Task 4 and 5: Characterization of Incidents and Spills

Size: px
Start display at page:

Download "Task 4 and 5: Characterization of Incidents and Spills"

Transcription

1 Gateway Pacific Terminal Vessel Traffic and Risk Assessment Study Task 4 and 5: Characterization of Incidents and Spills Prepared for Pacific International Terminals, Inc. Prepared by The Glosten Associates, Inc. in collaboration with Environmental Research Consulting, Inc. Northern Economics, Inc. File No March 2013 Rev. P0 Consulting Engineers Serving the Marine Community 1201 Western Avenue, Suite 200, Seattle, Washington TEL FAX

2 Contents References... iii Executive Summary... iv Scope of Work per Professional Services Agreement... v Section 1 Introduction Forecasting Contaminant Outflow Definitions of Terms Acronyms, Abbreviations, Parameters, and Variables Incident Types... 4 Section 2 Contaminant Outflow Model Objective Methodology Scenario Spill Volumes (SV v,a,i,l ) Scenario and Case Parameters Monte Carlo Simulation Post-Processing Monte Carlo Simulation Data Programming Environment Section 3 Input Data Traffic Days (TD v,a,l ) Incident Rates (IR v,a,i,l ) Vessel Capacities (VC f,v ) Tankers Cargo Ships Base Traffic Bulk Carriers GPT-Calling Bulk Carriers Tank Barges Tugboats Passenger and Fishing Vessels Fishing Vessels greater than 60 ft Spill Probabilities (SP f,v,i ) Outflow Percentage Probabilities (OP v,i ) Bunker Outflow Percentage for All Vessels for Other Non-Impact Spills Gateway Pacific Terminal VTS Study i The Glosten Associates, Inc.

3 3.5.2 Cargo Oil Outflow Percentage for Tankers and Tank Barges for Other Non-Impact Spills Dry Cargo Outflow Percentage for Bulk Carriers for Other Non-Impact Spills Dry Cargo Outflow Percentage for Bulk Carriers for Collision Spills Section 4 Results Interpreting Cumulative Distribution Functions Most Likely Geographic Location Where Spills May Occur (Task 4) Total Number of Incidents Total Number of Spills Number of Spills by Subarea Vessel Capacities Total Annual Oil Outflow Annual Oil Outflow by Subarea Total Bulk Outflow Bulk Outflow by Subarea Potential Size of Contaminant Release from a GPT-Calling Vessel (Task 5) GPT-Calling Vessel Incidents by Incident Type GPT-Calling Vessel Oil Spill Size GPT-Calling Vessel Oil Spill Distribution by Incident Type GPT-Calling Vessel Bulk Spill Size GPT-Calling Vessel Bulk Spill Distribution by Incident Type Section 5 Summary of Results Appendix A Characterization of Casualty Consequences (Task 5), Gateway Pacific Terminal Vessel Traffic and Risk Assessment Study, Environmental Research Consulting, 14 March Gateway Pacific Terminal VTS Study ii The Glosten Associates, Inc.

4 References 1. Task 2: Traffic Analysis, The Glosten Associates, Inc. in collaboration with Northern Economics, Inc., 22 March Task 3: Accident Probability Statistics, The Glosten Associates, Inc. in collaboration with Northern Economics, Inc., and Environmental Research Consulting, Inc., 22 March Revised Project Information Document, Gateway Pacific Terminal, Whatcom County, Washington, Pacific International Terminals, Inc., March Significant Ships, The Royal Institution of Naval Architects, Gateway Pacific Terminal VTS Study iii The Glosten Associates, Inc.

5 Executive Summary This Vessel Traffic and Risk Assessment Study (VTS) is being conducted by The Glosten Associates (Glosten) for the proposed Gateway Pacific Terminal (GPT) to be located at GPT/Cherry Point in Washington State. The purpose of the study is to assess the potential risks posed by new traffic that the proposed terminal will bring to the Puget Sound. Current vessel traffic levels and forecasted traffic levels are considered, including tugs and GPT-calling vessels. The area studied includes the designated Puget Sound vessel transit lanes, the maneuvering area near the planned GPT project at GPT/Cherry Point, the local anchorage areas, and the transit routes for tugs assisting GPT. Plans call for 487 total annual visits for the anticipated GPT-calling traffic at full throughput level in 2026 (Reference 3). Of the total vessel calls, it is projected that there will be 318 Panamax and 169 Capesize (up to 180,000 DWT) vessels. The GPT-calling vessels will be utilizing the established traffic lanes between Cape Flattery and Cherry Point. A computational model was developed to estimate the potential contaminant outflow (bunkers, cargo oil, and dry cargo) in the study area, with and without GPT operations. The difference in these outflow results is the predicted contaminant outflow due to the addition of GPT. A Monte Carlo method algorithm is implemented to quantify uncertainties and variability in the inputs used in the analysis. The Monte Carlo algorithm calculates the number of incidents, number of spills, and volume of outflows for 10,000 stochastic results, each of which models the total annual traffic activity. The contaminant outflow model is presented in Section 2. The various data sources used by the model are discussed in Section 3. Quantitative results of the model are presented in Section 4. A summary of results is presented in Section 5. A majority of GPT-calling vessel incidents are predicted to be in the other non-impact incidents category. However, the average annual oil spill volumes are predicted to be fairly evenly distributed between allisions, groundings, and other non-impact incidents, with less volumes from collisions and bunker errors. Average annual bulk spill volumes are predicted to be dominated by collisions, which are much rarer but have higher bulk outflow consequences than transfer errors and other nonimpact incident types. It is assumed that bulk outflow will not occur in the study area as a result of grounding or allision. This assumption is discussed in Appendix A. The model predicts much higher average oil outflow than median oil outflow from GPT-calling vessels due to rare predicted occurrences of very oil large spills. Oil spills due to collisions, allisions, and groundings are predicted to occur infrequently, but result in large oil outflows when they do occur. Spills due to bunker, transfer, and other non-impact incidents are predicted to occur more frequently, but result in small oil outflows when they do occur. The model predicts much higher average dry cargo outflow than median dry cargo outflow from GPT-calling vessels due to rare predicted occurrences of very large dry cargo spills. Dry cargo spills due to collisions are predicted to occur infrequently, but result in large dry cargo outflows when they do occur. Other non-impact incidents are predicted to occur more frequently, but result in small dry cargo outflows when they do occur. Gateway Pacific Terminal VTS Study iv The Glosten Associates, Inc.

6 Scope of Work per Professional Services Agreement 1 3. Determines the risk of accident involving GPT-calling vessels that may result in contaminant release. Accidents shall include collision, allision, power groundings and drift groundings. In evaluating these risks the study should consider all vessel traffic and reasonably foreseeable increases and decreases in vessel traffic along the entire pathway followed by vessels between Cherry Point and Buoy J, including but not limited to vessels calling in British Columbia, GPT-calling vessels, and vessels calling at the BP Cherry Point Refinery, Conoco Phillips Ferndale Refinery, Alcoa-Intalco Works, and any other reasonably foreseeable future marine terminal facilities in the Cherry Point area. 4. Determines the most likely geographic location where accidents, as defined in 3, may occur. 5. Determines the potential size of a contaminant release from an accident, as defined in 3. 1 Exhibit A, Scope of Services Tasks 4 and 5, Professional Services Agreement between Pacific International Terminals, Inc. and the Glosten Associates, Gateway Pacific Terminal Vessel Traffic and Risk Assessment Study, Effective Date June 18, Gateway Pacific Terminal VTS Study v The Glosten Associates, Inc.

7 Section 1 Introduction 1.1 Forecasting Contaminant Outflow A contaminant outflow model was developed to quantify the effects of the proposed Gateway Pacific Terminal (GPT). The model uses historical incident and traffic data to predict the rate at which incidents that may result in contaminant outflow occur in the forecasted years. Traffic volumes by vessel type are projected for different geographic regions throughout the system, and the composition of vessels that make up each vessel type are projected in terms of size and structural integrity (i.e., single- or double-hulled). Because of uncertainty in projections and variability in historical data, a Monte Carlo simulation is employed to generate a probabilistic set of potential contaminant outflow outcomes. The results are then post-processed so that comparisons can be made between forecast contaminant outflow with and without operation of GPT. Section 2 describes the Monte Carlo contaminant outflow model in detail. Section 3 describes the data that is used by the model. Section 4 presents the results of the simulation. A summary of results is presented in Section Definitions of Terms Definitions for the terms used in this study are provided as follows. Activity Type (a) Deadweight Tonnage Forecast Year (f) GPT (g) GPT-Calling Tugboats Incident Incident Rate A scenario parameter. The four (4) project-specific activity categories are: 1. Underway 2. Maneuvering 3. Docked 4. Anchored The measure of the amount of weight that a ship may carry, including cargo, bunkers (fuel), ballast water, fresh water, dirty water, provisions, crew, etc. A case parameter. The two (2) project-specific forecast years are: A case parameter. The two (2) project specific GPT indices are: 1. Without GPT 2. With GPT Tugboats are defined as GPT-calling tugboats during the time they are transiting to or from the GPT dock and while docking GPT-calling vessels. An event deemed by the US Coast Guard to have the potential for an oil spill. A spill may or may not have occurred. The number of incidents per vessel traffic day. IRs are defined for a given combination of scenario parameters: vessel type (v), activity type (a), incident type (i), and location (l). Gateway Pacific Terminal VTS Study 1 The Glosten Associates, Inc.

8 Incident Type (i) Location (l) Monte Carlo Simulation Parameter Poisson Distribution Probability Distribution Random Number Random Variable Regression Analysis A scenario parameter. The six (6) project-specific incident categories are: 1. Collision 2. Allision 3. Grounding 4. Transfer Error 2 5. Bunker Error 2 6. Other Non-Impact Incident 3 A scenario parameter. The seven (7) project-specific sub-areas, as shown in Figure 1, are: 1. Strait of Juan de Fuca West 2. Strait of Juan de Fuca East 3. Rosario Strait 4. Haro Strait Boundary Pass 5. Cherry Point 6. Saddle Bag 7. Guemes Channel Fidalgo Bay The process of calculating a sufficient number of stochastic results to produce high-resolution probability distributions of cumulative oil outflow for a given scenario or scenarios. An attribute with a set of prescribed, possible values for selection. A probability distribution used to describe rare events that occur independent of the time of last occurrence. A function describing the likelihood of each possible outcome. A number in the domain (0, 1) that is generated in order to sample a value of a probability distribution. A variable that is described as a probability distribution and sampled using a random number. Interpolation of data in order to estimate a value that is not explicitly available or given. Stochastic Result One possible contaminant outflow result; obtained by sampling all 1,008 scenarios (v,a,i,l combinations) once each. Scenario Study Area Subarea Traffic Day A combination of parameters present during a particular incident, as defined in Table 1 of this report, which includes: vessel type (v), activity type (a), incident type (i), and location (l). The geographic bounds of the area considered in the study. The area covered by all locations (l), as shown in Figure 1. See Location. 24 hours of time in the study area. Traffic days may be further defined with respect to the type of vessel (v), the activity (a), and/or the location (l). 2 Transfer Errors and Bunker Errors are grouped in Appendix A under the name Transfer Error. 3 Other Non-Impact Incidents are named Other Non-Impact Error in Appendix A. Gateway Pacific Terminal VTS Study 2 The Glosten Associates, Inc.

9 Vessel Capacity Vessel Type (v) VTS Vessel The capacity of a given vessel type for a given contaminant type (oil or bulk), including cargo and bunkers (fuel). A scenario parameter. The six (6) project-specific vessel categories are: 1. Tanker 2. Tank barge 3. Bulk carrier 4. General cargo ship 5. Tugboat 4 6. Passenger and Fishing Vessel 4 A vessel belonging to one of the project-specific vessel types. Figure 1 Project Study Area Showing Subareas (Locations), as referenced from Study Area definition 1.3 Acronyms, Abbreviations, Parameters, and Variables Definitions for the acronyms, abbreviations, parameters, and variables used in this study are provided as follows. a CDF DWT ERC f g Scenario parameter defining activity type Cumulative Distribution Function Deadweight Tonnage Environmental Research Consulting Case parameter defining forecast year Case parameter defining with or without GPT 4 Tugboat and passenger and fishing vessel types are grouped in Appendix A under the name Other. Gateway Pacific Terminal VTS Study 3 The Glosten Associates, Inc.

10 GPT I i IR l LOA NEI NI SP TD v VC VTS Gateway Pacific Terminal Incident Scenario parameter defining incident type Incident Rate Scenario parameter defining location (subarea) Length Overall Northern Economics, Inc. Number of Incidents Spill Probability Traffic Days Scenario parameter defining vessel type Vessel Capacity Vessel Traffic Study 1.4 Incident Types The six (6) incident types are collision, allision, grounding, transfer error, bunker error, and other non-impact incident. Collisions, allisions, and groundings are impact incidents. Drift and powered groundings are included as they occurred historically, but they are not differentiated in the incident rates. Transfer errors, bunker errors, and other non-impact incidents are nonimpact incidents. The other non-impact incident type includes the following causes: equipment failures, fires, explosions, operator errors, and structural failures. Historical incidents with unknown cause are also assigned to the other non-impact incident type. Gateway Pacific Terminal VTS Study 4 The Glosten Associates, Inc.

11 Section 2 Contaminant Outflow Model 2.1 Objective The objective of the contaminant outflow model is to determine the quantities and locations of oil and bulk cargo outflow for the following four (4) traffic volume cases within the vessel traffic study area (Table 1). Table 1 Case GPT VTA Traffic Volume Cases Forecast Year (f) With GPT (g) No Yes No Yes Oil outflow is from petroleum cargos and bunkers. Bulk cargo outflow is from all bulk commodity types. 2.2 Methodology Scenario Spill Volumes (SV v,a,i,l ) Total contaminant outflow for a given year is determined by summing all the individual spills that occur in that year. Determination of the quantity and volume of individual spills is accomplished by breaking the system into scenarios that represent each potential occurrence of oil and bulk outflow, and sampling each scenario to determine if that scenario results in any spills of oil cargo, bulk cargo, bunker fuel, or some combination thereof. Scenarios are defined by six (6) vessel types (v), four (4) activity types (a), six (6) incident types (i), and seven (7) locations (l), as defined in Section Thus there are = 1,008 scenarios for each traffic volume case (Table 1). These 1,008 scenarios are assumed to include all combinations of the scenario variables that may significantly contribute to the quantity of contaminants spilled. Other vessel types that are not included in the scenario set, such as pleasure boats, are assumed to have no significant contribution to the total annual quantity of contaminants spilled. Total annual contaminant outflow is thereby defined by the summation of spill volume for each scenario (SV v,a,i,l ), as shown in Equation 1. Total Annual Outflow(f,g) v 1 a 1 i 1 l 1 f, g, v, a, i, l 7 SV 1 Spills may or may not occur as the result of incidents that have the potential to result in a spill. It is necessary, therefore, to determine the rate at which incidents occur and the probability that a spill occurs, given an incident, for each scenario. Historical incident rates per vessel traffic day for each scenario (IR v,a,i,l ) are developed as detailed in the Task 3 report of this project (Reference 2). Spill probabilities for each (forecast year, vessel type, incident type) combination (SP f,v,i ) are described in Section 3.4. The model is concerned with yearly contaminant outflow, as opposed to daily contaminant outflow, so it is necessary to determine the forecasted yearly average incident rate (λ f,g,v,a,i,l ). Gateway Pacific Terminal VTS Study 5 The Glosten Associates, Inc.

12 This is accomplished by multiplying the historical incident rate by the forecast number of traffic days for the given scenario, as defined by Equation 2. Forecast traffic days (TD f,g,v,a,l ) for the four traffic volume cases are developed as detailed in the Task 2 report of this project, Reference 1. IR TD 2 To determine the number of incidents that occur for a given scenario, it is assumed that incidents are rare and occur independently of the time since the last incident, and thus follow the Poisson distribution. Equation 3 defines the Poisson distribution for the number of incidents (NI v,a,i,l ) that will occur for a forecasted yearly average incident rate (λ f,g,v,a,i,l ). NI e P( NI) 3 ( NI )! By defining P(NI v,a,i,l ) as a random number between 0 and 1 and solving for NI v,a,i,l, the number of incidents for the given scenario is determined. For each incident that occurs, the probability that it results in a spill is given by the Spill Probability (SP f,v,i ). By generating a random number between 0 and 1 and comparing it with (SP f,v,i ), it is determined whether or not a spill occurs. If a spill does occur, it is necessary to determine the spill volume (SV f,g,v,a,i,l ). Spill volume is the product of Outflow Percent (OP v,i ) and Vessel Capacity (VC f,v ), as defined in Equation 4. SV OP VC 4 Outflow percentages (OP v,i ) for each combination of vessel type and incident type are developed and presented in Appendix A. Vessel Capacities (VC f,v ) are described in Section Scenario and Case Parameters The variables described in Section are dependent on parameters for each scenario. The taxonomy of project scenarios is summarized in Table 2. Table 2 Project Scenario Parameters Vessel Type (v) Activity Type (a) Incident Type (i) Location (l) 1. Tanker 1. In-transit 1. Collision 1. Strait of Juan de Fuca West 2. Tank Barge 2. Maneuvering 2. Allision 2. Strait of Juan de Fuca East 3. Bulker 3. At dock 3. Grounding 3. Rosario Strait 4. General Cargo 4. At Anchor 4. Transfer 4. Haro Strait and Boundary Pass Error 5. Tug 5. Bunker Error 5. Cherry Point 6. Passenger or Fishing Vessel 6. Other Non- Impact 6. Saddlebag (including Vendovi Anchorages 7. Guemes Channel and Fidalgo Bay Gateway Pacific Terminal VTS Study 6 The Glosten Associates, Inc.

13 Certain variables are also dependent on case parameters 5, as presented in Table 3. The four GPT Traffic Volume Cases from Table 1 are defined analytically by variables dependent on the four combinations of case parameters. Table 3 Project Case Parameters Forecast Year (f) With or Without GPT (g) Without GPT With GPT An example of a Case 4 traffic volume scenario is as follows: In year 2026, with an operational Gateway Pacific Terminal, a tanker, while underway, has the potential for a collision in Strait of Juan de Fuca East. The parameters that define the variables used to determine whether a spill occurs in this scenario are identified by the following indices, as given in Table 2 and Table 3: f = 2, g = 2, v = 1, a = 1, i = 1, l = Monte Carlo Simulation The variables described in Section that are used to determine total contaminant outflow are not deterministic and are, thus, probabilistically distributed, meaning that each stochastic sample for each variable will return a different value within bounded ranges and with probabilities defined by distribution parameters 6. Therefore, one summation across all scenario outflows ( SV v,a,i,l ) will result in one stochastic result of total contaminant outflow volume. In order to understand the uncertainty inherent in the prediction of potential outflow volumes and the likelihood that each outflow volume will occur, it is necessary to calculate total contaminant outflow many times. The Monte Carlo method is thus employed to build a probability distribution of possible solutions of total annual contaminant outflow. Each solution of total contaminant outflow is called a stochastic result. Each of the 1,008 scenarios are calculated for 10,000 stochastic results, for a total of 1,008 x 10,000 = 10,080,000 (10.08 million) calculations of scenarios potentially resulting in spills, for each traffic volume case (Table 1). The Monte Carlo simulation cycles through each case parameter and project scenario parameter, building a database of incidents and spills identified by these parameters, as detailed in Figure 2. For each scenario in each case, the Monte Carlo simulation generates random numbers to determine the sample values as summarized in Table 4. 5 For instance, Vessel Capacity (VC f,v ) is dependent on scenario parameter vessel type (v) and case parameter forecast year (f), since the average size of a given vessel type will change over time. In addition OP v,i is also dependent on the case parameter, f, since the proportion of double-hull vessels will increase over time. 6 For example, Outflow Percentage (OP v,i ) is bounded by 0% to 100%, with a mean typically skewed towards the lower end of the bounds, since only on rare occasion does a spill result in the outflow of a majority of the total vessel capacity. Gateway Pacific Terminal VTS Study 7 The Glosten Associates, Inc.

14 Table 4 Traffic Days (TD) Summary of Random Variables generated by Monte Carlo Simulation Number of Incidents (NI) Vessel Capacity (VC) Vessel Hull Type (SH/DH) Spill Probability (SP) Outflow Percentage (OP) The unit of time describing the number of days per year a vessel is engaged in a given activity type (a) in a given location (l). The number of annual incidents that occur for a given scenario incident rate (IR) and traffic days (TD). The capacity of the vessel for a given forecast year (f) and vessel type (v). Several other random numbers are sampled to determine the vessel capacity, depending on vessel type (v), as detailed in Section 3.3. For a given vessel type (v), whether the vessel is single-hulled or doubled hulled. For a given incident, whether a spill occurs. For a given spill, what percentage of the vessel capacity is spilled. Gateway Pacific Terminal VTS Study 8 The Glosten Associates, Inc.

15 Figure 2 Monte Carlo Simulation Flow Diagram Gateway Pacific Terminal VTS Study 9 The Glosten Associates, Inc.

16 2.2.4 Post-Processing Monte Carlo Simulation Data Changes in the forecasted quantity of contaminants spilled in the subject area with the addition of GPT are estimated by comparing the probability distributions of oil and bulk outflow for the four traffic volume cases. Because the model is set up so that all incidents and spills are stored in a database with their respective case and scenario parameters, the data can be postprocessed in any number of ways. Cumulative distribution functions (CDFs) are created to show the probability that quantities of a given phenomenon will happen. By creating a CDF of annual oil outflow in Cherry Point, for instance, the 50 th percentile (median) oil outflow volume in Cherry Point can be determined. By creating CDFs of oil outflow in Cherry Point for each of the four traffic volume cases, the median values for each traffic volume case can be extracted, allowing for a quantitative comparison of median annual oil outflow in Cherry Point with and without GPT. CDFs of outflow volume are created by summing all of the outflow volumes in each stochastic result, and sorting these totals by magnitude. This sorted array represents the CDF, where the index of each value, divided into the total number of stochastic results, represents the probability that the outflow volume will be that value or less. In an array of 10,000 stochastic results, therefore, the 5,000 th value represents the 50 th percentile (median) outflow volume. This methodology can be applied to any combination of parameters that were tracked, such as number of incidents per subarea. Results from the simulation are presented in Section Programming Environment The Monte Carlo simulation is programmed using the Python(X,Y) distribution of the Python programming language. Input data from Microsoft Excel is read into the program with xlrd. Random numbers are generated and cumulative distribution functions are interpolated using SciPy. Results are plotted using MatPlotLib. Gateway Pacific Terminal VTS Study 10 The Glosten Associates, Inc.

