been used. The coastal risk figures below do not include the Maui platforms or the FPSO.

9.1.1

Spill Size vs. Return Period

The first is the return period for a spill exceeding a given size. For example, for Area 1 (East Northland), representing the section of coastline between Cape Reinga and Whangarei, a spill exceeding 50 tonnes can be expected to occur approximately once every 87 years on average and a spill exceeding 5,000 tonnes once every 5,530 years on average. Similarly for Area 16 (Westport and Northwest Nelson), representing the area of coast between Cape Farewell and Westport, a spill exceeding 50 tonnes can be expected approximately once every 170 years. A spill exceeding 5,000 tonnes is very unlikely to occur because the likelihood of vessels operating in that area which carry that quantity of oil is extremely low. Spill Exceedance vs. Return Period curves for coastal spills are presented in Appendix G. The relative return period for a 1,000-tonne spill in each coastal area is shown in Figure 9-2. The estimated spill sizes for return periods of 100 years, 500 years and 1,000 years in each area are given in Table 9-3. These spill sizes are not calculated directly in the model but have been estimated from interpolation of return period plots, so there is an additional level of uncertainty in the figures, over and above that associated with the modelling itself. Hence they must be interpreted with care.

9.1.2

Risk Weighted Average Spill rate

The second measure of risk is the risk-weighted average spill rate, both for total oil spilled into the water and total oil reaching the shore. The risk-weighted spill rate for each coastal area is summarised in Table 9-1 and shown graphically in Figure 9-3. For Area 1 these are 12 tonnes per year and 4 tonnes per year respectively. For Area 16, they are 0.8 and 0.2 tonnes per year. Within each coastal area, the contribution for each vessel type can be determined and these summed over all coastal areas. The spill rate gives a measure of total oil spilled averaged over a long period, and is useful where a risk measure is required that is independent of spill size. It is noted that the individual coastal areas represent significantly different lengths of coastline, for example Area 13 (Taranaki) – less than 100 km, compared with Area 5c (Napier to Cape Palliser) – about

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

9-1

Results of Spill Likelihood Analysis

SECTION 9

300 km, so this must be recognised in any comparisons of spill rate by area. The Taranaki area has a relatively low average spill rate, perhaps unexpectedly. This reflects the following factors: •

the Maui platforms and the FPSO are not included in these figures;

•

the relative decline in production of oil and condensates since 1998 and the resulting drop in tanker traffic; and

•

over half of the tanker visits to New Plymouth were uplifting export cargoes and so in most cases their movements are counted in the adjacent coastal area instead (Area 12b).

9.1.3

Spill Exposure

An additional measure of relative spill potential, or exposure was calculated for each coastal area. This is expressed as tonne-years and represents oil carried “on the water” as opposed to oil spilled “in the water”. The exposure values for each coastal area are given in Table 9-1. Figure 9-4 shows the exposure graphically, ranked highest to lowest. Figure 9-5 shows the relationship between spill rate and exposure for each area.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

9-2

TABLE 9-1: COASTAL SPILL RISK SUMMARY

Return periods in years for spill size exceeding:

Area

Risk-weighted average spill rate in tonnes/year

Exposure tonne-years

50 t

100 t

1000 t

5000 t

10000t

All spills

Persistent Oils

Spills on shore

Oil carried 24415

EAST COAST

1

East Northland

87

333

463

5529

6477

12.2

9.1

4.0

2

East Auckland

41

358

673

14855

19561

6.3

2.6

3.2

8837

3

East Waikato (Great Barrier - Tauranga)

75

276

576

11490

15123

6.2

3.1

2.1

10856

4

BOP (Tauranga - East Cape)

143

414

627

12543

16515

5.1

2.8

1.5

9618

5a

East Cape - Gisborne

265

724

1125

22316

29379

2.8

1.6

0.9

5345

5b

Gisborne - Napier

122

593

1085

22217

29249

3.3

1.6

1.0

5953

5c

Napier - Cape Palliser

98

203

551

10762

14162

6.3

3.3

1.9

11829

42640

55598

2.5

1.3

1.7

3349

1.1

0.4

1.0

1181

6

Cook Strait

136

267

1959

7

Marlborough Sounds

135

348

128659

8

East Marlborough/North Canterbury

65

175

709

10455

13709

6.2

2.6

1.5

11575

9

South Canterbury

104

243

1421

28743

37826

3.2

1.3

1.0

5788

10a

Timaru - Dunedin

151

905

2665

63974

84416

1.7

0.7

0.5

2948

10b

Dunedin - Bluff

102

510

4288

63417

83593

1.8

0.5

0.5

3068

11

Southland West/Fiordland South Coast

171

449

35858

1309324

1697501

0.8

0.1

0.3

1246

12a

West Northland - Manukau

98

714

8459

43272

56545

1.6

0.3

0.3

3033

12b

113

501

6867

22139

28602

2.1

0.5

0.4

4162

13

Manukau - New Plymouth Taranaki (Maui - New Plymouth) 1

360

969

10338

145313

190328

0.7

0.2

0.2

1142

14

Wanganui/West Manawatu

128

277

1017

19517

25533

3.8

1.9

0.9

6780

15

Tasman Bay

172

1827

5713

197551

261151

1.0

0.3

0.4

1461

16

Westport/Northwest Nelson

172

375

277907

0.8

0.1

0.2

1243

17

West Coast/Fiordland

77

201

23964

1.9

0.2

0.3

2831

71.4

34.4

23.9

126659

WEST COAST

Total Notes: 1 Excludes FPSO

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\5000 - TECHNICAL\Model\Summary File: Summary tables.xls Sheet: Coastal Summary

