Port performance benchmarking and efficiency analysis

... 88 Table 4.5 Port Efficiency of DEA-CCR, DEA-BCC and FDH in 2005 89 Table 4.6 Port Efficiency of DEA-CCR, DEA-BCC and FDH in 2007 90 Table 4.7 Port Efficiency of DEA-CCR, DEA-BCC and FDH in 2009... resources Also, port operators can use the information from performance analysis to improve their port planning and operations 1.2 Difficulties in Port Performance Measurement and Benchmarking In... identify port performance indicators relevant to the activities of vessels, cargo and terminals Through the analysis of ports efficiency using identified indicators, insights on port performance benchmarking

PORT PERFORMANCE BENCHMARKING AND EFFICIENCY ANALYSIS YIN LU (B.Eng., Southeast University) A THESIS SUBMITTED FOR THE DEGREE OF MASTER OF ENGINEERING DEPARTMENT OF CIVIL AND ENVIRONMENTAL ENGINEERING NATIONAL UNIVERSITY OF SINGAPORE 2014 DECLARATION I hereby declare that the thesis is my original work and it has been written by me in its entirety. I have duly acknowledged all the sources of information which have been used in the thesis. This thesis has also not been submitted for any degree in any university previously. Yin Lu 18 Nov, 2014 ACKNOWLEDGEMENTS The author would like to express her deep and sincere thanks and appreciation to her supervisor, Dr. Ong Ghim Ping Raymond, for his patient guidance, invaluable advice, constant support and encouragement throughout the course of this research. Special thanks are extended to fellow research mates, Ms. Sou Weng Sut, Mr. Zhang Lei, Dr. Yang Jiasheng, Ms. Fu Rao and Ms. Lim Emiko for their kind help and the friendship. Gratitude is also accorded to Mr. Goh Joon Kiat, Mr. Mohammed Farouk, Mr. Foo Chee Kiong, Mrs. Yu-Ng Chin Hoe and Mrs. Yap-Chong Wei Leng of the Transportation Engineering Laboratory for the kind assistance and support they have provided. Last but not least, the author would like to express her heartfelt gratitude and thanks to her parents for their utmost support, tremendous care and encouragement given to the author in her work. i TABLE OF CONTENTS ACKNOWLEDGEMENTS .................................................................................. i TABLE OF CONTENTS ..................................................................................... ii SUMMARY ........................................................................................................... v LIST OF TABLES .............................................................................................. vii LIST OF FIGURES ............................................................................................. ix CHAPTER 1: INTRODUCTION ........................................................................ 1 1.1 Background Information ............................................................................... 1 1.2 Difficulties in Port Performance Measurement and Benchmarking ............. 2 1.3 Significance of Port Performance and Efficiency Study .............................. 5 1.4 Objectives ..................................................................................................... 6 1.5 Organization of Thesis .................................................................................. 7 CHAPTER 2: LITERATURE REVIEW ........................................................... 9 2.1 Performance Metrics and Index Methods ..................................................... 9 2.1.1 Financial Metrics and Financial Productivity Measures ...................... 12 2.1.2 Physical Productivity Measurements ................................................... 14 2.1.3 Total Factor Productivity Measurements ............................................. 16 2.2 Port Impact Studies ..................................................................................... 18 2.2.1 Port Economic Impact Study ............................................................... 18 2.2.2 Port Trade Efficiency Studies .............................................................. 21 2.3 Frontier Approaches.................................................................................... 25 2.3.1 Parametric Approaches ........................................................................ 27 2.3.2 Non-Parametric Approaches ................................................................ 32 2.4 Studies on Ship Turn-around Time ............................................................. 38 2.5 Research Needs and Scope of work ............................................................ 45 2.6 Summary ..................................................................................................... 47 CHAPTER 3: METHODOLOGY .................................................................... 49 3.1 Methodology Adopted in Research ............................................................ 49 3.2 DEA Technique for Measuring Port Efficiency ......................................... 51 3.2.1 Theory of Data Envelopment Analysis ................................................ 51 3.2.2 Alternative DEA Models ..................................................................... 55 3.2.2.1 CCR Model ................................................................................... 56 3.2.2.2 BCC Model ................................................................................... 58 ii 3.3 FDH Model for Measuring Port Efficiency ................................................ 63 3.4 Probability Models ...................................................................................... 65 3.4.1 Count Data Models .............................................................................. 66 3.4.1.1 Poisson Regression Model ............................................................ 66 3.4.1.2 Negative Binomial Regression Model .......................................... 69 3.4.1.3 Poisson Regression Model with Normal Heterogeneity ............... 70 3.4.2 Duration Models .................................................................................. 71 3.4.3 T Test on Individual Regression Coefficients ...................................... 76 3.4.4 Temporal Stability Test on Regression Models ................................... 77 3.5 Summary ................................................................................................ 77 CHAPTER 4 PORT EFFICIENCY ANALYSIS WITH DEA AND FDH .... 79 4.1 Introduction ................................................................................................. 79 4.2 Empirical Setting ........................................................................................ 79 4.2.1 Model Specification ............................................................................. 79 4.2.2 Ports and Analysis Period .................................................................... 80 4.2.3 Input and Output Variables .................................................................. 81 4.2.3.1 Input Variables .............................................................................. 81 4.2.3.2 Output Variables ........................................................................... 83 4.3 Port Efficiency Analysis of Global Ports .................................................... 85 4.3.1 Global Port Efficiency Analysis .......................................................... 85 4.3.2 Individual Port Efficiency Analysis on a Global Scale ....................... 93 4.3.3 Return to Scale for Global Ports .......................................................... 94 4.3.4 Evaluating Effectiveness of DEA and FDH Models in Port Efficiency Analysis ........................................................................................................ 95 4.3.5 Case Study of Selected Ports ............................................................. 101 4.4 Summary ................................................................................................... 106 CHAPTER 5: MODELING OF CONTAINER SHIP TURN-AROUND TIME IN PORTS USING PROBABILITY MODELS ................................. 108 5.1 Introduction ............................................................................................... 108 5.2 Empirical Setting ...................................................................................... 108 5.2.1 Data Sources ...................................................................................... 109 5.2.2 Definition of Variables ...................................................................... 109 5.2.3 Models in Study ................................................................................. 110 5.3 Modeling Results of Container Ship Turn-around Time .......................... 113 5.3.1 Results of Count Data Models ........................................................... 113 iii 5.3.1.1 Selection of Appropriate Model Form of Count Data Models ... 113 5.3.1.2 Model Estimation for Poisson Regression Model with Heterogeneity .......................................................................................... 114 5.3.2 Results of Hazard-based Duration Models ........................................ 118 5.3.2.1 Selection of Appropriate Model Form of Duration Models ....... 118 5.3.2.2 Model Estimation for Generalized Gamma Model ..................... 122 5.3.3 Elasticity Analysis using Generalized Gamma Ship Turn-around Time Model .......................................................................................................... 128 5.3.4 Temporal Stability of Ship Turn-around Time Model....................... 129 5.3.5 Comparison between Poisson Regression Model with Heterogeneity and Generalized Gamma Model ................................................................. 130 5.5 Summary ................................................................................................... 132 CHAPTER 6 PORT EFFICIENCY ANALYSIS CONSIDERING SHIP TURN-AROUND TIME................................................................................... 134 6.1 Introduction ............................................................................................... 134 6.2 Concept of Container Port Production ...................................................... 134 6.2.1 Container Port Production and Operations ........................................ 134 6.2.2 Considering Variables Affecting Port Efficiency in DEA Analysis .. 135 6.3 Empirical Setting ...................................................................................... 136 6.3.1 Ports and Analysis Period .................................................................. 136 6.3.2 Input and Output Variables ................................................................ 136 6.3.3 Models Considered in Study .............................................................. 137 6.4 Results of the Efficiency Analysis and Interpretation .............................. 139 6.4.1 Throughput as Single Output in DEA Models ................................... 139 6.4.2 Ship Turn-around Time as Single Output in DEA Models ............... 142 6.4.3 DEA Models with Multiple Outputs .................................................. 145 6.4.4 Comparison between Single and Multiple Output DEA Models ...... 148 6.5 Summary ................................................................................................... 150 CHAPTER 7 CONCLUSIONS AND RECOMMENDATIONS .................. 152 7.1 Major Findings of Research ...................................................................... 152 7.2 Recommendations for Further Research ................................................... 153 REFERENCES .................................................................................................. 155 APPENDIX I Port Infrastructure Dataset (2001 to 2011) ............................ 177 APPENDIX II Ship Turn-around Time Dataset (2012 to 2013) .................. 183 APPENDIX III Port Infrastructure Dataset (2012 to 2013) ......................... 207 iv SUMMARY In this increasingly competitive landscape of port industry, it is important for port operators to constantly review the performance of their ports so that they can keep their competitive advantage. Within such a competitive environment, it is important to have a reliable measurement of port performance so that useful advice can be drawn to port operators or managers to improve their port efficiency. Various practical and theoretical approaches were conducted in the past to study the performance of ports, but there is still no consensus to date on an unified method to benchmark port performance. Moreover, ship turn-around time is an important indicator that reflects the service quality of a port and this consideration is seldom made in most of the past studies in the literature. This study therefore assesses the capabilities of different non-parametric approaches to measure port efficiency and examines port efficiency with consideration to ship turn-around time. Data envelopment analysis (DEA) and free disposal hull (FDH) are used to evaluate port efficiency in this study due to its ability to analyze multiple outputs and inputs concurrently. A comparative study between the FDH and DEA methods are made and average efficiency of 61 global container ports are analyzed at the aggregate level. It was found that FDH lacks the sensitivity to analyze port efficiency compared to the DEA models. DEA is more stringent in determining efficient ports and should be used as a preferred method in port efficiency studies. v Three count data models (Poisson regression model, negative binomial regression model and Poisson regression model with normal heterogeneity) and five duration models (exponential, Weibull, log-logistic, log-normal and generalized gamma model) are applied to model ship turn-around time. Poisson regression model with normal heterogeneity and generalized gamma model are found to be the two most appropriate in modeling ship turn-around time respectively compared with the other two count data models and four duration models. The estimated ship turnaround time by the two models is presented. It was found that the estimated ship turn-around time in the generalized gamma model provides a much better fit to actual data. The efficiency of 61 ports in the analysis period (2012 to 2013) is finally presented considering ship turn-around time as output measure in the DEA models. Port efficiency is determined based on single-output-measure and multiple-output-measures DEA-CCR and DEA-BCC models. The result suggests that there is a need to consider both throughput and ship turn-around time in port efficiency studies. vi LIST OF TABLES Table 2.1 Literature review on parametric approaches to the port sector 30 Table 2.2 Literature review on applying DEA to the port sector 39 Table 3.1 Shape of Generalized Gamma Hazard Function 76 Table 4.1 International Ports Considered in Models 81 Table 4.2 Descriptive Statistics of the Input and Output Variables during the Analysis Period 84 Table 4.3 Port Efficiency of DEA-CCR, DEA-BCC and FDH in 2001 87 Table 4.4 Port Efficiency of DEA-CCR, DEA-BCC and FDH in 2003 88 Table 4.5 Port Efficiency of DEA-CCR, DEA-BCC and FDH in 2005 89 Table 4.6 Port Efficiency of DEA-CCR, DEA-BCC and FDH in 2007 90 Table 4.7 Port Efficiency of DEA-CCR, DEA-BCC and FDH in 2009 91 Table 4.8 Port Efficiency of DEA-CCR, DEA-BCC and FDH in 2011 92 Table 4.9 Comparison of Efficiency Results between DEA and FDH Models 98 Table 4.10 Physical Facilities Utilization of Kaohsiung Port Estimated by DEA and FDH Models 104 Table 4.11 Physical Facilities Utilization of Rotterdam Port Estimated by DEA and FDH Models 105 Table 5.1 Variables Considered in Models 111 Table 5.2 Descriptive Statistics of Variables Considered in Study 112 Table 5.3 Models Considered in Study 112 Table 5.4 Estimation Results for Count Data Models using Data in 2012 116 Table 5.5 Estimation Results for Count Data Models using Data in 2013 Table 5.6 Estimation Results for Count Data Models using Data vii 116 in Analysis Period 116 Table 5.7 Estimation Results of Variable Coefficients in Poisson Regression Model with Normal Heterogeneity 117 Table 5.8 Duration Models for Ship Turn-around Time using Data in 2012 124 Table 5.9 Duration Models for Ship Turn-around Time using Data in 2013 125 Table 5.10 Duration Models for Ship Turn-around Time using Data in Analysis Period 126 Table 5.11 Estimation Results of Variable Coefficients in Generalized Gamma 127 Table 5.12 Elasticity of Variables in Generalized Gamma Model 129 Table 5.13 Temporal Stability Tests for Generalized Gamma Models 130 Table 5.14 Comparison of Ship Turn-around Time between Probability Models and Observation 131 Table 6.1 Descriptive Statistics of the Input and Output Variables Considered in DEA Models 138 Table 6.2 Models Considered in Study 137 Table 6.3 Efficiency Estimations in DEA models when Throughput is the Single Output 140 Table 6.4 Efficiency Estimations in DEA models when Ship Turn-around Time is the Single Output 143 Table 6.5 Port Efficiency using DEA Models with Multiple Outputs 146 Table 6.6 Summary of Efficient Ports in DEA Models Applied with Single Output and Multiple Outputs 149 viii LIST OF FIGURES Figure 2.1 Definition of Technical Efficiency 11 Figure 2.2 Illustration of a Production Frontier 26 Figure 2.3 Production Frontier Comparison of DEA models and FDH model 33 Figure 3.1 Flow Chart of Research Methodology in Thesis 50 Figure 3.2 DMU and Homogeneous Units 52 Figure 3.3 Illustration of Efficiency and Productivity 54 Figure 3.4 Production Frontier in CCR Model 56 Figure 3.5 Production Frontier in BCC Model 56 Figure 3.6 Production Frontier and Inefficiency in DEA 62 Figure 3.7 A Cost Frontier of the DEA and FDH model 64 Figure 4.1 Average Efficiency for all Container Ports by CCR, BCC and FDH 86 Figure 4.2 Percentage of Efficient Ports by Throughput Category in 2001 100 Figure 4.3 Number of Efficient Ports by Throughput Categories in 2001 100 Figure 4.4 Efficiency of Singapore, Shanghai and Hong Kong during 2001-2011 102 Figure 4.5 Efficiency of Kaohsiung Port during 2001-2011 102 Figure 4.6 Efficiency of Rotterdam Port during 2001-2011 105 Figure 5.1 Integrated Hazard Function of Model D to H in Analysis Period 119 Figure 5.2 Comparison of Ship Turn-around Time Obtained from Probability Models and Observed Values 131 ix CHAPTER 1: INTRODUCTION 1.1 Background Information Due to globalization of world’s economy, shipping and seaborne liner industries have experienced huge and rapid growth in the past decade. In particular, container transportation has become increasingly important in international trade. Since the 1990s, more than 90% of international cargo moves through seaports, and 80% of seaborne cargo moves in containers (Ramani, 1996). Compared to other traditional modes of transportation, container shipping has numerous technical and economic advantages. Containers can be loaded and unloaded, stacked and transported efficiently over long distances without being opened, transport costs have been dramatically reduced. Containerization has also reduced congestion in ports, significantly shortened shipping time and reduced losses from damage and theft (Marc, 2013). Standing at the crucial interface between sea and inland transportation, container ports form a crucial link in the overall trading chain and therefore play a vital role in the supply chain. One distinctive feature of container port industry today is that the competition between container ports has become much more intensive than ever before. Previously, port markets play a monopolistic role as a result of its exclusive and irreplaceable geographical location. However in recent years, market structure has drastically changed due to the fast growth of intermodal and international container transportation, resulting in port markets facing intense competition. The monopolistic nature of many container ports become virtually non-existent and 1 traditionally dominant ports are forced to compete regionally and globally. For example, Cullinane et al. (2004) has noted that the port of Shenzhen in Mainland China has been threatening the position of Hong Kong as the dominant hub in the South China region. Such intense competition between container ports results in the interest of port operators to improve their efficiency. Port efficiency, which measures the utilization of port resources, is of importance to contribute a nation's international competitiveness (Wang et al., 2002). The analysis of port efficiency allows port operators to compare performance of different ports. This allows them to enhance operations and produce as much as outputs with limited resources. Also, port operators can use the information from performance analysis to improve their port planning and operations. 1.2 Difficulties in Port Performance Measurement and Benchmarking In the literature, there have been extensive studies that focus on port performance measurement and benchmarking (Ashar, 1997; Cullinane, 2002; Bichou and Gray, 2004). Topics such as individual performance metrics, performance measurement frameworks, relationship between performance systems and the port environment are studied by many researchers (Bendall and Stent, 1987; Frankel, 1991; Talley, 1994; Fourgeaud, 2000). Ashar (1997) and Cullinane (2002) suggested that a combination of inputs (e.g. labor, various types of equipment, land) and multiple outputs (containers, cargo, ships) can be used as partial productivity measurements to evaluate port performance. Cullinane and Wang (2006) argued that one weakness of partial productivity measures is that it is difficult to evaluate 2 the overall impact of multiple variables on port performance. Therefore, some researchers have focused on developing a total factor productivity measure to evaluate port performance (Kim and Sachish, 1986; Talley, 1994). For example, Talley (1994) used the shadow price of port throughput per profit dollar as the single performance indicator to evaluate port performance. No consensus on a single framework for port performance benchmarking has been established to date. Bichou (2006) reviewed the most practical and theoretical approaches to port performance measurement benchmarking over the last three decades and summarized the core differences in these studies(Roll and Hayuth,1993; Christmann and Taylor, 2001; Tongzon 2001; Valentine and Gray, 2001; Langen, 2002; Wang et al., 2002; Barros, 2003; Cullinane et al., 2004; Harahap et al., 2005):  Fundamental differences on the principle to define and classify port performance, i.e. whether port performance is shown by efficiency, productivity, utilization, effectiveness or other economic concepts (Wang et al., 2002);  Fundamental differences on benchmarking contexts measured by individual or combined indicators, such as container throughput, ship working rate or ship calls (Roll and Hayuth,1993; Tongzon 2001; Cullinane et al., 2004);  Perceptual differences among multi-institutional port stakeholders, such as operator, regulator, customer and other participants and the resulting impact on the objective, design and implementation of performance frameworks and analytical model (Christmann and Taylor, 2001; Harahap et al., 2005); 3  Boundary-spanning complexities of port operational dimension, such as the types of ships serviced, terminals managed, systems operated and spatial dimension, such as port cluster, port, terminal, quay system and yard system resulting in confusion on what to benchmark against and how to measure (Langen, 2002);  Dissimilarities exist in both space and time of the studied ports, resulting in the different institutional models, functional scopes and strategic orientations (Valentine and Gray, 2001; Barros, 2003). Cullinane (2002) argued that there is a lack of systematic and unified approach to measure the performance of ports with different inherent characteristics. Langen (2004) claimed that although the port is a cluster of economic activities where a large number of firms provide products and services and together create different port products, ports are often dissimilar in characteristics. Even within a single port, the potential port-related activities can change over time. Therefore, it is not easy to determine a standard method with appropriate indicators to benchmark the performance of ports with different characteristics. As a simplification to the complex problem, many recent studies (Tongzon, 2001; Park and De, 2004) have chosen to analyze the performance of port terminals since they are the most essential component of ports because the quay transfer operations and yard operations in the terminal fundamentally decide the efficiency of a port (Cullinane et al., 2005; Langen, 2007). Port throughput is one of the most widely used port performance indicators (Tongzon, 2001; Wang and 4 Cullinane, 2006). The growth in throughput is regarded as a direct evidence of port’s performance. Although throughput is an important indicator evaluating a port’s overall performance, it may not be sufficient to measure the economic impact of a port on the region. Other performance measures such as the port value added as percentage of regional GDP and profitability of firms in port may be better to measure impact of port to regional economy although they are not able to measure port efficiency. The United Nations Conference on Trade and Development (UNCTAD, 1976) suggests port financial indicators such as tonnage worked, berth occupancy revenue per ton of cargo and labor expenditure as measures of port performance from the economic perspective. In recent years, increasing attention has been paid to service measures that reflect the performance of port operations. This includes waiting time and service time of arrived ships and ship working rate. 1.3 Significance of Port Performance and Efficiency Study Performance measurement is important to organizational development. Dyson (2000) claimed that performance measurement plays an essential role in evaluating production at its current and future state. By appropriately measuring performance, the system within an organization can be tweaked to move towards a desired direction through analyzing behavioral responses and understanding the impact of various performance measures on port efficiency. However, misspecified performance measures will lead the organization to the wrong direction and will cause unintended negative consequences. 