17 Section 3 Input Data The contaminant outflow model forecasts probability distributions for oil and bulk outflow in the study area based on historical data and projections of vessel traffic. The following sections describe how these data and projections are obtained and used by the model. 3.1 Traffic Days (TD v,a,l ) Traffic days are the number of days spent, per year, for each combination of vessel type, activity type, and location. Northern Economics, Inc. (NEI) developed probability distributions of traffic days for every (vessel type, activity type, location) combination, for the four traffic volume cases (Table 1). The results are presented in Reference 1. NEI provided a database of 10,000 stochastic results of vessel traffic days for each v, a, and l combination, which are read into the contaminant outflow model. 3.2 Incident Rates (IR v,a,i,l ) Incident rates are the rate at which incidents occur for each scenario; i.e., for each combination of vessel type, activity type, incident type, and location. Generally, the incident rate for each scenario is determined by dividing the number of incidents that occurred during the historical study period ( ) by the traffic days for that vessel-activity-location combination, with exceptions for those scenarios for which there is insufficient data to support a non-zero incident rate. Incident rates are tabulated in an Excel spreadsheet, which is read in by the contaminant outflow model. A comprehensive discussion of incident rate formulation is given in the Task 3 report (Reference 2). 3.3 Vessel Capacities (VC f,v ) Vessel capacities are the maximum amounts of bunker (fuel) oil, cargo oil, and bulk cargo that each vessel in the system can carry. Tens of thousands of study vessels transit within the system each year and it is impossible to know the exact distribution of vessels in the system in the forecasted years. Thus, vessel capacities for each vessel type are random variables with the capacity of each vessel type being described by a probability distribution function. Changes in the capacity distributions of certain vessel types are also anticipated for future years. Various techniques are employed to account for trends and sudden anticipated changes in the capacity distributions of vessels in the system. The methods used to determine vessel capacities for each vessel type and forecast year are detailed below Tankers NEI provided deadweight tonnage (DWT) for every tanker that transited through the system in Tankers were split into two sub-categories (product and crude) due to significant differences in size and transit frequency. Figure 3 shows the DWT distributions for tanker subtypes in the system in Gateway Pacific Terminal VTS Study 11 The Glosten Associates, Inc.

18 Figure 3 Tanker DWT Distribution, 2010 (Data source: Northern Economics, Inc. 2013) NEI also provided formulae for the weighted average deadweight tonnage of product tankers (Equation 5) and crude tankers (Equation 6), and the forecast numbers of product and crude tanker vessel traffic days in the system, which were then converted to percentages (Table 5). Average DWT(Product) = 6,482 x ln(year 1998) + 49,895 5 Average DWT(Crude) = 5,685.4 x ln(year 1998) + 105,505 6 Table 5 Tanker Traffic Breakdown by Subtype Product 56% 57% 60% Crude 44% 43% 40% Finally, Environmental Research Consulting (ERC) provided regression equations for estimating tanker bunker capacity (Equation 7) 7 and tanker cargo capacity (Equation 8) 8 in gallons, based on DWT (Appendix A). Bunker Capacity(Tanker) = x DWT + 106,942 (gallons) 7 Cargo Capacity(Tanker) = x DWT (gallons) 8 To sample a tanker capacity for a given year, a random number is generated to determine the tanker subtype (product or crude), with probabilities of returning a given subtype in a given forecast year shown in Table 5. Another random number is generated to randomly select a deadweight tonnage of that tanker subtype from the database of tankers in the system in The DWT is then extrapolated to the forecast year by multiplying the sampled DWT by the ratio of average 2010 DWT to average forecast year DWT using Equation 5 or 6, depending on the tanker subtype. Finally, the bunker and cargo capacities are derived from Equations 7 and 8. The model results for tanker capacity distributions for product and crude tankers are presented in Section Cargo Ships NEI provided deadweight tonnage (DWT) for every cargo ship that transited through the system in Cargo ships were split into two sub-categories, container and general cargo, 7 Based on adjustment for 70% bunker capacity, as noted in Appendix A. 8 Based on adjustment for 98% cargo oil capacity, as noted in Appendix A. Gateway Pacific Terminal VTS Study 12 The Glosten Associates, Inc.

19 due to significant differences in size and transit frequency. Figure 4 shows the DWT distributions for cargo ship subtypes in the system in Figure 4 Cargo Ship DWT Distribution, 2010 (Data source: Northern Economics, Inc. 2013) NEI also provided formulae for the weighted average DWT of container ships (Equation 9) and general cargo ships (Equation 10), and the forecast numbers of container and general cargo ships in the system, which were then converted to percentages (Table 6). Average DWT(Container) = 6,534.9 x ln(year 1998) + 39,156 9 Average DWT(General Cargo) = x ln(year 1998) + 21, Table 6 Cargo Ship Traffic Breakdown by Subtype Container 87% 84% 81% General Cargo 13% 16% 19% Finally, ERC provided a regression equation for estimating cargo ship bunker capacity in gallons, based on DWT, Equation 11 9 (Appendix A). Bunker Capacity(Cargo Ship) = 27,545 x DWT 64,922 (gallons) 11 To sample a cargo ship bunker capacity for a given year, a random number is generated to determine the ship subtype, container or general cargo, with probabilities of returning a given subtype in a given forecast year shown in Table 6. Another random number is generated to randomly select a DWT of that cargo ship subtype from the database of cargo ships in the system in The DWT is then extrapolated to the forecast year by multiplying the sampled DWT by the ratio of average 2010 deadweight to average forecast year DWT, using Equation 9 or 10, depending on the cargo ship subtype. Finally, the bunker capacity is derived from Equation Base Traffic Bulk Carriers NEI provided deadweight tonnage (DWT) for every bulk carrier (bulker) that transited through the system in Bulkers were split into two sub-categories, grain and non-grain, due to significant differences in size and transit frequency. Figure 5 shows the DWT distributions for bulker subtypes in the system in Based on adjustment for 70% capacity, as noted in Appendix A. Gateway Pacific Terminal VTS Study 13 The Glosten Associates, Inc.

20 Figure 5 Bulk Carrier DWT Distribution, 2010 (Data source: Northern Economics, Inc. 2013) NEI also provided formulae for the weighted average DWT of grain bulkers (Equation 12) and non-grain bulkers (Equation 13), and the forecast numbers of grain and non-grain in the system, which were converted to percentages (Table 7). Average DWT(Grain) = x ln(year 1998) + 62, Average DWT(Non Grain) = x ln(year 1998) Table 7 Bulker Traffic Breakdown by Subtype Grain 43% 36% 27% Non-Grain 57% 64% 73% A regression equation for estimating bulk carrier bunker capacity based on DWT was formulated using information from 21 bulkers of various sizes, including Capesize and Panamax vessels, as shown in Figure 6. The data points circled in green are vessels that actually transited through the system in The Capesize vessels are those in the upper right-hand corner of the figure. The gap that exists between approximately 80,000 and 180,000 DWT is because very few deadweight tankers are built in this size range, for economic reasons. The least-squared regression line shown in Figure 6 is used to estimate bulker bunker capacity in gallons. An adjustment factor of 70% is then applied to the equation describing the line to match the convention in Appendix A. The regression equation is shown as Equation 14. Gateway Pacific Terminal VTS Study 14 The Glosten Associates, Inc.

21 Figure 6 Bulker Bunker Capacity versus DWT (Data source: Environmental Research Consulting 2013) Bunker Capacity(Bulker) = x DWT + 218,520 (gallons) 14 Finally, ERC provided a regression equation for estimating bulker cargo capacity in cubic feet, Equation 15, based on DWT (Appendix A). Cargo Capacity(Bulker) = x DWT (ft 3 ) 15 To sample a bulker bunker or cargo capacity for a forecast year, a random number is generated to determine the bulker subtype, grain or non-grain, with probabilities of returning a given subtype in a given forecast year shown in Table 7. Another random number is generated to randomly select a DWT of that bulker subtype from the database of bulkers in the system in The DWT is then extrapolated to the forecast year by multiplying the sampled DWT by the ratio of average 2010 deadweight to the average forecast year DWT, using Equation 12 or 13, depending on the bulker subtype. Finally, the bunker and cargo capacities are derived from Equations 14 and 15. The model results for bulk carrier capacity distributions for bunker and cargo are presented in Section GPT-Calling Bulk Carriers Data from Reference 3 was used to determine the DWT range of GPT-calling bulk carriers (GPT bulkers) in the forecast years, as summarized in Table 8. Table 8 GPT Bulker Traffic Breakdown by Subtype (Data source: Northern Economics, Inc. 2013) Panamax Capesize Size Range (DWT) 65,000-85, , , Calls (% of total) 144 (65%) 77 (35%) 2026 Calls (% of total) 318 (65%) 169 (35%) To sample a GPT bulker bunker or cargo capacity for a forecast year, a random number is generated to determine the GPT bulker subtype (Panamax or Capesize), with probabilities of returning a given subtype in a given forecast year equal to the percentage of calls shown in Gateway Pacific Terminal VTS Study 15 The Glosten Associates, Inc.

22 Table 8. Another random number is generated to randomly determine a DWT of that GPT bulker subtype within the size range given in Table 8 for that GPT bulker subtype. It is assumed that there is a uniform distribution of DWT within in size range, meaning that every value within the DWT range has an equal probability of being randomly selected. Finally, the bunker and cargo capacities are derived from Equations 14 and 15. The model results for bulk carrier capacity distributions for bunker and cargo are presented in Section Tank Barges A comprehensive study of tank barges operating in the study area in 2012 was conducted, and a database of characteristics of those 26 vessels was compiled. Length times beam times depth and capacity for 18 tank barges with available capacity data were plotted against each other, and a least-squared regression line was fit to the data, the equation of which was used to estimate the capacities of the remaining eight tank barges. An adjustment factor of 98% was applied to the tank barge capacities to match the convention in Appendix A. Based on expert judgment and interviews with tank barge owners, when possible, it was estimated when each tank barge would reach the end of its service life, to account for a changing distribution over time. It was assumed that no new tank barges will begin operating in the system in the future, as tank barges are being phased out as obsolete technology. Of those tank barges in the system, it is expected that between 2010 and 2016 several are expected to be taken out of service, but between 2016 and 2026 no others will reach the end of their service life. Therefore, the 2016 and 2026 tank barge capacity distributions are identical. A summary of tank barge sizes is shown in Figure 7. Figure 7 Tank Barge Cargo Capacity Distribution (Data source: Northern Economics, Inc. 2013) It is important to note that tank barges are towed by tugboats (tugs), which are also at risk for oil outflow. Oil outflow from tugs is accounted for separately Tugboats As part of their vessel traffic study (Reference 1), NEI provided a comprehensive database of tugboats that transited the system from In total, there were 668 tugs accounting for 76,929 transits through the system. It would be impractical to obtain capacity information on all 668 tugs, so a representative distribution of tugboat bunker capacities was developed based on the tugs in the system. This was accomplished by sorting the tugs by number of transits Gateway Pacific Terminal VTS Study 16 The Glosten Associates, Inc.

23 from and obtaining capacity information for tugs accounting for a significant percentage of total tug traffic in the system. In all, 24 tug bunker capacities were obtained for tugs that accounted for 18,246 (24%) of the tugboat traffic. An adjustment factor of 70% is then applied to these capacities to match the convention in Appendix A. The tug bunker capacity distribution of this representative database is summarized in Figure 8. Figure 8 Tugboat Bunker Capacity Distribution (Data source: Glosten 2013) To obtain a tugboat bunker capacity, the capacity of one of the tugs in the representative database is randomly selected. It is assumed that there will be negligible change in the capacity distribution of tugs between 2010 and GPT-Calling Tugboats Tugboats are defined as GPT-calling tugboats during the time they are docking GPT-calling vessels. It is assumed that the bunker capacities of GPT-calling tugboats follow the same distribution of all tugboats in the system, as described in Section The difference, as far as the contaminant outflow model is concerned, is that additional tugboats transit the system as a result of Gateway Pacific Terminal operations Passenger and Fishing Vessels Passenger and Fishing Vessels are composed of three vessel subtypes: cruise ships, passenger ferries, and fishing vessels greater than 60 feet in length overall (LOA). Due to significant differences in size and transit frequency, bunker capacity distributions were developed for each of these subtypes. To sample a Passenger and Fishing Vessel bunker capacity, a random number is generated to determine the Passenger and Fishing Vessel subtype (cruise, ferry, or fishing vessel), with probabilities of returning a given subtype in a given forecast year shown in Table 9, as provided by NEI as part of their vessel traffic study (Reference 1). Gateway Pacific Terminal VTS Study 17 The Glosten Associates, Inc.

24 Table 9 Passenger and Fishing Vessel Traffic Breakdown by Subtype (Data source: Northern Economics, Inc. 2013) Cruise Ship 29% 35% 38% Passenger Ferry 38% 37% 39% Fishing Vessel 33% 28% 23% Depending on which Passenger and Fishing Vessel subtype is randomly selected, one of the methods in the following sections is used to return a bunker capacity for that subtype Cruise Ships NEI provided deadweight tonnage (DWT) for every cruise ship that transited through the system in Figure 9 shows the DWT distribution for cruise ships in the system in Figure 9 Cruise Ship DWT Distribution, 2010 (Data source: Northern Economics, Inc. 2013) The yearly periodical Significant Ships (Reference 4) was used to build a database of cruise ship characteristics. DWT and bunker capacity for 23 cruise ships were plotted against each other, and a least-squared regression line was fit to the data, as shown in Figure 10. The data points circled in green are vessels that actually transited through the system in An adjustment factor of 70% was applied to the equation describing the line to match the convention in Appendix A. The regression equation is shown as Equation 16. Gateway Pacific Terminal VTS Study 18 The Glosten Associates, Inc.

25 Figure 10 Cruise Ship Bunker Capacity versus DWT (Data source: Environmental Research Consulting 2013) Bunker Capacity(Cruise Ship) = x DWT + 69,910 (gallons) 16 To obtain a cruise ship bunker capacity, the DWT of one of the cruise ships in the NEI database of 2010 cruise ship transits is randomly selected. The bunker capacity is then derived using Equation 16. It is assumed that there will be negligible changes in the DWT distribution and relationship between DWT and bunker capacity of cruise ships between 2010 and Passenger Ferries Washington State Ferries is the largest passenger and automobile ferry fleet in the US, and its vessels account for a significant portion of the traffic in the study area. Therefore, the bunker capacity distribution of its current fleet was assumed to represent the bunker capacity distribution of all passenger ferries in the study area. Vessel capacities were obtained from a phone interview with Washington State Ferries Chief Naval Architect Cotty Fay. An adjustment factor of 70% was then applied to these capacities to match the convention in Appendix A. The passenger ferry bunker capacity distribution of this representative database is summarized in Figure 11. Figure 11 Passenger Ferry Bunker Capacity Distribution Gateway Pacific Terminal VTS Study 19 The Glosten Associates, Inc.

26 To obtain a passenger ferry bunker capacity, the capacity of one of the ferries in the representative database is randomly selected. It is assumed that there will be negligible change in the capacity distribution of passenger ferries between 2010 and Fishing Vessels greater than 60 ft NEI provided a database of 6,710 recorded fishing vessel transits of vessels greater than 60 feet Length Overall (LOA) between 2008 and This database of fishing vessels was assumed to represent the entire distribution of fishing vessel LOA for the system. The fishing vessel LOA distribution of this representative database is summarized in Figure 12. Figure 12 Fishing Vessel greater than 60 feet Length Overall Distribution (Data source: Northern Economics, Inc. 2013) Various sources were used to compile a database of fishing vessel characteristics. LOA and bunker capacity for 16 fishing vessels of various sizes were plotted against each other, and a least-squared regression curve was fit to the data, as shown in Figure 13. The data points circled in green are vessels that actually transited through the system between 2008 and An adjustment factor of 70% was applied to the equation describing the curve to match the convention in Appendix A. The regression equation is shown as Equation 17. Gateway Pacific Terminal VTS Study 20 The Glosten Associates, Inc.

27 Figure 13 Fishing Vessel Bunker Capacity versus LOA Bunker Capacity(Fishing Vessel) = x LOA (gallons) 17 To obtain a fishing vessel bunker capacity, the LOA of one of the fishing vessels in the NEI database of fishing vessel transits is randomly selected. The bunker capacity is then derived using Equation 17. It is assumed that there will be negligible changes in the LOA distribution and relationship between LOA and bunker capacity of fishing vessels between 2010 and Spill Probabilities (SP f,v,i ) When an incident occurs, it is necessary to determine whether the incident results in a spill. This is accomplished by assigning a spill probability to each forecast year, vessel type, incident type (f,v,i) combination, and sampling for a spillwith a random number. ERC provided spill probabilities for the project vessel types based on vessel type, incident type, and an additional factor, which is number of hulls, either single or double. ERC also provided the probabilities of having a single or double hull for each vessel type for each forecast year. By randomly sampling the number of hulls of a vessel for a given forecast year based on hull type probability, the appropriate spill probability can then be randomly sampled for the given vessel type and incident type, thus returning the result of either a spill or no spill for a given (f,v,i) combination. The method for determining if an incident results in a spill is summarized in Figure 14. Spill probability data for each (f,v,i) combination are presented in Appendix A. Gateway Pacific Terminal VTS Study 21 The Glosten Associates, Inc.

28 Incident Occurs Randomly sample for number of hulls P(Number of Hulls) Single Hull Randomly sample for spill SP(SH) Double Hull Randomly sample for spill SP(DH) Spill No Spill Spill No Spill Figure 14 Flow diagram for determination of whether a spill occurs given an incident For incidents involving tankers and bulk carriers, there are independent probabilities of a bunker spill and a cargo spill. Each probability is randomly sampled, and if both samples result in a spill, then the total spill volume is the sum of the bunker spill and the cargo spill. 3.5 Outflow Percentage Probabilities (OP v,i ) When a spill occurs, it is necessary to determine the quantity of oil or bulk outflow. When a spill occurs for a given vessel type, incident type (v,i) combination, a cumulative distribution function (CDFs) is randomly sampled to return a percentage of outflow for the (v,i) combination. The capacity of the vessel type (v) is then randomly sampled using the appropriate method from Section 3.3. The spill volume (SV f,v,i ) for a given spill equals the product of outflow percentage and vessel capacity (Equation 4). Based on historical spill records, ERC developed CDFs of bunker and cargo outflow as percentages of vessel bunker and cargo capacities for vessel type and incident type (v,i) combinations (Appendix A). The outflow percentage curves from Appendix A that are used in the contaminant outflow model are listed in Table 10. Table 10 Outflow Percentage Curves from Appendix A used in Contaminant Outflow Model Vessel Type(s), (v) Commodity Incident Type(s), (i) Tanker Cargo Oil Transfer Error Tank Barge Cargo Oil Transfer Error Tanker, Bulker, Cargo Vessel Bunker Oil Transfer Error Tug, Passenger Vessel, and Fishing Vessel Bunker Oil Transfer Error Bulker Dry Cargo Transfer Error Single Hull Tanker Cargo Oil Impact Incidents Double Hull Tanker Cargo Oil Impact Incidents Single Hull Tank Barge Cargo Oil Impact Incidents Double Hull Tank Barge Cargo Oil Impact Incidents All Vessel Types Bunker Oil Impact Incidents Gateway Pacific Terminal VTS Study 22 The Glosten Associates, Inc.

29 Additional CDFs were developed for prediction of oil or bulk outflow for (v,i) combinations not listed in Table 10, as described in Sections through Bunker Outflow Percentage for All Vessels for Other Non-Impact Spills A bunker outflow percentage cumulative distribution function for all vessel types and the other non-impact spill category is provided by ERC in Appendix A, based on worldwide spill data. However, it was found that outflow percentages in this curve drastically exceed historical spill percentages in the study area. An alternative CDF curve was developed using historical incident data from the study area provided by ERC (presented in the Task 3 Report, Reference 2). Bunker spill volumes of other non-impact incidents from this database that resulted in spills were used to construct a bunker outflow percentage cumulative distribution function. For tankers, it was unknown whether the amount spilled was from bunkers, cargo, or both. It was therefore assumed that both were spilled, with the amount of bunkers and cargo oil spilled being proportional to the bunker and cargo oil capacity of the vessel. The historical database of 429 incidents in the study area contains no incidents of 100% bunker oil outflow (total loss). To capture the theoretical possibility of a total loss, it is assumed that the next incident will be a total loss. This results in a probability of 1 / ( ) = that the spill volume will be less than a total loss but greater than the next largest spill. Above the cumulative probability of = , cumulative probability approaches unity as outflow percentage approaches 100%. This CDF is illustrated in Figure 15. Figure 15 CDF of Bunker Outflow Percentage for All Vessels Other Non-Impact Spills Gateway Pacific Terminal VTS Study 23 The Glosten Associates, Inc.

30 3.5.2 Cargo Oil Outflow Percentage for Tankers and Tank Barges for Other Non-Impact Spills Cargo oil outflow percentage cumulative distribution functions for tanker and tank barge other non-impact spills are provided by ERC in Appendix A, based on worldwide spill data. However, it was found that outflow percentages in these curves drastically exceed historical spill percentages in the study area. An alternative CDF was therefore developed using historical incident data from the study area, as provided by ERC to develop incident rates (as presented in the Task 3 Report, Reference 2). Tanker and tank barge cargo spill volumes of other non-impact incidents from this database that resulted in spills were used to construct cargo oil outflow percentage cumulative distribution functions. For tankers, it was unknown whether the amount spilled was from bunkers, cargo, or both. It was therefore assumed that both were spilled, with the amount of bunkers and cargo oil spilled being proportional to the bunker and cargo oil capacity of the vessel. Worldwide historical spill data shows that the theoretical worst-case outflow percentage has been 12.8% for a tanker and 30% for a tank barge for spills due to other non-impact incidents (Appendix A). Tanker and tank barge spill data were aggregated to create an outflow percentage curve, but to account for different theoretical worst-case outflow percentages, the CDFs for tankers and tank barges diverge at their maximum possible outflow percentage. The historical database of 429 incidents in the study area contains no incidents of maximum theoretical cargo oil outflow for tankers or tank barges. To capture the theoretical maximum outflow, it is assumed that the next incident that will enter the database will be a maximum outflow event. This results in a probability of 1 / ( ) = that the spill volume will be less than the maximum theoretical outflow but greater than the next largest spill. Above the cumulative probability of = , cumulative probability approaches unity as outflow percentage approaches the maximum theoretical outflow. The CDFs for tanker and tank barge cargo oil other non-impact spill outflow percentages are shown in Figure 16 and Figure 17, respectively. Gateway Pacific Terminal VTS Study 24 The Glosten Associates, Inc.