20/08/2004

Figure 9-4: Tonne-year and Spill Rate summary by Coastal Area and Source 30000

14.0

12.0

25000

Large crude carriers Crude/condensat e tankers

Exposure in tonne-years

Coastal product tankers 10.0 20000 8.0

Foreign product tankers Small craft Fishing Small

15000 6.0 10000

5000

0 1

2

3

4

5a

5b

5c

6

7

8

9

10a 10b 11 12a 12b 13

14

15

16

17

Fishing Large

4.0

Coastal cargo and passenger

2.0

International cargo & passenger Container shipping

0.0

Spill Rate (tonnes/year)

Coastal Area S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\5000 - TECHNICAL\Model\Coastal File: Coastal model4.xls Sheet: Figure 9-4

20/08/2004

Results of Spill Likelihood Analysis

9.2

SECTION 9

Port Spill Risk

Similar risk measures have been used for port spill risks. However, unlike the coastal model, it has been assumed that all oil spilled in the water reaches the shore. Port spill risks are summarised in Table 9-2. For Whangarei / Marsden Point (Port 2), the estimated return period for a spill exceeding 50 tonnes is around 66 years and for a 5,000 tonne spill, 945 years. For Lyttelton (Port 8), the estimated return periods are 44 years and 2350 years respectively. As for coastal areas, Spill Exceedance vs. Return Period curves have also been developed and these are presented in Appendix G. The estimated spill sizes for return periods of 100 years, 500 years and 1,000 years in each port are given in Table 9-3. As noted in Section 9.1.1, these spill sizes are estimated by interpolation and so have an inherent margin of error. The risk-weighted average spill rates are 33 tonnes/year and 27 tonnes/year respectively for the above two ports. Again, contributions by vessel type and source can be extracted.

9.3

Total Spill Risk

Port risks and coastal risks can be combined to give an overall national risk. This is presented as a Spill Exceedance vs. Return Period curve in Figure 9-5. The contribution from the FPSO is also included. The 1998 results have also been plotted on this curve for comparison. In addition, historical spill data for New Zealand during the period 1998 – 2003 (as discussed in Section 5.1) is also shown.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

9-8

TABLE 9-2: PORT SPILL RISK SUMMARY (INCLUDING FPSO)

Return periods in years for spill size exceeding: Port

Risk-weighted

Risk-weighted

spill rate

spill rate tonnes/year

EAST COAST

50 t

100 t

1000 t

5000 t

10000t

tonnes/year All Oils 1

Persistent Oils

1

Whangarei/Marsden Pt

66

215

302

945

1524

32.7

20.2

2

Auckland

19

61

96

3719

4947

82.3

18.1

3

Tauranga

56

122

210

4778

6534

15.6

8.1

4

Gisborne

175

2436

5609

1.4

0.3

5

Napier

80

217

389

12407

16860

8.4

4.4

6

Wellington

24

35

269

2325

3105

26.9

11.5

7

Picton

37

56

6091

6.5

4.2

8

Lyttelton

44

102

225

2352

3154

27.4

8.2

9

Timaru

113

524

1155

14117

19382

10.7

1.4

10

Otago

86

412

872

12065

16451

5.7

2.0

11

Southport

94

532

1175

14675

19856

4.6

1.5

0.7

0.2

558

4021

5606

8.4

3.3

WEST COAST 12

Manukau

329

974

13

Westgate

263

302

14

FPSO

63

64

228

10182

14223

15.1

3.2

15

Nelson

57

251

662

13291

18056

7.3

2.8

16

Westport

160

795

1.2

0.1

254.8

89.8

Total Notes: 1. See section 8.3.2 regarding the significance and interpretation of these figures

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\5000 - TECHNICAL\Model\Summary File: Summary tables.xls Sheet: Port Summary

20/08/2004

Table 9-3: Spill Size Exceedance by Port and Coastal Area

Port Output (estimated spill size in tonnes) Port 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

EAST COAST Whangarei/Marsden Pt Auckland Tauranga Gisborne Napier Wellington Picton Lyttelton Timaru Otago Southport WEST COAST Manukau Westgate FPSO/Maui Nelson Westport