5 The performance of a port can influence the economic growth of a region greatly because ports connect the sea transport and inland transport modes. They are also crucial providers for the activities of vessels, cargo and inland transport. A port with good performance provides satisfactory service for ships and efficient cargo operations and contributes to the economic development of a region. Inefficient operations cause wastage of resources. Analysis on port efficiency provides operators with clear ideas about the extent to which a port’s resources are employed and helps them to compare their advantages and disadvantages. Measurement of port performance improves port development and maintains its competitiveness in an increasingly competitive commercial environment. Therefore, it is meaningful to first conduct a comprehensive study to identify port performance indicators relevant to the activities of vessels, cargo and terminals. Through the analysis of ports efficiency using identified indicators, insights on port performance benchmarking on an international scale can be obtained. 1.4 Objectives In this thesis, performance benchmarking of global container ports using efficiency analysis is studied. There are three important objectives of this thesis: 1. To determine an appropriate method to evaluate port’s efficiency using nonparametric approaches. 2. To develop appropriate probability models to relate ship turn-around time to the characteristics of ships and ports. 6 3. To study port efficiency with consideration to the ship turn-around time using an improved non-parametric approach. 1.5 Organization of Thesis Chapter 1 provides the background of the study of port performance benchmarking and efficiency analysis and highlights the objectives of the current research work. Chapter 2 reviews the existing literature on the measures of port performance study and the relevant research studies on ship turn-around time in port industry. The concepts of performance metrics and index methods, economic impact studies and frontier approaches are described and the applications of classical operation strategies and logistic process simulation in the port industry considering ship turn-around time are discussed. The needs of current research are highlighted based on the limitations of past studies and the scope of this research work is defined. Chapter 3 presents in detail the methodology used in this research work. The concept and formulation of non-parametric approaches include the FDH, DEACCR and DEA-BCC models used to estimate port efficiency are described. Probability models include three count data models and five duration models that used to study the ship turn-around time are discussed, followed by the description of the T test on individual regression coefficients and the temporal stability test. Chapter 4 evaluates the efficiency of 61 international ports using DEA-CCR, DEA-BCC and FDH models. A comparative study between the FDH and DEA 7 method are conducted focusing on the analysis of the average efficiency of ports at the aggregate level, individual port efficiency and identifying factors affecting port efficiency. Chapter 5 explores the relationship between ship turn-around time and port’s infrastructure, ship’s characteristics and other factors using probability models. Three count data models based on discrete probability analysis and five duration models based on continuous probability analysis were evaluated in order to select the model provide the best fit. The temporal stability and the elasticity of variables of the selected model are further analyzed to understand the impacts of variables on ship turn-around time. Chapter 6 discusses the efficiency results of container ports considering ship turnaround time in DEA models. Efficiency results of 61 world’s leading container ports are determined based on single-output-measure and multiple-outputmeasures DEA-CCR and DEA-BCC models. Container throughput and ship turnaround time are considered as output measures in DEA models to evaluate port efficiency. Chapter 7 summarizes the main conclusions of this study and provides recommendations and directions for further research. 8 CHAPTER 2: LITERATURE REVIEW This chapter shall present a review of the literature on a few major aspects of this research. Methods related to the study of port performance and efficiency are first introduced. Three broad approaches that can be used to study port benchmarking performance are introduced: (1) performance metrics and index methods, (2) port impact studies and (3) frontier approaches (Bichou, 2006). Research studies on ship turn-around time in port industry are then presented, including the applications of ship turn-around time in port classical operation strategies and port logistic process simulations. 2.1 Performance Metrics and Index Methods Performance measurement in ports and terminals begins with identifying individual metrics at different functional or operational levels. A performance metric can be used to evaluate the performance of ports. It is expressed numerically in order to quantify the attributes of a port and allow for comparing the performance between different ports. Performance metrics include input measures (such as time, cost and resource), output measures (such as production, throughput and profit) and composite measures (such as productivity, efficiency, utilization, profitability and others). To evaluate the performance of an object, a performance metric can be a single measure or a combination of any of the three measures. Composite measures are usually expressed by the ratio of output to input, with the objective to maximize the output within the given input or minimize input while satisfying the required amount of output. Each composite index can be further broken down into two or more components on the basis of 9 approach, typology and the dimensions of performance. For example, in the production economics literature (Aigner and Chu, 1968; Afriat, 1972), efficiency encompasses at least three dimensions: technical efficiency, allocative efficiency and distributional efficiency. Technical efficiency reflects the ability to produce the maximum level of output without requiring more inputs or to reduce the input to the minimum given the same output. Allocative efficiency considers the costs or profits of production and reflects the ability to allocate inputs optimally with a minimum cost of outputs, for a given input price and technology. On the other hand, the distributional efficiency is related to the choice of consumers or welfare optima. It refers to the effectiveness with which a social benefit reaches its intended beneficiaries. The definition of technical efficiency can be simply illustrated in Figure 2.1. Points A, B and C represent three different producers; x-axis represents inputs and y-axis denotes outputs respectively. The productivity of point A is measured by the ratio DA/OD and the efficiency of point A is measured by the ratio of the productivity of point A to that of point B* with the maximum output given the same input, as shown in Eq. (2.1) Technical efficiency  AD / OD B* D / OD (2.1) where B* is the point with the maximum output given the same input as A. Technical efficiency reflects the ability to maximize the output within a given amount of inputs (output-oriented) or to minimize the input but given the same 10 output (input-oriented). In case of point A in Figure 2.1, productivity can be improved by moving from point A to point B without changing input. productivity of C productivity of B and E Output y C B F productivity of A G output of A E A D O input of A Input x Figure 2.1 Definition of Technical Efficiency Source: Derived from Coelli et al (1998, p. 5) When the monetary information of input and output, i.e. price, cost and revenue in each producer (such as point A, B and C in Figure 2.1) is given, allocative efficiency can be estimated based on either the assumption of profit maximization or cost minimization. For example, given output of A as shown in Figure 2.1, the allocative efficiency of point A can be estimated based on the assumption of profit maximization, as shown in Eq. (2.2) Allocative efficiency  AD  revenue B* D  revenue (2.2) or based on the assumption of cost minimization, as shown in Eq. (2.3) E *G  cost Allocative efficiency  AG  cost (2.3) where B* is the point with the maximum output given the same input as A and E * is the point with the minimum input producing the same output as A. 11 Port performance measurement research has shifted from the utilization and effectiveness dimensions of port performance to the efficiency dimension due to a lack of uniformity on standard productivity (Bichou, 2006). An efficiency measure is defined as the ratio of actual quantity of output to the actual quantity of input. Depending on the range and nature of the selected inputs and outputs, financial productivity measures and physical productivity measurements can be defined. Physical indicators generally focus on the measurement of quay transfer operations in the terminal and are mainly concerned with ship-related parameters, such as ship turn-around time, berth occupancy rate, working time at berth. Financial productivity measures usually focus on assessing cost or profit of a port’s throughput. Measures include charge per twenty foot equivalent (TEU), total income and expenditure related to net registered tonnes (NRT) or gross registered tonnes (GRT) (Bichou and Gray, 2004). There are single factor productivity indicators (SFP), partial factor productivity indicators (PFP) and total factor productivity indices (TFP) for use as performance metrics and their selection depends on whether single or multiple-input and output models are used to evaluate port efficiency. 2.1.1 Financial Metrics and Financial Productivity Measures Financial metrics use monetary values of inputs and outputs to estimate port performance. Financial performance measurement is rooted to the concept of profitability i.e. the difference between a firm's total revenue and total costs. The financial productivity of a port is defined as the ratio between revenue and cost, shown in Eq. (2.4) to (2.6). 12 Financial Productivity  Revenue Cost (2.4) Revenue = revenue earned from a services to cargo (handling rates, warehousing, consolidation, etc.) and services to ship (mooring, pilotage, wharf dues, bunkers, ship repair, etc.) (2.5) Cost = total cost of capital, labor, time (expressed in cost/monetary unit) and other expenses (2.6) Financial ratios are applied and the most comprehensive and cited study is the annual survey of financial performance of US public ports (MARAD, 2003). Common measures for financial performance include return on investment or assets, short-term liquidity and capital structure. Conventional financial ratios are not suitable for port performance measurement and benchmarking for a number of reasons. Bichou (2006) argued that financial performance has little correlation with the effective and efficient use of port resources as higher profitability can be driven by price inflation or other external conditions rather than by efficient utilization or productivity. Moreover, the focus on short-term profitability when using financial ratios is not consistent with the nature and goals of long-term investments. This is because dissimilarity exists between various costing and accounting systems when one wants to compare ports from different countries. Even within a single country, port financing and institutional structures, such as ownership, landlord and tool are hardly comparable. Other aspects that influence the financial performance of a port 13 include price and access regulation, statutory freedom, access to private equity and market power (Bichou, 2006). Because financial productivity measures are incapable of measuring port efficiency, physical productivity measures are considered to be more reliable in evaluating port efficiency. 2.1.2 Physical Productivity Measurements Single productivity indicator (SFP) is defined as the ratio of a single output quantity to a single input quantity as shown in Eq. (2.7): SFP  Output Input (2.7) Typical input includes resources such as labor, capital and land while the output quantity may be the cost drivers of the measured activities or resources. It is usually difficult to obtain data of cost drivers in ports and they are usually replaced by physical productivity measures, such as container throughput (Cullinane et al., 2005). Partial productivity indicator (PFP) compares a subset of outputs to a subset of inputs when multiple outputs and inputs are involved. For example, PFP ratios include crane throughput per machine hour when evaluating port performance and is defined as: Crane throughput per machine hour  throughput of all cranes (2.8) No. of cranes  total working hour where throughput of all cranes (a subset of throughput) is the output, the number of cranes and total working hour ( a subset of physical facilities) are the inputs. 14 Examples of PFP ratios in ports include gang or worker output per man-hour and quay or berth throughput per square-meter capacity. SFP and PFP measures try to capture the change in productivity caused by a single factor or a subset of factors respectively. They are both focused on a single or partial form of input and output. There are many studies in the literature that uses physical productivity measurements that falls under the category of SFP or PFP (UNCTAD, 1976; Bendall and Stent, 1987; Monie, 1987; Frankel, 1991; Talley, 1994; Fourgeaud, 2000). Talley (1994) used the shadow price of port throughput per profit dollar as the single performance indicator to evaluate port performance. Bendall and Stent (1987) suggested that throughput is the appropriate output measure and input measures should consider factors related to time, capital and labor when estimating the port productivity. However, in literature related performance metrics measures, many studies only provide ‘snap-shot’ measurements for a single port operation (discharging, storage, loading, distribution, etc.) or port facility (berth, crane, warehouse, etc.) (Bendall and Stent, 1987; Fourgeaud, 2000). For example, Fourgeaud (2000) suggested that the technical capacity of a terminal can be measured by the ‘snap-shot’ performance, i.e. the average number of ship calls and the average flow volume over a standard period time. Port authorities had used the container throughput in 20-foot equivalent units (TEUs) to rank container ports and terminals worldwide and this is a ‘snap-shot’ measurement of port performance. Port performance measured by container throughput can be misleadingly since it is assumed that throughput is equal to efficiency or productivity. 15 In some studies, composite metrics may be used as physical productivity measurements to evaluate port performance (Drewry Shipping Consultants, 1997; Commission, 1998). This includes the number of containers per hour versus the size of ship (Drewry Shipping Consultants, 1997) and the net crane rate by liner shipping trade (Commission, 1998). Connectivity and accessibility to land transportation modes is also an important port productivity indicator. Cargo dwell time (the total time the cargo unloaded from a ship to its departure from the port), may be used in conjunction with time-based utilization metrics such as average ship service time and berth occupancy rate. A utilization ratio compares the input actually used against that of available resources and is defined as: Utilization  used inputs actual inputs (2.9) where inputs can be the physical facilities of a port, such as the number of berth, terminal area, the number of quay cranes and yard cranes. For example, the utilization of quay cranes in a port is the number of working cranes versus total number of cranes. However, both utilization metrics and single productivity measure are not suitable for performance studies as port performance cannot be assessed based on a single value or measure (Ashar, 1997; Cullinane, 2002). In a typically complex port operation system, SFP and PFP indicators are considered to be incomplete when measuring performance. 2.1.3 Total Factor Productivity Measurements Total factor productivity (TFP) combines multiple inputs and outputs into port performance measurement by using an aggregate index or using indices estimated 16 from cost or production functions. TFP synthesizes the productivity index by assigning weights that reflect the relative significance of costs and production components as shown in Eq. (2.10). M TFP  a Y m 1 K m m b X k 1 k (2.10) k where am and bk are the weights, M is the number of outputs and K is the number of inputs. The output weights and input weights must each sum to one. It is important to choose proper weights for inputs and outputs in practice. A basic assumption in TFP measures is that output and input market achieve productive efficiency (i.e. output price = marginal cost and input prices = marginal product value) so that the weight are estimated by output and input share in total revenue and cost respectively (Estache, 2004). Primarily, a TFP index can be obtained directly from data without needing statistical estimation from a production or cost function. However, this requires information on output and input data, namely the price, revenue share and cost. When the data is unavailable, estimation of weights from production functions or econometric models may be used. Past studies on port efficiency have made use of TFP to measure port performance. Kim and Sachish (1986) used labor and capital as input and throughput in metric tons as output to develop a composite TFP index to measure port performance. Talley (1994) suggested that a shadow price variable should be used as a TFP index for evaluating a port’s performance with respect to its 17 economic optimum throughput. Sachish (1996) developed a linear programming model with an objective function to minimize deviations between calculated and actual productivity to obtain a total productivity index. Lawrance and Richards (2004) developed a decomposition method for a total productivity index to calculate the distribution of the benefits from productivity improvements between customers, labor and shareholders in an Australian container terminal. The main advantage of TFP measurements is that overall impacts of the changes in multiple inputs on total output are shown. However, the results of TFP depend largely on the definition of weights and the technique used to estimate the weights and as such, different results may be obtained. 2.2 Port Impact Studies Port impact studies investigate the relationship between port trade and the regional economic impacts. Port impact studies literature typically involve: port economic impacts and port trade efficiency studies (Bichou, 2006). 2.2.1 Port Economic Impact Study Port economic impact study is an important aspect of determining the regional economic influence of a port. It is useful in determining the capital and operating budgets for publicly-owned port facilities and any decision of local governmental agencies to construct port facilities is often preceded by a port economic impact study (Waters, 1977; Yochum, 1987). In port economic impact studies, ports are considered as economic catalysts for their neighboring regions as the aggregation of port activities and services generates benefits and socio-economic wealth. For example, the volume of import or export cargoes transported to the hinterland can 18 be affected by port performance. In this aspect, port performance is measured in terms of its ability to produce maximum output and economic wealth. Davis (1983) discussed the economic impacts on the port region resulting from market demand and supply that directly affect trade volume through a port. Rodrigue et al.(1997) studied the relationship between economic changes and transport geography. Maritime systems are being investigated from the perspectives of transport supply and demand, containerization and spatial diffusion and the adaptative capacity of transport networks. Langen (2002) applied the concept of clustering to maritime industries in the Netherlands and identified four agglomeration economies that attract firms to cluster, namely a joint labor pool, a broad supplier and customer base, knowledge spillovers, and low transaction costs. Much of the past research on port economic impact studies is based on inputoutput analysis (I-O) (MARAD, 1978; Hamilton et al., 2000; Boske and Cuttino, 2001). I-O is a method of systematically quantifying the mutual interrelationships among the various sectors of a complex economic system. It is expressed by a set of linear Equations where the outputs of various branches in the economy are calculated based on an empirical estimation of inter-sector transactions, as shown below: y1  a1 x1  a2 x2  ...  an xn y2  b1 x1  b2 x2  ...  bn xn (2.11) ... ... ym  c1 x1  c2 x2  ...  