31 Figure 16 CDF of Cargo Oil Outflow Percentage for Tanker Other Non-Impact Spills Figure 17 CDF of Cargo Oil Outflow Percentage for Tank Barge Other Non-Impact Spills Gateway Pacific Terminal VTS Study 25 The Glosten Associates, Inc.

32 3.5.3 Dry Cargo Outflow Percentage for Bulk Carriers for Other Non-Impact Spills A dry cargo outflow percentage cumulative distribution function for bulk carrier transfer errors and ranges for transfer error spill sizes and other non-impact spill sizes were provided by ERC based on worldwide spill data. These spill size ranges are shown in Table 11. Table 11 Amounts of Dry Cargo Spillage by Cause (Data source: ERC 2013, Appendix A) Incident Type Range (tons) Transfer Error Other Non-Impact It is assumed that the outflow percentage distribution of other non-impact spills has the same shape as the transfer error outflow percentage distribution curve. The other non-impact CDF was developed by fitting the transfer error CDF curve shape to the other non-impact range provided in Table 11. The dry cargo outflow percentage curve for bulk carrier other nonimpact spills is shown in Figure 18. Figure 18 CDF of Dry Cargo Outflow Percentage for Bulk Carrier Other Non-Impact Spills Dry Cargo Outflow Percentage for Bulk Carriers for Collision Spills It is assumed that there is a zero probability of a grounding or allision resulting in dry cargo outflow in the study area, as noted in Appendix A. Outflow due to collision is assumed to be a possibility, but there is no historical data on the outflow percentage spills due to bulk carrier collision incidents, so several assumptions are made. It is assumed that the maximum outflow is one quarter (1/4) of one cargo hold. The average tanker has five cargo holds, so the worstcase dry cargo outflow due to collision is assumed to be 1/4 x 1/5 = 1/20 = 5%. The distribution is assumed to have a cumulative probability of zero at zero percent outflow, and a cumulative probability of one at maximum cargo outflow, so an elliptical distribution shape was assumed. The dry cargo outflow percentage curve for bulk carriers in collision spills is described by Equation 18 and shown in Figure 19. Gateway Pacific Terminal VTS Study 26 The Glosten Associates, Inc.

33 P(OP v=3,i=2 ) ( OP 0.05) 18 Figure 19 CDF of Dry Cargo Outflow Percentage for Bulk Carrier Collision Spills Gateway Pacific Terminal VTS Study 27 The Glosten Associates, Inc.

34 Section 4 Results 4.1 Interpreting Cumulative Distribution Functions A selection of results is presented in the following sectionsin the form of cumulative distribution functions (CDFs). A cumulative distribution function shows the level of probability that any given quantity will not be exceeded (the probability of non-exceedance). For example, see Figure 20, Total Annual Number of Incidents. The graph reads that in 2016 without GPT, there is a probability of 0.9 that the number of incidents in Case 1 will be less than or equal to 37. Another way to interpret this is that there is a 90% chance that, in 2016 and without GPT, there will be 37 incidents or less. The inverse is also true: there is a 10% chance that there will be more than 37 incidents. Multiple CDFs can be presented in the same figure, which allows for comparisons to be made. In Figure 20, for instance, all four traffic volume cases are shown. The predicted increase in total annual incidents due to the addition of GPT at any given probability level is the difference between the appropriate curves. The graph shows that the 90 th percentile annual number of incidents increases from 37 without GPT in 2016 (Case 1) to 41 with GPT in 2016 (Case 2). 4.2 Most Likely Geographic Location Where Spills May Occur (Task 4) Total Number of Incidents Figure 20 shows the cumulative distribution function of total yearly incidents for the entire study area for each traffic volume case. Because the Poisson distribution is used to sample for the number of incidents in each scenario, the number of annual incidents is always returned as an integer value. Annual average number of incidents is a non-integer. The simulation results for 2016 show that an increase in the average number of incidents of 12% is predicted throughout the system with the addition of GPT [( )/29.03 = 12%]. For 2026, an increase in the average number of incidents of 13% is predicted. Gateway Pacific Terminal VTS Study 28 The Glosten Associates, Inc.

35 Figure 20 CDF of Total Annual Number of Incidents (All Subareas) Total Number of Spills Figure 21 shows the cumulative distribution function of total yearly spills throughout the system for each traffic volume case. The simulation results show that for 2016 an average annual increase in the number of spills of 16% is predicted throughout the system with the addition of GPT [( )/11.08 = 16%]. For 2026, a 15% increase in the average annual number of spills is predicted. Gateway Pacific Terminal VTS Study 29 The Glosten Associates, Inc.

36 Figure 21 CDF of Total Annual Number of Spills (All Subareas) Number of Spills by Subarea Table 12 shows the average number of spills (both oil and dry bulk spills) predicted by the contaminant outflow model for each subarea in 2016, with and without GPT. The simulation shows that the most likely geographic location where a spill from a GPT-calling vessel will occur is Cherry Point. The simulation predicts an average increase in number of spills for the entire study area of 16% due to the addition of GPT. Table 12 Average Number of Spills per Subarea, Cases 1 and 2 Case 1: 2016 Case 2: 2016 Change due to GPT without GPT with GPT Magnitude Percent Strait of Juan de Fuca West % Strait of Juan de Fuca East % Rosario Strait % Haro Strait and Boundary Pass % Cherry Point % Saddlebag % Guemes Channel and Fidalgo Bay % Total % Table 13 shows the average number of spills predicted by the contaminant outflow model for each subarea in 2026, with and without GPT. The simulation shows that the most likely geographic location where a spill from a GPT-calling vessel will occur is Cherry Point. The simulation predicts an average increase in number of spills throughout the system of 15% due to the addition of GPT. Gateway Pacific Terminal VTS Study 30 The Glosten Associates, Inc.

37 Table 13 Average Number of Spills per Subarea, Cases 3 and 4 Case 3: 2026 Case 4: 2026 Change due to GPT without GPT with GPT Magnitude Percent Strait of Juan de Fuca West % Strait of Juan de Fuca East % Rosario Strait % Haro Strait and Boundary Pass % Cherry Point % Saddlebag % Guemes Channel and Fidalgo Bay % Total % Table 14 and 15 show the 50 th percentile (median) of spills predicted by the contaminant outflow model for each subarea in 2016 and 2026, respectively, with and without GPT 10. Table 14 50th Percentile (median) Annual Number of Spills per Subarea, Cases 1 and 2 Case 1: 2016 Case 2: 2016 Change due to GPT without GPT with GPT Magnitude Percent Strait of Juan de Fuca West % Strait of Juan de Fuca East % Rosario Strait Haro Strait and Boundary Pass Cherry Point % Saddlebag % Guemes Channel and Fidalgo Bay % 10 Note that the sum of number of spills per subarea for each case for a given probability (percentile) does not necessarily equal the median number of spills across the entire study area for that case. This is a normal statistical phenomenon. An intuitive way to understand this phenomenon is to consider the 99 th percentile number of spills. It is highly unlikely that the 99 th percentile number of spills for each subarea will all occur in the same year, so the 99 th percentile number of spills across all subareas will intuitively be less than the sum of the 99 th percentile of each subarea. Gateway Pacific Terminal VTS Study 31 The Glosten Associates, Inc.

38 Table 15 50th Percentile (median) Annual Number of Spills per Subarea, Cases 3 and 4 Case 3: 2026 Case 4: 2026 Change due to GPT without GPT with GPT Magnitude Percent Strait of Juan de Fuca West % Strait of Juan de Fuca East % Rosario Strait Haro Strait and Boundary Pass Cherry Point % Saddlebag % Guemes Channel and Fidalgo Bay % Figure 22 through Figure 28 show the cumulative distribution functions of predicted number spills per subarea. Because the Poisson distribution is used to sample for the number of incidents in each scenario, the number of annual incidents, and thus spills, is always calculated as an integer value. Figure 22 CDF of Annual Number of Spills in Strait of Juan de Fuca West Gateway Pacific Terminal VTS Study 32 The Glosten Associates, Inc.

39 Figure 23 CDF of Annual Number of Spills in Strait of Juan de Fuca East Figure 24 CDF of Annual Number of Spills in Rosario Strait Gateway Pacific Terminal VTS Study 33 The Glosten Associates, Inc.

40 Figure 25 CDF of Annual Number of Spills in Haro Strait and Boundary Pass Figure 26 CDF of Annual Number of Spills in Cherry Point Gateway Pacific Terminal VTS Study 34 The Glosten Associates, Inc.

41 Figure 27 CDF of Annual Number of Spills in Saddlebag Figure 28 CDF of Annual Number of Spills in Guemes Channel and Fidalgo Bay Gateway Pacific Terminal VTS Study 35 The Glosten Associates, Inc.

42 4.2.4 Vessel Capacities Spill volume is a function of vessel capacity. As the compositions and utilizations of vessels in the system for the forecast years are uncertain, the distribution of vessel capacities is uncertain. For each incident in the model a vessel capacity is sampled using the methods discussed in Section 3.3. Example results from the simulation are shown for tankers, as they have the potential for the greatest oil outflow. The resultant vessel cargo oil capacity distribution for crude tankers is shown in Figure 29, and for product tankers is shown in Figure 30. Figure 29 Crude Tanker Cargo Capacity Results Figure 30 Product Tanker Cargo Capacity Results Bulk carriers are also of particular interest, so the model results for bulk carrier capacity distributions are shown in Figure 31 (bunker capacity) and Figure 32 (dry cargo capacity). The presence of Panamax and Capesize bulkers is illuminated by the bimodality of the distributions. Gateway Pacific Terminal VTS Study 36 The Glosten Associates, Inc.

43 Figure 31 Bulker Bunker Capacity Results Figure 32 Bulker Cargo Capacity Results Total Annual Oil Outflow Table 16 shows the average, 50 th percentile (median), and 95 th percentile oil outflow predicted by the model across all subareas for 2016 with and without GPT. Table 16 Predicted 2016 Total Annual Oil Outflow, Cases 1 and 2 (in gallons) Case 1: 2016 Case 2: 2016 Change due to GPT without GPT with GPT Magnitude Percent Average 22,020 25,111 3,091 14% Median 909 1, % 95th Percentile 56,876 72,097 15,221 27% Table 17 shows the average, 50 th percentile (median), and 95 th percentile oil outflow predicted by the model across all subareas for 2026 with and without GPT. Gateway Pacific Terminal VTS Study 37 The Glosten Associates, Inc.

44 Table 17 Predicted 2026 Total Annual Oil Outflow, Cases 3 and 4 (in gallons) Case 3: 2026 Case 4: 2026 Change due to GPT without GPT with GPT Magnitude Percent Average 28,384 33,237 4,853 17% Median 1,931 2, % 95th Percentile 76,890 99,394 22,504 29% Figure 33 shows the cumulative distribution function of total annual volume of oil outflow throughout the system for each traffic volume case. The predicted increase in annual oil outflow due to GPT in 2016 at any given probability level is the difference between Case 1 and Case 2 at that probability level. For instance, the 50 th percentile (median) oil outflow is predicted to increase by 320 gallons (1, = 320). The predicted increase in annual oil outflow for 2026 is found by comparing Case 3 and Case 4. Figure 33 Predicted CDF of Total Annual Volume of Oil Outflow (All Subareas) Figure 34 and Figure 35 show the predicted distributions of average annual spill volume by incident type for 2016 and 2026, respectively. Gateway Pacific Terminal VTS Study 38 The Glosten Associates, Inc.

45 Figure 34 Predicted 2016 Average Annual Oil Spill Volume by Incident Type Figure 35 Predicted 2026 Average Annual Oil Spill Volume Distribution by Incident Type Annual Oil Outflow by Subarea Table 18 shows the average annual oil outflow predicted by the model for each subarea for 2016 with and without GPT. The geographic location of greatest oil outflow from GPT-calling vessels is predicted to be Cherry Point. Table 18 Average Annual Oil Outflow per Subarea, Cases 1 and 2 (in gallons) Case 1: 2016 Case 2: 2016 Change due to GPT without GPT with GPT Magnitude Percent Strait of Juan de Fuca West 5,467 5, % Strait of Juan de Fuca East 6,410 7,443 1,033 16% Rosario Strait % Haro Strait and Boundary Pass % Cherry Point 5,287 6,611 1,324 25% Saddlebag 1,908 2, % Guemes Channel and Fidalgo Bay 2,723 2, % Total 22,020 25,111 3,091 14% Gateway Pacific Terminal VTS Study 39 The Glosten Associates, Inc.

46 Table 19 shows the average annual oil outflow predicted by the model for each subarea for 2026 with and without GPT. The geographic location of greatest oil outflow from GPT-calling vessels is predicted to be Cherry Point. Table 19 Average Annual Oil Outflow per Subarea, Cases 3 and 4 (in gallons) Case 3: 2026 Case 4: 2026 Change due to GPT without GPT with GPT Magnitude Percent Strait of Juan de Fuca West 5,971 5, % Strait of Juan de Fuca East 8,713 10,233 1,520 17% Rosario Strait % Haro Strait and Boundary Pass % Cherry Point 7,646 9,431 1,785 23% Saddlebag 1,774 2,878 1,104 62% Guemes Channel and Fidalgo Bay 3,641 4, % Total 28,384 33,237 4,853 17% Figure 36 through Figure 42 show the cumulative distribution functions of oil outflow per subarea. Figure 36 Predicted CDF of Total Annual Volume of Oil Outflow in Strait of Juan de Fuca West Gateway Pacific Terminal VTS Study 40 The Glosten Associates, Inc.

47 Figure 37 Predicted CDF of Total Annual Volume of Oil Outflow in Strait of Juan de Fuca East Figure 38 Predicted CDF of Total Annual Volume of Oil Outflow in Rosario Strait Gateway Pacific Terminal VTS Study 41 The Glosten Associates, Inc.

48 Figure 39 Predicted CDF of Total Annual Volume of Oil Outflow in Haro Strait and Boundary Pass Figure 40 Predicted CDF of Total Annual Volume of Oil Outflow in Cherry Point Gateway Pacific Terminal VTS Study 42 The Glosten Associates, Inc.

49 Figure 41 Predicted CDF of Total Annual Volume of Oil Outflow in Saddlebag Figure 42 Predicted CDF of Total Annual Volume of Oil Outflow in Guemes Channel and Fidalgo Bay Total Bulk Outflow Bulk spills are rarely recorded, so the outflow results are based on extremely limited data, as discussed in Appendix A. Table 20 shows the average, 50 th percentile (median), and 95 th Gateway Pacific Terminal VTS Study 43 The Glosten Associates, Inc.

50 percentile bulk outflow predicted by the model across all subareas for 2016, with and without GPT. Table 20 Total Yearly Bulk Outflow, Cases 1 and 2 (in cubic feet) Case 1: 2016 Case 2: 2016 Change due to GPT without GPT with GPT Magnitude Percent Average 2,416 8,792 6, % 50th Percentile th Percentile 14,832 59,472 44, % Table 21 shows the average, 50 th percentile (median), and 95 th percentile bulk outflow predicted by the model across all subareas for 2026, with and without GPT. Table 21 Total Yearly Bulk Outflow, Cases 3 and 4 (in cubic feet) Case 3: 2026 Case 4:2026 Change due to GPT without GPT with GPT Magnitude Percent Average 2,797 16,765 13, % 50th Percentile th Percentile 20,025 99,503 79, % Figure 43 shows the cumulative distribution function of total annual bulk outflow throughout the system for each traffic volume case. The predicted increase in annual bulk outflow due to GPT in 2016 at any given probability level is the difference between Case 1 and Case 2 at that probability level. For instance, the 95 th percentile (median) bulk outflow increases by two (2) cubic feet (2 0 = 2). The predicted increase in annual oil outflow for 2026 is found by comparing Case 3 and Case 4. Figure 43 Predicted CDF of Total Annual Bulk Outflow (All Subareas) Gateway Pacific Terminal VTS Study 44 The Glosten Associates, Inc.

51 4.2.8 Bulk Outflow by Subarea Table 22 shows the average bulk outflow for 2016, with and without GPT. The results show that significant increases in bulk outflow are predicted in the Strait of Juan de Fuca East, Cherry Point, Saddlebag, and Guemes Channel and Fidalgo Bay subareas. The geographic location of greatest bulk outflow from GPT-calling vessels is predicted to be Cherry Point. Table 22 Average Annual Bulk Outflow per Subarea, Cases 1 and 2 (in cubic feet) Case 1: 2016 Case 2: 2016 Change due to GPT without GPT with GPT Magnitude Percent Strait of Juan de Fuca West % Strait of Juan de Fuca East 1,156 3,041 1, % Rosario Strait Haro Strait and Boundary Pass % Cherry Point 168 2,134 1,966 1,170% Saddlebag 149 1,786 1,637 1,099% Guemes Channel and Fidalgo Bay % Total 2,416 8,792 6, % Table 23 shows the average bulk outflow for 2026, with and without GPT. The results show that significant increases in bulk outflow are predicted in the Strait of Juan de Fuca East, Cherry Point, Saddlebag, and Guemes Channel and Fidalgo Bay subareas. The geographic location of greatest bulk outflow from GPT-calling vessels is predicted to be Cherry Point. Table 23 Average Annual Bulk Outflow per Subarea, Cases 3 and 4 (in cubic feet) Case 3: 2026 Case 4: 2026 Change due to GPT without GPT with GPT Magnitude Percent Strait of Juan de Fuca West 736 1, % Strait of Juan de Fuca East 1,180 4,989 3, % Rosario Strait % Haro Strait and Boundary Pass % Cherry Point 156 4,651 4,495 2,886% Saddlebag 235 4,178 3,943 1,679% Guemes Channel and Fidalgo Bay 461 1,542 1, % Total 2,797 16,765 13, % The high average bulk commodity outflow volumes are due to infrequent collision events that can result in significant outflow volumes. Figure 44 through Figure 50 show the cumulative distribution functions of bulk outflow per subarea. Gateway Pacific Terminal VTS Study 45 The Glosten Associates, Inc.

52 Figure 44 Predicted CDF of Total Annual Bulk Outflow in Strait of Juan de Fuca West Figure 45 Predicted CDF of Total Annual Bulk Outflow in Strait of Juan de Fuca East Gateway Pacific Terminal VTS Study 46 The Glosten Associates, Inc.

53 Figure 46 Predicted CDF of Total Annual Bulk Outflow in Rosario Strait Figure 47 Predicted CDF of Total Annual Bulk Outflow in Haro Strait and Boundary Pass Gateway Pacific Terminal VTS Study 47 The Glosten Associates, Inc.

54 Figure 48 Predicted CDF of Total Annual Bulk Outflow in Cherry Point Figure 49 Predicted CDF of Total Annual Bulk Outflow in Saddlebag Gateway Pacific Terminal VTS Study 48 The Glosten Associates, Inc.

55 Figure 50 Predicted CDF of Total Annual Bulk Outflow in Guemes Channel and Fidalgo Bay 4.3 Potential Size of Contaminant Release from a GPT-Calling Vessel (Task 5) GPT-Calling Vessel Incidents by Incident Type Figure 51 and Figure 52 show the predicted distributions of GPT-calling vessel average annual number of incidents by incident type for 2016 and 2026, respectively. Figure 51 Predicted 2016 GPT-Calling Vessel Average Annual Incidents by Incident Type Gateway Pacific Terminal VTS Study 49 The Glosten Associates, Inc.

56 Figure 52 Predicted 2026 GPT-Calling Vessel Average Annual Incidents by Incident Type GPT-Calling Vessel Oil Spill Size When a spill does occur, the size of the spill is dependent on the capacity of the vessel and the percentage of the vessel capacity that is spilled. The addition of GPT will introduce new Panamax and Capesize bulk carriers and tugboats into the system, as described in Sections and , respectively. The distribution of oil spill sizes of GPT-calling bulk carriers and tugs are shown in Figure 53 and Figure 54, respectively. No change in spill size distribution is anticipated between 2016 and 2026, as the vessel size distribution of GPT-calling vessels is assumed to remain constant. Figure 53 CDF of GPT-Calling Bulker Oil Spill Size Gateway Pacific Terminal VTS Study 50 The Glosten Associates, Inc.

57 Figure 54 CDF of GPT-Calling Tug Oil Spill Size GPT-Calling Vessel Oil Spill Distribution by Incident Type Figure 55 and Figure 56 show the predicted distributions of GPT-calling vessel average annual oil spill volume by incident type for 2016 and 2026, respectively. Figure 55 Predicted 2016 GPT-Calling Vessel Average Annual Oil Spill Volume Distribution by Incident Type Gateway Pacific Terminal VTS Study 51 The Glosten Associates, Inc.

58 Figure 56 Predicted 2026 GPT-Calling Vessel Average Annual Oil Spill Volume Distribution by Incident Type GPT-Calling Vessel Bulk Spill Size The distribution of bulk spill sizes of GPT-calling bulk carriers is shown in Figure 57. No change in spill size distribution is predicted between 2016 and 2026, as the vessel size distribution of GPT-calling vessels is assumed to remain constant. Figure 57 CDF of GPT-Calling Bulker Bulk Spill Size Gateway Pacific Terminal VTS Study 52 The Glosten Associates, Inc.

59 4.3.5 GPT-Calling Vessel Bulk Spill Distribution by Incident Type Figure 58 and Figure 59 show the predicted distributions of average GPT-calling vessel average annual dry cargo spill volume by incident type for 2016 and 2026, respectively. It is assumed that bulk outflow will not occur in the study area as a result of grounding or allision. This assumption is discussed in Appendix A. Figure 58 Predicted 2016 GPT-Calling Vessel Average Annual Dry Cargo Spill Volume Distribution by Incident Type Figure 59 Predicted 2026 GPT-Calling Vessel Average Annual Dry Cargo Spill Volume Distribution by Incident Type Gateway Pacific Terminal VTS Study 53 The Glosten Associates, Inc.