Return Period (years) 100 500 1000 56 2513 5475 1121 2310 2728 83 2142 2476 5 57 68 57 1781 2118 290 2213 2977 202 225 254 100 2288 3020 34 97 521 52 194 689 51 96 489 <5 11 390 61 10

63 618 1080 417 77

111 2205 1230 2021 0

Coastal Output (estimated spill size in tonnes) Coastal Area EAST COAST 1 2 3 4 5a 5b 5c 6 7 8 9 10a 10b 11 12a 12b 13 14 15 16 17

East Northland East Auckland East Waikato (Great Barrier - Tauranga) BOP (Tauranga - East Cape) East Cape - Gisborne Gisborne - Napier Napier - Cape Palliser Cook Strait Marlborough Sounds East Marlborough/North Canterbury South Canterbury Timaru - Dunedin Dunedin - Bluff Southland West/Fiordland South Coast WEST COAST West Northland - Manukau Manukau - New Plymouth Taranaki (Maui - New Plymouth) Wanganui/West Manawatu Tasman Bay Westport/Northwest Nelson West Coast/Fiordland

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\5000 - TECHNICAL\Model\Summary File: Table 9-3 check Sheet: Table 9-3 Rev 1

Return Period (years) 100 500 1000 53 1265 2242 59 368 2024 56 501 2067 14 375 2044 <5 76 571 20 90 698 51 256 705 10 238 347 16 200 202 66 412 2022 41 264 424 5 73 167 50 99 235 <5 138 204 50 31 <5 <5 5 <5 60

83 100 61 293 60 200 203

205 219 113 935 77 201 210

14/01/2005

Estimated Return Period in Years for Spill Exceeding Given Size

Figure 9-5: Oil Spill Risk Summary (Combined Port, Coastal & FPSO)

100,000.0

10,000.0

1,000.0 1998 Results 2004 All Oils NZ Actual Spills 1998 - 2004

100.0

10.0

1.0

0.1 1

10

100

1,000

10,000

100,000

Spill size in tonnes

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\5000 - TECHNICAL\Model\Summary File: Summary tables.xls Sheet: Fig. 9-5

20/08/2004

Discussion and Conclusions of Spill Likelihood Modelling 10

SECTION 10

Discussion and Conclusions of Spill Likelihood Modelling

While the primary focus of the risk assessment is to provide a profile of the current spill risk and how it might change in the next six years, in order to inform the review of the national Spill Response Strategy, it is also useful to compare the results with those of the 1998 risk assessment. This section discusses the results of the spill likelihood modelling presented in the previous section, and addresses the following: •

What the results tell us – general overview

•

Comparison with 1998 results

•

Conclusions and recommendations

10.1

General Comments on Results

10.1.1 Benefits of Approach The risk model provides a means of estimating the return periods for oil spills of different sizes for each coastal area and port as well as nationally, based on a consistent set of rules and an input base of data. While there are some deficiencies in the data and some shortcomings may be identified in the model itself, the use of a defined logic and a stated set of assumptions throughout makes the approach transparent and consistent. It is therefore a useful tool for comparing different ports, coastal areas (where appropriate) and sources of oil spills and identifying the contribution that each makes to the overall picture. It is also valuable for determining the relative frequency of occurrence of small spills and large spills.

10.1.2 Spill Return Period As expected, the return period curve (Figure 9-5) displays a high incidence of smaller spills and a very low incidence of large spills. The bottom end of the curves (up to around 50 tonnes) largely reflect cargo transfer operations and small vessel and fishing vessel activity. The top end of the curves (above about 2000 tonnes) reflect tanker operations whereas the intermediate range represents the coastal and international shipping. Bunkering spills will generally never exceed 20 tonnes and 90% will be less than 1 tonne. For port risks, the estimated return periods (Tables 9-2 and 9-3) reflect the type of shipping activity. For 50 and 100-tonne spills, the port with the lowest return period (i.e. the highest expected frequency of occurrence) is Wellington, for a 1000-tonne spill, Auckland, and for 5000 tonnes and 10,000 tonnes, Marsden Point.

10.1.3 Spill Rate The combined spill rate for ports, coastal areas and the FPSO is shown in Column A of Figure 10-1. The total spill rate for ports is around 3 – 3.5 times higher than for the combined coastal areas, reflecting the

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

10-1

Discussion and Conclusions of Spill Likelihood Modelling

SECTION 10

greater risks associated with vessel movements in and out of harbours. The contribution to the overall spill rates from bunkering and cargo transfer are not significant. Overall, tankers contribute just under 30% of the overall spill rate (combined ports, coastal areas and the FPSO), and international cargo and passenger vessels about the same. Fishing vessels contribute around 13% and small craft around 21%. Coastal cargo and passenger vessels, plus the FPSO make up the remaining 9%, roughly equally split. Non-persistent oils (including petrol, condensates and gas oils) contribute around 62% of the total spill rate, but the split between persistent and non-persistent oils varies significantly across the vessel categories.