cn xn 19 where x ( x1 , x2 … xn )represents the relevant input variables in ports and y ( y1 , y2 … ym ) represents the outputs related to port economy. a ( a1 , a2 … an ), b ( b1 , b2 … bn ) and c ( c1 , c2 … cn ) are the coefficients between outputs and inputs. The US Maritime Administration (MARAD) adopted the I-O method and developed the software package Port Economic Impact Kit (Port Kit) to measure the impacts of ports and port-related activities on a region’s economy (MARAD, 1978). It is perhaps the most comprehensive and regularly updated input-output port model, which was firstly published in 1970s and has become the standard model for evaluating economic impacts of US ports (Boske and Cuttino, 2001). Hamilton et al.(2000) developed a software to evaluate the economic impact of existing rural inland waterways ports and terminals in US. Input-output models have also been applied to assess the impacts of existing port facilities (Moloney and Sjostrom, 2000) and to justify future port investments (Le Havre Port, 2000). The gravity model can also be used to model trade flows and analyze its economic impact on inland cities (Wilson et al., 2003). The basic structure of the impedence function in the gravity model is defined in Eq. (2.12). ln(VIJt )  b1 ln(100  TARIFFIJt )  b2 ln x1x2 ...xn  b3 ln( DISTIJ )  eJIt (2.12) where the b term ( b1 , b2 , b3 ) are coefficients, I is importer and J is the exporter. t denotes trading years; VIJ is the value of manufactures exports from country J to country I ; TARIFFIJt denotes the applied ad valorem tariff specific to trading partners I and J in year t . The term x ( x1 , x2 ,..., xn ) denotes the factors related 20 to port economy; DISTIJ is the geographic distance between capital cities I and J and eJIt is the error term. The gravity model assumes that the amount of trade between two countries increases with the size of country (measured by the national incomes) and decreases with the transport cost (measured by distance) (Tinbergen, 1962). Khadaroo and Seetanah (2008) applied the gravity model to evaluate the importance of transport infrastructure in determining the tourism attractiveness of destination, taking into consideration the number of ports in each country. The main disadvantage of using input-output models and gravity models in port economic impact studies is that they are not suitable for benchmarking port performance. This is because each port-country has its own economic structure and a separate inter-sectoral configuration. In addition, the data relevant to the port economic impact studies by input-output models and gravity models (such as the profit, price and cost of cargo, transport and labor cost) are limited. 2.2.2 Port Trade Efficiency Studies Port trade efficiency has recently been of importance to researchers due to the growing importance of understanding the role of ports in trade facilitation. Better trade facilitation allows improved efficiency in administration and procedures as well as enhanced logistics at ports and customs (Wilson et al., 2003). In most port trade efficiency studies, port efficiency is often studied in conjunction with evaluated in relation to transport and logistics costs (Clark et al., 2004; Haddad et al., 2010) 21 Relevant literature in the field of port trade efficiency includes research works by Hofmann (2001), Micco and Pérez (2001), Fink et al. (2002), De and Ghosh (2003), Sanchez et al. (2003), Clark et al. (2004) and Haddad et al. (2010). Most of these studies focus on evaluating the impact of port efficiency on maritime transport cost. Computable general equilibrium models (CGE) and principal component analysis (PCA) are the two types approaches that have been widely applied. Computable general equilibrium models (CGE) are useful tools for understanding and managing the changes in a structure or system. CGE models incorporate production at a level of aggregation that permits the analysis of structural change and captures the essential interdependent nature of production, demand and trade within a general equilibrium framework (Dio et al., 2001). Over the last decade, CGE models have become increasingly popular with applications across different sectors (Devarajan and Rodrik, 1991; Buckley, 1992; Kim and Hewings, 2003). Devarajan and Rodrik (1991) used the CGE model to study the economic impacts of trade reform policies on Cameroon. The marginal cost in the CGE model is defined in Eq. (2.13), 1 3 w rP k MC  ( ) ( m )m  ( k )k   INT j ( N  X ) a m1  m  j (2.13) where MC is the marginal cost, wm is the wage of labor group m and INT j is the intermediate input purchase of sector j . N is the number of firms; X is the output per firm; P k is the price of capital goods and a , r and  m are assumed parameters. 22 Dio et al.(2001) used the CGE model to analyze efficiency improvements at Japanese ports, finding that the technological efficiency improvements in ports results in reduced cost of shipping transportation and growing national GDP. The production of port X in the CGE model is calculated by Eq. (2.14), X i  ai (bi Li i 1 i  (1  bi ) Ki i 1 i ) i i 1 i , (2.14) where X i denotes gross domestic output for sector i , Li represents labor in sector i and K i is capital used in sector i . i is the elasticity of substitution between labor and capital for sector i . ai and bi are assumed parameters. Clark et al. (2004) studied the relationship between port efficiency and transport cost using the CGE model and observed that the inefficiency in ports increases handling costs and reduces maritime trade. Haddad et al. (2010) used the CGE model to simulate the impacts of increases in port efficiency on the transport network system in Brazil and noted that improvement in port efficiency may attract more trade with other countries. CGE model has also been applied to quantify benefits of improved port efficiency on trade facilitation (APEC, 1999) and to study the impact of anti-competitive practices on port and transport services (Fink et al., 2002a). Principal component analysis (PCA) can also be used to evaluate the impact of port efficiency on maritime transport cost. Sanchez et al. (2003) examines the determinants of waterborne transport costs with emphasis on the efficiency at port 23 level by the PCA, finding that more efficient seaports are associated with lower freight costs. The cost of maritime transport in the PCA model is defined as: ijk   ( I , J , k )  mc(i, j, k ) (2.15) where ijk is the cost of maritime transport per unit of weight of product k as transported between points i and j , i is the port of origin located in country I and j is the port of destination located in country J .  is the markup and mc is the marginal costs. Both marginal costs ( mc ) and markup (  ) are assumed functions of factors dependent on the port. De and Ghosh (2003) employed PCA to study the relationship between port performance and port traffic. It was found that higher efficiency induces higher traffic at most of India ports and suggested that government should give priority to improve port performance by enhance facilities. Tongzon and Heng (2005) used PCA to investigate the quantitative relationship between port ownership structure and port efficiency and found that private sector participation in the port industry may improve port operation efficiency. CGE and PCA models are not suitable for measuring the performance of multiinput and multi-output port production systems. In addition, CGE models also have other limitations. For example, the assumptions of perfect competition between ports and the freely move of capital and labor between different sectors in a port are inconsistent with the actual port industry structure in practice. 24 2.3 Frontier Approaches There are many frontier-based methods that can be used to assess port efficiency. A frontier denotes the lower or upper limit to a boundary efficiency range (Farrell, 1957; Roll and Hayuth, 1993; Liu, 1995). Typically, a statistical central tendency approach is employed to evaluate performance of an average unit or firm. A central tendency in statistics is a central value or a typical value for a probability distribution (Weisberg, 1992) and can be calculated for a finite set of values to indicate the tendency of quantitative data to cluster around some central value (Dodge, 2003; Upton and Cook, 2008). The simplest measure of central tendency is the arithmetic mean, which is defined by Eq. (2.16): central tendency  1 n  ai n i 1 (2.16) where n is the number of units. Unlike the statistical approach, the frontier approach focus on evaluating the efficiency through the estimation or calculation of an efficiency frontier. Under this approach, units are deemed to be efficient when they operate on the production or cost frontier and inefficient units are found either below or above the frontier. Inefficient units operate below the frontier in a production frontier and operate above it in the situation of a cost frontier. Frontier approaches often used in efficiency analysis as its concept is consistent with the economic theory of behavior optimization (Bauer, 1990). The technical efficiency of inefficient units can be interpreted by its distance away from the frontier allowing a relative comparison of economic units when performing benchmarking port performance. 25 Figure 2.2 illustrates the concept of a production frontier. Units A, B, C, D and E represent five different producers. The x-axis represents inputs and the y-axis denotes outputs. Efficient units B, C and E together constitute the production frontier while inefficient units A and D are below the frontier. It is obvious from the figure that unit B can produce more outputs than unit A while using use the same amount of inputs. This means that unit A is inefficient. Output y C B Frontier A E D O F Input x Figure 2.2 Illustration of a Production Frontier Farrell (1957) proposed analyzing economic efficiency using deviations from an idealized frontier isoquant. In order to estimate the degree to which an individual unit’s actual operation deviates from an efficient frontier, for example AF , the BF precise location of the frontier has to be determined. Two methods can be used to locate the frontier, namely parametric methods or non-parametric methods. Parametric methods assume a particular functional form of variables while nonparametric approaches do not need such a pre-defined production function. Nonparametric approaches often require mathematical programming and one of the 26 commonly applied techniques is data envelopment analysis (DEA). Parametric method, on the other hand, makes use of econometric methods to estimate the statistical frontier production function. 2.3.1 Parametric Approaches Parametric or econometric approaches require a functional form to statistically estimate the frontier given a group of input and output observations. Frontiers can be either deterministic or stochastic, depending on the assumptions made. In both cases, inefficiencies are reflected by the error term which is essentially the deviations from the frontier. Deterministic frontier assumes that the deviations are exclusively due to a certain economic inefficiency (Coto et al. 2000). In a frontier cost function, it is defined as: efficiency  f ( wp , y p ;  ) C p  1 exp(u p ) u p  0 (2.17) where C p is the cost of the p -th firm, w is the price vector of the inputs, y is the output vector, f (w, y,  ) represents the minimum cost and u p represents the deviations of the cost of each firm from the minimum cost. Early parametric frontier models (Aigner and Chu, 1968; Afriat,1972) were deterministic in nature as analyzed economic units are assumed to commonly share a fixed form of frontier. Researchers believed that this assumption is an oversimplification and the validity of the frontier is being compromised (Coto et al., 2000; Cullinane et al., 2002; Cullinane and Song, 2006). The assumption which uses u p as the fixed form of frontier representing the economic 27 inefficiency does not take into consideration the possible exogenous factors (such as random shock) and endogenous factors (such as inefficiency) associated to an economic unit’s observed performance. Stochastic frontier model is therefore considered to be an enhancement over deterministic frontier model and is based on the concept that deviations from either a production frontier or cost frontier are probably not entirely under the control of the studied economic units (Greene, 1993). It takes into account the random and uncontrolled factors that may affect the production and costs of a firm. Therefore, the error term (with random and uncontrolled effects) have to be considered in the frontier cost function, shown in Eq. (2.18): C p  f (wp , y p ;  ) exp( p ),  p  v p  u p (2.18) where v p represents the random effect and u p for economic inefficiency. The error term consists of two elements: a component which captures the inefficiency effects related to the stochastic frontier and another (symmetric) component that allows for the random variation of the frontier across firms. Measurement error such as statistical ‘noise’ as well as random shocks outside the control of firms can therefore be captured. The stochastic frontier models not only permit the evaluation of technical inefficiency, but also allow the study of random shocks outside the control of firms. Table 2.1 provides an illustration of the major applications of parametric approaches in port research. Of all the studies presented in the table, the two most 28 commonly used functional forms were the log-linear Cobb–Douglas form and the quadratic form. Liu (1995) applied a set of panel data of 28 commercially important ports in the UK to test for the hypothesis that private sector ports are inherently more efficient than those in the public sector. It was found that ownership cannot be identified as an important factor of production and there is no evidence that private sector ports are more efficient than ports with other ownership types. Cullinane et al. (2002) found that the transformation of ownership from public to private sector improves economic efficiency, based on his panel data from 15 Asia ports or terminals. The size of a port or terminal is noted to be closely correlated with its efficiency (Cullinane et al., 2002). Coto et al. (2000) found that the most efficient ports are often those which are smaller in size and managed under a more centralized regime. One main concern on the application of parametric models in port performance studies is its requirement of a pre-defined frontier function. The structure of port production may limit the econometric estimation of a production or cost function to the level of a single port or terminal. This is considered to be suitable for international port benchmarking (Braeutigam et al.,1984; Kim and Sachish, 1986). Furthermore, the use of parametric frontier function is also not suitable for the multi-input and multi-output port systems (Bichou, 2006). 29 Table 2.1 Literature review on parametric approaches to the port sector Model Functional Variables form Stochastic Cobb-Douglas Billing for services output production Translog Labor input frontier Capital input Port size Port location Port ownership Capital intensity Time trend Author Data Liu (1995) Panel 28 UK ports 1983-1990 Coto et al. (2000) Panel 27 Spanish ports 1985-1989 Stochastic cost frontier Cobb-Douglas Translog Cullinane et al. (2002) Panel 15 Asia ports or terminals 1993-1998 Stochastic production frontier Log-linear Cobb– Douglas Cullinane and Song (2003) Panel and crosssectionsl data 65 observations from UK and Korea 1978-1996 Stochastic production frontier Log-linear Cobb– Douglas Tons of merchandise or TEUs handled output Labor price Capital price Intermediate input price Time trend Annual container throughput in TEUs output Terminal quay length input Terminal area input The number of pieces of cargo handling equipment employed input Labor inputs: the total remuneration of directors and total salaries paid to employees Capital inputs: the net book value of fixed equipment, buildings and land terminal operations Terminal output: the turnover derived from the provision of terminal services 30 Main findings Fail to identify ownership as an important factor of production and the evidence does not establish a clear-cut pattern of efficiency in favor of one or other type of ownership The most efficient ports are those which are smaller in size and managed under a more centralized regime. The size of a port or terminal is closely correlated with its efficiency and the transformation of ownership from public to private sector improves economic efficiency In Korean and UK container terminal industries, improved productive efficiency has followed the implementation of privatization and deregulation Table 2.1 Literature review on parametric approaches to the port sector (Continued) Model Functional Variables Main findings form Jara et al. Panel Stochastic Quadratic Dependent variables: Total annual Liquid bulk and non(2003) 26 Spanish ports cost frontier Form cost, labor, amortization and other containerized general cargo 1985-1995 expenses respectively show the lowest and Explanatory variables: five output largest marginal cost and port components and three indices for input specialization is not appropriate prices in terms of infrastructure Tovar et al. Panel data Stochastic Quadratic Tons of containerized general cargo The operation of port terminals (2003) 3 port terminals cost frontier Form Tons of general cargo for port should be analyzed by means of 1990-1999 Tons of ro-ro cargo for port multiproduct theory and the Non port worker personal price calculation of cost indicators Ordinary port work price such as marginal cost, economies Special port worker price of scale are key tools to port Intermediate input price regulators Capital price Firm-specific dummy variable Temporal trend Author Data 31 2.3.2 Non-Parametric Approaches Unlike parametric models, non-parametric approaches use linear programming to determine the efficiency frontier. Two mathematical programming techniques are commonly applied to estimate the efficient frontier. They are data envelopment analysis (DEA) and free disposal hull (FDH). Both DEA and FDH solve a series of linear programming problems and determine the optimal solution that maximizes each decision-making unit’s (DMU) efficiency (defined as ratio of its weighted output to input). The rationale behind the two techniques is that, in seeking to solve the problem by assigning DMUs different value weights to their inputs and outputs, each DMU will be given the optimal combination of weights that guarantees them a most favorable position compared with other DMUs. The efficiency frontier is determined by some of these DMUs and by measuring the relative distance an individual observation lies away from the efficient frontier. The efficient frontier of DEA models (such as CCR, BCC) and FDH are illustrated by Figure 2.3. The DEA-CCR, DEA-BCC and FDH models are defined by Eq. (2.19) to (2.24). (2.19) max U U , Subject to Uy 's  Y '   0 (2.20) X '   x 's  0 (2.21)   0 (DEA-CCR) (2.22) 32 e '  1 (DEA-BCC) (2.23) s {0,1} (FDH) (2.24) where inputs are xs  ( x1s , x2 s ...xms )  Rm , producing outputs ys  ( y1s , y2 s ... yns )  Rn . The row vectors xs and ys form the s th rows of the data matrices X and Y , respectively. s  (1 , 2 ...s )  Rs is a non-negative vector that forms the linear combinations of the s DMUs and e  (1,1,...,1) is a suitably dimensioned vector of unity values. Output y The production frontier of DEA-CCR E C The production frontier of DEA-BCC D B H The production frontier of FDH G Efficient units in DEA-CCR, DEA-BCC and FDH F A O Input x Figure 2.3 Production Frontier Comparison of DEA models and FDH model Non-parametric approaches are widely applied to measure the efficiency of units with multiple outputs and inputs. Wang et al. (2003) used both FDH and DEA techniques to analyze the efficiency of container terminals from 28 important international ports and concluded that a combination of both techniques can be of 33 great significance to help port authorities make strategic decisions. Cullinane et al. (2005) evaluated the efficiency of the world’s important container ports using FDH and DEA and noted that the results are different in DEA and FDH models. It is shown in their studies that a lack of sensitivity in the FDH model when measuring the efficiency of DMUs. This is because the underlying logic and step function solution algorithm of FDH assumes a strong disposability of input and output, i.e. outputs can always remain feasible with any increase of the inputs or with given inputs it is still possible to reduce the outputs. This means that some DMUs may be regarded as efficient units when it is actually not efficient. As illustrated in Figure 2.3, units B, D and F are efficient in FDH but are considered as inefficient in DEA models. Many studies in the literature have assessed the efficiency of container ports or terminals using DEA models. DEA was first applied by Roll and Hayuth (1993) in their study to evaluate port performance and efficiency. Since then many researches have applied DEA for port research. A detailed review of DEA applications to port economic efficiency research is summarized in Table 2.2 which provides a review of the major studies undertaken to date (Estache et al., 2002; Panayides et al., 2009). Literature on the application of DEA models in port performance and efficiency studies may include the following five broad considerations:  DEA-CCR model (Valentine and Gray, 2001), DEA-BCC model (Martinez-Budria et al.,1999), a combination of both DEA-CCR and DEA-BCC models (Poitras et al., 1996; Tongzon, 2001; Barros and 34 Athanassiou, 2004; Park and De, 2004; Cullinane et al., 2005; Cullinane and Wang, 2006) and their extensions such as the Additive Model (Tongzon, 2001), the extended CCR and BCC models (Park and De, 2004; Barros, 2006) and the super-efficiency model (Barros, 2006).  Input-oriented models (Barros, 2003; Barros and Athanassiou, 2004; Park and De, 2004) and output-oriented models (Cullinane et al., 2004; Cullinane et al., 2005).  Aggregate port operations (Barros and Athanassiou, 2004) and single port function (Cullinane et al., 2004).  DEA results as a single analysis or DEA followed by a second-stage analysis, such as the regression modeling of port production (Bonilla et al., 2002; Turner et al., 2004; Bichou, 2011; Wanke, 2013).  