60 Section 5 Summary of Results The numbers of annual oil and dry cargo spills in the study area are predicted to increase with the addition of the Gateway Pacific Terminal due to increased bulk carrier and tugboat traffic in the system. The increase in spills due to GPT vessels is predicted to be concentrated in eastern Strait of Juan de Fuca, Cherry Point, Saddlebag, and Guemes Channel and Fidalgo Bay. The geographic location where the most GPT vessel spills are predicted to occur is Cherry Point. Total annual oil and bulk outflow in the study area is predicted to increase with the addition of GP. The increased numbers of spills result from the increased bulk carrier and tugboat traffic in the system due to the introduction of GPT-calling vessel traffic. The magnitude of the increase in total annual oil and bulk outflow is predicted to be proportional to the quantity and size of the vessel traffic introduced into the system by GPT, which are Panamax and Capesize bulk carriers and tugboats. The increase in oil and bulk outflow due to GPT vessels is predicted to be concentrated in eastern Strait of Juan de Fuca, Cherry Point, Saddlebag, and Guemes Channel and Fidalgo Bay. The geographic location where the most oil outflow from GPT-calling vessels is predicted to occur is Cherry Point. A majority of GPT-calling vessel incidents are predicted to be in the other non-impact incidents category. However, the average annual oil spill volumes are predicted to be fairly evenly distributed between allisions, groundings, and other non-impact incidents, with less volumes from collisions and bunker errors. Average annual bulk spill volumes are predicted to be dominated by collisions, which are much rarer but have higher bulk outflow consequences than transfer errors and other nonimpact incident types. It is assumed that bulk outflow will not occur in the study area as a result of grounding or allision. This assumption is discussed in Appendix A. The model predicts much higher average oil outflow than median oil outflow from GPT-calling vessels due to rare predicted occurrences of very oil large spills. Oil spills due to collisions, allisions, and groundings are predicted to occur infrequently, but result in large oil outflows when they do occur. Spills due to bunker, transfer, and other non-impact incidents are predicted to occur more frequently, but result in small oil outflows when they do occur. The model predicts much higher average dry cargo outflow than median dry cargo outflow from GPT-calling vessels due to rare predicted occurrences of very large dry cargo spills. Dry cargo spills due to collisions are predicted to occur infrequently, but result in large dry cargo outflows when they do occur. Other non-impact incidents are predicted to occur more frequently, but result in small dry cargo outflows when they do occur. Gateway Pacific Terminal VTS Study 54 The Glosten Associates, Inc.

61 Appendix A Characterization of Casualty Consequences (Task 5), Gateway Pacific Terminal Vessel Traffic and Risk Assessment Study, Environmental Research Consulting, 14 March Gateway Pacific Terminal VTS Study The Glosten Associates, Inc.

62 Gateway Pacific Terminal Vessel Traffic and Risk Assessment Study Characterization of Casualty Consequences (Task 5) Prepared by Dagmar Schmidt Etkin, PhD Environmental Research Consulting 41 Croft Lane Cortlandt Manor, NY March 2013

63 Contents Contents... 2 List of Tables... 3 List of Figures... 4 Purpose... 5 Terminology... 5 Nomenclature... 5 Equation Variables... 6 Calculations for the Probability of Spillage... 7 Probability of Oil Cargo Spillage... 7 Probability of Bunker Spillage, P(BS) Special Issue of Tanker Bunker and/or Cargo Spillage Combined Probabilities of Hull Type and Spillage Probability Probabilities of Spillage in Tanker Transfer Operations Calculations for Vessel Oil Capacity Approaches to Estimating Oil Cargo Capacity for Tankers Development of Formula for Actual Amount of Oil on Fully Laden Tanker Oil Cargo Capacity (K o ) for Tankers and Tank Barges Bunker Capacity (K b ) for Tankers, General Cargo Vessels, Bulk Carriers, Dry Cargo (Grain) Capacity (K d ) Calculation of Spill Volume Probability Distributions Oil Cargo Oil Cargo Spill Volume Distributions Oil Outflow Probability for Tankers in Impact Accidents Oil Outflow Probability for Tankers in Other, Non-Impact Incidents Oil Outflow Probability for Tank Barges in Impact Accidents Calculation of Spill Volume Probability Distributions Bunker Fuel Bunker Spill Volume Distributions from Impact Accidents Oil Outflow for Tanker Oil-Cargo Transfer Incidents Spill Volumes from Tank Barge Oil Cargo Transfer Incidents Bunker Outflow from Transfer Errors in General Cargo Vessels, Tankers, and Bulk Carriers Bunker Outflow from Transfer Errors in Other Vessels Calculation of Spill Volume Probability Distributions Dry Cargo Dry Cargo (Grain) Spill Volume Distributions Dry Cargo Sweeping as an Input References ERC GPT Study: Characterization of Likely Accidents and Consequences

64 List of Tables Table A: Equation Variables... 6 Table 1: Variables for Probability of Oil Cargo Spillage... 8 Table 2: Cargo Spill Probabilities for Tankers and Tank Barges... 8 Table 3: Probabilities of Hull Types for Tankers and Tank Barges by Year... 9 Table 4: Combined Oil Cargo/Hull Spill Probabilities for Tankers and Tank Barges by Year Table 5: Variables for Probability of Bunker Spillage for All GPT VTS Vessels Table 6: Bunker Spill Probabilities for All GPT VTS Vessels Table 7: Application of Double-Hulls for Bunker Tank Percentages to Future Projections Table 8: Combined Bunker Spillage and Hull Type Probabilities by Year Table 9: Stowage Factors for Dry Cargo on Bulk Carriers Table 10: Variables for Probability Distributions of Oil Cargo Spillage Volume Table 11: Oil Outflow Probability for Double-Hull Tankers in Impact Accidents Table 12: Oil Outflow Probability for Single-Hull Tankers in Impact Accidents Table 13: Oil Outflow Probability for Single or Double-Hull Tankers in Other, Non-Impact Incidents Table 14: Oil Outflow Probability for Single-Hull Tank Barges in Impact Accidents Table 15: Oil Outflow Probability for Double-Hull Tank Barges in Impact Accidents Table 16: Outflow Probability: Single/ Double-Hull Tank Barges in Other, Non-Impact Incidents Table 17: Bunker Outflow Probability from All Vessel Impact Accidents Table 18: Bunker Outflow Probability from All Vessel Other Non-Impact Incidents Table 19: Oil Cargo Outflow Probability from Tanker Transfer Errors Table 20: Oil Cargo Outflow Probability from Tank Barge Transfer Errors Table 21: Bunker Outflow Probability from Tankers, Bulk Carriers, and General Cargo Vessels due to Transfer Errors during Bunkering Operations Table 22: Bunker Outflow Probability from Other Vessels: Transfer Errors during Bunkering Table 23: Dry Cargo Incidents (Spills and Potential Spills) in US Waters Table 24: Details of Bulk Carrier Dry Cargo Incidents in US Waters ( ) Table 25: Dry Cargo Incidents Reported in US Waters Table 26: Causes of Dry Cargo Incidents in US Waters Table 27: Bulker Incident Rates Using Data Table 28: Bulker Incident Rates Using Data Table 29: Dry Cargo Incidents (US ) Analysis Table 30: Amounts of Dry Cargo Spillage by Cause Table 31: Dry Cargo Outflow Probability from Transfer Errors ERC GPT Study: Characterization of Likely Accidents and Consequences

65 Table 32: Dry Cargo Outflow Probability from Other, Non-Impact Errors Table 33: Amounts of Dry Cargo Inputs from Washdown Operations List of Figures Figure 1: Areas of GPT Study Zone with Potential for Dry Cargo Sweepings Inputs Figure 2: Size Distribution of Dry Cargo Washdown Inputs Figure 3: Cumulative Probability Distribution of Dry Cargo Washdown Amounts ERC GPT Study: Characterization of Likely Accidents and Consequences

66 Gateway Pacific Terminal Vessel Traffic and Risk Assessment Study Study Characterization of Casualty Consequences Purpose The purpose of this report is to provide necessary data and algorithms for the development of the Monte Carlo traffic risk modeling effort associated with the Gateway Pacific Terminal Traffic and Risk Assessment Study, as well as for the fulfillment of Task 5 Characterization of Casualty Consequences. Terminology Nomenclature Actual bunker fuel load: the physical capacity of the vessel s bunker fuels reduced to 70% as that is the largest actual amount of bunker fuel typically carried on a vessel in actual practice. 1 Actual oil cargo load: the physical capacity of the vessel s cargo tanks reduced to 98% (or 93.6% of deadweight tonnage) as that is the largest actual amount of oil cargo typically carried on a vessel. Allision: an incident in which a moving object strikes a stationary object (e.g., when a vessel strikes a pier or another vessel that is anchored or docked). Bunker hull type: the type of hull (single or double) on the bunker fuel tanks of a general cargo vessel, bulk carrier, or tanker. Bunker: includes all types of bunker fuel (Bunker A, Bunker B, Bunker C, No. 6 fuel oil, intermediate fuel oil IFO), as well as diesel fuel (No. 2 fuel oil), and marine gas oil. Bunkering: the transfer of bunker fuels from one vessel to another or from a stationary facility (storage tank) to a vessel. Cargo hull type: the type of hull (single or double) on the cargo tanks of a tanker or tank barge Collision: an incident in which two moving vessels strike each other. Crude tanker: a tank ship (tanker) that is between 67,000 and 125,000 DWT and usually carries crude oil rather than refined products. Cumulative probability 2 : the probability that a value (e.g., oil outflow of a certain percentage) will be less than or equal to that value. For example, if the cumulative probability of an oil outflow of 80% of the oil cargo is 95%, it means that there is a 95% chance that an oil outflow will be of 80% oil cargo or less. There is only a 5% chance that the oil outflow percentage will be larger. This is similar to the term percentile. The 95 th percentile spill is that spill volume for which there is only a 5% chance that the spill will be larger. Dry cargo: bulk commodities carried by bulk carriers, including coal, grain, sand, stone, etc. Deadweight tonnage (DWT): the weight (in long tons 3 ) that a vessel can carry, including oil (or other) cargo, bunker fuel, stored water, ballast (when not cargo-laden), crew, and miscellaneous minor contributors to weight. On an oil tanker, 97.5% of DWT is available for oil cargo, 2% for bunker fuel, and 0.5% for stored water. 1 The derivation of this adjustment is described later in this report. 2 This is distinct from an alternative use of this term in statistical practice which means the probability of multiple events occurring at the same time. 3 1 short ton = 2,000 pounds (lbs); 1 long ton = 2,240 lbs; 1 metric ton (tonne) = 2,205 lbs; 1 long ton = metric ton (tonne). 5 ERC GPT Study: Characterization of Likely Accidents and Consequences

67 GPT VTS Vessels: vessels for which there are sufficient traffic data and that are therefore included in the analysis of vessel traffic risk. Impact accident: an incident involving a collision, allision, or grounding. Incident: an occurrence with a vessel that leads to the potential for spillage of oil or dry cargo or actual spillage. Oil transfer: any movement of oil cargo and/or bunkers from one vessel to another or from a stationary facility (storage tank) to a vessel. Other Vessels: this category includes only GPT VTS vessels not included in the other categories of tanker, bulker, tank barge, or general cargo cruise ships, regularly-scheduled ferries, tugboats (tugboats and towboats), and fishing vessels of 60 feet or larger. Other, Non-Impact Error: the category of vessel incidents that excludes impact accidents (allisions, collisions, and groundings) and transfer errors, but includes a variety of other causes, such as equipment failures, operations errors, structural failures, sinking, mechanical failures, intentional discharges, unintended discharges and leakages, and unknown causes. Outflow percentage: the percentage of the adjusted cargo or bunker capacity on board the vessel that will be released or spilled with a particular incident. Product tanker: a tank ship (tanker) that is between 22,000 and 67,000 DWT and usually, but not necessarily, carries refined products rather than crude oil. Articulated tank barges (ATBs) and integrated tank barges (ITBs) are in the product tanker size category. R 2 : the coefficient of determination is a value between 0 and 1 that describes how closely a regression curve (derived equation) fits the data. Based on the proportion of data variability that is accounted for in the statistical model (derived equation), a high R 2 means that the equation fits well and will more accurately predict future outcomes. Spill volume: the amount of spillage (for oil, this is in gallons; for dry cargo, this can be in cubic feet or a weight measurement). Tankers: tankers are tank ships that carry oil (crude or refined product) as cargo, including integrated tug barges (ITBs) and articulated tug barges (ATBs). Tank Barge: a barge carrying oil cargo that may or may not be attached to a tug (towboat or tugboat) at the time of the incident. The analytical results apply only to the tank barge (oil spillage, probabilities) and not to the tug. Tugs are separately accounted for under the category Other Vessel. Equation Variables Table A: Equation Variables Variable Description Potential Values P(x) Probability of event x CS, cargo spillage BS, bunker spillage CH x, cargo hull type BH x, bunker hull type SV x, spill volume O x, outflow CS Cargo spillage - BS Bunker spillage - 6 ERC GPT Study: Characterization of Likely Accidents and Consequences

68 Table A: Equation Variables Variable Description Potential Values Vx Vessel of type x Values for x: t, tanker pt, product tanker ct, crude tanker tb, tank barge b, bulk carrier g, general cargo o, other vessel y Year y = 1 for year 2010, y = 2 for year 2011,, y = 17 for year 2026 I x Incident with cause x Values for x: c, collision a, allision g, grounding cag, all impact accidents combined o, other, non-impact t, transfer error ot, all non-impact incidents combined CH x Cargo hull type Values for x: d, double hull s, single hull BH x Bunker hull type Values for x: d, double hull s, single hull DWT Deadweight tonnage - GRT Gross registered tonnage - Length Vessel length (ft) - K x SV x Vessel capacity (actual load) Spill volume Values for x: o, oil cargo b, bunker fuel d, dry cargo Values for x: o, oil cargo b, bunker fuel d, dry cargo O x Outflow percentage 4 o, oil cargo b, bunker fuel Values for x: d, dry cargo Calculations for the Probability of Spillage The probability of spillage is the probability that given an incident there will be a spill of any volume (from very small to very large). This probability does not indicate the volume of spillage, which is calculated in a separate step. The probability of cargo spillage is related to the variables of vessel type, incident cause, and hull type. Since the probability of hull type will change over time, it will be necessary to incorporate a year-dependent probability of hull type for both oil cargo spillage and bunker spillage. Probability of Oil Cargo Spillage The relevant variables for determining the probability of oil cargo spillage, P(CS), are shown in Table 1 and Equations 1 and 2. Oil cargo spillage can only occur from tank vessels tankers and tank barges. 4 Percentage of vessel adjusted capacity. 7 ERC GPT Study: Characterization of Likely Accidents and Consequences

69 Table 1: Variables for Probability of Oil Cargo Spillage Variable Product Tanker, V pt Vessel Type, V x Crude Tanker, V ct Tank Barge, V tb Cargo Hull, CH 5 Single Hull, CH s Double Hull, CH d Allision, I a Collision, I c Incident Cause, I x Grounding, I g Other, Non-Impact, I o Transfer Error, I t Values P( cs) f ( V, I, CH ) P( CH ) f ( y) x x x x [1, 2] The spill probabilities for each vessel/incident cause/hull combination are shown in Table 2. The probabilities of spillage in this table for collisions, allisions, and groundings are derived from outflow models on tankers and tank barges that were developed by naval architects and engineers working on behalf of the International Maritime Organization (IMO) 6 to estimate the probability of spillage given various types of vessel accidents, as well as a more recent study that conducted regression analyses on US Coast Guard vessel casualty data to investigate the effect of double-hulls on spillage rates. 7 The spillage rates for other, non-impact errors and transfer errors are derived from data in National Research Council (NRC) studies and studies conducted by ERC for the US Army Corps of Engineers. 8 Table 2: Cargo Spill Probabilities for Tankers and Tank Barges 9 Vessel Type Incident Cause Hull 10 Cargo Spill Probability in Incident, P(CS) Collision (I c ) Single (CH s ) 0.68 Double (CH d ) 0.15 Single (CH Allision (I a ) s ) 0.68 Double (CH d ) 0.15 Product Single (CH Tanker Grounding (I g ) s ) 0.91 Double (CH (V pt ) d ) 0.18 Single (CH Other, Non-Impact Error (I o ) s ) Double (CH d ) Transfer Error (I t ) Double (CH d ) Single (CH s ) Collision (I c ) Single (CH s ) 0.81 Double (CH d ) 0.19 Single (CH Crude Allision (I a ) s ) 0.81 Double (CH Tanker d ) 0.19 Single (CH (V ct ) Grounding (I g ) s ) Double (CH d ) Other, Non-Impact Error (I o ) Single (CH s ) 0.40 Double (CH d ) Single or double hull on cargo tanks for tankers (tank ships) and tank barges. Note that articulated tank barges (ATBs) and integrated tank barges (ITBs) are considered tankers. 6 Rawson 1998; NRC 1998; NRC 2001; IMO Yip et al. 2011b. 8 NRC 1998; NRC 2001; Etkin et al Based on Yip et al. 2011b; Rawson 1998; NRC 1998; NRC 2001; IMO 1995; Etkin et al For tank vessels, hull refers to cargo hull. For all other vessels hull refers to bunker tank hull. 8 ERC GPT Study: Characterization of Likely Accidents and Consequences

70 Table 2: Cargo Spill Probabilities for Tankers and Tank Barges 9 Vessel Type Incident Cause Hull 10 Cargo Spill Probability in Incident, P(CS) Transfer Error (I t ) Single (CH s ) Double (CH d ) Collision (I c ) Double (CH d ) Single (CH s ) Allision (I a ) Single (CH s ) 0.76 Double (CH d ) 0.13 Tank Barge Single (CH (V tb ) 11 Grounding (I g ) s ) Double (CH d ) Other, Non-Impact Error (I o ) Double (CH d ) Single (CH s ) Transfer Error (I t ) Single (CH s ) 0.92 Double (CH d ) 0.92 The probabilities of a tanker or tank barge having a single or double hull are show in Table 3 and in Equations 3 and 4. Table 3: Probabilities of Hull Types for Tankers and Tank Barges by Year Years (y) Double Hull, P(CHd) Single Hull, P(CHs) 2010 (y = 1) (y = 2) (y = 3) (y = 4) (y = 5) (y = 6) (y = 7) (y = 8) (y = 9) (y = 10) (y = 11) (y = 12) (y = 13) (y = 14) (y = 15) (y = 16) (y = 17) P CH E y E y E y E y y y R ( s) ( d ) P CH E y E y E y E y y y R [3, 4] Note that for the category Tank Barge, the incident rate relates only to the tank barges themselves, which may or may not occur while there is a tug (towboat or tugboat) associated with tank barge. Incidents involving tugs are included under Other Vessels. In those cases, the tug (towboat or tugboat) may be operating independently or have a tank barge or other barge attached to it. 9 ERC GPT Study: Characterization of Likely Accidents and Consequences

71 The probabilities of spillage shown in Table 2 and the probabilities of hull type in Table 3 are combined in Table 4 as in Equation 5. In Table 4, the spill probabilities are apportioned by hull type (single or double). P( CS) [ P( CH ) P( CS) ] [ P( CH ) P( CS ) ] [5] y s y CH d y CH Table 4: Combined Oil Cargo/Hull Spill Probabilities for Tankers and Tank Barges by Year Year Vessel Incident Cause Type Product Tanker (V pt ) Crude Tanker (V ct ) Tank Barge (V tb ) s Collision Allision Grounding Other, Non-Impact Transfer Error Collision Allision Grounding Other, Non-Impact Transfer Error Collision Allision Grounding Other, Non-Impact Transfer Error Probability of Bunker Spillage, P(BS) The relevant variables for determining the probability of bunker 13 spillage, P(BS), are shown in Table 5 and Equations 6 and 7. A transfer error of bunker fuel is also called a bunkering error. Table 5: Variables for Probability of Bunker Spillage for All GPT VTS Vessels Variable Values Tanker, V t Tank Barge, V tb Vessel Type, V x Bulk, Vb General Cargo, V g Other, V o Bunker Hull, BH 14 Single Hull, BH s Double Hull, BH d Allision, I a Collision, I c Incident Cause, I x Grounding, I g Other, Non-Impact, I o Transfer Error, I t d 12 The probabilities for the years 2015 through 2026 are the same due to the complete programmatic replacement of single hulls with double hulls for tankers. 13 The term bunker is used for all fuel types Bunker A, Bunker B, Bunker C, Intermediate Fuel Oil (IFO), diesel (No. 2 fuel), gasoline, etc. 14 Single- or double hull on bunker tanks for all vessels including tankers, except for tank barges, which do not have bunker tanks. 10 ERC GPT Study: Characterization of Likely Accidents and Consequences

72 P( BS) f ( V, I, BH ) P( BH ) f ( y) x x x x [6, 7] The probabilities of bunker spillage by vessel type, cause, and hull type are shown in Table 6. The hull configuration for bunker tanks is also independent of the hull configuration of the cargo tanks. That is, there can be a double-hull on the cargo tanks and only a single-hull on the bunker tanks. The schedules for implementation of double-hulls on cargo and bunker tank are different (see Tables 3 and 7). The probabilities in Table 6 are based on bunker tank outflow modeling conducted for IMO 15 and studies conducted on US oil spills. 16 Tankers, bulk carriers, and general cargo vessels have been assigned the same bunker spill probabilities as the previous analyses conducted on bunker spillage probabilities do not differentiate between different vessel types. For the vessels in the other vessels category, there is no difference between spillage probabilities in double and single hulled tanks. These vessels are not covered under the regulations that will mandate double hulls on bunker tanks. There will therefore be no difference in double and single hulls for these vessels. Tank barges have been assigned probabilities of 0.00 for bunker spillage as they do not have bunker fuels on board. Table 6: Bunker Spill Probabilities for All GPT VTS Vessels 17 Vessel Type Incident Cause Hull Bunker Spill Probability Collision (I c ) Single (BH s ) Double (BH d ) Allision (I a ) Double (BH d ) Single (BH s ) Tankers (V t ) 18 Grounding (I g ) Double (BH d ) Single (BH s ) Other, Non-Impact Error (I o ) Double (BH d ) Single (BH s ) Transfer Error (I t ) Double (BH d ) Single (BH s ) Collision (I c ) Double (BH d ) Single (BH s ) Allision (I a ) Single (BH s ) 0.00 Double (BH d ) 0.00 Tank Barge Single (BH (V tb ) 19 Grounding (I g ) s ) Double (BH d ) Other, Non-Impact Error (I o ) Double (BH d ) Single (BH s ) Transfer Error (I t ) Single (BH s ) 0.00 Double (BH d ) 0.00 Bulk Carriers Single (BH Collision (I (V b ) c ) s ) 0.05 Double (BH d ) Michel and Winslow 1999, 2000; Barone et al Etkin and Michel 2003; Herbert Engineering et al Based on Etkin and Michel 2003; Michel and Winslow 1999, 2000; Barone et al. 2007; Herbert Engineering et al. 2003; Barone et al Product and crude tankers are treated as a combined category only as there are no differences in bunker spillage probabilities between the two vessel sub-categories. 19 Note that since tank barges do not carry bunker fuel, the probability for bunker fuel spillage is zero. The probability for the tug (towboat or tugboat) towing the tank barge is separately handled in the Other Vessel category. 11 ERC GPT Study: Characterization of Likely Accidents and Consequences