10.2

Comparison with 1998

Changes in oil industry and general shipping activity since the 1998 risk assessment and the outlook for the future have already been discussed in Section 4. It must be emphasised that the oil spill risk assessment is an ongoing process, with the aim over time of improving the characterisation of the risk so as to better understand it, while at the same time actively working to reduce that risk. In any comparison of the quantitative results for 2004 with the previous work, it is therefore necessary to distinguish between: •

Actual changes in the risk e.g. due to changes in risk creating activity, improved mitigation and other risk factors

•

Apparent changes e.g. more accurate input data and changes in the method of analysis

The distinctions are not necessarily clear cut, so caution must be exercised in drawing conclusions from comparisons with the previous work. The sensitivity analysis in Section 11.1 assesses in more detail the effect on the results of changes to the main inputs to the model. The key changes in the model that may affect the outcome, are as follows: 1.

Updated transit times for the coastal areas based on actual steaming times used for coastal tankers – generally reducing transit times slightly in most areas, and better reflecting reality;

2.

Updated estimates of average oil carried. However, for the non-tanker categories, this is still based on fairly limited information.

3.

Splitting large coastal areas containing intermediate ports into smaller areas - enabling a more accurate count of coastal movements;

4.

Addition of extra vessel categories, in particular container vessels and small fishing – adding sources of spill risk not previously included;

The effects of these changes in the model are summarised in general terms in Table 10-1 and discussed in more detail in subsequent sections.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

10-2

Discussion and Conclusions of Spill Likelihood Modelling

SECTION 10

Table 10-1: Effect of Improved Data and Model Changes Model Input

Change to modelling parameters

Effect on output

Coastal area transit times

Revised to better reflect reality – reduced slightly in most areas. Transit times for container vessels reduced over all areas to reflect greater speed.

Reduces coastal shipping incident rates, reducing spill rate.

Average oil carried

Lower value adopted for small craft, also small fishing vessels. Slightly lower value for international cargo vessels.

Spill rate is a direct function of oil carried

Vessel movements

Generally more accurate counts available for all categories – but more fishing and small vessels now included.

Incident rate and spill rate are a direct function of movements

Additional coastal areas

Improves accuracy of vessel movement count. Does not affect overall coastal transit times.

Avoids some overcounting of vessel movements.

Additional vessel categories

Better data on fishing and small vessel counts. Fishing and small vessel numbers up, little change to international cargo & passenger vessels.

Effect on spill rate and return period depends on relative influence of total movements and average oil carried.

Figure 10-1 shows the overall spill rates by source and shows the results of sensitivity testing (discussed in Section 11.1). Column A Results for the 2002/2003 year – “as is” now; essentially this is a reflection of the current spill risk in terms of average risk-weighted spill rate for all oils and all sources. Column B Original results from 1998 – “as was” then. The differences between A and B reflect changes in activity for all vessel categories, changes to some modelling parameters and the inclusion of container vessels as a new category, addition of a small fishing vessel category and the way that fishing vessels and small craft have been counted. Columns C, D and E reflect the effect of changes to specific modelling parameters since 1998: Column C

1998 inputs with revised coastal transit times.

Column D

1998 inputs with revised estimates of average oil carried for the various vessel categories.

Column E

1998 inputs with 2002/2003 movement data.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

10-3

Figure 10-1: Average Spill Rate Summary 500.0

E 450.0

15.2

400.0

Spill Rate (tonnes/yr)

350.0

A

145.2

15.1

300.0 69.5

250.0 43.4

200.0 150.0

15.1

Small craft

B 15.2 13.6 16.5 15.5

86.2 90.3

C

0.0

Fishing

D 61.0

15.2 13.4 15.6 15.1

15.2 6.5 16.5 15.5

15.8

80.7

82.3

113.1

100.0 50.0

FPSO

34.5

59.7

55.8

48.4

31.0 6.1 22.1

17.4 16.4 18.3

16.5 15.6 17.8

24.1 15.8 20.7

2004 Output

1998 Output - Inputs Unchanged

45.9

Coastal cargo and passenger International cargo & passenger Coastal product tankers Foreign product tankers Crude/condensate tankers Large crude carriers

23.4 7.0 19.0

1998 Output - Transit 1998 Output - Average 1998 Output - Vessel Time oil movements Data Source

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\5000 - TECHNICAL\Model\Summary File: Summary tables.xls Sheet: Fig. 10-1