Cross-sectional data and panel data (Cullinane et al., 2004; Cullinane et al., 2005; Min and Park, 2005; Rios and Maçada, 2006). Despite the fact that the use of DEA in port efficiency studies is more common than FDH, some researchers have argued that FDH prevails over DEA in terms of ‘data fit’(Tulkens, 1993; Vanden Eeckaut et al.,1993). Inefficient observations in FDH can be projected onto an orthant spanned by a single efficient producer which is weakly dominating in both cost (or production) and outputs. This single producer can be interpreted to function as a role model for the inefficient unit which is not available in DEA (Borger and Kerstens, 1996). However, FDH lacks the sensitivity in identifying inefficiency among similar units for the number of efficient observations on FDH is typically larger than on DEA (Lovell and 35 Vanden Eeckaut,1994). As such, it is fair to say that both DEA and FDH have their respective strengths and weaknesses. A comparative study of these two approaches may provide greater insight into the intricacies involved when measuring production efficiency. There are some advantages of the non-parametric approaches when performing efficiency analysis as compared to parametric approaches. Non-parametric approaches are suitable for measuring efficiency of observations with multiple inputs and outputs, as well as providing information about the sources of the relative efficiency. It is also not necessary to pre-define the functional relationship between variables, which does not subject the analysis to subjective weighting and randomness. This means that there is no need to impose a specific cost or production function in non-parametric approaches or assume a functional form. Furthermore, rather than to benchmarking ports against a statistical measure or an exogenous standard, DMUs are benchmarked against a real ‘best’ unit in nonparametric approaches. This makes non-parametric approaches particularly attractive for port efficiency studies. On the other hand, there are some limitations associated with non-parametric approaches. For example, DEA or FDH models do not allow for measurement errors and stochastic factors which are typically considered in parametric approaches. They cannot provide information on statistical significance or confidence intervals due to the lack of statistical functional form. Although a second stage analysis based on regression modeling can be adopted to resolve this issue, the regression assumption of data interdependency and the imposition of a 36 functional form in some degree deprive the major advantage in DEA or FDH models of not requiring to assume a deterministic functional form. Another drawback of DEA stems from the sensitivity of efficiency scores to the choice and weights of variables. A DMU can be considered to be efficient simply because of its patterns of inputs and outputs. Most applications of DEA models in port efficiency studies assumed constant efficiency over time that ignores the incremental impact of port investment. As such, the model favors ports with no investment in new equipment or facilities during the study period. In addition, input saving or output increase potentials identified under DEA models are not always achievable in port operational settings, particularly when small amounts of an indivisible input or output unit are involved in. Other issues related frontier approaches (both parametric and nonparametric) in port literature are highlighted as follows: (a) In all of the studies mentioned above (deterministic or stochastic frontier analysis, DEA and FDH), the determinants (inputs and outputs) of port performance were either subjective or were obtained from empirical analysis. (b) There is still no consensus on an approach which is best suited for analyzing port performance. Both (a) and (b) partly explains the inconsistent results in port performance analysis performed in the literature. For example, Coto et al.,(2000) in their study on the relationship between port size and efficiency found that the most efficient ports are often those which are smaller in size. Martinez-Budria et al. (1999), on 37 the other hand, suggested that there is no direct relationship between port size and efficiency. 2.4 Studies on Ship Turn-around Time Much research efforts related to ship turn-around time are focused on its working time for loading and unloading operations at the berth (Goodchild and Daganzo, 2004; Imai et al., 2005; Golias et al., 2009). This time is considered to be an important component of operational efficiency in container ports (Ramani, 1996; Kia et al., 2000).The objective to improve port efficiency is to minimize ship turnaround time. Past research on this topic can be classified into two categories: classical operation strategies and logistic process simulation. Operational strategy studies includes: double cycling application (Goodchild and Daganzo, 2004), where the double cycling method is demonstrated that it can reduce ship turnaround time by unloading and loading a ship simultaneously; and stowage planning and sequencing (Hamedi et al., 2007) where the stowage planning is optimized at the same time improving the efficiency of crane utilization. In addition, Kim (2002) proposed a method of determining the optimal amount of storage space and the optimal number of transfer cranes for handing import containers, which minimized the terminal operation cost. 38 Author Roll and Hayuth (1993) Table 2.2 Literature review on applying DEA to the port sector Model Parameters Domain Data DMUs Outputs Inputs Entire world Fictitious and 20 ports Container Size of labour force crossthroughput Annual investment per sectional, Service level port single period User satisfaction The uniformity of facilities Ship calls and cargo Poitras et al. (1996) Australian Crossand sectional international 23 ports TEU berth hour Total number of containers handled per year MartinezBudria et al. (1999) Spain Time series (1993-1997) 26 ports in 5 year span Total cargo moved through the docks Tongzon (2001) Australia Crosssectional 21 ports Cargo throughput Ship working rate 39 Main findings DEA is a promising and easily adaptable approach to providing relative efficiency ratings for port performance studies Mix of 20-foot and 40-foot containers, average delays, difference between the berth time and gross working time, number of containers lifted per quay crane hour, number of gantry cranes, frequency of ship calls, average government port charges per container Labour expenditures Depreciation charges Miscellaneous expenditures Port production can follow different returns to scale, i.e., constant, increasing and decreasing returns to scale Capital (number of berths, cranes, tugs), Labour (number of stevedore gangs), Land (size of terminal areas), Length of delay The ports of Melbourne, Osaka, Rotterdam and Yokohama are the four most inefficient ports in the sample, based on the assumptions of constant and variable returns to scale Ports of high complexity offer higher comparative efficiency levels while ports of low complexity display a negative evolution in global efficiency levels Author Domain Valentine Entire and world Gray (2001) Table 2.2. Literature review on applying DEA to the port sector (Continued) Model Parameters Main findings Data DMUs Outputs Inputs Cross21 ports Total tonnes Total length of berth Port performance is related to the sectional throughput Container berth length ownership of ports and its Number of organizational structure by the containers application of DEA Barros (2003) Portugal Panel data 11 ports Park and De (2004) Korea Crosssectional 11 ports Cullinane Worldwide Time series et al. (1992-1999) (2004) 25 ports Ships Movement of freight Gross gauge Break-bulk cargo Containerized freight Solid bulk and liquid bulk Productivity Cargo throughput, Number of ship calls Profitability, revenue Marketability Overall efficiency Customer satisfaction Throughput (TEU) 40 Labour (number of workers) Capital (book value of the assets) None of the 11 port authorities achieved total productivity improvements within the study period, while almost all ports achieved improvements in technical efficiency without technological change Productivity Berthing capacity, cargohandling capacity Profitability, revenue Cargo throughput, number of ship calls Marketability Overall efficiency Berthing capacity, cargohandling capacity Land factor Total quay length, terminal area Equipment factor Number of quay gantry cranes, yard gantry cranes, straddle carriers The efficiency of a port can be divided into four stages, namely productivity, profitability, marketability and overall efficiency and each stage should be analyzed by DEA model Efficiency of the different container ports could fluctuate over time and production scale is not a major source of inefficiency Author Domain Barros and Greece and Athanassiou Portugal (2004) Table 2.2. Literature review on applying DEA to the port sector (Continued) Model Parameters Main findings Data DMUs Outputs Inputs Balanced 6 ports Ships Number of workers More than half of the studied panel Movement of freight Book value of assets ports operated at a high level of data Total cargo handled pure technical efficiency and the Containers loaded dominant source of efficiency and was scale unloaded Cullinane et al. (2005) Worldwide Time series (1992-1999) 25 ports Container throughput (TEU) Terminal length and area Quayside gantry Yard gantry Straddle carrier Total length of quay Number of cranes Size of hard areas Size of labour force There is no clear-cut relation between privatization or the ownership of port and efficiency Min and Park (2005) Korea Time series 11 container terminals in 4 year span Cargo throughput Wang and Cullinane (2006) Pan European Crosssectional 104 terminals Container throughput (TEU) Terminal length Terminal area Equipment costs Many ports with a relative efficiency below 10% and large production scale was more likely to be accompanied by higher efficiency scores Cullinane et al. (2006) Worldwide Crosssectional 57 Container throughput (TEU) Terminal length and area Number of quayside gantry cranes, yard gantry cranes and straddle carriers Large ports with more than one million TEUs throughputs are found to be scale inefficient and small ports are scale efficient 41 The advancements in information technology and terminal operations would lead to the improvements of terminal efficiency Table 2.2. Literature review on applying DEA to the port sector (Continued) Model Parameters Data DMUs Outputs Inputs Time series 23 terminals TEUs handled Number of cranes (2002Average number of Number of berths 2004) containers handled per Number of employees hour per ship Terminal area Amount of yard equipment Author Rios and Maçada (2006) Domain North America (Brazil, Argentina, Uruguay) Barros (2006) Italy Balanced 24 ports panel data (years 2003-2004) Hung et al. (2010) Asia-Pacific Crosssectional 31 ports Liquid bulk Dry bulk Number of ships Number of passengers Number of containers(with TEU) Number of containers(with no TEU) Total sales Container throughput 42 Main findings 60% of the terminals are efficient in the during study period and Zarate, Rio Cubatao and Teconvi are served as reference for inefficient terminals more often than other terminals Number of personnel Value of capital invested Size of operating costs Large ports, more containerized ports and ports with fewer employees tend to have higher efficiency Terminal area Terminal length Ship-shore container gantry cranes Container berths The overall technical inefficiencies of Asian container ports are primarily due to pure technical inefficiencies rather than to scale inefficiencies Author Domain Bergantino Southern And Musso European (2011) Table 2.2. Literature review on applying DEA to the port sector (Continued) Model Parameters Data DMUs Outputs Inputs Balanced 18 ports Movement of cargo Dimension of quay panel data Number of terminals (years Area of the port used for 1995-2007) handling freight Cargo handling equipment Main findings Environmental factors and general economic conditions do have a significant effect on port efficiency Wanke et al. (2011) Brazilian Crosssectional 25 terminals Aggregate throughput Loaded shipments Number of berths Number of trucks Terminal area The efficiency of Brazilian terminals depends mostly on the level of shipment consolidation and private terminals tend to be more efficient than government owned terminals Yun et al. (2011) China Crosssectional 13 Ports Cargo throughput Container throughput Port cargo increasing rate Port container increasing rate Direct hinterland GDP increasing rate Production berths length Production berth quantities Production 10000-toncapacity port berth quantity GDP of direct hinterland The proportion of tertiary industry in port city GDP Port hinterland’s highway length Port hinterland’s railway length Majority of port in China are through a rapid growth and have a great room for development 43 In logistic simulation, Ramani (1996) presented an interactive simulation model to reduce the ship turn-around time by optimally utilizing the port resources. Lee et al. (2003) simulated the logistics planning model of a container terminal in view of supply-chain management. Demirci (2003b) focused on the loading or unloading operations of vehicles and presented a port simulation model which analyzes the investment needed for handling equipment. In addition, automated guided vehicle systems are simulated to reduce delay in ship operations of container terminals (Liu et al., 2000; Kim and Bae, 2004; Liu et al., 2004). Most of these work utilizes queuing theory and stochastic models (Daganzo, 1990), simulation (Lai and Leung, 2000) or classical operation research techniques such as routing, network, and scheduling problems (Kim and Kim, 2002) to apply ship turn-around time as an efficiency indicator for container ports. Such approaches assume that the arrival pattern of ships is a stochastic process described by some type of probability distribution (Fararoui, 1989) and the exponential distribution of ship inter-arrival time is the most commonly used approximation. As a result many researchers have simply assumed the turnaround time to be following the exponential distribution without proof. For example, Kozan (1997) assumed in his analytical queue model for port operations that ship turn-around time is exponentially distributed. Other researchers such as Gaver (1959) had assumed this distribution in his transient Markov chain analysis of arrival in batches and Burke (1975) solved a single-server queuing system with an arbitrary service time using the renewal theorem. 44 Few studies have sought to prove that ship turn-around time follows a specified distribution and found out the relationship between ship turn-around time and other factors of ports. Kia et al. (2002) studied the container terminal in Melbourne with the statistical data derived from 1999 real-time operations and concluded the exponential distribution fits the ships inter-arrival time well. With the data of 297 ships arrived at the Trabzon Port, the average inter-arrival time for each type of ship was proven to be following an exponential distribution (Demirci, 2003a). However, such models are only valid for the single port where the model is developed for and could not be transferred to other international ports. Moreover, the models, being a simplistic probability distribution fitting on existing ship dwell time data, offers no indication on the factors that can affect ship turn-around time in the ports. Recognizing this need, Edmond et al. (1976) studied the relationships between ship size, handing rate and ship turn-around time at UK ports. In their work, they found that the dwell time of container ships are extremely varied and the relationships between dwell time and either ship size or cargo handled can be rather complex. 2.5 Research Needs and Scope of work Based on the extensive literature review provided in this chapter, it is noted that although various practical and theoretical approaches have been conducted to study the efficiency of ports, there is still no consensus on an acknowledged method to benchmark port performance. Ship turn-around time as an important indicator that reflects the efficiency of quay crane load and unloading operations 45 in a port is seldom considered in most of the studies. To address these issues, following areas have been identified as needing more research efforts:  To assess the capabilities of different non-parametric approaches to measure port efficiency and to benchmark global container ports.  To develop an appropriate model to study ship turn-around time and analyze the relationship between ship turn-around time and port and ship characteristics.  To study port efficiency with consideration to the ship turn-around time. Data envelopment analysis (DEA) and free disposal hull (FDH) are used to evaluate port efficiency in this study due to its ability to analyze multiple outputs and inputs concurrently. In addition, it does not require a pre-defined functional formulation among the multiple performance indicators. Ship turn-around time is studied because it is perhaps most critical part in service measure-related port efficiency. A shorter ship turn-around time in a port is analogous to a better level of service. Based on the objectives of the research study, the following tasks are identified:  To determine an appropriate method to evaluate port efficiency based on the application of the DEA-CCR, DEA-BCC and FDH models in global container ports. A comparative study between the FDH and DEA method needs to be done focusing on the analysis of the average efficiency of ports at the aggregate level, individual port efficiency and identifying factors affecting port efficiency. 46  To develop discrete probability models to relate the ship turn-around time and variables of ports and ships. Poisson regression model, negative binomial regression model and Poisson regression model with normal heterogeneity are studied.  To develop continuous probability models to relate the ship turn-around time and variables of ports and ships. Exponential, Weibull, log-logistic, log-normal and generalized gamma models are studied.  To analyze port efficiency with consideration to ship turn-around time using an improved non-parametric approach. Port efficiency is evaluated using single output measure and multiple output measures models. 2.6 Summary This chapter reviews the existing literature on the measures of port performance and the relevant research studies on ship turn-around time in port industry. There are three broad categories approaches to benchmark port performance, namely performance metrics and index methods, port impact studies and frontier approaches. Each of the approach has its own advantages and limitations and there is still no consensus to date on an unified method to benchmark port performance. Performance metrics and index methods are relatively easy to perform but has its limitation when selecting appropriate inputs or outputs. Economic impact studies evaluate the economic impacts of ports on their hinterlands and are useful in determining the capital and operating budgets for local governmental agencies. However, it is very data-intensive and requires detailed information such as the profit of cargo and costs of transport and labor. 47 Frontier approaches use an assumed efficiency frontier to estimate port efficiency which is suitable for benchmark port performance. Its main limitation lies in its assumption that may be inconsistent with the actual operations in the port industry. The study of ship turn-around time in port industry main includes the applications on classical operation strategies, logistic process simulation as well as the queuing theory and stochastic models. Few studies have sought to prove that ship turnaround time follows a specified distribution and found out the relationship between ship turn-around time and other factors of ports. The current research emphasizes the importance to study port efficiency with consideration to the ship turn-around time. 48 CHAPTER 3: METHODOLOGY 3.1 Methodology Adopted in Research As introduced in Chapter 1, there are three objectives in this study: (1) to determine an appropriate method to evaluate port’s efficiency using nonparametric approaches, (2) to develop appropriate probability models to relate ship turn-around time to the characteristics of ships and ports, and (3) to study port efficiency with consideration to the ship turn-around time using an improved non-parametric approach. Figure 3.1 shows the research methodology adopted in this research study. For the first objective, a comparative study using the DEA-CCR, DEA-BCC and FDH models to evaluate the efficiency of global container ports needs to be conducted. The concept and formulation of the DEA-CCR, DEA-BCC and FDH models are described in Sections 3.2 and 3.3. For the second objective, three count data models and five duration models are applied to model ship turn-around time. The concept and formulation of the count data models (such as Poisson regression model, negative binomial regression model and Poisson regression model with normal heterogeneity) and the duration models (such as exponential, Weibull, log-logistic, log-normal and generalized gamma models ) are described in Section 3.4. For the last objective, port efficiency is discussed based on the single output measure and multiple output measures using the improved non-parametric 49 Port Performance Benchmarking and Efficiency Analysis Stage I: To determine an appropriate non-parametric approach to evaluate the efficiency of ports Data Envelopment Analysis  DEA-CCR  DEA-BCC Efficiency ranking based on constant or variable returns to scale Free Disposal Hull Efficiency ranking based on non-convex production possibility set Stage II: To develop an appropriate probability model to relate ship turn-around time to ship and port factors Count data regression models  Poisson regression model  Negative binomial regression model  Poisson regression model with normal heterogeneity Hazard-based duration models  Exponential  Weibull  Log-logistic  Log-normal  Generalized gamma Ship turn-around time based on discrete probability analysis Ship turn-around time based on continuous probability analysis Stage III: To develop an improved non-parametric approach taking ship turn-around time into consideration Port efficiency ranking based on single or multiple outputs Figure 3.1 Flow Chart of Research Methodology in Thesis 50 approach. The estimated ship turn-around time from the probability model is used in the non-parametric method for efficiency analysis. 3.2 DEA Technique for Measuring Port Efficiency This section shows the methodology related to the use of DEA models to evaluate the efficiency of global container ports. As one of the most important non-parametric approaches in the literature, DEA is widely used in estimating port efficiency (Roll and Hayuth, 1993; Cooper et al., 2007). It calculates efficiency scores on the basis of the multi-variate frontier estimation. Unlike traditional regression approach, DEA does not need to assume a production function between inputs and outputs but instead construct a piecewise linear line representing the production frontier. The production frontier is determined using observations with best efficiency. 3.2.1 Data Envelopment Analysis Two significant concepts on performance measurement are used in data envelopment analysis: productivity and efficiency. The productivity of a producer can be loosely defined as the ratio of output(s) to input(s) while efficiency can be defined as the relative productivity over time or space (Wang et al., 2002). In this study, efficiency is a measure of performance of a port relative to other ports that has the “best” performance. DEA measures the efficiency of a decision making unit (DMU) given multiple inputs and/or multiple outputs (Wang et al., 2002). In DEA, the individual unit under study is known as a decision making unit (DMU) (Charnes et al., 1978) or 51 the Unit of Assessment (Thanassoulis, 2001). Each DMU with m inputs and n outputs in DEA models can be shown in Figure 3.2 and the total number of DMUs is s . The input and output data for DMU j are ( x1 j , x2 j ,…, xmj ) and ( y1 j , y2 j ,…, ynj ), respectively. The input data matrix X and the output data matrix Y are in Eq. (3.1) and (3.2), where X is an ( m  s ) matrix and Y an ( n  s ) matrix.  