73 Table 6: Bunker Spill Probabilities for All GPT VTS Vessels 17 Vessel Type Incident Cause Hull Bunker Spill Probability Allision (I a ) Single (BH s ) Double (BH d ) Grounding (I g ) Double (BH d ) Single (BH s ) Other, Non-Impact Error (I o ) Double (BH d ) Single (BH s ) Transfer Error (I t ) Double (BH d ) Single (BH s ) Collision (I c ) Double (BH d ) Single (BH s ) Allision (I a ) Single (BH s ) 0.05 Double (BH d ) 0.02 General Cargo Single (BH Grounding (I Vessels (V g ) g ) s ) Double (BH d ) Other, Non-Impact Error (I o ) Double (BH d ) Single (BH s ) Transfer Error (I t ) Double (BH d ) Single (BH s ) Collision (I c ) Double (BH d ) Single (BH s ) Allision (I a ) Single (BH s ) 0.05 Double (BH d ) 0.05 Other Vessels 20 Single (BH Grounding (I (V o ) g ) s ) Double (BH d ) Other, Non-Impact Error (I o ) Double (BH d ) Single (BH s ) Transfer Error (I t ) Single (BH s ) 0.92 Double (BH d ) 0.92 The probabilities of vessels in Table 6 having a single or double hull are show in Table 7. The exceptions are tank barges, which do not have bunker tanks, 21 and vessels in the Other Vessels category, which will not likely have double hulls within the study period through Table 7: Application of Double-Hulls for Bunker Tank Percentages to Future Projections Years (y) Probability of Double Hull (BH d ) Probability of Single Hull (BH s ) 2010 (y = 1) (y = 2) (y = 3) (y = 4) (y = 5) (y = 6) (y = 7) (y = 8) (y = 9) (y = 10) (y = 11) (y = 12) (y = 13) (y = 14) (y = 15) Includes only GPT VTS vessels not in other categories of tanker, bulk, tank barge, or general cargo. 21 This is referring only to the tank barge and not its associated tug. 12 ERC GPT Study: Characterization of Likely Accidents and Consequences

74 Table 7: Application of Double-Hulls for Bunker Tank Percentages to Future Projections Years (y) Probability of Double Hull (BH d ) Probability of Single Hull (BH s ) 2025 (y = 16) (y = 17) The above probabilities can also be expressed as the fitted Equations 8 and (The same R 2 applies to both equations.) P( BH ) y P( BH ) y R 2 d s [8, 9] Special Issue of Tanker Bunker and/or Cargo Spillage For tankers only, the spill of oil cargo is a separate event from the spillage of bunker fuel. There are separate probabilities that a bunker spill will occur with an impact and that cargo spill will occur with an impact. They are independent events. For all incident causes, there is a higher probability of oil cargo spillage than for bunker spillage. Transfer errors are treated differently, as there are two separate events for bunkering operations and cargo transfer operations. Combined Probabilities of Hull Type and Spillage Probability The probabilities of bunker spillage shown in Table 6 and the probabilities of bunker hull type in Table 7 are combined in Table 8 as in Equation 10 with the spill probabilities are apportioned by hull type (single or double). P( BS) [ P( BH ) P( BS) ] [ P( BH ) P( BS ) ] [10] y s y BH d y BH The results in Table 8 are only for impact accidents (collisions, allisions, and groundings) and other, nonimpact accidents but not transfer error incidents involving tankers. Transfer errors involving tankers are handled separately. Probabilities of Spillage in Tanker Transfer Operations There are two types of oil transfer operations that occur with tankers oil cargo transfers and bunker fuel transfers (bunkering). If there is a reported error (i.e., a transfer error ) in a cargo transfer operation, there is a probability of 0.92 that there will be a spill. Likewise, if there is an error during bunker spill operations, there is also a probability of 0.92 that there will be a spill. The events of oil cargo transfer errors and bunker transfer errors themselves have different probabilities. In the course of 16 years ( ), there have been 27 transfer error incidents involving tankers in the GPT study area. 23 One of those incidents involved bunker spillage during bunkering operations. The other 26 incidents involved the spillage of oil cargo during transfer operations. For both oil cargo transfer- s d 22 Based on R 2, the coefficient of determination is a value between 0 and 1 that describes how closely a regression line (derived equation) fits the data. Based on the proportion of data variability that is accounted for in the statistical model (derived equation), a high R 2 means that the equation fits well and will more accurately predict future outcomes. 23 Etkin ERC GPT Study: Characterization of Likely Accidents and Consequences

75 error related incidents and bunker transfer-related incidents there appeared to be no issue of both bunker fuel and oil cargo spilling during transfers. This is because oil cargo transfer operations are generally conducted separately from bunkering operations. The relative rate of bunkering errors to cargo transfer errors can be applied to the overall incident rate. That is, 1/27 or 3.7% of transfer errors in tankers results in the spillage of bunker fuel and 96.3% of transfer errors results in the spillage of oil cargo. It is important to note that the probability distribution functions of oil volume spilled from tankers due to oil cargo transfer operations and bunker transfer operations are different. 14 ERC GPT Study: Characterization of Likely Accidents and Consequences

76 Table 8: Combined Bunker Spillage and Hull Type Probabilities by Year Vessel Cause Collision Allision Tanker (V t ) Grounding Other, Non-Impact Collision Allision Tank Grounding Barge (V tb ) Other, Non-Impact Transfer Collision Allision Bulk (V b ) Grounding Other, Non-Impact General Cargo (V g ) Other (V o ) Transfer Collision Allision Grounding Other, Non-Impact Transfer Collision Allision Grounding Other, Non-Impact Transfer ERC GPT Study: Characterization of Likely Accidents and Consequences

77 Calculations for Vessel Oil Capacity Since the spillage or outflow is determined as a percentage of the amount of oil on board the vessel as a function of its volumetric capacity (either for oil cargo or bunker fuel), the capacity of each vessel in the system must be estimated based on vessel type and size, typically deadweight tonnage (DWT). Approaches to Estimating Oil Cargo Capacity for Tankers In general, there is a distinction between the vessel s true capacity, i.e., the volumetric capacity of its cargo tanks and the actual amount of oil that is on board a fully-laden tanker in practice. Two professors on shipping practices, Niko Wijnolst, Chairman of the European Network of Maritime Clusters, and Tor Wergeland stated in their textbook on shipping 24 that, in practice, loading rates for crude oil carriers vary from 80% to 97% of the deadweight tonnage (DWT) for a fully laden tanker. The authors state that utilization in practice hardly exceeds 95%, but could be as low as 65%. (Note that this is a fully laden tanker not one that has off-loaded a portion of its cargo at one port and proceeds to the next with less than the original amount.) In two significant dynamic collision risk modeling studies, the figure of 91% is utilized. 25 For outflow modeling purposes, IMO uses 98% of volumetric capacity of the cargo tanks. 26 The calculations for outflow are based on these values. In official records of a vessel s cargo capacity (e.g., Clarkson Register, Lloyds Register, American Bureau of Shipping) the cargo capacity of a tanker is reported as 98% of the volumetric capacity of its cargo tanks. Professors Wijnolst and Wergeland 27 write that, typically, 2.5% of the deadweight tonnage of a vessel is used for storage of water and bunker oil, with bunker oil assumed to be about 2%. 28 This would mean then that if one was using 95% deadweight tonnage maximum loading value, % could be subtracted for the bunker fuel and oil, giving a high value of 92.5% DWT that is actually oil cargo. Note also that even if one is using the 98% full tank, when one is calculating the amount of oil on board the vessel from its DWT, one has to subtract 2.5% of the DWT for bunker fuel and stored water. Development of Formula for Actual Amount of Oil on Fully Laden Tanker As a practical matter, for the Monte Carlo simulation and other aspects of the current study, the oil on board of tankers, which represents the worst-case discharge potential for the vessels, must be derived as a function of some measure of vessel size. Deadweight tonnage is the most appropriate measure of vessel size for these purposes. Deadweight tonnage (DWT) of a tanker is the total weight that a vessel can carry. This includes the oil cargo, bunker fuels, stored water, ballast (when the vessel is in ballast rather than laden), and miscellaneous other smaller loads, including the crew. On an oil tanker, clearly the vast majority of DWT is taken up by the oil cargo when the tanker is laden. Using the rule of thumb of Wijnolst and Wergeland 24 Wijnolst and Wergeland Eide et al. 2007; Behrens et al National Research Council 1998, Wijnolst and Wergeland This also bears out in analyses of known bunker capacities and deadweight tonnages as in Etkin and Michel Based on Eide et al and Behrens et al ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

78 (1997) that 2% of DWT is bunker fuel, and 0.5% of DWT is stored water, this leaves 97.5% of DWT for oil cargo alone. The actual percentage may be somewhat less depending the contribution of the other minor factors of crew and miscellaneous loads. 30 The remaining 97.5% DWT is then the theoretical maximum capacity of the tanker for oil cargo. This can then be further broken down depending on the assumption of capacity. This would need to be applied to any formulae or algorithms that are working directly with the capacity of tanks rather DWT. If one begins with the assumption of 98% full cargo tanks, 31 this needs to be converted to a percentage of DWT as in Equations to estimate the actual cargo load: K ( long tons) DWT o K ( tonnes) DWT o K ( gallons) DWT o [11, 12, 13] Using this formula, the WCD for the 125,000 DWT tanker is million gallons. While the regulatory basis for limiting the maximum amount of oil cargo transported through Puget Sound is clearly based on a limit of 125,000 DWT as per federal regulations 32, that is by the tanker s tonnage, from Department of Ecology s perspective, the state authorities operate on the assumption of 33 million gallons being the maximum oil that can be transported. 33 The same approach is applied to tank barges for which there are no known algorithms for calculating actual cargo loading. Oil Cargo Capacity (K o ) for Tankers and Tank Barges Equation 14 is for estimating oil cargo capacity. 34 Where K o = actual tank ship cargo load (in gallons) 35 DWT = deadweight tonnage of the vessel K DWT [14] o 30 It is assumed that this is less than 0.5% since it is not even mentioned in the calculations of Wijnolst and Wergeland (1997) and others (Behrens et al. 2003; Eide et al. 2007). 31 National Research Council 1998, CFR (Code of Federal Register) b 33 This is based on multiple discussions with Ecology for various contracted projects, as well as their own literature. For example, the 21 September 2006 Ecology Press Release concerning the commencement of rescue tug service states barges or tankers can carry up to 33 million gallons of oil. 34 Based on Etkin 1999; Etkin and Michel 2003; Etkin et al. 2009; French-McCay et al. 2008; State of WA JLARC Note that tones have been converted to gallons using a standard conversion of 294 gallons/tonne, which is for most oils. Lighter oils (e.g., diesel) will have more gallons per tonne, whereas heavier oils (e.g., Bunker C) will have fewer gallons per tonne as they are more dense. 17 ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

79 For tankers transiting Puget Sound there is a federal regulatory limit of 125,000 DWT for tankers, 36 which translates to a 35.7-million-gallon capacity per Equation 14. The cargo capacity of a tank barge may be calculated with Equation 14. However, if the actual carrying capacities are available for the tank barges for all current and future tank barges transiting Puget Sound, these data may be used. It is important to adjust the carrying capacity to account for only 98% capacity (or 93.5% DWT) as this is generally the maximum amount of oil cargo that is on board the vessel rather than 100% of the capacity. Bunker Capacity (K b ) for Tankers, General Cargo Vessels, Bulk Carriers, Again, for oil outflow modeling purposes only, IMO uses 98% of volumetric capacity as the maximum assumed bunker load on a vessel. 37 In actual practice, however, the expert advice has been that bunker tanks are never more than 70% full in practice. 38 The recommended formulae for estimating bunker capacity for GPT study vessels are Equations 15 and 16.. These formulae were derived from regressions of known bunker volumes (corrected to 70%) for the vessel types tankers and general cargo vessels. The Glosten Associates has developed its own equation for the purpose of estimating bunker capacity in vessels as the regression developed from bunker vessels in the incident data did not include vessels of a capacity above 44,000 DWT. K ( V ) 5.086DWT 106,924 R b 2 2 b( b) ,172 2 t K V DWT DWT R [15, 16] Where K b (V t ) = bunker tank capacity of tankers (in gallons) adjusted for 70% capacity 39 K b (V g ) = bunker tank capacity of general cargo vessels 40 (in gallons) DWT = deadweight tonnage CFR (Code of Federal Register) b 37 Barone et al. 2007; Michel and Winslow 1999, This is the value that was used in the US Army Corps study (Etkin and Michel 2003), as well as studies for Puget Sound (Etkin 2001; Etkin et al. 2009; French-McCay et al. 2008) and other parts of the US (Etkin 2003, 2003). ERC has not seen any other mention of the actual percentage of bunker tank capacity that is filled with bunker fuel. These assumptions were applied to all of the aforementioned studies on the Puget Sound (Etkin 2001; Etkin et al. 2009; French-McCay et al. 2008; Etkin et al. 2005; French-McCay et al. 2005, 2006a, 2006b, 2006c, 2006d), as well as US-wide studies (Etkin 2002, 2003). 39 In other studies conducted by ERC with Herbert Engineering, Inc., adjustments were made to bunker tank capacity as it is common practice that bunker tanks are rarely filled to more than 70% capacity even when full (Etkin and Michel 2003). 40 Based on data available for container ships. 18 ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

80 Dry Cargo (Grain) 41 Capacity (K d ) The capacity for dry cargo in bulk carriers depends on the configuration of the vessel (e.g., numbers of holds). The amount of dry cargo that each vessel will carry when fully loaded will vary by the commodity being carried. The reason for this is that the weight of the cargo is of importance in determining the optimal and safe loading of the vessel. Because the density of dry bulk commodities varies considerably (Table 9), the volume of actual loading relative to the physical volumetric capacity of the bulk carrier holds will vary. There are different stowage factors (SF) for different commodities, as shown in Table 9. Stowage factor is the inverse of the density of the cargo (Equations 15 17): t Density 3 m 1 SF Density SF m t 3 [15, 16, 17] Table 9: Stowage Factors for Dry Cargo on Bulk Carriers Specific Weight Stowage Factor Typical % Commodity long Capacity tons/m 3 kg/m 3 long m 3 / m 3 / ft 3 / ton/ft 3 long ton metric ton long ton Filled 42 Anthracite Coal Bituminous Coal Lignite Coal Petroleum Coke Potash , Wheat/Millet % Corn % Barley Taconite , % Wood Chips Sulfur Stone , Sand , In the shipping industry, the term grain is used as a proxy for dry cargo as a means of representing the bulk volume that can be accommodated in bulker holds. It does not necessarily imply grain as wheat, barley, rice, etc. This measure can be used for other commodities, including coal. The International Code for the Safe Carriage of Grain in Bulk is commonly called the International Grain Code was adopted by the IMO Maritime Safety Committee by resolution MSC.23(59). It applies to ships regardless of size, including those of less than 500gt, engaged in the carriage of grain in bulk, to which part C of chapter VI of the 1974 SOLAS Convention, as amended, applies (A 1.1). 42 Filled in only for those commodities for which data were available. 43 Taconite is a type of iron ore that contains more than 15% iron interlayered with quartz, chert, or carbonate. 19 ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

81 For example, for iron ore, cargo holds typically hold 25% of their capacity. For grains, the holds will typically hold 87.5% of their capacity. 44 The space required to stow one ton of wood chips requires six times the space required for one ton of iron ore. Loading pattern is also an important factor for bulk carriers. The patterns vary based on the nature of the cargo and whether there is a homogeneous load (all one commodity) or heterogeneous (more than one type of commodity being carried) and the pattern of unloading planned. There are three typical loading patterns utilized on bulk carriers 45, as shown in Figures 1 3. The homogeneous loading pattern (Figure 1) is used for lighter cargoes like grain or coal. The alternate hold loading pattern (Figure 2) is most often used with high-density cargoes to raise the center of gravity. The block loading pattern (Figure 3) is most typically employed when the vessel is partly loaded. Figure 1: Homogeneous Loading Pattern for Bulk Carrier Figure 2: Alternate Hold Loading Pattern for Bulk Carrier Figure 3: Block Loading Pattern for Bulk Carrier The loading pattern and precise measurements of cargo loading for each bulk carrier type and dry cargo commodity combination are complex 46 and beyond the scope of this project. For the purposes of the GPT VTS, it is important to estimate the dry cargo load of the bulk carriers that may be transiting the study area in future so that estimates of potential spillage might be made. The volumetric dry cargo capacity (as grain capacity) can be calculated from deadweight tonnage as in Equation 18: K 1.4DWT [18] Where K d = volumetric bulker capacity (as grain capacity) in cubic meters. d This capacity assumes that the vessel is completely full with grain (with a stowage factor of 1.4 m 3 /metric ton). The actual load on board a bulk carrier when fully loaded will vary with the bulk commodity. 44 Personal communications with bulk carrier industry experts and on-line communities. 45 Ship Structure Committee ( Lotsberg See IACS ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

82 Initially, GPT 47 will be shipping mostly low-sulfur, low ash coal, which corresponds most closely with lignite coal with an average density long tons per cubic foot, giving it a stowage factor of 47.7 cubic feet per long ton or 1.35 cubic meters per metric ton (tonne), which is slightly lower than the grain capacity, as in Equation 19: K 1.35DWT [19] coal This capacity should then be further adjusted to take into account that the cargo holds are rarely completely full but rather about 84% full 48 when the vessel is filled with coal with a homogeneous holding pattern (Figure 1) as in Equations 20 and 21: adjustedk adjustedk coal coal 0.84 (1.35 DWT ) 1.134DWT [20, 21] Calculation of Spill Volume Probability Distributions Oil Cargo If a spill of oil cargo does occur, it will involve a volume (from very small to very large) based on the type of vessel, including hull type, and the accident cause. Based on historical data, a distribution of probabilities is assigned to the spill volumes. Generally, smaller spills are more common and very large spills are rare. Oil Cargo Spill Volume Distributions The relevant variables for determining the probability distributions of oil cargo spillage volume, P(Svc), are shown in Table 10. Oil cargo spillage can only occur from tank vessels tankers and tank barges. The probabilities of hull type for tankers and tank barges by year were shown in Table 3. Table 10: Variables for Probability Distributions of Oil Cargo Spillage Volume Variable Values Tanker, V Vessel Type, V t x Tank Barge, V tb 49 Single Hull, CH Cargo Hull, CH s x Double Hull, CH d Allision, I a Collision, I c Incident Cause, I x Grounding, I g Other, Non-Impact, I o Transfer Error, I t 47 Based on information in the Gateway Pacific Terminal Vessel Traffic and Risk Assessment Study Request for Proposal 31 March The 84% was derived by multiplying the 87.5% known for grains (SF of 1.4) with the ratio of the coal stowage factor and the grain stowage factor. The coal is slightly heavier than the grain and would thus fill up somewhat less of the hold to achieve the same weight. 49 Single or double hull on cargo tanks for tankers (tank ships) and tank barges. Note that articulated tank barges (ATBs) and integrated tank barges (ITB)s are considered tankers. 21 ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

83 Cargo oil spill volume is the percentage outflow of the cargo (O o ) times the oil cargo capacity (Co), as in Equation 22. SVo Oo K o [22] The percentage outflow (O o ) is a function of the vessel type (V x ), incident type (I x ), and cargo hull (CH x ), as in Equations 23 and 24. O f ( V, I, CH ) o x x x P( CH ) f ( y) x [23, 24] Oil Outflow Probability for Tankers in Impact Accidents Oil outflow probabilities differ somewhat by hull type for tankers. The probability distribution of percentage of outflow for double-hull tankers involved in impact accidents is as shown in Table 11. The probability distribution of percentage of outflow for single-hull tankers involved in impact accidents is as shown in Table 12. The percentage oil outflow probabilities are based on international studies of the amount of oil actually spilled compared with the reported amount of oil cargo on the tanker, 50 which was in turn, adjusted to derive the same based probability density function of spill volumes based on 98% of volumetric cargo capacity rather than the original known cargo amounts as per Equation 14. The approach was verified by existing oil outflow models developed for IMO. 51 Table 11: Oil Outflow Probability for Double-Hull Tankers in Impact Accidents % Actual Cargo Outflow Probability P(O o ) 52 Cumulative Probability 0.002% % % % % % % % % Table 12: Oil Outflow Probability for Single-Hull Tankers in Impact Accidents % Actual Cargo Outflow Probability P(O o ) 52 Cumulative Probability 0.002% % % % % % % % % Etkin 2001; Etkin 2002; Etkin 2003; Etkin and Neel 2001; Etkin and Michel 2003; Etkin et al Rawson 1998; Yip et al. 2011b; NRC 1998; NRC Based on Etkin 2001; Etkin 2002; Etkin 2003; Etkin and Neel 2001; Etkin and Michel 2003; Etkin et al. 2009; Rawson 1998; Yip et al. 2011b; NRC 1998; NRC ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