20/08/2004

Discussion and Conclusions of Spill Likelihood Modelling

SECTION 10

10.2.1 Return Period Following the 1998 risk assessment, the MSA adopted a 1% PEL (Probability of Exceedance Level) as a baseline for local response capability. The 1% PEL spill size at that time was estimated to be around 7,000 tonnes, by extrapolation of the return period vs. spill size curve. Re-examination of the 1998 results has confirmed errors in the plotting of this information and the results have been recalculated, giving a revised return period of around 3,500 tonnes previously. The 2004 results indicate that the 1% PEL spill size is now a little under 3,500 tonnes. As noted, the return period is not calculated in the model directly, but determined by extrapolation from a curve, so the resulting value will have some level of uncertainty. Within the model, spills of greater than 2,100 tonnes are solely a function of tanker activity, so any change in the 1% PEL is a reflection in the change in actual tanker traffic since 1998, differences in the way tanker movements have been counted and changes to transit times for the coastal areas used in the 2004 model. While the average cargo carried by large crude carriers (LCCs) visiting Marsden Point has increased by around 10% since 1998 and the number of visits is also slightly up (15%), these vessels only contribute around 9% of all port visits by tankers over a year and 6% of the total time spent annually by tankers in New Zealand in coastal waters. The number of movements by other tankers, particularly the coastal fleet and crude/condensate tankers has decreased significantly, so the fact that the 1% PEL spill has not changed significantly over the period is not unexpected.

10.2.2 Spill Rate Columns A and B in Figure 10-1 allow for comparison between the spill rate output for the 2004 model and 1998 study respectively. While the spill rate output changed for all categories of vessels, the change is significant for small craft and fishing vessels. Large changes have also occurred for coastal product tankers and foreign product tankers. The changes are a result of the combined effect of the differences between the models in the average oil carried, coastal transit time and vessel movements (both actual movements and what has been included/excluded in the count), as discussed in Section 11.1. The overall spill rate for 2002/2003 is about 5% lower than in 1998, when these changes in the model are taken into account.

10.3

Conclusions and Recommendations

The 2004 risk assessment gives an updated, and we believe, a more reliable picture of the likelihood of an oil spill in New Zealand waters than the previous study. It also includes better information on fishing vessels and smaller vessel activity and their contribution to the overall spill risk. Overall, it should provide greater insight into the patterns of shipping activity and the relative contribution to oil spill risk from the different risk creators, as well as giving an improved picture of the geographical spread of spill risk.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

10-5

Discussion and Conclusions of Spill Likelihood Modelling

SECTION 10

However, there are two remaining areas of significant uncertainty that should be addressed prior to the next risk assessment being carried out: •

The number of vessel movements through each coastal area.

•

The quantity of oil (fuel) carried on average by each category of vessel.

These items are not really an issue for tankers, where the overall numbers of coastal movements in a year is small (and can be quite accurately estimated) and where the quantity of oil being carried is relatively easily ascertained. However, for other foreign vessels, which account for over 8,500 vessel-days in coastal waters per annum, the lack of hard information is significant. Nor are they so significant for coastal cargo and passenger vessels which contribute 50% more port movements than foreign vessels, but only one third the number of vessel-days and carry on average only a fraction of the fuel. In order to improve the accuracy of the risk assessment, the MSA should consider focussing efforts on improved reporting and data collection from these foreign vessels. Our two key recommendations are as follows: 1.

The MSA should develop a suitable data collection system using standardised format. This would need to be based on existing port reporting requirements for vessel movements and bunkering. Each port could then provide to MSA monthly/yearly data which could feed straight into a risk assessment model.

2.

Each port should conduct localised risk assessments focused on reducing the risk and frequency of oil spills, as well as producing scenario-based contingency planning and exercises to test existing assumptions about oil spill response equipment, tactics and management.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

10-6

Sensitivity Analysis and Data Limitations 11

SECTION 11

Sensitivity Analysis and Data Limitations

This section supplements Section 10 and provides a more detailed assessment of the spill likelihood model, specifically addressing. •

The sensitivity of the model results to various inputs (sensitivity analysis) and the implications of this.

•

The quality and limitations of the data used to derive inputs for the modelling.

11.1

Sensitivity Analysis

To allow a more solid basis for interpreting the 2004 results, a sensitivity analysis was carried out. This gave an indication of the effects of changing key model inputs on the return period and spill rate output, the results of which are explained below. There were two separate procedures that comprised the model sensitivity analysis, as follows: •

Effects of changes in the average oil carried on the current (2004) model output. This was done as there was some inherent uncertainty in the values adopted for the average oil carried, particularly for the international cargo and passenger shipping category. This value is a key input to the model.

•

Effects of changing the inputs in the 1998 model from 1998 values to 2004 values for the transit time, oil carried and vessel movements. This was carried out to determine how current values for these inputs and the degree to which the changes in the analysis affect the risk output.

The method and results of each procedure follow.

11.1.1 Average oil carried - 2004 model The average oil carried for all vessels other than oil tankers was derived from an analysis of bunker data and vessel GRT. The method and results of this analysis are given at Appendix F. Notably bunker data was not readily available for the international cargo and passenger vessels category resulting in less confidence in the estimates of the average oil carried used for this category in the model. A value of 2,100 tonnes was specified for both container shipping and international cargo and passenger vessels. The effects of changing this input on the risk output, including the degree to which the return period output and the risk-weighted average spill rate are affected, were therefore analysed for these classes of vessel.