x11  x X   21    xm1 x12  y11  y Y   21    yn1 y12 x22 xm 2 x1s   x2 s    xms  (3.1) y1s   y2 s    yns  (3.2) y22 yn 2 input output m inputs … DMU 1 … n outputs m … DMU 2 … n … n … m … DMU s Figure 3.2 DMU and Homogeneous Units 52 When studying port performance or efficiency, the observed port is considered to be a DMU in DEA model and its efficiency score is calculated by the ratio of the productivity of itself to that of other DMUs in the production frontier. In DEA models, the DMU j to be evaluated is designated as DMU o ( o  1, 2,..., s ). The productivity of DMU o can be determined by calculating the ratio between the virtual output with weight ur and the virtual input with weight vi (Cooper et al., 2007) as defined by Eq. (3.3): Productivity = virtual output u1 y1o  u2 y2o  ...  un yno  virtual input v1 x1o  v2 x2o  ...  vm xno (3.3) where the virtual output and input are the estimated value based on the efficiency frontier that defined as the ratio of the sum of weighted outputs to the sum of weighted inputs. The definition of efficiency and productivity is illustrated in Figure 3.3. DMUs A, B and C represent three different producers; x-axis represents inputs and y-axis denotes outputs respectively. Thus the productivity of DMU A is measured by the ratio DA/OD while the efficiency of DMU A is measured by the ratio of the productivity of DMU A to that of DMU B, i.e. AD / OD . The efficiency of DMU BD / OD A shows the productivity can be potentially improved by moving from DMU A to DMU B with an unchanged input. For example, if a DMU represents a port, output y is container throughput and input x is the number of quay-cranes in port. Port A can become as efficient as port B by without requiring more quay-cranes. 53 In this case port A must improve the operational efficiency of quay-cranes in order to load and unload more containers and obtain as much throughout as port B. productivity of C productivity of B and E Output y F B (efficient unit) Optimal C Scale productivity of A output of A A (inefficient unit) E D O input of A Input x Figure 3.3 Illustration of Efficiency and Productivity The efficiency defined in Figure 3.3 is known concept of technical efficiency and the technical efficiency of a unit is calculated by Eq. (3.4): technical efficiency  unit A[Output (s) Input (s)] efficient units[Output (s) Input (s)] (3.4) Technical efficiencies can be input-oriented or output-oriented. This means that producer can either minimize inputs while satisfying the given output levels (input-oriented, i.e. moving from A to E) or maximize outputs without requiring more input values (output-oriented, i.e. moving A to B) through technology improvement. Therefore, it is possible to determine a curve which is known as the production frontier and indicated by OF in Figure 3.3. All DMUs on this 54 production frontier are technically efficient (such as E, C, B in Figure 3.3) while DMUs below the frontier (such as A in Figure 3.3) are technically inefficient. Scale efficiency relates a possible divergence between actual and ideal production size (Wang et al., 2002). It can be explained by the productivity of DMU C in Figure 3.3 where C has the maximum possible productivity. Thus, scale efficiency is defined as: scale efficiency  unit A[Output ( s) Input ( s)] maximum[Output ( s) Input ( s)] (3.5) 3.2.2 Alternative DEA Models The two most widely used DEA models are the DEA-CCR (Charnes et al., 1978) and DEA-BCC models (Banker et al.,1984). The key difference between CCR and BCC models is that the CCR model assumes a constant returns to scale (CRS) while the BCC model assumes a variable returns to scale (VRS). CRS implies that a change in the amount of the inputs will lead to a similar change in the amount of outputs (shown in Figure 3.4) and all observed production combinations can be scaled up or down proportionally. BCC model, on the other hand, allows for VRS and is graphically represented by a piecewise linear convex frontier (Wang and Cullinane, 2006) as shown in Figure 3.5. 55 Output CCR Input Figure 3.4 Production Frontier in CCR Model Output Production Frontier BCC Input Figure 3.5 Production Frontier in BCC Model 3.2.2.1 CCR Model The CCR model is a fractional programming (FP) problem to calculate the variables of input weights vi (i=1,…, m ) and output weights ur (r=1,…, n ) of DMU o as shown in Eq. (3.3) based on a constant returns to scale. The efficiency of each DMU need to be measured once and hence need in total s optimizations, 56 one for each DMU j to be evaluated. The CCR model can be expressed by Eq. (3.6) to (3.9). ( FPo ) Max  u1 y1o  u2 y2o   un yno v1 x1o  v2 x2o   vm xmo (3.6) subject to u1 y1 j  u2 y2 j   un ynj v1 x1 j  v2 x2 j   vm xmj 1 ( j  1, , s) (3.7) v1 , v2 , , vm  0 (3.8) u1 , u2 , , un  0 (3.9) The constraint in Eq. (3.7) shows that the ratio of ‘virtual output’ ( u1 y1o  u2 y2o   un yno ) to ‘virtual input’ ( v1 x1o  v2 x2o   vm xmo ) cannot exceed 1 for every DMU. This is based on the assumption that production output should not exceed its input. The objective of this fractional programming (FP) problem in Eq. (3.6) is to obtain the input weights vi and output weights ur that can maximize the production ratio of DMU o , the DMU being evaluated. It can be seen from the Equations that the optimal value  * is 1 under these constraints. Mathematically, the non-negativity constraints in Eq. (3.8) and Eq. (3.9) are not sufficient to guarantee a positive value in Eq. (3.7). For port efficiency analysis, all outputs and inputs should have nonzero worth as reflected by a positive value in ur and vi . 57 The above FP [Eq. (3.6) to (3.9)] can be replaced by the following linear programming (LP) problem, shown in Eq. (3.10) to (3.14) (Cooper et al.,2007): ( LPo ) Max   u1 y1o  u2 y2o   un yno (3.10) subject to v1 x1o  v2 x2o   vm xmo  1 u1 y1 j  u2 y2 j   un ynj  v1 x1 j  v2 x2 j  (3.11)  vm xmj ( j  1, , s) (3.12) v1 , v2 , , vm  0 (3.13) u1 , u2 , , un  0 (3.14) This transformation from FP to LP in the CCR model makes computation easier and feasible (Wang et al., 2002). In the above linear programming problem, the optimal solution of ( LPo ) is designated as v  v* , u   * with the optimal objective value  * . DMU o is considered to be efficient if  * =1 and there exists at least one optimal ( v* ,  * ), with v* >0 and  * >0. Otherwise, it is considered to be inefficient. 3.2.2.2 BCC Model Unlike the CCR model, the BCC model assumes a variable returns to scale (VRS) (Banker et al., 1984). In the BCC model, the production frontier is spanned by the convex hull of existing DMUs and with the piecewise linear and concave characteristic (see Figure 3.5), allowing the model to distinguish between 58 technical and scale inefficiencies. The mathematical linear programming Equations are shown in Eq. (3.15) to (3.19). Max   u1 y1o  u2 y2o   un yno  c0 (3.15) subject to v1 x1o  v2 x2o  (u1 y1 j  u2 y2 j   vm xmo  1 (3.16)  un ynj )  c0  v1 x1 j  v2 x2 j   vm xmj ( j  1, , s) (3.17) v1 , v2 , , vm  0 (3.18) u1 , u2 , , un  0 (3.19) c0 represents the situation for returns to scale and is free in sign. If c0  0 , it indicates a situation where returns to scale is increasing; if c0  0 , it shows a decreasing returns to scale; if c0  0 , it shows a constant returns to scale (i.e. the CCR model). The variable c0 in the BCC model has its practical meaning in the evaluation of port performance. In the constraint (3.17), when the input value v1 x1 j  v2 x2 j  ( j  1,  vm xmj ( j  1, , s) is 0, the output u1 y1 j  u2 y2 j   un ynj , s) would also be 0 in the CCR model, which is impossible in reality. This situation can be avoided by introducing a negative c0 in the BCC model. Technical inefficiency and scale inefficiency are often analyzed from the DEA model results (Cooper et al., 2007). Technical inefficiencies include inefficient use of labor and inefficient operations while scale inefficiencies are caused by the 59 disadvantageous port operating conditions. To analyze the technical and scale inefficiencies of a DMU, the CCR score and BCC score are compared. The production possibility set in the CCR model is based on the assumption of CRS where the radial expansion and reduction of all observed DMUs and their nonnegative combinations are possible. The CCR score is therefore known as (global) technical efficiency. On the other hand, BCC model assumes that the convex combinations of all observed DMUs form the possibility set. The BCC score is hence known as (local) pure technical efficiency (PTE) (Cooper et al.,2007). The scale efficiency a DMU can be defined using the ratio of CCR and BCC scores, shown in Eq. (3.20) SE   *CCR  *BCC (3.20) where  *CCR is the score in the CCR model and  *BCC is the score in the BCC model. SE is not greater than one and the relationship in Eq. (3.20) demonstrates a decomposition of efficiency as  *CCR   *BCC  SE (3.21) where  *CCR shows the technical efficiency,  *BCC reflects the pure technical efficiency and SE is the scale efficiency. The decomposition in Eq. (3.21) shows whether inefficiency caused by inefficient operations or by disadvantageous conditions or by both. For example, if a DMU is fully efficient (100%) in both CCR and BCC models, it means that the DMU 60 operates in the most productive scale size. If a DMU is efficient in the BCC model but inefficient in the CCR model, it reflects local efficiency but not globally efficiency. This lack of global efficiency is due to the scale size of the DMU (such as limited port facilities and infrastructure). The measurement of technical and scale inefficiencies can be conceptually explained by Figure 3.6. The figure shows a number of production units with the same inputs but producing different amounts of outputs (output 1 and output 2) in Figure 3.6. Production units A to F form the production frontier and they are all efficient DMUs. Other units in the figure are inefficient and are ‘enveloped’ by the frontier. Technical efficiency reveals how outputs can be proportionally improved to the production frontier. For example, technical efficiency of production unit H is measured by the ratio of OH: OD and this ratio reflects the technical inefficiency of H with respect to D. Scale inefficiency can be explained by units A and B. Although both units are all on the production frontier, unit A can still improve Output 1 to the same level of output as unit B. For example, if Output 1 is the number of arrived ships in a port, A can improve its scale efficiency by constructing more berths, enlarging port area and reduce ships service charges to attract more ships’ arrival. 61 Production Frontier Output 2 A B C D E F H Technical inefficiency O Output 1 Figure 3.6 Production Frontier and Inefficiency in DEA CCR and BCC model both have dual input and output orientations. In general, the input-oriented model focuses on minimizing inputs while maintaining the same level of output and the output-oriented model focuses on maximizing the outputs while at the same time keeping the level of inputs constant. In the study of port performance benchmarking and efficiency, output-oriented models are typically employed (Cullinane et al.,2004; Cullinane et al., 2005). The LP Equations of CCR and BCC models can now expressed in the form of output-orientated models, formulated as: (3.22) max U U , subject to Uy 's  Y '   0 (3.23) 62 X '   x 's  0 (3.24)   0 (DEA-CCR) (3.25) e '  1 (DEA-BCC) (3.26) where inputs are xs  ( x1s , x2 s ...xms )  Rm , producing outputs ys  ( y1s , y2 s ... yns )  Rn . The row vectors xs and ys form the s th rows of the data matrices X and Y shown in Eq. (3.1) and (3.2), respectively. s  (1 , 2 ...s )  Rs is a non-negative vector that forms the linear combinations of the s DMUs and e  (1,1,...,1) is a suitably dimensioned vector of unity values. 3.3 FDH Model for Measuring Port Efficiency This section shows the methodology of the FDH model in order to evaluate the efficiency of global container ports with comparison to the DEA models. Free Disposal Hull (FDH) is another non-parametric approach that had been used in the literature to measure the efficiency of DMUs under the condition of multiple outputs and inputs (Deprins et al., 1984; Vanden Eeckaut, 1993). Unlike the DEA models, FDH assumes a non-convex (staircase) production possibility set instead of the convex production possibility set in the DEA models as shown in Figure 3.7. In Figure 3.7, the cost frontier of both DEA and FDH models are developed for the case of one output. It is shown that the cost frontier of the FDH model is made up by DMUs A, B, C, D and E while the cost frontier of the DEA model is the dashed line ABCE. In the FDH model, each observed cost and output combination spans one orthant, positive in the cost level and negative in the 63 output. An important characteristic of the FDH model as stressed by Lovell and Vanden Eeckaut (1994) is that inefficient observations are projected onto an orthant spanned by a single efficient producer which is weakly dominating in both cost and outputs. For example, in Figure 3.7 the inefficient observation F is dominated by C and D as well as by G, which is itself inefficient. F is projected onto point F’ situated on the orthant spanned by C, which is one of the dominating observations. This single producer C can therefore be interpreted to function as a role model for the inefficient unit. In the DEA model, F is projected to point F’’, which is a linear combination of observations B and C. No such unique role model such as the single producer C is available in the DEA model and inefficient observations are projected onto a fictitious linear combination of efficient observations. In the FDH model, it ensures that the efficiency evaluation of DMUs is only affected by the performance of observed DMUs. Cost E F D G F’ C DEA F’’ B A FDH O Output Figure 3.7 A Cost Frontier of the DEA and FDH model 64 Eq. (3.22) to (3.26) express the output-oriented CCR and BCC model in vectormatrix notation. FDH can be simplified by using the mixed integer programming formulation, as shown in Eq. (3.27) to (3.30). (3.27) max U U , subject to Uy 's  Y '   0 (3.28) X '   x 's  0 (3.29) s {0,1} (FDH) (3.30) Computation of efficiency scores of the DEA and FDH models can be performed using the software package DEA-Solver professional version 9.0 (Cooper et al., 2007). 3.4 Probability Models This section introduces the methodology of discrete and continuous probability models in order to study the relationship between ship turn-around time and the characteristics of ports and ships. The concept and formulation of count data models based on discrete probability modeling of ship turn-around time are described in Section 3.4.1. Duration models based on continuous probability modeling of ship turn-around time are discussed in Section 3.4.2. Container ship turn-around time is the total time taken by the ship from its arrival at the harbor to its departure. It is an important parameter reflects the efficiency of 65 ports for container operations as the quay transfer operation fundamentally decides the efficiency of a port and the most critical part of ship turn-around time is the working time for loading and unloading containers (Ramani, 1996; Cullinane et al., 2005). A shorter ship turn-around time in a port is analogous to a better level of service, which is an important factor influencing port performance and efficiency. Probability models are adopted in the thesis to study the distribution of ship turn-around time. This include count data models (Poisson Regression, Negative Binomial and Poisson Regression with normal heterogeneity) and duration models (exponential, Weibull, generalized gamma, log-logistic and log-normal). 3.4.1 Count Data Models Count data is a type of data in which the observations can take only the nonnegative integer values. The most common methods applied to study the count data are Poisson and negative binomial regression models. Poisson regression model has been applied to a wide range of transportation research (Washington et al., 2010). Ship turn-around time is non-negative and may be regarded as an integer value (such as 1 minute, 10 minutes, 100 minutes and so on) for simplicity. In this case, count data models can be used to simulate the probability function of ship turn-around time. 3.4.1.1 Poisson Regression Model For a discrete random variable, Y , observed over a period of length Ti and observed frequencies, yi , i  1,..., n ,where yi is a nonnegative integer count, i is 66 both the mean and variance of yi per unit of time, with regressors xi , the Poisson regression model is EXP(Ti i )(Ti i ) yi Prob (Y  yi xi )  , yi =0,1,…; yi ! (3.31) E (Y Ti xi )  i (3.32) with the restrictive equal dispersion property that Var[Y xi ]  E[Y xi ]  i (3.33) Poisson regression models are estimated by specifying the Poisson parameter i as the function of explanatory variables. In modeling ship turn-around time, explanatory variables are related to ships such as ship size and ports, such as berth length, the number of quay-cranes, port location, etc. The most common relationship between explanatory variables and the Poisson parameter is the loglinear model, i  EXP( xi ) (3.34) or, equivalently LN (i )   xi where, (3.35) xi is a vector of explanatory variables and  is a vector of estimable parameters. This model is estimable by standard maximum likelihood methods, with the likelihood function given as L(  )   i EXP[ EXP(  xi )][ EXP(  xi )] yi yi ! 67 (3.36) The log of the likelihood function is simpler to manipulate and more appropriate for estimation, n LL(  )  [ EXP(  xi )  yi  xi LN ( yi !)] (3.37) i 1 In order to have a complete understanding of parameter estimation results, partial effects are computed to evaluate the relative impact of each variable on the dependent variable, i.e. the ship turn-around time. The partial effect is defined as the effect on the conditional mean of y of a change in one of the regressors, say xi . It can provide a good approximation to the amount of change in y that will be produced by one unit change in xi (Cameron and Trivedi, 2009). In Poisson regression model, the partial effects are E[ yi xi ] xi  i  (3.38) To access the fitness of the Poisson regression model to observed data, the value 2 of R p is computed based on standardized residuals, 2  y  ˆ  i   i  ˆ i 1  i   1  2 n  yi  y     i 1   y  n Rp 2 (3.39) where the numerator reflects the residual and the denominator reflects the sum of square errors of observations. The more R p 2 approaches to one, the better the model fit to the actual data. 68 3.4.1.2 Negative Binomial Regression Model The negative binomial regression model is an extension of the Poisson regression model where the mean of the sample data is allowed to differ from the variance. The data are said to be under-dispersed when E[ yi ]  VAR[ yi ] or over-dispersed when E[ yi ]  VAR[ yi ] . A disturbance term  i is introduced into the Poisson regression model to account for dispersion effects. Therefore the negative binomial model can be derived as, i  EXP( xi   i ) (3.40) where EXP(  i ) is a gamma-distributed error term with mean one and variance  2 (thus allowing the variance and mean are not equal). The probabilities in the negative binomial model are given as, 1 ((1  )  yi )  1   Prob (Y  yi xi )    (1  ) yi !  (1  )    i    (1  )  i  yi (3.41) where  is the connection between the two models and the Poisson model results if  is zero. It is with the property that, Var[ yi ]  E[ yi ][1   E[ yi ]]  E[ yi ]   E[ yi ]2 (3.42) where (.) is a gamma function. The likelihood function is given as, 1 ((1  )  yi )  1   L(i )     (1  ) yi !  (1  )  i  i   i    (1  )  i  yi (3.43) A test for over- or under-dispersion is provided by Cameron and Trivedi (1990) in the Poisson Regression Model. It is hypothesized that the mean of ( y  E[ y])2  E[ y] is zero and the testing framework is built as 69 H 0 : Var[ yi xi ]  ui (3.44) H1 : Var[ yi xi ]  ui   g (ui ) A simple linear regression is estimated where zi is regressed on wi in order to conduct this test, zi  wi  ( yi  ui )2  yi (3.45) ui 2 g (ui ) (3.46) ui 2 They suggest two possibilities for this liner regression zi  bwi , where g (ui )  ui (3.47) and g (ui )  ui 2 (3.48) Under the null hypothesis of equal dispersion, the statistics g (ui ) have limiting  2 distributions with one degree of freedom. The critical  2 value at 5% significance level at one degree of freedom is 3.84 and the null hypothesis would be rejected if g (ui ) is greater than 3.841. 3.4.1.3 Poisson Regression Model with Normal Heterogeneity The Poisson regression model is modified by adding a disturbance term  to allow individual heterogeneity. The distribution of  in this model is hypothesized to follow a normal distribution where,  x,   EXP( x   ) (3.49)  x ~ N[0,  2 ] (3.50) Conditioned on  , the mean and variance of y is equal to: 70 EXP( x)  EXP( ) (3.51) The term EXP( ) has a lognormal distribution. Based on the properties of the lognormal distribution, the mean and variance of y are: E[ y x]  EXP( x)  EXP(1 2 2 ) (3.52) Var[ y x]  EXP(  x)  EXP(1 2  2 )  [ EXP(  x)]2  [ EXP(2 2 )  EXP( 2 )]  E[ y ]  {1  E[ y]( EXP( 2 )  1)} (3.53) If  is one, the model becomes the Poisson model. In order to determine if heterogeneity should be considered, the Vuong test is performed. P H  mi  log  i 0   Pi H1  (3.54)  1  n  n    mi   n  i 1   n  m  V Sm 2 1 n m  m      i  n  i 1 (3.55) where mi is the ratio of the logs of the probabilities for the i th observation under the null and alternative hypotheses. m is the mean, S m is the standard deviation and n is the sample size. V is asymptotically standard normal distributed and the critical V value at a 95% significance level is ±1.96. 