84 Outflow modeling has demonstrated that the volumes of outflows for the very largest incidents would be reduced by 50% with double hulls. 52 For Puget Sound, the largest tanker spill volume of 34 million gallons from a single-hulled tanker would result in spillage of 17 million gallons from a double-hulled tanker. The smaller spillage volumes would not be affected. Note also that this is independent of the probability of spillage occurring with an impact accident. Double hulls on tankers accomplish two things reduction of the probability of any spillage occurring in the first place, and reduction of the volume of spillage for the very largest incidents by 50%. This is not the case for double hulls on bunker tanks, for which there is a reduction in the probability of spillage occurring in an impact accident, but there is no reduction in spillage volume with large incidents. 52 Oil Outflow Probability for Tankers in Other, Non-Impact Incidents The hull type does not affect the probability of non-impact accident outflows. The probability of percentage outflow for single-hull and double-hull tankers involved in Other, Non-Impact incidents is as shown in Table 13. There is no difference between single- and double-hulled tankers for these types of incidents. The percentage oil outflow probabilities are based on international studies of the amount of oil actually spilled compared with the reported amount of oil cargo on the tanker, 53 which was in turn, adjusted to derive the same based probability density function of spill volumes based on 98% of volumetric cargo capacity rather than the original known cargo amounts. Table 13: Oil Outflow Probability for Single or Double-Hull Tankers in Other, Non-Impact Incidents % Actual Cargo Outflow Probability P(O o ) 54 Cumulative Probability 0.012% % % % % % Oil Outflow Probability for Tank Barges in Impact Accidents The probability of percentage of outflow for single-hull tank barges 55 involved in impact accidents (collisions, allisions, and groundings) is as shown in Table 14. The percentage oil outflow probabilities are based on international studies of the amount of oil actually spilled compared with the reported amount of oil cargo on the tanker, 56 which was in turn, adjusted to derive the same based probability density function of spill volumes based on 98% of volumetric cargo capacity rather than the original known cargo amounts. Table 14: Oil Outflow Probability for Single-Hull Tank Barges in Impact Accidents % Actual Cargo Outflow Probability P(O o ) 57 Cumulative Probability 0.001% Etkin 2001; Etkin 2002; Etkin 2003; Etkin and Neel 2001; Etkin and Michel 2003; Etkin et al Based on Etkin and Michel 2003; Etkin 2001; Etkin Note that the oil outflow only comes from the tank barge itself. Tugs (towboats and tugboats) are separately tracked under Other Vessels. 56 Etkin 2001; Etkin 2002; Etkin 2003; Etkin and Neel 2001; Etkin and Michel 2003; Etkin et al Based on Etkin and Michel 2003; Etkin 2001; Etkin ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

85 Table 14: Oil Outflow Probability for Single-Hull Tank Barges in Impact Accidents % Actual Cargo Outflow Probability P(O o ) 57 Cumulative Probability 0.01% % % % % % % % % % The probability distribution of percentage of outflow for double-hull tank barges involved in impact accidents is as shown in Table 15. Table 15: Oil Outflow Probability for Double-Hull Tank Barges in Impact Accidents % Actual Cargo Outflow Probability P(O o ) 58 Cumulative Probability 0.001% % % % % % % % % % % The probability distribution of percentage of outflow for single-hull and double-hull Tank Barges 59 involved in involved in Other Non-Impact incidents is as shown in Table 16. There is no difference between single- and double-hulled tank barges for these types of incidents. The percentage oil outflow probabilities are based on international studies of the amount of oil actually spilled compared with the reported amount of oil cargo on the tanker, 60 which was in turn, adjusted to derive the same based probability density function of spill volumes based on 98% of volumetric cargo capacity rather than the original known cargo amounts. Table 16: Outflow Probability: Single/ Double-Hull Tank Barges in Other, Non-Impact Incidents % Actual Cargo Outflow Probability P(O o ) 61 Cumulative Probability % % % % Based on Etkin 2001; Etkin 2002; Etkin 2003; Etkin and Neel 2001; Etkin and Michel 2003; Etkin et al. 2009; Rawson 1998; Yip et al. 2011b; NRC 1998; NRC Note that the oil outflow only comes from the tank barge itself. Tugs (towboats and tugboats) are separately tracked under Other Vessels. 60 Etkin 2001; Etkin 2002; Etkin 2003; Etkin and Neel 2001; Etkin and Michel 2003; Etkin et al Etkin 2001, 2002, ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

86 Table 16: Outflow Probability: Single/ Double-Hull Tank Barges in Other, Non-Impact Incidents % Actual Cargo Outflow Probability P(O o ) 61 Cumulative Probability 0.01% % % % % % % % % % Calculation of Spill Volume Probability Distributions Bunker Fuel If a spill of bunker fuel does occur, it will involve a volume (from very small to very large) based on the type of vessel, including hull type, and the accident cause. Based on historical data, a distribution of probabilities is assigned to the spill volumes. Generally, smaller spills are more common and very large spills are rare. Bunker Spill Volume Distributions from Impact Accidents Note that in the modeling, for tankers, it is assumed that the volume of spillage is for either bunker fuel or oil cargo, not a summation of both, as the probability of both spilling simultaneously is very small. Spill volume is derived by multiplying the oil outflow percentage times the capacity as in Equation 25. SVb Ob K b [25] The probability distribution of percentage of outflow for all vessels (except tank barges, which have no bunker fuel) involved in impact accidents is as shown in Table 17. Note that there is no difference between double- and single-hulled vessels with regard to oil outflow percentage. The probability that a spill will occur is reduced by the presence of a double hull. This is addressed in the spill probability algorithms. The percentage oil outflow probabilities are based on international studies of the amount of oil actually spilled compared with the estimated or reported amount of bunker tanks in vessels at their full (i.e., 70% full) capacity. 62 The approach was verified by oil outflow modeling conducted for IMO. 63 Table 17: Bunker Outflow Probability from All Vessel Impact Accidents % Actual Bunker Outflow Probability P(O b ) 64 Cumulative Probability 0.01% % % % % Etkin 2001; Etkin 2002; Etkin 2003; Etkin and Neel 2001; Etkin and Michel 2003; Etkin et al Michel and Winslow 1999, 2002; Barone et al. 2007; Yip et al. 2011a. 64 Etkin and Michel 2003; Etkin 2001; Etkin 2002; Herbert Engineering et al. 2003; Michel and Winslow 1999, 2002; Barone et al. 2007; Yip et al. 2011a. 25 ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

87 Table 17: Bunker Outflow Probability from All Vessel Impact Accidents % Actual Bunker Outflow Probability P(O b ) 64 Cumulative Probability 10% % % % % The probability distribution of percentage of outflow for all vessels (except tank barges, which have no bunker fuel) involved in Other Non-Impact Incidents is as shown in Table 18. The percentage oil outflow probabilities are based on international studies of the amount of oil actually spilled compared with the reported amount of oil cargo on the tanker, 65 which was in turn, adjusted to derive the same probability density function of spill volumes based on 70% of volumetric bunker capacity rather than the original known bunker amounts. Table 18: Bunker Outflow Probability from All Vessel Other Non-Impact Incidents % Actual Bunker Outflow Probability P(O b ) 66 Cumulative Probability 0.001% % % % % % % % % % % % % % Oil Outflow for Tanker Oil-Cargo Transfer Incidents The probability distribution of percentage of outflow for tankers involved in transfer error incidents is as shown in Table 19. Note that there is no difference between double- and single-hulled tankers with regard to oil outflow from transfer errors. The percentage oil outflow probabilities are based on international studies of the amount of oil actually spilled compared with the reported amount of oil cargo on the tanker, 67 which was in turn, adjusted to derive the same based probability density function of spill volumes based on 98% of volumetric cargo capacity rather than the original known cargo amounts. Table 19: Oil Cargo Outflow Probability from Tanker Transfer Errors % Actual Bunker Outflow Probability P(O o ) 68 Cumulative Probability % % % Etkin 2001; Etkin 2002; Etkin 2003; Etkin and Neel 2001; Etkin and Michel 2003; Etkin et al Etkin 2001, 2002, Etkin 2001; Etkin 2002; Etkin 2003; Etkin and Neel 2001; Etkin and Michel 2003; Etkin et al Based on analyses conducted in Etkin 2001, 2002, 2003; Etkin and Neel 2001; Etkin ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

88 Table 19: Oil Cargo Outflow Probability from Tanker Transfer Errors % Actual Bunker Outflow Probability P(O o ) 68 Cumulative Probability % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % Spill Volumes from Tank Barge Oil Cargo Transfer Incidents The probability distribution of percentage of outflow for tankers and tank barges involved in transfer error incidents is as shown in Table 20. The percentage oil outflow probabilities are based on international studies of the amount of oil actually spilled compared with the reported amount of bunker tanks in vessels. 69 Table 20: Oil Cargo Outflow Probability from Tank Barge Transfer Errors % Actual Cargo Outflow Probability P(O o ) 70 Cumulative Probability % % % Etkin 2001; Etkin 2002; Etkin 2003; Etkin and Neel 2001; Etkin and Michel 2003; Etkin et al Based on analyses conducted in Etkin 2001, 2002, 2003; Etkin and Neel 2001; Etkin ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

89 Table 20: Oil Cargo Outflow Probability from Tank Barge Transfer Errors % Actual Cargo Outflow Probability P(O o ) 70 Cumulative Probability 0.004% % % % % % % % % Bunker Outflow from Transfer Errors in General Cargo Vessels, Tankers, and Bulk Carriers The probability distribution of percentage of outflow for general cargo vessels, tankers, and bulk carriers involved in transfer error incidents during bunkering (fueling) operations 71 is as shown in Table 21. The percentage oil outflow probabilities are based on international studies of the amount of oil actually spilled compared with the reported amount of bunker tanks in vessels. 72 Table 21: Bunker Outflow Probability from Tankers, Bulk Carriers, and General Cargo Vessels due to Transfer Errors during Bunkering Operations % Actual Bunker Outflow Probability P(O b ) 73 Cumulative Probability % % % % % % % % % % % % % % % % Bunker Outflow from Transfer Errors in Other Vessels The probability distribution of percentage of outflow for other vessels involved in transfer error incidents during bunkering (fueling) operations is as shown in Table 22. The percentage oil outflow probabilities are based on international studies of the amount of oil actually spilled compared with the reported amount of bunker tanks in vessels Also referred to as bunkering errors. 72 Etkin 2001; Etkin 2002; Etkin 2003; Etkin and Neel 2001; Etkin and Michel 2003; Etkin et al Based on analyses conducted in Etkin 2001, 2002, 2003; Etkin and Neel 2001; Etkin Etkin 2001; Etkin 2002; Etkin 2003; Etkin and Neel 2001; Etkin and Michel 2003; Etkin et al ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

90 Table 22: Bunker Outflow Probability from Other Vessels: Transfer Errors during Bunkering % Actual Bunker Outflow Probability P(O b ) 75 Cumulative Probability 0.001% % % % % % % % % % % % % Calculation of Spill Volume Probability Distributions Dry Cargo Spillage of dry cargo from bulkers has not been regularly or systematically recorded by Washington Department of Ecology or by the US Coast Guard for the years for Washington State or any other part of the US. Reports to the National Response Center (and the US Coast Guard) for the entire US has only been sporadic. Much of this is due to lack of information on the impacts of these spills and the fact that spillage of dry cargo residues has long been considered to be part of the routine operations of loading, unloading, and transport of bulk dry cargo on bulker vessels. This has led to irregular reporting of incidents. In US Coast Guard (and National Response Center) records there have been occasional reports of spills of dry bulk cargo from either barges or bulkers, as shown in Table 23. The fact that so few incidents were reported does not necessarily indicate that these spills do not occur, but rather that when they do they are considered minor or inconsequential because of the nature of the commodities and the lack of stringent regulations requiring reporting, or the lack of knowledge of existing regulations. Table 23: Dry Cargo Incidents (Spills and Potential Spills) in US Waters Year Number of Incidents Barges Bulkers Total Based on analyses conducted in Etkin 2001, 2002, 2003; Etkin and Neel 2001; Etkin Includes inland navigable waterways. From ERC spill databases. 29 ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

91 Table 23: Dry Cargo Incidents (Spills and Potential Spills) in US Waters Year Number of Incidents Total Total In all of the US, there have been only 53 dry cargo incidents with spillage or potential spillage reported in 26 years (or 2 per year), of which only 16 were from bulkers rather than freight barges (or 0.62 per year). Details of these incidents are shown in Table 24. Table 24: Details of Bulk Carrier Dry Cargo Incidents in US Waters ( ) Date Location Waterway Vessel Name GRT DWT Cause 77 Material Amount Spilled (lbs) Navigable American 11/30/1986 Lorain, OH 15,396 36,171 Transfer Coke Dust 20 Waters NEC Mariner Sault Ste Marie, H. Lee Taconite 5/7/1987 St. Marys River 21,815 35,019 Transfer 1 MI White Dust Henry 5/10/1989 Buffalo, NY Buffalo River 7,041 13,910 Transfer Coal Dust 5 Steinbrenner Unknown 5/14/1989 Duluth, MN Lake Superior Unknown Coal Dust 3 (Underway) Traverse City, William R. 8/28/1991 Lake Michigan MI Roesch 9,639 19,800 Transfer Coal Dust 50 Navigable Kaye E. Taconite 7/30/1992 Dearborn, MI 11,949 25,345 Transfer Waters NEC Barker Dust 2 Thousand St. Lawrence 5/27/1995 Tadoussac 20,634 28,800 Unknown Taconite Islands, NY River (Underway) Dust 500 6/11/2001 Bristle, PA Delaware River Unknown Transfer Cement 1 San Francisco, San Francisco 9/26/2001 CA Bay Andros 35,447 64,843 Transfer Lime 1,250 Newport News, The unknown 12/13/2002 James River 22,354 37,939 Transfer Coal Dust VA Thornhill (minor) New Orleans, Mississippi unknown 7/10/2004 Sanaga 17,784 28,215 Dumping Coal Dust LA River (minor) 77 While no incidents were directly attributed to transfer errors the incidents noted as transfer were reported to have been at a dock loading/unloading facility at the time of the incident and it is therefore assumed that there may have been transfer operations going on at the time of the reported incident or prior to the report of evidence of material in the water. 30 ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

92 Table 24: Details of Bulk Carrier Dry Cargo Incidents in US Waters ( ) Date Location Waterway 10/7/ /9/2004 3/10/2005 9/25/2006 Port Everglades, FL Dutch Harbor, AK San Francisco, CA New Haven, CT Vessel Name GRT DWT Cause 77 Material Amount Spilled (lbs) Atlantic Ocean Unknown Grounding Cement 0 Pacific Ocean San Francisco Bay Long Island Sound Selendang Ayu Unknown 39,775 72,937 Grounding Equipment Failure (Underway) Grain (Soybeans) 132,000,000 Concrete Dust Barkald 28,924 49,463 Sinking Coal 4/14/2010 Portland, OR Columbia River Hellenic Sea 36,448 65,434 Transfer Coal Dust unknown (minor) unknown (minor) unknown (minor) The commodities spilled (or potentially spilled) have included those shown in Table 25. Fifty-eight percent of the incidents involved coal. One of the bulker incidents occurred in Washington waters, though outside of the area of the GPT study zones. In 2010, there was a spill of an unknown amount of coal at the United Harvest facility on the Columbia River in Portland, Oregon 78. The spill (or potential spill) causes have included those shown in Table 26. Table 25: Dry Cargo Incidents Reported in US Waters Commodity Barges Bulkers Total Cement Coal Concrete Fertilizer Grains Lime Limestone Sand Sulfur Taconite (Iron Ore) Wood Chips Total Table 26: Causes of Dry Cargo Incidents in US Waters Cause Barges Bulkers Total Number % Number % Number % Impact Accident 0 0% 2 13% 2 4% Other Non-Impact Error 32 86% 5 31% 37 70% Transfer Error 5 14% 9 56% 14 26% Total % % % Of the bulker incidents that occurred during the study period time ( ), two (20%) involved groundings. Considering the data over the years , the percentage of groundings is 13%. Note 78 Also reported as Kalama, Washington. 79 Includes inland navigable waterways. From ERC spill databases. 80 Includes inland navigable waterways. From ERC spill databases. 31 ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

93 that there were no reported incidents involving collisions or allusions that resulted in any spillage or potential spillage of dry cargo. There have been many collisions, allisions, and groundings of bulker vessels in US waters over the course of the 16-year time period, with an annual average of about 40 incidents per year that cause spillage of at least 50 gallons bunkers 81, while there have been only two groundings of bulkers have caused any spillage of dry cargo. That would translate to a probability of for a bulker grounding to cause spillage of dry cargo. There have been no incidents of collisions or allisions causing spillage or potential spillage of dry cargo. With no groundings of bulkers in Washington waters, this leaves a probability of zero for spills of dry cargo. Another more conservative approach is presented in Tables 27 and 28. These are estimates of probabilities of spillage by cause per transit based on the number of incidents that occurred in the US over 16 years and the approximate number of port visits. (It can be assumed that each port visit is one day.) It is conservatively estimated that all of the incidents described above caused some spillage even though the records do not necessarily indicate that. Table 27: Bulker Incident Rates Using Data GPT Cause Incidents in 16 years Estimated Transit- Days in 16 years 82 Probability of Incident Involving Spillage of Dry Cargo 83 Per Transit-Day Impact Accident 2 155, Other Non-Impact Error 4 155, Transfer Error 4 155, Table 28: Bulker Incident Rates Using Data GPT Cause Incidents in 16 years Estimated Transit- Days in 26 years 84 Probability of Incident Involving Spillage of Dry Cargo 85 Per Transit-Day Impact Accident 2 252, Other Non-Impact Error 5 252, Transfer Error 9 252, The probability of a dry cargo spill is independent of the probability of a bunker spill in bulk carriers. The amount of spillage from the historical data is uncertain in most cases because of the lack of follow-up after the initial reports, and difficulties in estimating the amount of spillage after material sank into the water. In one case, 500 pounds of taconite (iron ore) was reported to have spilled from a bulker in the St. Lawrence River in 1995 due to unknown causes. The paucity and inaccuracy of data on spill amounts of dry cargo presents a challenge for estimating a probability distribution of spill volumes. There are also no outflow models to estimate the amount of dry cargo that would be spilled in different types of accidents. The incidents in Table 24 have been broken down further to estimate the volume and percentage outflow for each incident. Incidents for which there was no vessel information were eliminated from the analysis. Minor spillages were conservatively (i.e., likely overestimating) assumed to be 20 lbs. 81 Etkin Based on estimated 9,700 port visits (transits) annually (from Etkin 2010). 83 This is independent of the probability that there will be an incident that could cause a bunker oil spill. 84 Based on estimated 9,700 port visits (transits) annually (from Etkin 2010). 85 This is independent of the probability that there will be an incident that could cause a bunker oil spill. 32 ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

94 Table 29: Dry Cargo Incidents (US ) Analysis Amount Spilled Vessel Name DWT Material Metric Cubic Cubic % Cargo Pounds % DWT Tons Meters Feet Holding American Mariner 36,171 Coke % % H. Lee White 35,019 Taconite % % Henry Steinbrenner 13,910 Coal % % William R. Roesch 19,800 Coal % % Kaye E. Barker 25,345 Taconite % % Tadoussac 28,800 Taconite % % Andros 64,843 Lime 1, % % The Thornhill 37,939 Coal % % Sanaga 28,215 Coal % % Selendang Ayu 72,937 Grain 132,000,000 59,928 77,307 2,745,562 82% 97% Barkald 49,463 Coal % % Hellenic Sea 65,434 Coal % % The typical amounts of reported (and verified) spillage by cause are shown in Table 30. Dry cargo spills are generally very small unless the entire vessel breaks up and/or sinks. The largest incident reported, that of the Selendang Ayu, involved a drift grounding in a storm in which the vessel broke up, spilled its bunkers, and released the bulk of its soybean cargo. There have been no reported incidents involving dry cargo spillage (or potential spillage) during collisions or allisions. Table 30: Amounts of Dry Cargo Spillage by Cause 86 Dry Cargo Amount Spilled 87 Short Tons Long Tons Metric Tons (Tonnes) Transfer Error Other Non-Impact Error Dry Cargo (Grain) Spill Volume Distributions Dry cargo spill volume is derived by multiplying the bulk outflow percentage times the capacity as in Equation 26. SVd Od adjustedk coal [26] Note that dry cargo capacities are adjusted for reduced capacity and assume coal as the commodity. The probability distribution of percentage outflow for transfer errors based on the extremely limited data available is shown in Table 31. Table 31: Dry Cargo Outflow Probability from Transfer Errors % Cargo Outflow Probability P(O d ) Cumulative Probability % % % % Includes inland navigable waterways. From ERC spill databases. 87 One short ton = 2,000 lbs. One long ton = 2,240 lbs. One metric ton = 2,204.6 lbs. 33 ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

95 A similar distribution could be used for other, non-impact error spills, albeit with the largest spillage adjusted to the largest spillage of that category as shown in Table 30. This would create a probability distribution of outflow as shown in Table 32. Table 32: Dry Cargo Outflow Probability from Other, Non-Impact Errors % Cargo Outflow Probability P(O d ) Cumulative Probability % % % % For collisions and allisions, there are no data whatsoever on which to base an outflow percentage probability distribution. There have been no recorded incidents in US waters between 1985 and 2010 in which dry cargo spilled during a collision or allision incident involving a bulk carrier, that is, not once in 26 years or 255,200 transit days for bulkers. There have been collisions and allisions in which bunker fuel spilled from a bulk carrier in US waters, but none in which any dry cargo spillage was reported. During 1995 through 2010 there was one collision and one allision involving bulk carriers in the GPT study area. Neither of these incidents resulted in the spillage of any bunker fuel, nor the spillage of any dry cargo. For groundings, there have been two reported incidents in which there was spillage or potential spillage of dry cargo. One involved no spillage (of cement), the other involved the spillage of 97% of its soybean cargo. The bulk carrier Selendang Ayu incident, previously mentioned, is noted for the spillage of bunker oil (and much less so for the spillage of soybeans) in a drift grounding in the Aleutian Islands after engine failure (and power loss) during high winds and heavy seas. The winds were reported to be Beaufort force 7 to 11 (near gale to violent storm), averaging force 9 (47 54 mph). This would create waves of feet up to 52 feet when at Beaufort In addition, due to the weather conditions and remoteness of the incident location, rescue tugs were unsuccessful in assisting the vessel. These are highly unlikely conditions for most of the GPT study area with regard to weather conditions. According to studies on Puget Sound conducted to determine the benefits of tug escorts for vessels, 89 there exists the potential for drift groundings within the GPT study area, but tug escorts would help to reduce the probability of drift groundings for all >300 GRT vessels by 65%. 90 The fact that only one grounding incident in which any dry cargo spillage was recorded in 26 years in all US waters or about once in an estimated 255,200 bulker transit days ( per transit day) demonstrates the highly unlikely probability of an event occurring in the GPT study area at all, let alone spilling a majority of its dry cargo. The fact that the weather conditions that led to the catastrophic drift grounding of the Selendang Ayu and its breakup and release of dry cargo in addition to its bunkers is highly unlikely in the GPT study area further supports that assumption. 88 Based on National Transportation Safety Board (NTSB) US Coast Guard 1999; The Glosten Associates et al US Coast Guard ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