Method The current value for the average oil carried of 2,100 tonnes, for both categories of vessels, is considered to be on the higher side of the representative volume carried. Therefore values of 2,500 tonnes and 1,200 tonnes were selected for the analysis. The effect of changing the container shipping input only, then the international cargo and passenger vessels input only, and finally changing both were recorded.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

11-1

Sensitivity Analysis and Data Limitations

SECTION 11

Results Effects on Return Period Generally, changing the international cargo and passenger vessels input had a larger effect than the change in container shipping input and can be attributed to the higher volume of port visits and coastal shipping movements for this vessel category. Container vessels also spend proportionately less time transiting coastal waters by virtue of their greater speed. Notably for return period impacts the greatest effect was realised by changing the inputs for both categories of vessel. Measurable change in return period occurred for spill size exceedance values of 100 tonnes to 2,000 tonnes with the largest change occurring for the spill size exceedance of 2,000 tonnes. These effects specifically, will be described further in the following discussion. Increasing the volume carried from 2,100 to 2,500 tonnes resulted in the combined port and coastal return period output decreasing from 26 years to 24 years for a 2,000 tonne spill. Decreasing the volume from 2,100 to 1,200 tonnes gave an increase in return period from 26 years to 151 years. Effects on port return period and coastal return period were analysed separately and the largest changes occurred for the coastal return period output. For the case where both shipping inputs were changed by decreasing the volume carried, the coastal return period increased from 85 years to 940 years and the port return period increased from 38 years to 190 years respectively. Effects on Spill Rate As the risk-weighted average spill rate for a given vessel type is directly proportional to the average oil carried for that vessel type, an increase or decrease in the average oil carried will result in a proportional increase or decrease in the risk-weighted average spill rate.

Conclusion The return period is a direct function of the spill frequency output and therefore the changes can be explained by considering the calculation method for spill frequencies. The spill frequencies are derived from the spill probability distribution and are based on a step-wise spill exceedance scale. The spill probability is a function of the average oil carried and spill size exceedance in tonnes and will be greater than zero if the average oil carried is greater than the spill size in question. As the average oil carried for container shipping and the international cargo and passenger vessel, of 2,100 tonnes, is close to the 2,000 tonne spill size exceedance step, these two vessel types are the major contributors to the combined port and coastal return period for a spill size exceedance of 2,000 tonne. By increasing the average oil carried (from 2,100 to 2,500 tonnes) there will be a marginal increase in the probability for spill size exceedance of 2,000 tonnes and therefore an increase in spill frequency and a corresponding decrease in return period. By decreasing the oil carried from 2,100 to 1,200 tonnes, the contribution to a spill size exceedance of 2,000 tonnes is omitted. The combined port and coastal return period will be correspondingly higher and will equate to the contribution to the 2,000 tonne spill size exceedance from the oil tanker class of vessels. S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

11-2

Sensitivity Analysis and Data Limitations

SECTION 11

The average oil carried for international cargo and passenger vessels is therefore a key input in determining both return period for spills in the critical 2000-tonne range (in terms of the current New Zealand based spill response capability) and the overall contribution of these groups of vessels to the total spill rate. Hence, effort should be focussed on improving reporting and data collection for these vessels before the next risk assessment is carried out.

11.1.2 Spill Rate Analysis As previously noted the risk-weighted average spill rate is a key measure of oil spill risk representing the expected oil spilled over a long period, averaged over all spill sizes and expressed on an annual basis. A sensitivity analysis was carried out to determine the effect that coastal transit time, average oil carried and vessel movements (both port visits and coastal movements) had on the risk-weighted average spill rate. The combined port and coastal spill rate output was used in the analysis. Figure 10-1 gives a summary of this analysis. Column A represents the output of the current 2004 model. The relative contribution from each vessel category is shown, including the FPSO. The analysis was done by comparing the original 1998 output (Column B in Figure 10-1) to the spill rate output resulting from changes made to the 1998 model inputs. A comparison was then made of the 2004 output to the 1998 output with all inputs changed in the 1998 model. It is noted that the 2004 input data for transit time and vessel movements had to be modified for inclusion in the 1998 model to account for differences in the number of vessel categories and the breakdown of coastal regions. A discussion of the impact of each input change follows.

Coastal transit time The effect of transit time changes was to reduce the average spill rate output for every vessel category. The vessel transit time is simply the estimated time taken to traverse a given coastal region. Generally where transit times have been updated between the 1998 and 2004 analyses, the change has been a reduction in the time. The collision, grounding and fire frequencies respectively, are a direct function of the transit time. Therefore a reduction in transit time will result in a reduction in these frequencies leading to a reduction in spill rate as calculated.