3.4.2 Duration Models Duration models are continuous probability models that have been used to study accident risk analysis (Mannering, 1993), travelers' activity behavior (Mannering et al., 1994) and other transportation-related problems (Jones et al., 1991; Hensher 71 and Mannering, 1994; Nam and Mannering, 2000). The variable of interest in duration models is the length of time that elapsed from the beginning of an event until its end. In the study of ship turn-around time, duration models can be employed to calculate the probability of a ship departing from the berth after a certain dwell time and relate it to port and ship characteristics. The hazard function is defined to analyze the probability of a ship’s departure from the berth, conditioned on the inability of ships to leave up to a certain dwell time, t s . Let F (ts ) be the cumulative distribution function of the ship turn-around time, t s , such that F (ts )  Pr[Ts  ts ] (3.56) where Ts is a random variable, and t s is some specified dwell time value. The survive function S (ts ) is the probability when S (ts )  Pr(Ts  ts )  1  F (ts ) (3.57) with the integrated hazard function ih(ts ) ih(ts )   log S (ts ) (3.58) The corresponding density function f (ts ) is f (ts )  dF (ts ) / dt (3.59) with hazard function h(t s ) 72 h(t s )  f (ts ) / [1  F (ts )] (3.60) where h(t s ) is the approximate rate at which container ships are departing from the berth after some specified dwell time, t s . Hazard functions can be assessed by evaluating the first derivative with respect to time. If dh(ts ) / dt  0 at some dwell time, t s , the hazard is increasing in duration. That is to say, the longer a container ship is in the berth, the more likely it is to leave soon (i.e. positive duration dependence). If dh(ts ) / dt  0 , the hazard is decreasing in duration, meaning the longer a container ship is in berth, the less likely it is to leave soon (i.e. negative duration dependence). Finally, if dh(ts ) / dt  0 , the hazard is constant and the likelihood of a container ship departure from the berth is independent of the dwell time spent in the berth (i.e. no duration dependence). In general, the expected modeling result would be dh(ts ) / dt  0 , indicating that the longer a ship has been served, the greater the probability of the ship will depart from the berth. To include covariates (i.e. factors will influence ship turn-around time) that affect duration time, a proportional hazards model which assumes that covariates act multiplicatively on some underlying or baseline hazard is used. This is represented as h(ts Z ) and is defined as, h(ts Z )  h0 (ts ) exp( Z ) (3.61) 73 where Z is a vector of covariates (such as the number of quay-cranes, port location, terminal area), h0 (ts ) denotes the baseline hazard, exp( Z ) is a commonly used functional form for covariate effects and  is a vector of estimable coefficients. The five most commonly used distributions in duration modeling are the exponential, Weibull, log-logistic, log-normal and generalized gamma distribution (Kiefer, 1988). The exponential distribution is the simplest for its hazard function is not a function of t s , with parameter   0 , its density and hazard function are f (ts )   exp(ts ) (3.62) h(ts )   (3.63) In this case, the probability a ship departs from the port is independent of the dwell time it spent in the berth. The Weibull distribution is a generalized version of the exponential distribution. With parameters   0 and P  0 , the Weibull distribution has the following density and hazard functions: f (ts )   P(ts ) P1 exp[(ts ) P ] (3.64) h(ts )   P(ts ) P1 (3.65) The Weibull’s hazard function implies that if P  1 , the hazard is monotone decreasing (the longer a container ship is in berth, the less likely it is to leave 74 soon). If P  1 , the hazard is monotone increasing (the longer a container ship is in the berth, the more likely it is to leave soon). If P  1 , the hazard is constant and is independent of ship turn-around time (i.e. the Weibull distribution reduces to the exponential distribution). The log-logistic distribution relaxes the monotonicity restriction on the hazard function. With parameters   0 and P  0 , its density and hazard functions are, f (ts )   P(ts P) P1[1  (ts ) P ]2 (3.66) h(ts )  [ P(ts ) P1 ] / [1  (ts ) P ] (3.67) This hazard implies that if P  1, the hazard is monotone decreasing to infinity; P  1 , the hazard is monotone decreasing from  ; and P  1 , the hazard is increasing from zero to maximum at ( an inflection point) and decreasing toward zero thereafter. ts  ( P  1)1/ P /  (3.68) The hazard functions of log-normal and generalized gamma distributions are not closed form. The hazard function of log-normal distribution follows the loglogistic distribution with P  1 , i.e. the hazard function is increasing from zero to maximum at t s and then decreases to zero. The density function of generalized gamma distribution is f (ts )  ( P)(ts ) P 1 exp[(ts ) P ] ( ) (3.69) 75 where parameters P and  determine the shape of hazard function. Glaser (1980) studied the properties of the generalized gamma hazard and found that the shape of its hazard function is determined by the value of P and  , as shown in Table 3.1. Table 3.1 Shape of Generalized Gamma Hazard Function (Glaser, 1980) Parameter Value Shape of Hazard Function Hazard function is monotone decreasing P  1  0 P 1 Hazard is an U-shaped function P 1 P  1  0 P 1 P 1 Hazard is an inverted U-shaped function Hazard function is monotone increasing P  1  0 P  1 and   1 P 1 P 1 Hazard function is a constant Hazard function is monotone decreasing Hazard function is monotone increasing 3.4.3 T Test on Individual Regression Coefficients To test the significance of each regressor in models, the hypothesis test is built as (Montgomery and Runger, 2010): H 0 : j  0 (3.73) H1:  j  0  j is the coefficient of each regressor. Test statistic T0 is obtained T0  ˆ j  0  2C jj  ˆ j (3.74) S .E.( ˆ j ) where ˆ j is the regression coefficient; C jj is the diagonal element of ( X T X )1 corresponding to ˆ j and S .E.( ˆ j ) is the standard error of the parameter. The null 76 hypothesis is rejected if t0  t 2,v . If H 0 : j  0 is not rejected, this indicates that the regressor x j can be deleted from the model. 3.4.4 Temporal Stability Test on Regression Models To test the temporal stability of the model, a likelihood ratio test can be performed (Washington et al., 2010). The likelihood ratio test statistic is given by: X 2  2[ LL(T )  LL(a )  LL( b )] (3.75) where LL( T ) is the log likelihood at convergence of the model applied with data during the whole studying period, LL(  a ) is the log likelihood at convergence of the model using the data of year a ; LL( b ) is the log likelihood at convergence of the model using the data of year b . This X 2 statistic is  2 distribution with the degree of freedom that v  va  vb  vT (3.76) where va and vb are the degree of freedom of time a model and time b model; vT is the degree of freedom of the total model. The null hypothesis is rejected if X 2   20.05,v , which means that the model is time dependent. 3.5 Summary This chapter presents the research methodology adopted in the thesis. Two most significant non-parametric approaches, DEA and FDH, are employed to evaluate the efficiency of global container ports. Both DEA and FDH models can measure the efficiency of DMUs under multiple outputs and inputs. The main difference is that FDH assumes a non-convex (staircase) production possibility while DEA 77 models assume the convex production possibility set of the efficiency frontier. Two DEA models are described in this chapter, namely the DEA-CCR model and the DEA-BCC model. The key difference between the CCR and BCC models is that the CCR model assumes a constant returns to scale (CRS) while the BCC model assumes a variable returns to scale (VRS) when estimate the efficiency of DMUs. Probability models such as duration models and count data models are developed to study the relationship between the ship turn-around time and the characteristics of ports and ships. Three count data models (Poisson regression model, negative binomial regression model and Poisson regression model) and five duration models (exponential, Weibull, log-logistic, log-normal and generalized gamma models) are considered. The concept and formulation of each model used in this research work is specifically described. Application of these models will be discussed in the following chapters. 78 CHAPTER 4 PORT EFFICIENCY ANALYSIS WITH DEA AND FDH 4.1 Introduction Data envelopment analysis (DEA) and free disposal hull (FDH), the two commonly used non-parametric approaches are widely used in evaluating port efficiency (Wang et al., 2003; Cullinane et al., 2005). Both DEA and FDH are suitable for port efficiency studies due to the ability to consider multiple inputs and outputs concurrently. It provides information on the relative efficiency of DMUs without requiring a pre-defined functional relationship between variables (Vanden Eeckaut, 1993; Roll and Hayuth, 1993; Bichou, 2006). From the research methodology presented in Figure 3.1 of Chapter 3, there is a need to ascertain an appropriate non-parametric approach to evaluate port efficiency. Both DEA and FDH models are developed in this chapter to assess the capabilities of non-parametric approaches in measuring port efficiency and to benchmark global container ports. A comparative study of the DEA-CCR, DEABCC and FDH models is discussed based on efficiency estimation of 61 international container ports during the analysis period (2001-2011). The efficient ports from the estimation results are discussed and the most appropriate nonparametric approach in evaluating port efficiency is determined. 4.2 Empirical Setting 4.2.1 Model Specification It is important to distinguish whether the model should be input-oriented or output-oriented. Input-oriented DEA models focus on minimizing all inputs 79 while producing a given level of output and output-oriented DEA models try to maximize outputs for a given level of inputs. Wang et al. (2003) proposed that input-oriented models are more closely related to container port operation and management while output-oriented models are associated with port planning and strategy issues. Cullinane et al. (2005) noted that it is important for container ports to frequently review their capacity so as to provide satisfactory services to port users and maintain their competitive edge. It is of significance for a port to understand if their existing facilities are under-utilized and output in terms of throughput has been maximized given the various input variables. For this purpose, output-oriented model is chosen to analyze port efficiency. 4.2.2 Ports and Analysis Period 61 international container ports are included in the efficiency analysis presented in this thesis as shown in Table 4.1. The analysis period adopted in the chapter is between 2001 and 2011. Ports were chosen from the top 100 world’s leading container ports in terms of annual throughput (TEUs) and port data were retrieved from the Containerization International Yearbooks (2001, 2003, 2005, 2007, 2009, and 2011). 39 ports were eliminated due to the lack of complete data required for the analyses presented in the chapter. 80 Region Asia Australasia Caribbean Central America Europe Middle East North America South America Table 4.1 International Ports Considered in Models Port Name Busan, Colombo, Dalian, Fuzhou, Guangzhou, Hong Kong, Inchon, Jawaharlal Nehru, Kaohsiung, Karachi, Keelung, Kobe, Laem Chabang, Lianyungang, Manila, Nagoya, Osaka, Qingdao, Shanghai, Singapore, Taichung, Tianjin, Tokyo, Xiamen, Yantai, Yokohama Brisbane, Melbourne Freeport, Kingston Balboa Algeciras, Antwerp, Barcelona, Bremen, Constantza, Duisburg, Felixstowe, Genoa, Gioia Tauro, Hamburg, La Spezia, Le Havre, Rotterdam, Southampton, St Petersburg, Valencia, Zeebrugge Dubai, Haifa Charleston, Houston, Long Beach, Los Angeles, Montreal, New York, Oakland, Seattle, Tacoma, Vancouver BC Buenos Aires 4.2.3 Input and Output Variables 4.2.3.1 Input Variables Similar to previous research studies (Wang et al., 2003; Cullinane et al. 2005), input variables can include the necessary physical facilities of container ports (for example, the number of berth, terminal area, storage capacity and number of equipment). Based on past studies on DEA analyses of port production efficiency (Valentine et al., 2001; Tongzon, 2001; Rios and Maçada, 2006; Yun et al., 2011), the physical characteristics of container berth are commonly used. For example, Valentine et al.(2001) and Yun et al. (2011) used the total length of berths as an input variable, while in other studies (Tongzon, 2001;Rios and Maçada, 2006), the number of berths is used as an input variable. Since the number of berths and berth length are 81 related to berth capacity, they were considered as input variables in this study. Furthermore, the average berth depth is also used as an input variable. The number of gantry cranes in a container port or terminal is usually treated as an input variable in DEA analysis. Poitras et al. (1996) used the number of gantry cranes as the sole variable to represent port equipment. Cullinane et al. (2004) combined the number of quay gantry cranes, yard gantry cranes and straddle carriers as one input variable. However, such a definition may be misleading as quayside gantry and yard gantry cranes serve entirely different functions (Cullinane et al., 2005). In this study, the number of quayside gantry cranes and the number of yard cranes and tractors are defined as two significant input variables. Yard cranes and tractors include large rubber-tyred gantry (RTG) cranes, rail-mounted gantry (RMG) cranes as well as other mobile cranes and tractors. The input variables in this chapter include the number of berths, the total berth length, average berth depth, terminal area, storage capacity, the number of quayside gantry cranes and the number of yard cranes and tractors. Table 4.2 summarizes the statistics of these input variables over the analysis period. It was found that port physical facilities have increased during the analysis period, indicating the rapid development and construction of global container ports in these years. For example, the average number of berth has increased from 14 in 2001 to 20 in 2011 and the maximum terminal area has increased from 5,863,612 m2 to 9,441,323 m2 during the analysis period. 82 4.2.3.2 Output Variables Cullinane and Wang (2006) noted that container throughput is the most widely accepted container terminal output variable. Almost all previous research studies have treated container throughput as an output variable (Roll and Hayuth, 1993; Tongzon, 2001; Cullinane et al., 2004; Park and De, 2004), since it relates closely to the need for cargo-related facilities and services. This variable is the primary basis upon which container ports are compared, especially in the aspect of estimating relative size of container ports, investment magnitude and activity levels of ports. In this chapter, container throughput is chosen as the sole port output variable and this variable is defined as the total number of containers loaded and unloaded in 20-foot equivalent units (TEUs) in a year. Table 4.2 presents the descriptive statistics of throughput during the analysis period. It can be seen from the table that port throughput has been increasing steadily from 2001 to 2011 with the exception of 2009 during the economic crisis. 83 Table 4.2 Descriptive Statistics of the Input and Output Variables during the Analysis Period Output Variable Throughput (TEU) Input Variables Berth (no.) Total Average Berth Berth Length Depth (m) (m) 2001 Terminal Area (m2) Storage Capacity (TEU) Quayside Yard Gantry Cranes Cranes and (no.) Tractors (no.) 159 1798 1 10 23 337 26 337 Max 17,900,000 82 13,597 19.3 5,863,612 175,148 Min 206,449 1 180 8.3 31,000 790 Mean 2,467,696 14 3,730 12.2 1,414,876 41,579 Standard 3,098,814 12 2,877 1.8 1,297,605 36,514 Deviation 2003 Max 20,449,000 79 13,597 20 6,292,100 202,050 118 Min 206,449 2 410 7.9 37,000 790 1 Mean 3,057,299 15 3967 12.3 1,506,622 47,711 24 Standard 3,748,693 14 2954 1.9 1,342,654 42,377 22 Deviation 2005 Max 23,192,200 79 14,414 16.5 7,641,815 363,155 131 Min 550,462 2 381 7.9 14,000 250 1 Mean 3,819,932 18 4,575 12.4 1,955,900 63,534 29 Standard 4,616,168 15 3,203 1.8 1,741,898 67,673 26 Deviation 2007 Max 27,932,000 75 18,346 16.5 8,569,837 813,700 159 Min 901,000 2 381 7.8 14,000 250 1 Mean 4,774,568 19 5,096 12.5 2,176,182 97,670 34 Standard 5,599,187 16 3,677 1.8 1,930,325 129,603 31 Deviation 2009 Max 25,866,600 73 18,346 15.5 8,569,837 1,209,800 159 Min 607,483 2 410 7.8 175,000 5,950 1 Mean 4,427,996 20 5,394 12.6 2,231,340 119,288 35 Standard 5,228,962 16 3,970 1.7 1,976,393 189,907 32 Deviation 2011 Max 31,739,000 82 21,896 16.0 9,441,323 2,788,326 204 Min 556,694 2 410 7.8 175,000 5,950 4 Mean 5,295,073 20 5,449 12.8 2,342,660 200,923 40 Standard 6,404,021 17 4,081 1.5 2,027,007 419,985 36 Deviation Reference: Containerization International Yearbooks (2001, 2003, 2005, 2007, 2009, and 2011) 84 1482 10 363 329 1,636 2 402 348 1,577 2 436 370 2,086 19 453 397 1,943 24 467 403 4.3 Port Efficiency Analysis of Global Ports 4.3.1 Global Port Efficiency Analysis The average efficiency of ports is calculated to analyze the performance level of global ports during the analysis period (Cullinane et al., 2005; Cullinane and Wang, 2006). Efficiency within the port industry can be computed using the average efficiency and the results are shown in Figure 4.1. Efficiency results presented in the figure are based on DEA-CCR, DEA-BCC and FDH models obtained from the software DEA-Solver-PRO 9.0 (Cooper et al., 2007). Tables 4.3 to 4.8 shows the individual port efficiency scores and the aggregated average global port efficiency estimated from DEA-CCR, DEA-BCC and FDH during the analysis period. A value of 1 represents ideal efficiency. The average efficiency of ports estimated by different models is a reflection of their different assumptions on the efficiency frontier. Average efficiency of all ports in each year estimated the FDH model is the highest, followed by the DEA-BCC and DEACCR models. For example, Table 4.3 shows that average port efficiency in 2001 estimated by DEA-CCR, DEA-BCC and FDH are 0.5161, 0.6138 and 0.9717 respectively. Similar results were obtained for other years. The results are reasonable in accordance to the conceptual illustration shown in Figure 3.6 where the production frontiers are different for the three models. FDH assumes a nonconvex (stepwise function) production possibility set instead of the convex production possibility set in the DEA models, thereby allowing more ports to be estimated as efficient. On the other hand, the DEA-CCR model assumes a 85 constant returns to scale and the DEA-BCC model assumes a variable returns to scale. This means that if a port is CCR-efficient, it must be also BCC-efficient. Figure 4.1 depicts the average efficiency of all container ports estimated by DEACCR, DEA-BCC and FDH models during the analysis period. The general trend of average global port efficiency is downward during the analysis period. It indicates that the production of port fails to keep up with demand in the following years, i.e. most of the ports need to improve the container throughput and maintain efficient port operations. Figure 4.1 Average Efficiency for all Container Ports by CCR, BCC and FDH 86 Table 4.3 Port Efficiency of DEA-CCR, DEA-BCC and FDH in 2001 Port Throughput (TEU) FDH Efficiency DEA-CCR Efficiency DEA-BCC Efficiency Scale Efficiency Returns to Scale Singapore Shanghai Hong Kong Qingdao Taichung La Spezia Freeport Houston Barcelona Laem Chabang Yantai Kaosiung Gioia Tauro Dubai Guangzhou Algeciras Rotterdam Tacoma Busan Keelung Los Angeles Haifa Xiamen Colombo Tianjin Southampton Long Beach Lianyungang Manila Dalian Nagoya Jawaharlal Nehru Felixstowe New York Melbourne Karachi Bremen Tokyo Valencia Hamburg Balboa Antwerp Charleston Oakland Osaka Duisburg Le Havre Fuzhou Kingston Genoa Yokohama Incheon Zeebrugge Vancouver BC Buenos Aires Montreal Seattle Kobe Brisbane St Petersburg Constantza Average 15,520,000 6,340,000 17,900,000 2,640,000 1,100,000 974,646 860,000 983,451 1,411,054 2,312,439 290,000 7,540,000 2,488,332 3,501,820 1,730,000 2,151,770 6,102,000 1,320,274 8,072,814 1,815,854 5,183,520 839,000 1,290,000 1,726,605 2,010,000 1,160,976 4,462,971 502,300 2,296,151 1,210,000 1,872,272 1,573,677 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.8300 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.9266 0.9109 0.8994 0.8495 0.8316 0.6717 0.6573 0.6527 0.6447 0.6269 0.6024 0.6021 0.5845 0.5269 0.5117 0.5026 0.4855 0.4695 0.4303 0.4266 0.3930 0.3817 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.9526 1.0000 0.9098 0.8524 0.8296 0.6644 0.7880 0.6724 0.7658 0.6139 1.0000 0.6439 0.5717 0.5120 0.5322 0.5194 1.0000 1.0000 0.4485 0.4133 0.3891 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.9266 0.9562 0.8994 0.9337 0.9756 0.8097 0.9893 0.8283 0.9588 0.8186 0.9813 0.6021 0.9077 0.9216 0.9994 0.9444 0.9347 0.4695 0.4303 0.9512 0.9509 0.9810 Constant Constant Constant Constant Constant Constant Constant Constant Constant Constant Increasing Constant Increasing Increasing Increasing Increasing Increasing Increasing Constant Increasing Constant Increasing Increasing Increasing Constant Increasing Constant Increasing Increasing Increasing Increasing Increasing 2,800,000 3,316,275 1,276,476 455,000 2,896,381 2,535,841 1,506,805 4,688,669 448,565 4,218,176 1,528,034 1,643,585 1,502,989 340,000 1,525,000 418,000 888,941 1,526,526 2,303,780 610,000 875,926 1,146,577 1,011,748 989,427 1,315,109 2,010,343 460,844 360,899 206,449 1.0000 0.9478 1.0000 1.0000 1.0000 1.0000 1.0000 0.7847 1.0000 0.8859 0.9756 0.9380 0.9873 1.0000 0.9351 1.0000 1.0000 0.7249 0.8985 1.0000 0.9443 0.8357 1.0000 1.0000 0.8706 0.8453 0.8684 1.0000 1.0000 0.9717 0.3740 0.3722 0.3597 0.3564 0.3535 0.3517 0.3400 0.3345 0.3327 0.3303 0.3129 0.2955 0.2922 0.2904 0.2686 0.2632 0.2584 0.2583 0.2514 0.2472 0.2408 0.2234 0.2073 0.2037 0.1870 0.1563 0.1526 0.1467 0.1318 0.5161 1.0000 0.3723 0.4800 0.7961 0.3537 0.3521 0.3468 0.3347 0.4619 0.3443 0.3264 0.3033 0.3018 1.0000 0.2686 0.4563 0.3427 0.2932 0.2555 0.2912 0.2650 0.2295 0.2658 0.2305 0.1873 0.1564 0.1557 0.1949 0.9998 0.6138 0.3740 0.9997 0.7494 0.4477 0.9994 0.9989 0.9804 0.9994 0.7203 0.9593 0.9586 0.9743 0.9682 0.2904 1.0000 0.5768 0.7540 0.8810 0.9840 0.8489 0.9087 0.9734 0.7799 0.8837 0.9984 0.9994 0.9801 0.7527 0.1318 0.8695 Increasing Constant Increasing Increasing Constant Constant Constant Constant Increasing Increasing Increasing Increasing Increasing Increasing Constant Increasing Increasing Constant Constant Increasing Increasing Constant Increasing Increasing Constant Constant Increasing Increasing Increasing Note: efficiency score ‘1.0000’ equates to maximum efficiency and means efficient Returns to scale is determined the variable c0 in the DEA-BCC model (refer to section 3.2.2.2) 87 Table 4.4 Port Efficiency of DEA-CCR, DEA-BCC and FDH in 2003 Port Throughput (TEU) FDH Efficiency DEA-CCR Efficiency DEA-BCC Efficiency Scale Efficiency Returns to Scale Singapore Shanghai Hong Kong Guangzhou Dubai Haifa Freeport Houston Barcelona Gioia Tauro Keelung Duisburg Xiamen Tacoma Kaohsiung Algeciras Long Beach Tianjin Busan Rotterdam Qingdao Zeebrugge Los Angeles Fuzhou Incheon Melbourne Jawaharlal Nehru Nagoya Dalian Laem Chabang Hamburg Lianyungang Balboa Tokyo Colombo New York Southampton Manila Antwerp Karachi Charleston Genoa Bremen Valencia Felixstowe La Spezia Vancouver BC Kingston Le Havre Osaka Taichung St Petersburg Yokohama Oakland Seattle Montreal Brisbane Kobe Yantai Constantza Buenos Aires Average 18,100,000 11,280,000 20,449,000 2,761,700 5,151,958 1,069,000 1,057,879 1,243,866 1,652,366 3,148,662 2,000,707 500,000 2,331,000 1,738,068 8,840,000 2,515,908 4,658,124 3,015,000 10,407,809 7,106,779 4,239,000 1,012,674 7,178,940 590,000 821,071 1,721,067 2,268,989 2,073,995 1,670,000 3,181,050 6,138,000 502,300 448,565 3,313,647 1,959,354 4,067,812 1,377,775 2,552,187 5,445,436 455,000 1,690,846 1,605,946 3,189,853 1,992,903 2,500,000 1,006,641 1,539,058 1,137,798 1,977,000 1,609,631 1,246,027 649,812 2,504,628 1,923,136 1,486,465 1,108,837 639,570 2,045,714 290,000 206,449 590,677 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.8632 1.0000 1.0000 1.0000 0.7816 1.0000 1.0000 1.0000 0.7798 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.7786 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.7526 1.0000 0.9231 0.7476 0.9696 1.0000 1.0000 1.0000 0.8526 1.0000 0.8230 0.9711 0.9231 1.0000 0.8221 0.8309 0.8487 1.0000 1.0000 0.8253 0.9483 1.0000 1.0000 0.9580 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.8237 0.7991 0.7921 0.6960 0.6707 0.6631 0.6574 0.6370 0.5723 0.5495 0.5102 0.4972 0.4518 0.4369 0.4266 0.4091 0.3990 0.3854 0.3679 0.3643 0.3530 0.3435 0.3343 0.3255 0.3224 0.3183 0.3038 0.3009 0.3008 0.2643 0.2633 0.2585 0.2571 0.2486 0.2409 0.2322 0.2321 0.2310 0.2008 0.1993 0.1824 0.1709 0.1700 0.1523 0.1508 0.1468 0.1436 0.1308 0.0857 0.0850 0.0815 0.4613 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.6969 0.7411 1.0000 0.6895 0.6585 0.5876 0.5507 0.5281 0.7018 0.5113 0.5188 0.5890 0.7064 0.4839 0.3855 0.4971 0.3734 0.3729 1.0000 1.0000 0.3263 0.3351 0.3187 0.4648 1.0000 0.3251 1.0000 0.2672 0.2604 0.2625 0.2668 0.9998 0.2733 0.2393 0.3660 0.2035 0.1999 0.1995 0.2555 0.1865 0.1539 0.1543 0.1851 0.1746 0.1371 0.1534 0.9996 0.1358 0.5809 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.8237 0.7991 0.7921 0.9987 0.9050 0.6631 0.9534 0.9674 0.9740 0.9978 0.9661 0.7085 0.8836 0.8421 0.7243 0.5791 0.8246 0.9997 0.7401 0.9756 0.9466 0.3435 0.3343 0.9975 0.9621 0.9987 0.6536 0.3009 0.9253 0.2643 0.9854 0.9927 0.9794 0.9318 0.2409 0.8496 0.9699 0.6311 0.9867 0.9970 0.9143 0.6689 0.9115 0.9896 0.9773 0.7931 0.8225 0.9540 0.5587 0.0850 