96 Dry Cargo Sweeping as an Input Note: This section is added as the research team considers it to be important information on dry cargo inputs to the environment and an important part of the environmental impact assessment of GPT as relates to dry cargo shipping. While the inputs from dry cargo sweeping and washing of decks are more part of routine operations than spillage per se, it is likely to be of concern to stakeholders in the Puget Sound area. Implementation and enforcement on regulations on dry cargo sweeping inputs (as part of the EPA Vessel General Permits or as a separate guidance from the US Coast Guard) are likely to be implemented in Washington in the future. With bulkers there are continuous inputs of dry cargo due to sweeping and washing of decks that are part of standard operating procedures. 91 The only place in which these practices are at all curtailed are in the nearshore areas (three miles from the shore) 92, especially in the Great Lakes. 93 In the Puget Sound, this would mean potential inputs of dry cargo sweepings in the areas shown in Figure 4. Dry cargo sweeping practices are also somewhat controlled by the EPA Vessel General Permit regulations. In the proposed 2013 Vessel General Permit (Section 2.2.1) requires that vessel operations minimize the introduction of on-deck debris into deck washdown and runoff discharges. It is difficult to estimate what this really means in practice. Conservatively, one should assume that there is still dry cargo washdowns occurring with all trips, with an average of 300 pounds (0.15 tons) of cargo being washed down with each transit, but that the largest inputs are not occurring due to better practices. A comprehensive study on dry cargo inputs conducted in 2003 found the levels of input with each bulker transport as shown in Table 33 and Figures 5 and Table 33: Amounts of Dry Cargo Inputs from Washdown Operations Commodity Commodity Inputs (lbs) per Transit Minimum Input Maximum Input Mean Input 95 Median Input 96 Iron 10 44, Coal/Coke 10 66, Stone 10 2, Limestone 10 1, Salt Grain 10 4, Sand Millscale/Slag 10 2, Potash Gypsum 10 1, All Commodities 10 66, PMG and ERC Except for limestone carriers and those vessels for whom it would be a financial hardship not to do cargo sweepings in the nearshore 3-mile zone CFR 151 Docket No. USCG [Federal Register Volume 77, Number 146 (Monday, July 30, 2012) 94 PMG and ERC Mean input is the average input (i.e., the total amount of input divided by the number of input operations). 96 Median input is the input size for which 50% of inputs are the same size or smaller and 50% of inputs are larger. 35 ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

97 Figure 4: Areas of GPT Study Zone with Potential for Dry Cargo Sweepings Inputs Figure 5: Size Distribution of Dry Cargo Washdown Inputs 36 ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

98 Figure 6: Cumulative Probability Distribution of Dry Cargo Washdown Amounts References Barone, M., A. Campanile, F. Caprio, and E. Fasano The impact of the new MARPOL regulations on bulk carrier design: A case study. Proceedings of the Second International Conference on Marine Research and Transportation. 10 p. Behrens, H.L., Ø. Endresen, A. Mjelde, and C. Garmann, Environmental Accounting System for Norwegian Shipping EASNoS Phase 1. Det Norske Veritas (DNV) Report No Høvik, Norway. Eide, M.S., Ø. Endresen, Ø. Breivik, O.W. Brude, I.H. Ellingsen, K. Røang, J. Hauge, and P.O. Brett Prevention of oil spill from shipping by modeling of dynamic risk. Marine Pollution Bulletin Vol. 54: 1,619 1, 633. Etkin, D.S Oil Spill Response Reference Guide, Cutter Information Corp., Arlington, MA, 70 p. Etkin, D.S Analysis of Washington State Vessel and Facility Oil Discharge Scenarios for Contingency Planning Standards. Prepared for Washington Department of Ecology, Spills Program, Olympia, Washington. Contract No. C p. Etkin, D.S Analysis of past marine oil spill rates and trends for future contingency planning. Proc. of 25th Arctic & Marine Oilspill Prog. Tech. Sem.: Etkin, D.S Analysis of US oil spill trends to develop scenarios for contingency planning. Proc. of 2003 International Oil Spill Conference: ERC Report: Gateway Pacific Terminal Characterization of Likely Accidents and Consequences

FINAL REPORT: VTRA

FINAL REPORT: VTRA Table Contents Publication Information... vii Contact Information... vii PREFACE... 1 EXECUTIVE SUMMARY... 3 Description of Methodology... 4 Base Case and What-If Results... 7 Risk Mitigation and Historical

More information

Public Meeting. NEPA Environmental Review Process. To seek comments on the Draft EIS

Public Meeting. NEPA Environmental Review Process. To seek comments on the Draft EIS NEPA Environmental Review Process Publish Notice of Intent Public Meeting To seek comments on the Draft EIS Written comments must be postmarked and/or received by the Corps no later than August 6, 2014

More information

Assessment of Oil Spill Risk due to Potential Increased Vessel Traffic at Cherry Point, Washington. A Draft Proposal Submitted to

Assessment of Oil Spill Risk due to Potential Increased Vessel Traffic at Cherry Point, Washington. A Draft Proposal Submitted to Assessment of Oil Spill Risk due to Potential Increased Vessel Traffic at Cherry Point, Washington A Draft Proposal Submitted to Ms. Christine Butenschoen Contracts Administrator BP Cherry Point Refinery

More information

BP Cherry Point Dock Draft Environmental Impact Statement

BP Cherry Point Dock Draft Environmental Impact Statement BP Cherry Point Dock Draft Environmental Impact Statement May 2014 U.S. Army Corps of Engineers Seattle District This page is left blank intentionally. Draft EIS May 2014 COVER SHEET Draft Environmental

More information

RESOLUTION NO

RESOLUTION NO RESOLUTION NO. 2012-22 A RESOLUTION REQUESTING THAT CERTAIN POTENTIAL ON AND OFF-SITE IMPACTS ASSOCIATED WITH THE GATEWAY PACIFIC TERMINAL BE ANALYZED AS PART OF THE SEPA AND NEPA PROCESSES WHEREAS, Pacific

More information

Storage Tank Explosion Frequencies on FPSOs

Storage Tank Explosion Frequencies on FPSOs Storage Tank Explosion Frequencies on FPSOs John Spouge, Principal Consultant, DNV GL, Vivo Building, 30 Stamford Street, London SE1 9LQ Introduction This paper presents a new estimate of the frequency

More information

1. Introduction

1. Introduction TABLE OF CONTENTS SEPA EIS FACT SHEET EIS SUMMMARY DRAFT ENVIRONMENTAL IMPACT STATEMENT List of Tables List of Figures List of Acronyms 1. Introduction... 1-1 1.1. Project Overview...1-1 1.2. Purpose and

More information

Washington State Ferries Liquefied Natural Gas Project

Washington State Ferries Liquefied Natural Gas Project Overview After three- and- a- half years of analysis, evaluation and several detailed studies (WSF) will seek the U.S. Coast Guard s approval to use Liquefied Natural Gas (LNG) as propulsion fuel. WSF

More information

Dr. Konstantinos Galanis

Dr. Konstantinos Galanis Dr. Konstantinos Galanis Operations & Technical Senior Manager Seanergy Maritime Holdings Corp. To describe the different kinds of ships that are in common use The employment they are engaged in The kind

More information

Waterways 1 Water Transportation History

Waterways 1 Water Transportation History Waterways 1 Water Transportation History Water Transportation Propulsion History Human (oars, poles) - - 7,000-10,000 BC Wind (sails) - - 3,000 BC Steamboat invented - - 1787 AD First diesel-powered ship

More information

PACIFIC STATES/BRITISH COLUMBIA OIL SPILL TASK FORCE SPILL and INCIDENT DATA COLLECTION PROJECT REPORT July, 1997

PACIFIC STATES/BRITISH COLUMBIA OIL SPILL TASK FORCE SPILL and INCIDENT DATA COLLECTION PROJECT REPORT July, 1997 PACIFIC STATES/BRITISH COLUMBIA OIL SPILL TASK FORCE SPILL and INCIDENT DATA COLLECTION PROJECT REPORT July, 1997 A. Background on the Task Force The Pacific States/British Columbia Oil Spill Task Force

More information

TRANS MOUNTAIN PIPELINE SYSTEM NORTHWEST AREA COMMITTEE MEETING SEATTLE, WA. 13 TH FEBRUARY 2013 MICHAEL DAVIES

TRANS MOUNTAIN PIPELINE SYSTEM NORTHWEST AREA COMMITTEE MEETING SEATTLE, WA. 13 TH FEBRUARY 2013 MICHAEL DAVIES TRANS MOUNTAIN PIPELINE SYSTEM NORTHWEST AREA COMMITTEE MEETING SEATTLE, WA. 13 TH FEBRUARY 2013 MICHAEL DAVIES 1 Trans Mountain Pipeline Proposed Expansion Expand capacity to 890,000 bpd Customer contracts

More information

TERMPOL Review Process Report on the Enbridge Northern Gateway Project

TERMPOL Review Process Report on the Enbridge Northern Gateway Project TERMPOL Review Process Report on the Enbridge Northern Gateway Project Northern Gateway Project TERMPOL Review Process Report Table of Contents FOREWORD... ii GLOSSARY... iii 1. INTRODUCTION... 1 1.1 Project

More information

Comparing Roundabout Capacity Analysis Methods, or How the Selection of Analysis Method Can Affect the Design

Comparing Roundabout Capacity Analysis Methods, or How the Selection of Analysis Method Can Affect the Design Comparing Roundabout Capacity Analysis Methods, or How the Selection of Analysis Method Can Affect the Design ABSTRACT Several analysis methods have been proposed to analyze the vehicular capacity of roundabouts.

More information

Oil & Gas, Environmental, Regulatory Compliance, and Training

Oil & Gas, Environmental, Regulatory Compliance, and Training Harvey Consulting, LLC. Oil & Gas, Environmental, Regulatory Compliance, and Training Review of Draft Environmental Impact Statement For the Tesoro Savage Petroleum Terminal LLC Application for Site Certification

More information

Priority Setting Process For Hydrographic Surveys Page 1 of Parameter 1 - Quality of Survey Data Currently Available... 9

Priority Setting Process For Hydrographic Surveys Page 1 of Parameter 1 - Quality of Survey Data Currently Available... 9 Priority Setting Process For Hydrographic Surveys Page 1 of 25 Contents 1 Scope... 4 2 Related Standards... 4 Symbols (& Abbreviated Terms)... 5 4 Prioritising Parameters... 6 5 Weighting... 7 6 Issuing

More information

LNG as a Marine Fuel. Cliff Gladstein President Gladstein, Neandross & Associates

LNG as a Marine Fuel. Cliff Gladstein President Gladstein, Neandross & Associates LNG as a Marine Fuel Cliff Gladstein President Gladstein, Neandross & Associates cliff@gladstein.org www.gladstein.org Panel Overview Why are we talking about LNG? What do ports need to know about LNG?

More information

MINIMUM REQUIREMENTS FOR VESSELS BOUND FOR OR LEAVING PORTS OF THE BALTIC SEA STATES AND CARRYING DANGEROUS OR POLLUTING GOODS

MINIMUM REQUIREMENTS FOR VESSELS BOUND FOR OR LEAVING PORTS OF THE BALTIC SEA STATES AND CARRYING DANGEROUS OR POLLUTING GOODS CONVENTION ON THE PROTECTION OF THE MARINE ENVIRONMENT OF THE BALTIC SEA AREA HELSINKI COMMISSION - Baltic Marine HELCOM 19/98 Environment Protection Commission 15/1 Annex 7 19th Meeting Helsinki, 23-27

More information

Port of Long Beach Port Master Plan Overview December 22, 2008

Port of Long Beach Port Master Plan Overview December 22, 2008 Port of Long Beach Port Master Plan Overview December 22, 2008 PORT OF LONG BEACH PORT MASTER PLAN OVERVIEW INTRODUCTION In 1978 the California Coastal Commission certified the Port of Long Beach Port

More information

AMER BADAWI Vice President Columbia Grain, Inc.

AMER BADAWI Vice President Columbia Grain, Inc. AMER BADAWI Vice President Columbia Grain, Inc. World Freight Markets By: Amer Badawi Vice President Topics Drybulk Fleet Profile General Trends World Grain World Output World Output - 47% increase y-o-y

More information

Environmental Reviews and the State Environmental Policy Act (SEPA) Process. November 2014

Environmental Reviews and the State Environmental Policy Act (SEPA) Process. November 2014 Environmental Reviews and the State Environmental Policy Act (SEPA) Process November 2014 Energy Storage or Refining - Existing Facilities and Proposals BP Refinery (Cherry Point) Phillips 66 Refinery

More information

Detachment Chief United States Coast Guard Liquefied Gas Carrier National Center of Expertise

Detachment Chief United States Coast Guard Liquefied Gas Carrier National Center of Expertise Detachment Chief United States Coast Guard Liquefied Gas Carrier National Center of Expertise 2901 Turtle Creek Drive Port Arthur, TX 77642-8056 Phone: (409) 723-6559 Fax: (409) 718-3838 lgcncoe@uscg.mil

More information

Notice of Tariff Change Effective January 1, 2014

Notice of Tariff Change Effective January 1, 2014 Notice of Tariff Change Effective January 1, 2014 Port of Seattle Terminals Tariff No. 5 is revised as follows: RATE CHANGES: ITEM 1460 SMALL LOTS FEE A charge of $116.09 increased from $114.33 will be

More information

Electronic Chart Display and Information Systems for navigational safety in maritime transportation

Electronic Chart Display and Information Systems for navigational safety in maritime transportation Electronic Chart Display and Information Systems for navigational safety in maritime transportation Erik Vanem, Magnus S. Eide, Rolf Skjong Presented at the E-navigation Conference, Oslo 17 October 2007

More information

November 17. Brochure.

November 17. Brochure. November 17 Brochure www.nextmaritime.com Index About us Ship Agency Logistics Customs Marine Associates About us............................. THE PREFERRED WORLDWIDE SHIPPING AGENT. Next Maritime is a

More information

Panama Canal Authority Toll Tariffs approved by Cabinet Council and published on the Official Gazzette Implementation: April 1, 2016

Panama Canal Authority Toll Tariffs approved by Cabinet Council and published on the Official Gazzette Implementation: April 1, 2016 Panama Canal Authority Toll Tariffs approved by Cabinet Council and published on the Official Gazzette Implementation: April 1, 2016 Reformulation for full container vessels 2/ Neopanamax locks: for vessels

More information

Addressing GHG emissions from international maritime transport ICAO/IMO Side Event UNFCCC COP 20

Addressing GHG emissions from international maritime transport ICAO/IMO Side Event UNFCCC COP 20 Addressing GHG emissions from international maritime transport ICAO/IMO Side Event UNFCCC COP 20 Dr Edmund Hughes Head, Air Pollution and Energy Efficiency Marine Environment Division, 1 st December 2014

More information

WINDSOR PORT AUTHORITY

WINDSOR PORT AUTHORITY By-Law No. 1 WINDSOR PORT AUTHORITY a By-Law fixing the fees to be paid to enter or use the Port of Windsor May 1, 2017 By-Law No. 1 WINDSOR PORT AUTHORITY a By-Law fixing the fees to be paid to enter

More information

EPA Issues General Permit for Vessels

EPA Issues General Permit for Vessels EPA Issues General Permit for Vessels Background On December 18, 2008, EPA issued its Vessel General Permit ( VGP ) for discharges incidental to the normal operation of vessels. The permit responds to

More information

DET NORSKE VERITAS & ERM WEST, INC.

DET NORSKE VERITAS & ERM WEST, INC. DET NORSKE VERITAS & ERM WEST, INC. Aleutian Islands Risk Assessment Phase A Preliminary Risk Assessment TASK 1: Semi-quantitative Traffic Study Report Prepared For: National Fish and Wildlife Foundation

More information

CHAPTER 4 SHIP RECYCLING DEMAND FORECASTING

CHAPTER 4 SHIP RECYCLING DEMAND FORECASTING CHAPTER 4 SHIP RECYCLING DEMAND FORECASTING 4.1 SCOPE OF DEMAND FORECASTING IN SHIP RECYCLING Demand forecasting is an essential analytical procedure executed through mathematical tools or otherwise by

More information

Abstract. 1 Introduction

Abstract. 1 Introduction A prototype statistical approach of oil pollution in the Mediterranean Sea N.P. Ventikos & H.N. Psaraftis Department ofnaval Architecture & Marine Engineering, National Technical University of Athens,

More information

SAINT JOHN PORT AUTHORITY TARIFF NOTICES. Effective: January 1, 2018

SAINT JOHN PORT AUTHORITY TARIFF NOTICES. Effective: January 1, 2018 SAINT JOHN PORT AUTHORITY TARIFF NOTICES Effective: January 1, 2018 797052 v5 TABLE OF CONTENTS Notice Tariff Page Definitions 2 N - 1 Berthage 4 N - 2 Wharfage 6 N - 3 Harbour Dues 10 N - 4 Water Services

More information

Maritime Safety Committee s 89 th Session

Maritime Safety Committee s 89 th Session News Update American Bureau of Shipping September 2011 Vol.20, No.2 Maritime Safety Committee s 89 th Session 11 to 20 May 2011 http://www.eagle.org/eagleexternalportalweb / Resources / Regulatory Information

More information

A MODAL COMPARISON OF DOMESTIC FREIGHT TRANSPORTATION EFFECTS ON THE GENERAL PUBLIC EXECUTIVE SUMMARY. November 2007

A MODAL COMPARISON OF DOMESTIC FREIGHT TRANSPORTATION EFFECTS ON THE GENERAL PUBLIC EXECUTIVE SUMMARY. November 2007 A MODAL COMPARISON OF DOMESTIC FREIGHT TRANSPORTATION EFFECTS ON THE GENERAL PUBLIC EXECUTIVE SUMMARY November 2007 Prepared by CENTER FOR PORTS AND WATERWAYS TEXAS TRANSPORTATION INSTITUTE 701 NORTH POST

More information

PORT TARIFF 2017 / 2018

PORT TARIFF 2017 / 2018 PORT TARIFF 2017 / 2018 INCLUDING GENERAL TERMS AND CONDITIONS Port Tariff issued by the Port, Duly approved by the Board of Directors For Clarifications or questions, please refer to; Sohar Industrial

More information

STATE OF MICHIGAN DEPARTMENT OF ENVIRONMENTAL QUALITY LANSING. August 4, 2017

STATE OF MICHIGAN DEPARTMENT OF ENVIRONMENTAL QUALITY LANSING. August 4, 2017 RICK SNYDER GOVERNOR STATE OF MICHIGAN DEPARTMENT OF ENVIRONMENTAL QUALITY LANSING C. HEIDI GRETHER DIRECTOR VIA ELECTRONIC SUBMITTAL Dynamic Risk 1110, 333 11 Avenue SW Calgary, Alberta T2R 1L9 CANADA

More information

BERTHING PROBLEM OF SHIPS IN CHITTAGONG PORT AND PROPOSAL FOR ITS SOLUTION

BERTHING PROBLEM OF SHIPS IN CHITTAGONG PORT AND PROPOSAL FOR ITS SOLUTION 66 BERTHING PROBLEM OF SHIPS IN CHITTAGONG PORT AND PROPOSAL FOR ITS SOLUTION A. K. M. Solayman Hoque Additional Chief Engineer, Chittagong Dry Dock Limited, Patenga, Chittagong, Bnagladesh S. K. Biswas

More information

EXHIBIT A Planning Commission Recommended Cherry Point Amendments

EXHIBIT A Planning Commission Recommended Cherry Point Amendments Cherry Point Planning Commission Final Recommendations (//0) EXHIBIT A Planning Commission Recommended Cherry Point Amendments January, 0 Cherry Point Planning Commission Final Recommendations (//0) 0

More information

CITY OF VANCOUVER ADMINISTRATIVE REPORT. General Manager of Engineering Services in consultation with the Director of Current Planning

CITY OF VANCOUVER ADMINISTRATIVE REPORT. General Manager of Engineering Services in consultation with the Director of Current Planning CITY OF VANCOUVER ADMINISTRATIVE REPORT Date: September 1, 2005 Author: Dale Bracewell Phone No.: 604.871.6440 RTS No.: 5291 CC File No.: 8203 Meeting Date: September 20, 2005 TO: FROM: SUBJECT: Standing

More information

THE UK PORTS INDUSTRY:

THE UK PORTS INDUSTRY: THE UK PORTS INDUSTRY: Forecasting Growth & Developing Capacity by MDS Transmodal 205024_presentationv5 1. AGENDA Past trends, 1965-2004 Ports & Public policy - History and forthcoming review Forecasting

More information

HOW TO PREDICT CARGO HANDLING TIMES AT THE SEA PORT AFFECTED BY WEATHER CONDITIONS

HOW TO PREDICT CARGO HANDLING TIMES AT THE SEA PORT AFFECTED BY WEATHER CONDITIONS HOW TO PREDICT CARGO HANDLING TIMES AT THE SEA PORT AFFECTED BY WEATHER CONDITIONS Tatjana Stanivuk University of Split, Faculty of Maritime Studies Zrinsko-Frankopanska 38, 21000 Split, Croatia E-mail:

More information

Port of Prince Rupert: Value of Trade Analysis

Port of Prince Rupert: Value of Trade Analysis Port of Prince Rupert: Value of Trade Analysis Prepared for Prince Rupert Port Authority Prepared by InterVISTAS Consulting Inc. 17 February 2012 i Table of Contents 1. Introduction... 1 1.1 Port of Prince

More information

Expectations for Port Customers and Clients. Dan Sheehy -NYK Line AAPA Meeting October 23, 2008

Expectations for Port Customers and Clients. Dan Sheehy -NYK Line AAPA Meeting October 23, 2008 Expectations for Port Customers and Clients Dan Sheehy -NYK Line AAPA Meeting October 23, 2008 Areas for Review 1. Overview of NYK Line 2. Liner Trade Business 3. Demand versus Supply Outlook 4. Bunker

More information

Alaska Deep-Draft Arctic Ports Navigation Feasibility Study. U.S. Army Corps of Engineers Bruce Sexauer P.E. Chief of Planning, Alaska District

Alaska Deep-Draft Arctic Ports Navigation Feasibility Study. U.S. Army Corps of Engineers Bruce Sexauer P.E. Chief of Planning, Alaska District Alaska Deep-Draft Arctic Ports Navigation Feasibility Study U.S. Army Corps of Engineers Bruce Sexauer P.E. Chief of Planning, Alaska District Alaska Deep Draft Arctic Port System Study Purpose To identify

More information

Marine Protection Rules Part 100 Port Reception Facilities Oil, Noxious Liquid Substances and Garbage

Marine Protection Rules Part 100 Port Reception Facilities Oil, Noxious Liquid Substances and Garbage Marine Protection Rules Part 100 Port Reception Facilities Oil, Noxious Liquid Substances and Garbage MNZ Consolidation Marine Protection Rules ISBN 978-0-947527-29-7 Published by Maritime New Zealand,

More information

WebShipCost - Quantifying Risk in Intermodal Transportation

WebShipCost - Quantifying Risk in Intermodal Transportation WebShipCost - Quantifying Risk in Intermodal Transportation Zhe Li, Heather Nachtmann, and Manuel D. Rossetti Department of Industrial Engineering University of Arkansas Fayetteville, AR 72701, USA Abstract

More information

CITGO PETROLEUM CORPORATION CORPUS CHRISTI MARINE FACILITY REGULATIONS FOR VESSELS

CITGO PETROLEUM CORPORATION CORPUS CHRISTI MARINE FACILITY REGULATIONS FOR VESSELS CITGO PETROLEUM CORPORATION CORPUS CHRISTI MARINE FACILITY REGULATIONS FOR VESSELS THE EMERGENCY SIGNAL FOR THIS TERMINAL IS TWO LONG BLASTS SEPARATED BY A SHORT PAUSE ON THE DOCKSIDE PORTABLE AIR HORN.