Average oil carried As the spill rate is function of the average oil lost (that in turn is a direct proportion of the average oil carried), the changes to spill rate correspond directly to the changes made in the average oil carried. The changes in oil carried included a decrease for crude/condensate tankers, coastal product tankers, international cargo and passenger vessel, and small craft. There was an increase in oil carried for large crude carriers and foreign product tankers.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

11-3

Sensitivity Analysis and Data Limitations

SECTION 11

Vessel movements There was a significant change to the average spill rate output for fishing and small craft categories of vessels. As with transit time, the collision, grounding and fire frequencies respectively, are directly proportional to the vessel movements. Therefore an increase (or decrease) in vessel movements will be directly reflected in an increase (or decrease) in the spill rate. The better estimates of small craft numbers and the inclusion of small fishing vessels led to an increase in both port visits and coastal movements for these classes of vessel, when compared to the 1998 study. An increase in port visits and coastal movements is also the reason why there was a relatively large change in spill rate for international cargo and passenger vessel and coastal product tanker categories.

11.2

Quality and Limitations of Input Data

In general terms, the quality of the input data and the detail of the analysis should reflect the level of risk as well as the consequences associated with the activity it is applied to. While the numerical estimate of risk could be similar for a small spill and a large spill (for example, 1 tonne spilled once every year equals the same average spill rate as 10,000 tonnes once every 10,000 years), in terms of practical response measures, low consequence events can be much more easily dealt with. The quality of the input data at the bottom end of the consequence range (i.e. smaller vessels) has been much poorer in terms of vessel numbers, oil capacities, routes and activities than at the top end, where the number of vessels is small and the operations well documented (and controlled). Data provided on tanker movements and associated cargo transfers was quite detailed and complete so it has been possible to develop an accurate picture of bulk oil movements. However, it has been difficult to obtain accurate or complete information on numbers and movements of small craft, fishing vessels and bunkering operations. However, where the consequences are low, this is less of an issue, because if there is a minimum level of spill response capability (say 10 tonnes) in all major ports, then the frequency of these smaller spills is less critical in terms of preparedness. One area of significant uncertainty for both the 1998 risk assessment and the current one has been the volume of international cargo and passenger shipping along the coasts. While vessel port visits are reasonably well documented, there is no vessel tracking system operating in New Zealand, so it is difficult to estimate exactly how many vessels pass a given point on the coastline. However, as all the port companies have provided traffic data in electronic format this time around, it has been possible (with some data manipulation) to generate the number of movements for each coastal area for international and coastal cargo vessels, based on the number of port visits. A number of simplifying assumptions have been made in this calculation regarding routes between ports (within New Zealand and to/from New Zealand) and a significant amount of effort has been required to generate these estimates. The coastal movement counts are therefore considered to be more reliable this time, but are still only estimates and so subject to some uncertainty. As these vessels make a significant contribution (around

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

11-4

Sensitivity Analysis and Data Limitations

SECTION 11

30%) to the overall spill rate, efforts should be made to improve the estimates of coastal movements prior to the next risk assessment being carried out. Detailed data on fishing activity has been made available from MFish in the form of “catch effort” reported as fishing days, by statistical area, catch method and vessel size. As a result, the estimate of time spent by fishing vessels in each coastal area is considered to be much more accurate this time. The availability of more accurate data has also enabled the inclusion of small fishing vessels (6 – 24 m), for which limited information was available in the 1998 risk assessment. However, in the absence of a detailed survey of fishing vessels, which was not included in this work, it is difficult to establish representative values of average oil carried with any degree of confidence. The values assumed are considered to be conservative (on the high side). Similarly, data from the MSA’s SSM and SOP databases has enabled a far better estimate of the numbers of small commercial craft operating in New Zealand though assumptions have had to be made about the associated level of activity (in terms of port and coastal movements).

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

11-5

References

12

SECTION 12

References

Allen, A.A. and Dale, D.H. 1997. Oil Slick Classification: A System for the Characterization and Documentation of Oil Slicks. 1997 International Oil Spill Conference. DETR, 1999. Identification of Marine High Risk Areas (MEHRAs) in the UK, UK Department of the Environment, Transport and the Regions, December 1999. IMO, 1989? Analysis of Serious Casualties to Sea-going Tankers 1974 – 1988. International Maritime Organisation (IMO). ITOPF, 2004. Oil spill statistics – historical data, International Tanker Owners Pollution Federation (ITOPF), http://www.itopf.com. MSA, 2000. The New Zealand Marine Oil Spill Response Strategy 1999/2000. Maritime Safety Authority of New Zealand, 2000. MSA, 2001. Review of the Voluntary Vessel Routeing Code for Shipping in New Zealand Coastal Waters. Maritime Safety Authority of New Zealand, 2000. Woodward-Clyde (NZ) Ltd, 1998. Phase 2 of NZ Marine Oil Spill Risk Assessment – Report and Appendices (2 volumes), December 1998. Woodward-Clyde (NZ) Ltd, 2000. New Zealand Risk Assessment for Sea Carriage of Hazardous and Noxious Substances – Project Report. December 2000.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

12-1

Glossary of Terms

13

SECTION 13

Glossary of Terms

The following glossary is a summary of terms used within the text of this report. It contains some terms that have been developed for this risk assessment. Bunkering