0.6001 0.8309 Constant Constant Constant Constant Constant Constant Constant Constant Constant Constant Increasing Increasing Increasing Increasing Constant Increasing Increasing Increasing Constant Increasing Increasing Increasing Constant Increasing Increasing Increasing Increasing Constant Increasing Constant Constant Increasing Increasing Constant Increasing Constant Increasing Increasing Constant Increasing Constant Increasing Constant Increasing Increasing Increasing Constant Increasing Constant Constant Increasing Increasing Constant Constant Constant Increasing Increasing Constant Increasing Increasing Increasing Note: efficiency score ‘1.0000’ equates to maximum efficiency and means efficient Returns to scale is determined the variable c0 in the DEA-BCC model (refer to section 3.2.2.2) 88 Table 4.5 Port Efficiency of DEA-CCR, DEA-BCC and FDH in 2005 Port Throughput (TEU) FDH Efficiency DEA-CCR Efficiency DEA-BCC Efficiency Scale Efficiency Returns to Scale Singapore Shanghai Hong Kong Houston Dubai Gioia Tauro Freeport Rotterdam Xiamen Tianjin Kaohsiung Qingdao Guangzhou Algeciras Long Beach Lianyungang Busan Keelung Dalian Jawaharlal Nehru Haifa Los Angeles Fuzhou Colombo Nagoya Tokyo Southampton Incheon Hamburg Bremen Manila Zeebrugge Barcelona Melbourne Kingston Laem Chabang Valencia Duisburg New York Felixstowe Charleston Tacoma Karachi Antwerp Taichung Balboa Genoa Seattle Vancouver BC La Spezia St Petersburg Le Havre Yantai Yokohama Osaka Brisbane Montreal Oakland Constantza Kobe Buenos Aires Average 23,192,200 18,084,000 22,427,000 1,582,081 7,619,222 3,160,981 1,211,500 9,300,000 3,342,300 4,801,000 9,471,056 6,307,000 4,685,000 3,179,614 6,709,818 1,005,300 11,843,151 2,091,458 2,655,000 2,666,703 1,122,580 7,484,624 1,177,200 2,455,297 2,491,198 3,593,071 1,375,000 1,153,465 8,087,545 3,735,574 2,665,015 1,407,933 2,071,481 1,862,993 1,670,820 3,765,967 2,409,821 712,000 4,792,922 2,700,000 1,986,586 2,066,447 850,000 6,482,061 1,228,915 663,762 1,624,964 2,087,929 1,767,379 1,024,455 1,119,346 2,118,509 550,462 2,873,277 1,802,309 766,275 1,254,560 2,273,990 771,126 2,262,066 1,075,173 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.9308 1.0000 0.8465 0.9819 0.8531 1.0000 0.9928 1.0000 1.0000 0.7280 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.9293 1.0000 0.9100 1.0000 1.0000 0.9308 0.8090 0.7982 0.9354 0.8643 0.7137 0.7352 0.8557 0.7844 0.6424 0.8517 0.9359 1.0000 0.8144 0.9603 0.6500 1.0000 0.8309 0.7368 0.8254 0.9215 0.9306 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.7799 0.6886 0.6471 0.6016 0.5866 0.5847 0.5578 0.5374 0.5315 0.5293 0.5188 0.5016 0.4911 0.4843 0.4586 0.4576 0.4476 0.4393 0.4280 0.4061 0.4059 0.3847 0.3820 0.3559 0.3545 0.3445 0.3264 0.3202 0.3144 0.3116 0.3085 0.2961 0.2850 0.2838 0.2754 0.2619 0.2613 0.2596 0.2583 0.2473 0.2437 0.2357 0.2343 0.2277 0.2246 0.2227 0.2130 0.1799 0.1691 0.1488 0.4920 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.8032 0.6920 0.6607 0.6165 0.6637 0.5982 0.6083 0.5518 0.5478 0.5805 0.5247 0.5371 0.5095 0.4941 0.4614 0.5089 0.4661 0.4506 0.4288 1.0000 0.4103 0.3883 0.3983 0.3746 0.3585 0.3470 0.3409 0.3246 0.9998 0.3117 0.3185 0.3134 0.2999 0.2873 0.3008 0.2795 0.2643 0.2681 0.3009 0.2536 0.2446 0.2973 0.2353 0.2277 0.2418 0.2704 0.2141 0.1886 0.1702 0.1886 0.5266 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.9710 0.9951 0.9794 0.9758 0.8838 0.9774 0.9170 0.9739 0.9702 0.9118 0.9888 0.9339 0.9639 0.9802 0.9939 0.8992 0.9603 0.9749 0.9981 0.4061 0.9893 0.9907 0.9591 0.9501 0.9888 0.9928 0.9575 0.9864 0.3145 0.9997 0.9686 0.9448 0.9503 0.9878 0.9156 0.9370 0.9886 0.9683 0.8584 0.9752 0.9963 0.7928 0.9958 1.0000 0.9289 0.8236 0.9949 0.9539 0.9935 0.7890 0.9432 Constant Constant Constant Constant Constant Constant Constant Constant Constant Constant Constant Constant Increasing Constant Constant Increasing Constant Constant Constant Constant Constant Constant Constant Constant Constant Constant Increasing Constant Constant Constant Increasing Constant Constant Constant Constant Constant Constant Constant Constant Increasing Constant Constant Constant Constant Increasing Constant Constant Constant Constant Constant Constant Constant Increasing Constant Constant Constant Constant Constant Constant Constant Increasing Note: efficiency score ‘1.0000’ equates to maximum efficiency and means efficient Returns to scale is determined the variable c0 in the DEA-BCC model (refer to section 3.2.2.2) 89 Table 4.6 Port Efficiency of DEA-CCR, DEA-BCC and FDH in 2007 Port Throughput (TEU) FDH Efficiency DEA-CCR Efficiency DEA-BCC Efficiency Scale Efficiency Returns to Scale Singapore Shanghai Hong Kong Dubai Lianyungang Gioia Tauro Freeport Xiamen Qingdao Kaohsiung Tianjin Guangzhou Rotterdam Jawaharlal Nehru Colombo Busan Balboa Los Angeles Hamburg Algeciras Keelung Dalian Long Beach Antwerp Bremen Incheon Manila Nagoya Southampton Tokyo Felixstowe Yantai New York Barcelona Valencia St Petersburg Karachi Melbourne Yokohama Vancouver BC Kingston Zeebrugge Constantza Fuzhou Osaka Laem Chabang Le Havre Duisburg Houston Seattle Haifa Tacoma La Spezia Charleston Buenos Aires Taichung Oakland Kobe Montreal Genoa Brisbane Average 27,932,000 26,150,000 23,998,449 10,653,026 2,001,000 3,445,337 1,634,000 4,627,000 9,462,000 10,256,829 7,103,000 9,200,000 10,790,604 4,059,843 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.8048 0.7890 0.7695 0.6990 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.8318 0.7913 0.7801 0.7213 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.9675 0.9971 0.9864 0.9691 Constant Constant Constant Constant Constant Constant Constant Constant Constant Constant Constant Increasing Increasing Increasing 3,381,693 13,270,000 1,833,778 8,355,039 9,900,000 3,414,345 2,215,484 4,574,192 7,312,465 8,175,952 4,892,239 1,663,800 2,869,447 2,896,221 1,900,000 4,123,920 3,300,000 2,214,631 5,299,105 2,610,099 3,042,665 1,697,720 1,219,724 2,206,567 3,428,112 2,307,289 2,016,792 2,020,723 1,411,414 1,177,200 2,309,820 4,641,914 2,600,000 901,000 1,768,687 1,973,504 1,148,628 1,924,934 1,187,040 1,750,000 1,710,896 1,250,000 2,387,911 2,472,808 1,363,021 1,855,026 1,000,066 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.8705 1.0000 0.9556 1.0000 0.9722 1.0000 1.0000 0.8293 0.8643 1.0000 0.8643 1.0000 1.0000 1.0000 0.9287 0.8101 1.0000 0.9011 0.7844 1.0000 0.8862 0.7270 0.7583 0.9603 0.8904 0.9263 0.9100 0.8368 0.6800 0.8465 0.6623 0.6424 0.7846 1.0000 0.6500 0.8710 0.8392 0.8818 0.6182 0.6500 0.9148 0.6857 0.6702 0.5789 0.5334 0.5272 0.5257 0.5227 0.5055 0.4952 0.4515 0.4282 0.4281 0.4275 0.4171 0.4048 0.4014 0.3840 0.3833 0.3753 0.3714 0.3702 0.3205 0.3130 0.3087 0.2987 0.2824 0.2806 0.2791 0.2775 0.2705 0.2680 0.2614 0.2607 0.2567 0.2550 0.2461 0.2426 0.2404 0.2376 0.2327 0.2221 0.2199 0.2104 0.2095 0.1862 0.1813 0.1663 0.4832 0.7193 0.6826 0.6778 0.5420 0.5294 0.5266 0.5731 0.5181 0.5083 0.4657 0.4331 0.4682 1.0000 0.4207 0.4052 0.4102 0.9993 0.4034 0.3761 0.3776 0.3832 0.3400 0.3147 0.3614 0.3014 0.3043 0.3003 0.2871 0.2878 0.2796 0.2685 0.2682 0.2611 0.2645 0.2602 0.2661 0.2491 0.2586 0.2684 0.2380 0.3125 0.2275 0.2119 0.2102 0.2316 0.1963 0.1783 0.5162 0.9533 0.9818 0.8541 0.9841 0.9958 0.9983 0.9121 0.9757 0.9742 0.9695 0.9887 0.9144 0.4275 0.9914 0.9990 0.9785 0.3843 0.9502 0.9979 0.9836 0.9661 0.9426 0.9946 0.8542 0.9910 0.9280 0.9344 0.9721 0.9642 0.9675 0.9981 0.9746 0.9985 0.9705 0.9800 0.9248 0.9739 0.9296 0.8852 0.9777 0.7107 0.9666 0.9929 0.9967 0.8040 0.9236 0.9327 0.9441 Constant Constant Constant Constant Constant Constant Constant Constant Constant Constant Constant Constant Increasing Constant Constant Decreasing Increasing Constant Constant Increasing Increasing Increasing Constant Increasing Constant Constant Increasing Constant Constant Constant Increasing Constant Constant Increasing Increasing Constant Constant Constant Constant Constant Increasing Constant Increasing Increasing Increasing Constant Constant Note: efficiency score ‘1.0000’ equates to maximum efficiency and means efficient Returns to scale is determined the variable c0 in the DEA-BCC model (refer to section 3.2.2.2) 90 Table 4.7 Port Efficiency of DEA-CCR, DEA-BCC and FDH in 2009 Port Throughput (TEU) FDH Efficiency DEA-CCR Efficiency DEA-BCC Efficiency Scale Efficiency Returns to Scale Singapore Shanghai Hong Kong Guangzhou Lianyungang Freeport Xiamen Qingdao Tianjin Kaohsiung Dubai Gioia Tauro Balboa Colombo Busan Rotterdam Valencia Jawaharlal Nehru Dalian Algeciras Manila Los Angeles Incheon Zeebrugge Antwerp Bremen Tokyo Hamburg Felixstowe Yantai Duisburg Long Beach Fuzhou Melbourne Haifa New York Vancouver BC Keelung Southampton Karachi Houston Laem Chabang Barcelona Kingston Yokohama St Petersburg Osaka Nagoya Oakland Kobe Le Havre Buenos Aires Taichung Seattle La Spezia Tacoma Montreal Charleston Brisbane Genoa Constantza Average 25,866,600 25,002,000 21,040,096 11,190,000 3,020,800 1,702,000 4,680,355 10,280,000 8,700,000 8,581,273 11,124,082 2,857,438 2,011,778 3,464,297 11,954,861 9,743,290 3,653,890 4,061,343 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.9626 0.9459 0.7543 0.6969 0.6650 0.6506 0.5843 0.5723 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.9849 0.9459 1.0000 0.7303 0.6777 0.6586 0.6139 0.5844 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.9774 1.0000 0.7543 0.9543 0.9813 0.9879 0.9518 0.9793 Constant Constant Constant Constant Constant Constant Constant Constant Constant Constant Decreasing Constant Increasing Constant Constant Constant Constant Decreasing 4,552,000 3,043,268 2,815,004 6,748,994 1,559,425 2,328,198 7,309,639 4,578,642 3,810,769 7,007,704 3,100,000 2,342,262 1,006,000 5,067,597 1,177,000 2,236,633 1,140,000 4,561,528 2,492,107 1,577,824 1,400,000 1,307,000 1,797,198 4,537,833 1,800,214 1,692,811 2,555,000 1,341,850 1,843,067 2,112,738 2,045,211 2,247,024 2,240,714 1,412,462 1,193,943 1,584,596 1,046,063 1,545,853 1,247,690 1,181,353 918,998 1,533,627 607,483 0.9108 1.0000 1.0000 1.0000 0.9722 0.7879 0.9284 0.9738 0.8848 1.0000 1.0000 0.7286 1.0000 1.0000 1.0000 0.9903 1.0000 0.9821 0.6629 0.8395 0.8293 0.8500 0.8368 0.8765 0.9677 0.9141 0.7534 0.9287 0.8095 0.7735 0.7163 0.7690 0.7846 1.0000 0.7286 0.6800 0.7200 0.6623 0.9884 0.7846 0.7103 0.8340 0.8149 0.9081 0.5441 0.4912 0.4877 0.4604 0.4562 0.4210 0.4195 0.4139 0.4115 0.4036 0.3927 0.3873 0.3785 0.3721 0.3679 0.3539 0.3478 0.3447 0.3308 0.3068 0.2986 0.2934 0.2925 0.2821 0.2712 0.2521 0.2504 0.2491 0.2376 0.2356 0.2231 0.2213 0.2207 0.2174 0.2068 0.2046 0.1990 0.1986 0.1973 0.1677 0.1624 0.1340 0.1189 0.4764 0.5917 0.4923 1.0000 0.4702 0.4889 0.4251 0.4343 0.4306 0.4240 0.4066 1.0000 0.4307 1.0000 0.3816 0.9999 0.4076 1.0000 0.3520 0.3646 0.3281 0.3155 0.3139 0.2959 0.2932 0.2834 0.2616 0.2548 0.2495 0.2393 0.2371 0.2349 0.2213 0.2216 0.3294 0.2354 0.2327 0.2323 0.2178 0.2300 0.1795 0.1863 0.1347 0.1265 0.5434 0.9196 0.9978 0.4877 0.9792 0.9331 0.9904 0.9659 0.9612 0.9705 0.9926 0.3927 0.8992 0.3785 0.9751 0.3679 0.8683 0.3478 0.9793 0.9073 0.9351 0.9464 0.9347 0.9885 0.9621 0.9570 0.9637 0.9827 0.9984 0.9929 0.9937 0.9498 1.0000 0.9959 0.6600 0.8785 0.8792 0.8567 0.9118 0.8578 0.9343 0.8717 0.9948 0.9399 0.9063 Constant Decreasing Increasing Constant Constant Constant Constant Constant Decreasing Constant Increasing Decreasing Increasing Constant Increasing Increasing Increasing Constant Constant Decreasing Decreasing Decreasing Constant Constant Increasing Constant Decreasing Constant Decreasing Constant Constant Decreasing Constant Increasing Constant Constant Decreasing Decreasing Increasing Constant Constant Constant Decreasing Note: efficiency score ‘1.0000’ equates to maximum efficiency and means efficient Returns to scale is determined the variable c0 in the DEA-BCC model (refer to section 3.2.2.2) 91 Table 4.8 Port Efficiency of DEA-CCR, DEA-BCC and FDH in 2011 Port Throughput (TEU) FDH Efficiency DEA-CCR Efficiency DEA-BCC Efficiency Scale Efficiency Returns to Scale Singapore Shanghai Hong Kong Xiamen Qingdao Lianyungang Kaohsiung Balboa Tianjin Guangzhou Busan Dubai Dalian Gioia Tauro Manila Colombo Valencia Los Angeles Rotterdam Incheon Hamburg Bremen Jawaharlal Nehru Long Beach Antwerp New York Duisburg Algeciras Tokyo Fuzhou Zeebrugge Felixstowe Laem Chabang Yantai Melbourne Keelung Barcelona Karachi Nagoya Haifa Yokohama Houston St Petersburg Vancouver BC Freeport Kobe Buenos Aires Southampton Kingston La Spezia Oakland Taichung Osaka Le Havre Montreal Genoa Seattle Charleston Tacoma Brisbane Constantza Average 29,937,700 31,739,000 24,384,000 6,454,200 13,020,100 3,870,000 9,636,289 3,232,265 11,587,600 14,260,400 16,163,842 12,617,595 6,400,300 2,264,798 3,342,200 3,651,963 4,327,371 7,940,511 11,876,920 1,924,644 9,014,165 5,915,487 4,307,622 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.8878 1.0000 1.0000 0.8835 1.0000 1.0000 1.0000 0.9722 0.9974 0.9519 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.9639 0.9589 0.8908 0.7644 0.6642 0.5987 0.5852 0.5033 0.4850 0.4745 0.4646 0.4602 0.4585 0.4421 0.4366 0.4298 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.9987 0.8969 0.7655 0.6650 0.6606 0.5852 1.0000 0.4863 0.5082 0.4708 0.4647 0.4672 0.4509 0.4592 0.4361 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.9639 0.9601 0.9932 0.9986 0.9988 0.9063 1.0000 0.5033 0.9973 0.9337 0.9868 0.9903 0.9814 0.9805 0.9508 0.9856 Constant Constant Constant Constant Constant Constant Constant Increasing Constant Constant Constant Constant Constant Constant Increasing Constant Constant Constant Constant Increasing Constant Constant Constant 6,061,091 8,664,243 5,503,485 1,181,000 3,608,301 4,416,119 1,318,958 2,207,257 3,248,592 5,731,063 2,342,262 2,467,967 1,749,388 2,033,747 1,545,434 2,471,821 1,238,000 2,992,517 1,866,450 2,365,174 2,507,032 1,116,272 2,725,304 1,851,701 1,324,581 1,724,928 1,307,274 2,342,504 1,383,578 2,172,797 2,215,262 1,362,975 1,847,648 2,033,535 1,381,352 1,485,617 1,004,983 556,694 0.8746 0.9050 0.9821 1.0000 0.7869 0.9043 1.0000 0.7852 0.8782 0.8671 0.7286 0.9730 0.8053 0.8965 0.8571 0.7969 1.0000 0.7706 0.8368 0.9071 0.6600 0.8095 0.7615 0.7846 0.8117 0.9189 0.7556 0.7817 0.7286 0.7972 0.7786 0.9884 0.8345 0.6778 0.7846 0.6623 0.7133 0.7514 0.8893 0.4058 0.3892 0.3795 0.3718 0.3670 0.3669 0.3466 0.3391 0.3067 0.3026 0.2990 0.2971 0.2915 0.2626 0.2551 0.2514 0.2412 0.2411 0.2395 0.2362 0.2308 0.2240 0.2203 0.2158 0.2115 0.2090 0.2078 0.2042 0.1900 0.1888 0.1851 0.1815 0.1794 0.1705 0.1670 0.1495 0.1261 0.0924 0.4283 0.4168 0.4047 0.3887 1.0000 0.3700 0.3718 0.9998 0.3564 0.3298 0.3151 0.3260 0.3178 0.2996 0.2705 0.2664 0.2527 1.0000 0.2415 0.2422 0.2401 0.2554 0.2390 0.2237 0.2163 0.2257 0.2159 0.2297 0.2071 0.2138 0.1928 0.1879 0.2013 0.1814 0.1836 0.1763 0.1607 0.1373 0.1010 0.4799 0.9736 0.9617 0.9763 0.3718 0.9919 0.9868 0.3467 0.9515 0.9300 0.9603 0.9172 0.9349 0.9730 0.9708 0.9576 0.9949 0.2412 0.9983 0.9889 0.9838 0.9037 0.9372 0.9848 0.9977 0.9371 0.9680 0.9047 0.9860 0.8887 0.9793 0.9851 0.9016 0.9890 0.9286 0.9472 0.9303 0.9184 0.9149 0.9269 Constant Constant Constant Increasing Constant Constant Increasing Constant Constant Constant Constant Increasing Constant Constant Constant Constant Increasing Decreasing Constant Constant Constant Constant Decreasing Constant Constant Increasing Constant Constant Constant Constant Constant Increasing Constant Constant Constant Constant Constant Constant Note: efficiency score ‘1.0000’ equates to maximum efficiency and means efficient Returns to scale is determined the variable c0 in the DEA-BCC model (refer to section 3.2.2.2) 92 4.3.2 Individual Port Efficiency Analysis on a Global Scale The individual port efficiency is studied in detail to understand the cause of inefficiencies in different ports. By analyzing the efficiency score in the DEACCR and DEA-BCC model, port inefficiency caused by pure technical inefficiency and/or by scale inefficiency can be determined. If the efficiency score of a port is equal to 1 in the DEA-BCC model but less than 1 in the DEA-CCR model, then it reflects the port is technical efficient with scale inefficiencies. For example, Table 4.3 shows the efficiency scores of Yantai port in 2001 are 1 in the DEA-BCC model and 0.9266 in the DEA-CCR model, reflecting that scale inefficiencies exist in Yantai port. Such scale inefficiencies might be caused by the limited storage area or number of berths in the port. Table 4.4 shows that 10 out of 61 ports are CCR-efficient and 18 out of 61 ports are BCC-efficient in 2003. This means that there were 8 ports with scale inefficiencies in 2003. Table 4.5 shows the port efficiency values in 2005. Out of the 61 ports studied, 11 were efficient in the DEA-CCR model and 12 were efficient in the DEA-BCC model which means that only one port had scale inefficiencies in 2005. If the efficiency score of a port is less than 1 in both DEA-CCR and DEA-BCC models, it reflects the port has both technical and scale inefficiencies. Table 4.3 shows that the efficiency scores of Kaosiung port in 2001 were 0.9109 in the DEA-CCR model and 0.9526 in the DEA-BCC model, reflecting both pure technical and scale inefficiencies in the port. The technical inefficiencies could be the inefficient use of labor and inefficient operations (such as the loading and unloading process of quay-cranes and yard tractors). In 2001, 44 ports were 93 neither technical efficient nor scale efficient as shown in Table 4.3. The number of ports that were neither DEA-CCR efficient nor DEA-BCC efficient in 2003, 2005, 2007, 2009 and 2011 were 43, 49, 50, 46 and 50, respectively. 4.3.3 Return to Scale for Global Ports The scale properties of port production can be reflected by the results of ‘return to scale’ in the DEA-BCC model. Table 4.3 shows the scale properties of port production in 2001. Of the 50 ports with scale inefficiencies, 15 had constant returns to scale, 35 had increasing returns to scale, and no ports had decreasing returns to scale. It was found that ports with constant returns to scale’ tend to be large container ports with throughput more than 1 million TEU. Ports with increasing returns to scale tend to be small container ports with throughput less than 1 million TEU. For example, Yantai, Haifa and Karachi had throughput less than 1 million TEU in 2001 and they all had increasing returns to scale. Similar findings were found for the other years studied as shown in Table 4.4 to 4.8. The results presented in Table 4.4 to 4.8 suggest that small ports are more likely to increase their operational scale while large ports often face difficulty in ensuring further growth. This is because a small port usually faces less difficulty than a large port in gaining access to the capital resources for investing its infrastructure (Cullinane and Wang, 2006). Compared to the larger ports, the investment are smaller for small ports due to the lower level of port scale. Furthermore, small ports with lower throughput face less physical constraints on expansion. This is especially true for new ports or terminals which are in their early stages of evolution. On the other hand, larger ports face greater difficulty in 94 expanding their scale. This is due to the unavailability of land and higher operating cost. 4.3.4 Evaluating Effectiveness of DEA and FDH Models in Port Efficiency Analysis To assess the capabilities of the DEA and FDH models to measure port efficiency and to benchmark global container ports, port efficiency values estimated by the DEA-CCR, DEA-BCC and FDH models are shown in Table 4.9. Table 4.9 essentially summarizes the estimation efficiency of ports presented in Tables 4.3 to 4.8, showing the number of efficient ports in terms of production scale during the analysis period. FDH model ensures that the efficiency evaluation of DMUs is only affected by the performance of observed DMUs. There are two sources of efficiency in FDH as proposed by Eeckaut et al.(1993), namely ‘efficiency by domination’ and ‘efficiency by default’. The former means that some DMUs are estimated to be efficient in the FDH model as they are able to dominate other DMUs and are regarded as the reference set for the inefficient DMUs. The latter ‘efficiency by default’ means that some DMUs are estimated as efficient simply because they are not dominated by any other DMUs. The two sources of efficiency in FDH are in Table 4.9. In 2001, of the 61 ports, 46 ports were efficient estimated by FDH because there are no other ports can dominate them. These ports forms the efficiency frontier of the FDH model. Among the 46 efficient ports, most of them are regarded as ‘efficiency by default’ because they are not dominated by any other DMUs (such as Hong Kong, Rotterdam and Busan) and only a few ports are 95 considered as ‘efficiency by dominating’ because they can be the reference set for other inefficient ports. For example, Shanghai port dominates Hamburg, Antwerp and New York. The result that a much greater proportion of efficient ports are regarded as ‘efficiency by default’ indicates that FDH lacks the sensitivity in determining efficient ports. In fact, many of these efficient ports in the FDH model could be regarded as inefficient in the DEA models due to the more stringent requirement on the efficiency frontier. It is found from Table 4.9 and Figure 4.2 that the percentage of efficient ports across different throughput categories estimated by FDH shows no significant difference in results. However, the results estimated by the DEA-CCR and DEABCC models shows significant differences across throughput categories. For example, in 2001, the percentage of efficient ports in the “below 1,000,000 TEUs” throughput category and “above 10,000,000 TEUs” throughput category are 88% and 100% from the FDH model, compared to 41% and 100% in the DEA-BCC model and 18% and 100% in the DEA-CCR model. Figure 4.3 shows that the numbers of efficient ports in each category as estimated by the FDH model are much more than that estimated by the DEA models. This shows that DEA models tend to be selective in determining efficient ports. The results in Figure 4.2 and 4.3 also suggest that FDH lacks the sensitivity to analyze the efficiency of ports as compared to the DEA models. In practice, a port which is estimated to be efficient by the FDH model may have inefficiencies (Wang et al., 2003). This is because the port is likely to be regarded as ‘efficiency by default’ in the FDH model. This may cause the port lose the incentive to 96 improve its production efficiency although it may still be inefficient. This is a significant limitation of the FDH model. DEA, to some extent, overcomes this drawback by constructing a hypothetical convex hull to nest all the DMUs. In this way, some efficient DMUs in FDH may become inefficient in the DEA models. For example, in Table 4.9 for 2001, 12 FDH-efficient (=15 - 3) and 8 FDHefficient ports (= 15-7) in the “below 1,000,000 TEUs” throughput category become inefficient when estimated by DEA-CCR and DEA-BCC respectively. In 2007, 7 ports are estimated as FDH-efficient in the “above 10,000,000 TEUs” throughout category. They are Singapore, Shanghai, Hong Kong, Busan, Rotterdam, Dubai and Kaohsiung. However, Busan and Rotterdam are estimated as inefficient by both DEA-BCC and DEA-CCR models. 