More information

Pacific States/British Columbia Oil spill Task Force Data Dictionary. Revised 2007

Pacific States/British Columbia Oil spill Task Force Data Dictionary. Revised 2007 Pacific States/British Columbia Oil spill Task Force Data Dictionary Revised 2007 The U.S. members 1 of the Pacific States/British Columbia Oil Spill Task Force signed an agreement in 1997 to incorporate

More information

Vessel slowdown trial: Info session breakfast

Vessel slowdown trial: Info session breakfast Vessel slowdown trial: Info session breakfast Presented by: Vancouver Fraser Port Authority April 27, 2017 Morning Overview Welcome Overview of ECHO program and vessel slowdown trial Overview of industry

More information

Critical parameters in deriving fire fragility functions for steel gravity frames

Critical parameters in deriving fire fragility functions for steel gravity frames Proceedings of the 9 th International Conference on Structures in Fire (SiF 16) 8-10 June 2016, Princeton University, NJ, USA Critical parameters in deriving fire fragility functions for steel gravity

More information

PORT INFORMATION GUIDE NOTICE OF AMENDMENT

PORT INFORMATION GUIDE NOTICE OF AMENDMENT PORT INFORMATION GUIDE NOTICE OF AMENDMENT Date Issued: June 15 th, 2015 Date effective: July 15 th, 2015 Subject: This notice is being posted publicly to notify industry, stakeholders and the public of

More information

YES MARITIME SDN BHD

YES MARITIME SDN BHD YES MARITIME SDN BHD You, as our customer, will receive our focused attention with utmost integrity at all times. Efficient, reliable and cost effective services to our customers. Safety of personnel,

More information

LNG as a marine fuel in BC. West Coast Marine LNG Workshop 26 th June 2012

LNG as a marine fuel in BC. West Coast Marine LNG Workshop 26 th June 2012 LNG as a marine fuel in BC West Coast Marine LNG Workshop 26 th June 2012 LNG as a Marine Fuel» Marpol Annex VI - Emissions Legislation ECA s commence in 2015 Global sulphur cap in 2020 / 2025 Potential

More information

Bring your shipping operations into the 21st century

Bring your shipping operations into the 21st century See BEYOND Bring your shipping operations into the 21st century The shipping industry continues to face pressure from a variety of sources. Measures like the MRV directive are steering companies to adopt

More information

Failure to Act. Of current Investment Trends in. Airports, Inland Waterways, and Marine Ports. Infrastructure EXECUTIVE SUMMARY

Failure to Act. Of current Investment Trends in. Airports, Inland Waterways, and Marine Ports. Infrastructure EXECUTIVE SUMMARY Failure to Act The economic impact Of current Investment Trends in Airports, Inland Waterways, and Marine Ports Infrastructure EXECUTIVE SUMMARY EXECUTIVE SUMMARY Air and waterborne transportation infrastructure

More information

Balancing Risk and Economics for Chemical Supply Chain Optimization under Uncertainty

Balancing Risk and Economics for Chemical Supply Chain Optimization under Uncertainty Balancing Risk and Economics for Chemical Supply Chain Optimization under Uncertainty Fengqi You and Ignacio E. Grossmann Dept. of Chemical Engineering, Carnegie Mellon University John M. Wassick The Dow

More information

Woodfibre LNG Limited Response to SIGTTO LNG Ports and Risk Reduction Options

Woodfibre LNG Limited Response to SIGTTO LNG Ports and Risk Reduction Options Woodfibre LNG Limited Response to SIGTTO LNG Ports and Risk Reduction Options Introduction: The following is in response to the Environmental Assessment Office request for the Society of International

More information

Transport Canada Marine Transportation in the Canadian Arctic Presentation to the International Maritime Statistics Forum

Transport Canada Marine Transportation in the Canadian Arctic Presentation to the International Maritime Statistics Forum Transport Canada Marine Transportation in the Canadian Arctic Presentation to the International Maritime Statistics Forum Centre of Excellence in Economics, Statistics, Analysis and Research (CEESAR) May

More information

Maritime and Trade Capabilities for Financial Institutions. Maritime & Trade

Maritime and Trade Capabilities for Financial Institutions. Maritime & Trade Maritime and Trade Capabilities for Financial Institutions Maritime & Trade 2017 2017 2016 IHS Markit. All Rights Reserved. Introducing IHS Markit We are a global information and analytics company ENERGY

More information

MRV Verification Process

MRV Verification Process MRV Verification Process Trafi Ballast Water and MRV seminar 15 June 2017 Niklas Rönnberg Working together for a safer world Implementation schedule EIF: Entry into force 1. EIF 3. Monitoring plans to

More information

ER Monitoring Report (ER-MR)

ER Monitoring Report (ER-MR) Forest Carbon Partnership Facility (FCPF) Carbon Fund ER Monitoring Report (ER-MR) ER Program Name and Country: Reporting Period covered in this report: Number of net ERs generated by the ER Program during

More information

DOWNSTREAM PETROLEUM 2017 DOWNSTREAM PETROLEUM

DOWNSTREAM PETROLEUM 2017 DOWNSTREAM PETROLEUM DOWNSTREAM PETROLEUM Maintaining supply security and reliability KEY MESSAGES Australia s longer-term fuel supply security and transport energy needs will be best met through market measures including:

More information

B.C. TOWBOAT INDUSTRY CONFERENCE

B.C. TOWBOAT INDUSTRY CONFERENCE LN G T erm inals on C ana da s W est C oast Kitimat LNG LN G T erm inals on C ana da s W est C oast Presentation to 17 th B.C. TOWBOAT INDUSTRY CONFERENCE Whistler, BC Kitimat LNG Alfred Sorensen Kitimat

More information

The Enel Vetting system Shipping Safety and Marine risk assessment

The Enel Vetting system Shipping Safety and Marine risk assessment The Enel Vetting system Shipping Safety and Marine risk assessment Roma, 07 Novembre 2011 Enel Group Enel is Italy's largest power company, and Europe's second listed utility by installed capacity. It

More information

Thus, there are two points to keep in mind when analyzing risk:

Thus, there are two points to keep in mind when analyzing risk: One-Minute Spotlight WHAT IS RISK? Uncertainty about a situation can often indicate risk, which is the possibility of loss, damage, or any other undesirable event. Most people desire low risk, which would

More information

Crude Oil Quality Group

Crude Oil Quality Group Crude Oil Quality Group Dominic Ferrari VP West Coast Pipelines L B h CA Long Beach, CA February 26, 2009 Forward-Looking Statements Disclosure This presentation contains forward-looking statements, including,

More information

The Concept of the Development of Turkmenbashi International Seaport and the Marine Merchant Fleet till Year 2020 was approved with Resolution of the

The Concept of the Development of Turkmenbashi International Seaport and the Marine Merchant Fleet till Year 2020 was approved with Resolution of the The Concept of the Development of Turkmenbashi International Seaport and the Marine Merchant Fleet till Year 2020 was approved with Resolution of the Esteemed President of Turkmenistan CURRENT VIEW OVER

More information

MARITIME ECONOMICS RESEARCH PAPER

MARITIME ECONOMICS RESEARCH PAPER MARITIME ECONOMICS RESEARCH PAPER A Capesized vessel: by Shiptrade house, Nov.2016, 25 th. TITLE: A market forecasting report for the dry bulk capesize subsector Athens, January 2017 By: Ioannis Valmas

More information

Breakdown, Breakeven, Breakdance what s in store next?

Breakdown, Breakeven, Breakdance what s in store next? Breakdown, Breakeven, Breakdance what s in store next? Dr Adam Kent - Maritime Strategies International (MSI) 9th Annual Marine Money London Ship Finance Forum 24 th Jan 2018 Agenda Breakdown, Breakeven,

More information

Port Capacity Analysis in Long Beach. Matt Plezia, POLB and Vaibhav Govil, AECOM TRB Irvine, CA June 29, 2010

Port Capacity Analysis in Long Beach. Matt Plezia, POLB and Vaibhav Govil, AECOM TRB Irvine, CA June 29, 2010 Port Capacity Analysis in Long Beach Matt Plezia, POLB and Vaibhav Govil, AECOM TRB Irvine, CA June 29, 2010 1 About the Port of Long Beach 2 Second busiest container port in N. America 7 container terminals

More information

TARIFF REGULATIONS 2018 APDL ADMINISTRAÇÃO DOS PORTOS DO DOURO, LEIXÕES E VIANA DO CASTELO, S.A.

TARIFF REGULATIONS 2018 APDL ADMINISTRAÇÃO DOS PORTOS DO DOURO, LEIXÕES E VIANA DO CASTELO, S.A. TARIFF REGULATIONS 2018 APDL ADMINISTRAÇÃO DOS PORTOS DO DOURO, LEIXÕES E VIANA DO CASTELO, S.A. GENERAL STIPULATIONS Article 1 Scope of application APDL - Administração dos Portos do Douro, Leixões e

More information

ANNEX 1. RESOLUTION MSC.325(90) (adopted on 24 May 2012)

ANNEX 1. RESOLUTION MSC.325(90) (adopted on 24 May 2012) Annex 1, page 1 ANNEX 1 RESOLUTION MSC.325(90) (adopted on 24 May 2012) ADOPTION OF AMENDMENTS TO THE INTERNATIONAL CONVENTION FOR THE SAFETY OF LIFE AT SEA, 1974, AS AMENDED THE MARITIME SAFETY COMMITTEE,

More information

КМТФ. National Maritime Shipping Company KazMorTransFlot. Astana November 2013

КМТФ. National Maritime Shipping Company KazMorTransFlot. Astana November 2013 КМТФ National Maritime Shipping Company KazMorTransFlot Astana November 2013 GENERAL INFORMATION Date of established 04.12.1998. Shareholder «National company «KazMunayGas» JSC. The National Maritime Carrier

More information

ABS TECHNICAL PAPERS Jorge Ballesio, American Bureau of Shipping (ABS), Houston, USA

ABS TECHNICAL PAPERS Jorge Ballesio, American Bureau of Shipping (ABS), Houston, USA Evaluation of Classification Rules Related to Machinery for an Oil Tanker Robert Cross, ABSG Consulting, Houston, USA Jorge Ballesio, American Bureau of Shipping (ABS), Houston, USA Published in the proceedings

More information

Introduction of Gas Turbine- Powered, LPG Fueled Ship. Byeongyeol Baek Advanced Lead Engineer GE s Marine Solutions March 6 to 7, 2018

Introduction of Gas Turbine- Powered, LPG Fueled Ship. Byeongyeol Baek Advanced Lead Engineer GE s Marine Solutions March 6 to 7, 2018 Introduction of Gas Turbine- Powered, LPG Fueled Ship Byeongyeol Baek Advanced Lead Engineer GE s Marine Solutions March 6 to 7, 2018 Presentation Title February 9, 2 GE s Marine Solutions +1,400 Vessel

More information

ANNEXES. to the Commission Implementing Regulation

ANNEXES. to the Commission Implementing Regulation Ref. Ares(2016)3985800-28/07/2016 EUROPEAN COMMISSION Brussels, XXX [ ](2016) XXX draft ANNEXES 1 to 3 ANNEXES to the Commission Implementing Regulation on templates for monitoring plans, emissions reports

More information

Current Status on Safe Regulation of River Transport and Challenges in Myanmar. Yangon

Current Status on Safe Regulation of River Transport and Challenges in Myanmar. Yangon Current Status on Safe Regulation of River Transport and Challenges in Myanmar Zaw Myint Thein Director Department of Marine Administration Yangon 7.3.2014 1 1 Introduction 2 Duties and Functions of DMA

More information

Formal Safety Assessment - LNG -IMO/MSC86 Review

Formal Safety Assessment - LNG -IMO/MSC86 Review Formal Safety Assessment - LNG -IMO/MSC86 Review London June 1st 2009 Rolf Skjong, Dr Chief scientist, DNV Authors, Reviewers, Partners, Data, Etc Hazid: Authors: I Østvik (LMG), E. Vanem (DNV), F. Castello

More information

Diploma in shipping (mercantile diploma course)

Diploma in shipping (mercantile diploma course) Diploma in shipping (mercantile diploma course) 1: Compulsory for all mercantile diploma courses 15 ECTS 1.1: Organization Theory of science and Methods 10 ECTS 1.2: Economic operations and management

More information

Influence of corrosion-related degradation of mechanical properties of shipbuilding steel on collapse strength of plates and stiffened panels

Influence of corrosion-related degradation of mechanical properties of shipbuilding steel on collapse strength of plates and stiffened panels Towards Green Marine Technology and Transport Guedes Soares, Dejhalla & Pavleti (Eds) 2015 Taylor & Francis Group, London, ISBN 978-1-138-02887-6 Influence of corrosion-related degradation of mechanical

More information

Properties of LNG Don Juckett US Department of Energy February 12, 2002 LNG Workshop Solomons, MD. U.S. Department of Energy

Properties of LNG Don Juckett US Department of Energy February 12, 2002 LNG Workshop Solomons, MD. U.S. Department of Energy Properties of LNG Don Juckett US Department of Energy February 12, 2002 LNG Workshop Solomons, MD U.S. Department of Energy Background LNG in the US US Natural Gas imports and Exports Projections of Market

More information

AFFECTED ENVIRONMENT AND ENVIRONMENTAL IMPACTS

AFFECTED ENVIRONMENT AND ENVIRONMENTAL IMPACTS Callou t box: Callout box en d. CHAPTER 3 AFFECTED ENVIRONMENT AND ENVIRONMENTAL IMPACTS 3.0 INTRODUCTION This introductory chapter and subchapters describe the existing conditions (affected environments)

More information

Reference Manual on Maritime Transport Statistics

Reference Manual on Maritime Transport Statistics Reference Manual on Maritime Transport Statistics 1 2017 Reference Manual on Maritime Transport Statistics Version 4.0 (November 2017) Reference Manual on Maritime Transport Statistics 2 Reference Manual

More information

Maritime Energy Transport Regulatory Overview and Status

Maritime Energy Transport Regulatory Overview and Status Maritime Energy Transport Regulatory Overview and Status CAPT B. J. Hawkins, P.E. Office of Design and Engineering Standards (CG-ENG) US Coast Guard Headquarters Washington, DC Office of Design and Engineering

More information

SPPI for Freight Water Transport in Sweden

SPPI for Freight Water Transport in Sweden 24th Voorburg Group Meeting Oslo, Norway September 14th 18th 2009 SPPI for Freight Water Transport in Sweden Thomas Olsson Price Statistics Unit, Statistics Sweden Contents 1 DEFINITION OF THE SECTOR BEING

More information

HALIFAX PORT AUTHORITY BERTHAGE AND ANCHORAGE NOTICE SCHEDULE OF RATES. Item Description Rate

HALIFAX PORT AUTHORITY BERTHAGE AND ANCHORAGE NOTICE SCHEDULE OF RATES. Item Description Rate BERTHAGE AND ANCHORAGE NOTICE EFFECTIVE: January 1, 2017 NOTICE N-1 I 1. The berthage rates per gross registered tonne are: (a) for the first period of 12 hours or part thereof... $0.0502 (b) for the second

More information

Coastal and Marine Third Party

Coastal and Marine Third Party Coastal and Marine Third Party Damage Incident DATA The data below was compiled from PHMSA, Minerals Management Service, United States Coast Guard and the National Transportation Safety Board. Description

More information

Global Supply Chain Planning under Demand and Freight Rate Uncertainty

Global Supply Chain Planning under Demand and Freight Rate Uncertainty Global Supply Chain Planning under Demand and Freight Rate Uncertainty Fengqi You Ignacio E. Grossmann Nov. 13, 2007 Sponsored by The Dow Chemical Company (John Wassick) Page 1 Introduction Motivation

More information

THE REPUBLIC OF LIBERIA LIBERIA MARITIME AUTHORITY

THE REPUBLIC OF LIBERIA LIBERIA MARITIME AUTHORITY Office of Deputy Commissioner of Maritime Affairs THE REPUBLIC OF LIBERIA LIBERIA MARITIME AUTHORITY Marine Notice POL-012 Rev. 10/17 TO: SUBJECT: ALL SHIPOWNERS, OPERATORS, MASTERS AND OFFICERS OF MERCHANT

More information

04 November LNG Ready Capabilities

04 November LNG Ready Capabilities 04 November 2015 LNG Ready Capabilities Ships Fuel in Transition? Reminiscent of coal to oil. could it be oil to LNG next? Conventional wisdom: status quo (i.e. LSFO in ECA only) until IMO decides when

More information

IACS. Recommendation 74 A GUIDE TO MANAGING MAINTENANCE IN ACCORDANCE WITH THE REQUIREMENTS OF THE ISM CODE. (April 2001) (Rev.

IACS. Recommendation 74 A GUIDE TO MANAGING MAINTENANCE IN ACCORDANCE WITH THE REQUIREMENTS OF THE ISM CODE. (April 2001) (Rev. IACS Recommendation A GUIDE TO MANAGING MAINTENANCE IN ACCORDANCE WITH THE REQUIREMENTS OF THE ISM CODE (April 2001) (Rev.1, May 2008) IACS A GUIDE TO MANAGING MAINTENANCE April 2001/Rev.1 2008 IACS -

More information

3.6 GROUND TRANSPORTATION

3.6 GROUND TRANSPORTATION 3.6.1 Environmental Setting 3.6.1.1 Area of Influence The area of influence for ground transportation consists of the streets and intersections that could be affected by automobile or truck traffic to

More information

CANADA TRANSPORTATION ACT REVIEW INITIAL SUBMISSION FOCUSED ON THE GRAIN INDUSTRY. Infrastructure, Efficiency, Transparency

CANADA TRANSPORTATION ACT REVIEW INITIAL SUBMISSION FOCUSED ON THE GRAIN INDUSTRY. Infrastructure, Efficiency, Transparency CANADA TRANSPORTATION ACT REVIEW INITIAL SUBMISSION FOCUSED ON THE GRAIN INDUSTRY Infrastructure, Efficiency, Transparency December 31, 2014 Table of Contents Introduction and context... 2 Executive summary...

More information

Safety and Environment Management Plan

Safety and Environment Management Plan PORT OF HASTINGS DEVELOPMENT AUTHORITY Safety and Environment Management Plan Revision number: 0 POH-HSE-PLN-002 Date: 25 May 2017 Glossary and Abbreviations AMSA AQIS AS/NZS 4801:2001 Ballast water Bilge

More information

Distribution Restriction Statement Approved for public release; distribution is unlimited.

Distribution Restriction Statement Approved for public release; distribution is unlimited. CECW-EH-D Regulation No. 1110-2-1404 Department of the Army U.S. Army Corps of Engineers Washington, DC 20314-1000 Engineering and Design HYDRAULIC DESIGN OF DEEP-DRAFT NAVIGATION PROJECTS Distribution

More information

Volume to Capacity Estimation of Signalized Road Networks for Metropolitan Transportation Planning. Hiron Fernando, BSCE. A Thesis CIVIL ENGINEERING

Volume to Capacity Estimation of Signalized Road Networks for Metropolitan Transportation Planning. Hiron Fernando, BSCE. A Thesis CIVIL ENGINEERING Volume to Capacity Estimation of Signalized Road Networks for Metropolitan Transportation Planning by Hiron Fernando, BSCE A Thesis In CIVIL ENGINEERING Submitted to the Graduate Faculty of Texas Tech

More information

INLAND WATERWAYS TRANSPORTATION: Our Competitive Advantage. Delbert R Wilkins Canal Barge Company Big River Moves Leadership Forum April 15, 2013

INLAND WATERWAYS TRANSPORTATION: Our Competitive Advantage. Delbert R Wilkins Canal Barge Company Big River Moves Leadership Forum April 15, 2013 INLAND WATERWAYS TRANSPORTATION: Our Competitive Advantage Delbert R Wilkins Canal Barge Company Big River Moves Leadership Forum April 15, 2013 INLAND WATERWAYS TRANSPORTATION: Our Competitive Advantage

More information