Movement of oil from a port storage facility to vessel fuel tanks by hose or pipeline (can also occur ship to ship)

Bulk transfer

Movement of large volumes of oil cargo from ship to shore or shore to sip by hose or pipeline

Coastal waters

All sea from high water mark out to 12 nautical mile limit of the Territorial Sea of New Zealand

Collision

Vessel to vessel impact – usually resulting in damage to one or other of the vessels

Contact

Collision between a vessel and a wharf or other port structure

Director of Maritime Safety

The Director of Maritime Safety as defined in section 439 of the Maritime Transport Act 1994

DOC

Department of Conservation

Exclusive Economic Zone (EEZ)

All marine waters from the outer edge of the Territorial Sea seaward for 188 nautical miles until the 200 mile limit

Grounding

Action of a vessels hull which has impacted with the sea bed/land

Harbour

A place of shelter for ships

Maritime Safety Authority

The Authority, established under the Maritime Transport Act 1994 as a body corporate, owned by the Crown with perpetual succession. It has the responsibility for providing effective marine pollution prevention and an effective marine pollution response system at reasonable cost

MFish

Ministry of Fisheries

National Marine Oil Spill Contingency Plan

The marine oil spill response plan produced by the Director of Maritime Safety

Oil

Any petroleum in any form including crude oil, fuel oil sludge, oil refuse, and refined products (other than petrochemicals)

Oil Industry

Producers, refiners and marketers of oil, and associated carriers and service contractors

Oil Pollution Fund

A fund managed by the Maritime Safety Authority which receives its income from the oil pollution levy. It is used to provide money for New Zealand’s preparation for oil spill response and to meet the costs of clean-up where no spiller can be found to the costs

Oil Pollution Levy

A differential levy imposed on all vessels which carry oil as either cargo (tanker) or as fuel according to a formula based on the risk of an oil spill from their particular operation. Some offshore installations also pay a set levy based on their risk factor

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

13-1

Glossary of Terms

SECTION 13

Oil spill response

The entire process by which a marine oil spill is managed, including response planning, set-up, clean-up, and termination

Oil persistence

In relation to coastlines it is a relative measure of the period of time that oil will remain within or on a specific coastal type under normal conditions

Persistent oil

Oils and petroleum products such as crude oils, fuel oils and lubrication oils that, when spilt, remain after weathering in a residual form in the environment for an appreciable period

Physical hazard

A hazard presented to shipping in relation any coast or entry into ports, specifically: coastal geometry, water depth, prevailing wind and weather along

Port

Location of berthing and loading/unloading facilities provided for commercial and private shipping. Legal port boundaries are defined in the New Zealand Gazette and include eland, coastline and open water

Regional councils

Comprise all the current regional councils and those unitary authorities with the powers and functions of regional councils

Resource value

A relative measure of the value of natural, commercial, recreational and social values in a specific area

Response capability

The materials, equipment, personnel, training and organisation structure in place to respond to and manage oil spill response

Risk

The probability of an oil spill (of a specific size and type of oil from a given user group or activity), together with the cost of preparing for, responding to and cleaning up that spill, including preparation to reduce environmental,, economic and social impacts (MSA definition)

Trans-shipment

Transfer of material or oil from one ship to another at sea

Vessel Types

Categories of ships defined for the purposes of the oil spill risk models.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\MOSRA 2004 R001-C.DOC\13-JAN-05

13-2

Limitations

2

Limitations

URS New Zealand Limited (URS) has prepared this report for the use of the Maritime Safety Authority of New Zealand (MSA) in accordance with the usual care and thoroughness of the consulting profession. It is based on generally accepted practices and standards at the time it was prepared. No other warranty, expressed or implied, is made as to the professional advice included in this report. It is prepared in accordance with the scope of work and for the purpose outlined in MSA’s letter of engagement Ref. OS-0101-07 dated 12 December 2003 and URS’ proposal referenced in that letter. The methodology adopted and sources of information used by URS are outlined in this report. URS has made no independent verification of this information beyond the agreed scope of works and URS assumes no responsibility for any inaccuracies or omissions. No indications were found during our investigations that information contained in this report as provided to URS was false. This report was prepared between December 2003 and August 2004 and is based on the information available at the time of preparation. URS disclaims responsibility for any changes that may have occurred after this time. This report should be read in full. No responsibility is accepted for use of any part of this report in any other context or for any other purpose or by third parties. This report does not purport to give legal advice. Legal advice can only be given by qualified legal practitioners.

S:\ADV JOBS\48306 MSA\004 OIL SPILL RISK\6000 - DELIVERABLES\MAIN REPORT\FINAL\LIMITATIONS.DOC\14-JAN-05

2-1

URS New Zealand Limited Lambton House, Level 4 160 Lambton Quay, P.O. Box 3367, Wellington, New Zealand Direct: 64 4 496 3750 • Fax: 64 4 496 3755