97 Table 4.9 Comparison of Efficiency Results between DEA and FDH Models DEA-CCR model DEA-BCC model FDH model Container throughput Number Number Percentage Number Percentage Number Percentage (TEU) (1) of Ports of of Efficient of of of of Efficient (2) Efficient Ports Efficient Efficient Efficient Ports Ports [=(3)/(2)] Ports Ports Ports [=(7)/(2)] (3) (4) (5) [=(5)/(2)] (7) (8) (6) 2001 Below 1,000,000 17 3 18 7 41 15 88 1,000,000-2,000,000 21 2 10 2 10 14 67 2,000,000-3,000,000 11 2 18 5 45 9 82 3,000,000-6,000,000 6 0 0 0 0 2 33 6,000,000-10,000,000 4 1 25 1 25 4 100 Above 10,000,000 2 2 100 2 100 2 100 Total 61 10 18 17 28 46 75 2003 Below 1,000,000 11 0 0 4 36 10 91 1,000,000-2,000,000 22 4 18 4 18 14 64 2,000,000-3,000,000 10 1 10 5 50 8 80 3,000,000-6,000,000 10 2 20 2 20 8 80 6,000,000-10,000,000 4 0 0 0 0 1 25 Above 10,000,000 4 3 75 3 75 3 75 Total 61 10 16 18 30 44 72 2005 Below 1,000,000 6 0 0 0 0 1 17 1,000,000-2,000,000 19 2 11 2 11 8 42 2,000,000-3,000,000 15 0 0 1 7 7 47 3,000,000-6,000,000 9 3 33 3 33 8 89 6,000,000-10,000,000 8 3 38 3 38 6 75 Above 10,000,000 4 3 75 3 75 4 100 Total 61 11 18 12 20 34 56 98 Table 4.9 Comparison of Efficiency Results between DEA and FDH Models (Continued) DEA-CCR model DEA-BCC model FDH model Container throughput Number Number Percentage Number Percentage Number Percentage (TEU) (1) of Ports of of of of of of Efficient (2) Efficient Efficient Efficient Efficient Efficient Ports Ports Ports Ports Ports Ports [=(7)/(2)] (3) [=(3)/(2)] (5) [=(5)/(2)] (7) (8) (4) (6) 2007 Below 1,000,000 1 0 0 0 0 0 0 1,000,000-2,000,000 19 1 5 1 5 3 16 2,000,000-3,000,000 14 1 7 2 14 7 50 3,000,000-6,000,000 13 2 15 2 15 9 69 6,000,000-10,000,000 7 1 14 1 14 6 86 Above 10,000,000 7 5 71 5 71 7 100 Total 61 10 16 11 18 32 52 2009 Below 1,000,000 2 0 0 0 0 0 0 1,000,000-2,000,000 21 1 5 3 14 5 24 2,000,000-3,000,000 12 0 0 2 17 3 25 3,000,000-6,000,000 13 2 15 3 23 8 62 6,000,000-10,000,000 6 2 33 2 33 5 83 Above 10,000,000 7 5 71 5 71 7 100 Total 61 10 16 15 25 28 46 2011 Below 1,000,000 1 0 0 0 0 0 0 1,000,000-2,000,000 18 0 0 2 11 3 17 2,000,000-3,000,000 14 0 0 0 0 1 7 3,000,000-6,000,000 12 1 8 3 25 5 42 6,000,000-10,000,000 7 2 29 2 29 3 43 Above 10,000,000 9 4 44 4 44 9 100 Total 61 7 11 11 18 21 34 99 Figure 4.2 Percentage of Efficient Ports by Throughput Category in 2001 Figure 4.3 Number of Efficient Ports by Throughput Categories in 2001 100 4.3.5 Case Study of Selected Ports Six ports are selected for detailed analysis. They are Singapore, Shanghai, Hong Kong, Kaohsiung and Rotterdam. From Tables 4.3 to 4.8, it was found that Singapore, Shanghai and Hong Kong are the only three ports estimated as efficient by both DEA and FDH models during the entire analysis period, as shown in Figure 4.4.This result suggests that ports with high annual throughput have a higher tendency to be estimated as efficient. Such results are expected since container throughput is the sole output considered in the DEA and FDH models. As claimed by other studies (Wang, 1998; Clark et al., 2004; Gordon and Lucas, 2005), these ports are always found to be efficient in performance benchmarking analysis due to their competitive advantage in time, quality and cost over other ports (Clark et al., 2004). Singapore port has the prime location where ship traffic between Europe and Southeast Asia must pass. It is Asia’s main transshipment hub for shippers and the busiest port in the world in terms of shipping tonnage (Gordon and Lucas, 2005). Hong Kong port is at the center of the Asia Pacific region and has deep-water condition suitable for super-sized container ships (Wang, 1998). 101 Figure 4.4 Efficiency of Singapore, Shanghai and Hong Kong during 2001-2011 Figure 4.5 presents the efficiency of Kaohsiung Port during the analysis period. It shows that an obvious decrease of port efficiency occurred from 2001 to 2003 and the efficiency keeps as high as 1 after 2005. This is caused by the unsatisfactory container throughput in 2003. Figure 4.5 Efficiency of Kaohsiung Port during 2001-2011 102 Table 4.10 explains how Kaohsiung port could improve the efficiency of its production in 2003 by the application of DEA models and FDH if such analysis described in the chapter were performed. It can be seen from Table 4.10 that the throughput of Kaohsiung port needed to increase from 8,840,000 TEUs to 11,280,000 TEUs, 11,928,723 TEUs and 13,179,737 TEUs based on the FDH, DEA-BCC and DEA-CCR models if Kaohsiung port aims to be efficient. Utilization of physical facilities in Kaohsiung port have to be improved in order to improve port efficiency. For example, the utilization of berth length is only 44.5%, 51.3% and 52.7%, respectively estimated by the FDH, DEA-BCC and DEA-CCR models. In order to improve the utilization of berth length, Kaohsiung port needs to attract more ship’s arrival. Transport costs need to be reduced and the operational efficiency for loading and unloading containers needs to improve (Clark et al., 2004). 103 Table 4.10 Physical Facilities Utilization of Kaohsiung Port Estimated by DEA and FDH Models Physical Facilities in Utilization Rate Estimated by Models (%) Kaohsiung Port DEA-CCR DEA-BCC FDH (2003) Berth (no.) 19 63.2 57.9 52.6 Total Berth 5,122 52.7 51.3 44.5 Length (m) Terminal 1,493,906 69.2 65.5 56.8 2 Area (m ) Storage 69,511 100 100 86.3 Capacity (TEU) Quayside 25 100 100 80 Gantry Cranes (no.) Yard 327 72.2 78 61.5 Cranes and Tractors (no.) Throughput Actual Value Expected Value as Efficient Port (TEU) 8,840,000 13,179,737 11,928,723 11,280,000 Figure 4.6 shows the efficiency of Rotterdam Port during the analysis period. Rotterdam port was estimated to be FDH-efficient during the entire analysis period but was only estimated to be efficient in 2005 by the two DEA models. This is because Rotterdam is regarded as ‘efficiency by default’ in the FDH model. In the DEA-CCR and DEA-BCC models, Rotterdam is considered to be inefficient due to the lack of utilization of berth length, terminal area, quayside gantry cranes and yard cranes. 104 Figure 4.6 Efficiency of Rotterdam Port during 2001-2011 Table 4.11 Physical Facilities Utilization of Rotterdam Port Estimated by DEA and FDH Models Physical Facilities in Utilization Rate Estimated by Models (%) Rotterdam Port DEA-CCR DEA-BCC FDH (2007) Berth (no.) 27 100 100 100 Total Berth 10,505 59.3 58.1 100 Length (m) Terminal 5,635,900 39.4 38.8 100 2 Area (m ) Storage 61,000 100 100 100 Capacity (TEU) Quayside 87 82.7 82.7 100 Gantry Cranes (no.) Yard 1,577 37.7 38.9 100 Cranes and Tractors (no.) Throughput Actual Value Expected Value as Efficient Port (TEU) 10,790,604 14,022,808 13,831,511 10,790,604 105 4.4 Summary This chapter studies the efficiency of 61 international container through the use of output-oriented DEA-CCR and DEA-BCC models as well as the FDH model. Empirical settings of the DEA and FDH models (such as model specification, analysis period and input and output variables) were discussed. Throughput is the sole output variable considered in this chapter and is defined as the total number of containers loaded and unloaded in terms of 20-foot equivalent units (TEUs) in a year. The choice of input variables mainly focuses on the physical facilities of ports, such as the number of berths, the length of berths and the number of quayside gantry cranes. Efficiency of global ports during the analysis period from 2001 to 2011 is then studied. It is shown that the average efficiency of global ports estimated by the FDH model is always the highest, followed by DEA-BCC model and then DEACCR model. The general trend of average global port efficiency is downward during the analysis period, indicating that most of the ports need to improve the container throughput and maintain efficient port operations. Port inefficiencies in individual ports are found to be caused by pure technical inefficiency or by scale inefficiency. It was found that small ports are more likely to be motivated to increase their operational scale while large ports often face more difficulty in ensuring further growth. A comparison between the DEA and FDH models in evaluating port efficiency was studied. It was found that FDH lacks the sensitivity to analyze port efficiency compared to the DEA models. A port which is estimated to be efficient by the 106 FDH model is not necessarily better and this can cause issues when using FDH model results in port efficiency studies. DEA, however, is more stringent in determining efficient ports and should be used as a preferred method in port efficiency studies. A case study on five international container ports was studied to illustrate the use of the DEA and FDH models. It was found that Singapore, Shanghai and Hong Kong are both considered efficient by DEA and FDH models over the entire analysis period. This is because the three ports have the beneficial geographical location and competitive advantage excels in time, quality and cost over other ports. It was also found that the container throughput and utilization of physical facilities in Kaohsiung and Rotterdam port need to improve in order to become efficient ports. 107 CHAPTER 5: MODELING OF CONTAINER SHIP TURN-AROUND TIME IN PORTS USING PROBABILITY MODELS 5.1 Introduction The activities of arrived ships in ports are given much attention as it is thought to be an important element of port performance (Demirci, 2003b). The increasing size of container ships in recent years causes higher requirements on seaport terminals and associated equipment and facilities. Larger ships require longer docks, larger storage area, more cranes, deep water at the dock, and a capability to rapidly move containers from the terminal to truck or rail (McCray, 2008). Therefore, it is important to study the relationship between ship turn-around time and port infrastructure and ship characteristics so that the working time for loading and unloading containers can be reduced and a better level of shipping service can be provided. This chapter aims to develop an appropriate probability model to relate ship turnaround time to ship characteristics and port infrastructure. Both count data regression models and hazard-based duration models were developed using data from over 3800 ships in 61 international ports. The models were assessed for model fit and the most appropriate model for ship turn-around time was determined. 5.2 Empirical Setting Ship turn-around time is defined as the time a container ship arrived at an available berth of the port until the time it departs from the port. The choice of variables for potential inclusion in models was based on previous empirical works 108 by Edmond and Maggs (1976) and Tongzon (1995) and improved to include variables related to port infrastructure and performance. 5.2.1 Data Sources Ship turn-around time data was collected from a proprietary ship tracking website (Marine Traffic Services, 2012 and 2013) over the period from Aug 2012 to Dec 2012 and from Oct 2013 to Dec 2013. A total of 61 international ports were studied and they were listed in Table 4.1. Port-related data was obtained from the Containerization International Yearbooks (2012 and 2013) and websites of various port authorities. 5.2.2 Definition of Variables Variables used in the ship turn-around time model are generally classified into five categories: ship characteristics, shipping demand, port infrastructure, port performance and port location as listed in Table 5.1. Table 5.2 shows the descriptive statistics of the variables considered in this chapter. It can be seen from Table 5.2 that the average ship turn-around time was 1084 minutes during the analysis period (1163 minutes in 2012 and 986 minutes in 2013). The average berth occupancy rate was 39.8% in analysis period (42.2% in 2012 and 36.7% in 2013), which reflects that the berth occupancy rate was low for most of the ports. From the data, it can be seen that the number of arrived ships in Asia ports is more than that in non-Asia ports (62.5% of container ships arrived in Asia ports in analysis period, 71.1% in 2012 and 51.8% in 2013). The average ship length and height reflect the characteristics of container ships calling 109 at major ports. The average length of container ships was 168 meters in analysis period (175 meters in 2012 and 159 meters in 2013) and the average height of container ships was 25 meters in analysis period (26 meters in 2012 and 23 meters in 2013). The number of arrived ships calling at ports was 5724 in analysis period (6272 ships calling in 2012 and 5045 ships calling in 2013). 5.2.3 Models Building There are a total of eight models considered to study the relationship between ship turn-around time and the port and ship characteristics and they are listed in Table 5.3. Two types of probability models are considered in the study, i.e. count data model (Dionne et al., 1995; Maher and Summersgill, 1996; Quddus, 2008) and hazard-based duration model (Mannering, 1993; Hensher and Mannering, 1994; Mannering et al., 1994). Both models can estimate the probability of a specified ship turn-around time considering explanatory variables listed in Tables 5.1 and 5.2. The major difference is that the count data model (Models A to C) assumes ship turn-around time is discretized into finite period while the hazard-based duration model (Models D to H) assumes the time to be continuous. For the three count data models and five duration models, the modeling results of ship turn-around time are accomplished by the software package LIMDEP version 10 (Greene, 2012). 110 Ship’s characteristics Infrastructure Table 5.1 Variables Considered in Models Variable Description Ship turn-around time The time between a ship’s arrival and departure in a (STT) berth Ship gross tonnage Ship's overall internal volume (SGT) Ship size (SL,SH) Length of ship (SL); Height of ship (SH) Number of Berths Port’s total number of berth ( nB ) Berth length (BL) Port’s total length of all berths Port area (PA) Port’s total area Number of Cranes ( nC ) The total number of quay-cranes, transfer cranes (reach stackers, forklifts, straddle carriers, prime movers etc.) of a port Number of Ships ( nS ) Port’s total amount of arrived container ships in a year Shipping demand TEUs per ship (TPS) Port’s average number of containers loaded and unloaded(TEUs) of all arrived container ships in a year: TPS  AT nS Port performance Annual throughput (AT) Port’s total number of containers loaded and unloaded in 20-foot equivalent units (TEUs) in a year Berth occupancy rate (BOR) The ratio between the total time of all ships at berth and the total number of berths multiply by 360 days: BOR  Region Port location (PL) STT  nS 100% nB  360 days Whether the ship is arrived in Asia ports or not PL = 1(Asia); PL = 0 (Non-Asia) 111 Table 5.2 Descriptive Statistics of Variables Considered in Study (Standard deviation in parentheses) Variable Mean for variable in given period 2012 2013 Analysis period Number of container ships 2109 1702 3811 Number of ports 62 62 62 Ship turn-around time (STT) 1162.5 986.4 1083.9 (mins) (1301.6) (1033.6) (1192.5) Ship gross tonnage (SGT) (in 24455 20590 22728 100 ft3) (25609) (25247) (25517) Ship length (SL) (m) 175.5 158.6 167.9 (73.0) (78.9) (76.1) Ship height (SH) (m) 25.8 22.9 24.5 (9.6) (10.8) (10.2) 27 22 25 Number of berths ( nB ) (20) (21) (17) Berth length (BL) (m) 6499.5 5942.2 6250.6 (4880.8) (4275.9) (4628.1) Port area (PA) (in 1000 m2) 2439.8 2566.7 2496.4 (2085.1) (2163.9) (2121.3) 567 530 551 Number of cranes ( nC ) (442) (433) (438) 6272 5045 5724 Number of ship arrivals ( nS ) (6291) (5734) (6079) (ships/year) TEUs per ship (TPS) 1500.5 1697.3 1588.4 (TEU/ship) (792.3) (808.7) (805.5) Berth occupancy rate (BOR) 42.2 36.7 39.8 (%) (27.2) (27.5) (27.5) Location of port (PL) 0.711 0.518 0.625 (0.454) (0.500) (0.484) Note: Analysis period is from 2012 to 2013. Table 5.3 Models Considered in Study Model Type Model Description Discrete probability A Poisson regression model model B Negative binomial regression model C Poisson regression model with normal heterogeneity Continuous probability D Exponential model E Weibull F Log-logistic G Log-normal H Generalized gamma 112 5.3 Modeling Results of Container Ship Turn-around Time Modeling results of ship turn-around time in count data models and hazard-based duration models are presented in this section. 5.3.1 Results of Count Data Models 5.3.1.1 Selection of Appropriate Model Form of Count Data Models Based on the data collected from over 2000 ships in 61 international ports, count data models of ship turn-around time were developed. Three count data models (Models A to C) are employed. In order to determine the appropriate model form of the three models, model coefficients, McFadden R p 2 , over-dispersion test statistics and the Vuong test statistic were computed for the variables listed in Tables 5.1 and 5.2. A summary of the estimation results are shown in Tables 5.4 to 5.6. From the tables, the following observations can be made:  Based on the McFadden R p 2 , it is observed that the Poisson regression model with heterogeneity provides the best fit. The McFadden R p 2 value of the Poisson regression model with heterogeneity is 0.998 using ship turn-around time data in 2012, which is higher than that from the Poisson regression model (0.212) and negative binomial regression model (0.984). This is also consistent with the estimation result in 2013 (see Table 5.5) and the 2012-2013 analysis period (see Table 5.6).  Over-dispersion tests suggest that over-dispersion exists in data. This means that the use of the Poisson regression model is not suitable.  The Vuong statistics further shows the suitability of the Poisson regression 113 model with heterogeneity. The Vuong test result (29.9) for model developed using the 2012 dataset favors Poisson regression model with normal heterogeneity at a 95% significance level. Similar findings are observed for the 2013 ship turn-around time data (Table 5.5) and for the 2012-2013 analysis period (Table 5.6). Combining these findings, the Poisson regression model with normal heterogeneity is considered to be the most appropriate count data model in modeling ship turn-around time. 5.3.1.2 Parameter Estimation for Poisson Regression Model with Heterogeneity Table 5.7 shows the estimation results of the Poisson regression model with normal heterogeneity using the 2012, 2013 and analysis period’s data. It can be observed from the table that the number of berths ( nB ) and the berth occupancy rate (BOR) are the two most dominant predictors of ship turn-around time. The relationship between ship turn-around time and the characteristics of ships can be obtained from Table 5.7. It was found that container ships with larger gross tonnage (and hence internal volume can accommodate more containers) need more time for loading and unloading operations and hence a longer ship turnaround time. A longer ship length requires more quay-cranes to load and unload containers simultaneously and hence result in a shorter ship turn-around time. Higher ship requires a longer ship turn-around times for loading and unloading containers in more stacks. A positive coefficient for TEUs per ship (TPS) 114 indicates that ships with more containers require more time for loading and unloading operations, resulting in a longer ship turn-around time. The coefficients related to port infrastructure indicate that ports with larger area and more berths have a greater probability in servicing larger container ships and hence a higher chance of ships having a longer turn-around time. The coefficient of berth length (BL) is negative indicating that ships arriving at the port with longer length of berth have shorter ship turn-around times. It was also observed that ports with more cranes result in a shorter ship turn-around time. It was noted that for a given berth occupancy rate and number of berths, a higher number of ship arrivals is often associated with a shorter ship turn-around time The coefficient of berth occupancy rate (BOR) is positive in the model, indicating that longer ship turn-around times occurred with higher berth occupancy rate when the number of ships and berths of a port are constant. Finally, it was observed that ports in Asia have shorter ship turn-around times compared to ports in other continents. 115 Table 5.4 Estimation Results for Count Data Models using Data in 2012 Statistics Poisson Negative Binomial Log likelihood function LL(  ) -1052165 -16606 Restricted log likelihood LL(0) 2 -1335504 -1052165 - 566679 0.212 2071116 0.984 34411948 0.998 Over-dispersion test statistic for g (ui )  ui 18.2 - - Over-dispersion test statistic for 16.4 - - - - 29.9 2 McFadden R p g (ui )  ui Poisson with normal heterogeneity -35563 2 Vuong statistic Table 5.5 Estimation Results for Count Data Models using Data in 2013 Statistics Poisson Negative Binomial Log likelihood function LL(  ) -685871 -13181 Poisson with normal heterogeneity -27984 Restricted log likelihood LL(0) 2 -842100 -685871 - 312456 0.186 1345380 0.981 22791233 0.998 Over-dispersion test statistic for g (ui )  ui 13.2 - - Over-dispersion test statistic for 11.8 - - - - 24.3 2 McFadden R p g (ui )  ui 2 Vuong statistic Table 5.6 Estimation Results for Count Data Models using Data in Analysis Period Statistics Poisson Negative Binomial Log likelihood function LL(  ) -1764024 -29806 Poisson with normal heterogeneity -62793 Restricted log likelihood LL(0) 2 -2191173 -1764024 - 854297 3468436 57256667 McFadden R p 0.194 0.983 0.997 Over-dispersion test statistic for g (ui )  ui 22.1 - - Over-dispersion test statistic for 19.9 - - - - 37.9 2 g (ui )  ui 2 Vuong statistic Note: Analysis period is from 2012 to 2013. 116 Table 5.7 Estimation Results of Variable Coefficients in Poisson Regression Model with Normal Heterogeneity ( t -statistics in parentheses) Variable Poisson Regression Model with Normal Heterogeneity 2012 2013 Analysis period Constant 6.65 5.89 6.28 (65084.2) (47132.2) (81291.7) Ship gross tonnage (SGT) 0.76E-05 0.19E-05 0.5E-05 3 (in 100 ft ) (3685.8) (619.5) (3114.6) Ship length (SL) (m) -0.0032 -0.0003 -0.0018 (-4337.4) (-262.3) (-3265.7) Ship height (SH) (m) 0.0068 0.0062 0.0071 (3032.2) (2104.6) (4484.4) 0.025 0.029 0.027 Number of berths ( nB ) (10033.9) (8391.4) (13436.3) Berth length (BL) (m) -0.59E-04 -0.53E-04 -0.55E-04 (-4568.1) (-2984.6) (-5455.2) 2 Port area (PA) (in 1000 m ) 0.61E-04 0.31E-04 0.45E-04 (2982.2) (1239.9) (2921.7) -0.0002 -0.0001 -0.0001 Number of cranes ( nC ) (-2349.1) (-1025.6) (-2271.4) -0.73E-04 -0.66E-04 -0.72E-04 Number of ship arrivals ( nS ) (-8434.9) (-7147.2) (-) (ships/year) TEUs per ship (TPS) 0.0002 0.0002 0.0001 (TEU/ship) (6720.0) (6007.2) (8584.8) Berth occupancy rate (BOR) 0.015 0.016 0.016 (%) (16620.3) (12380.5) (22319.7) Location of port (PL) -0.37 -0.23 -0.31 (-6404.1) (-2870.0) (-6901.2) Standard Deviation of Heterogeneity Sigma 0.077 0.079 0.082 Number of observations 2109 1702 3811 Note: Analysis period is from 2012 to 2013. 117 5.3.2 Results of Hazard-based Duration Models 5.3.2.1 Selection of Appropriate Model Form of Duration Models Hazard-based duration models of ship turn-around time were developed and five duration models (Models D to H) in Table 5.3 are considered. In order to compare the five models, various diagnostic statistics such as log likelihoods are examined. The integrated hazard function introduced [Eq. (3.58)] can be used to diagnose model adequacy. If the model specification is valid, the integrated hazard function should exhibit a straight line emanating from the origin. Departure from this behavior might signal model misspecification (Greene, 2012). Figure 5.1 shows the integrated hazard function for the five models. The curvatures of the integrated hazard functions as seen from Figure 5.1 suggest that the Weibull, exponential and generalized gamma models are preferred over log-normal and log-logistic models. 118 (a) Model D (Exponential distribution) (b) Model E (Weibull distribution) (c) Model F (Log-logistic distribution) Figure 5.1 Integrated Hazard Function of Models D to H in Analysis Period 119 (d) Model G (Log-normal distribution) (e) Model H (Generalized gamma distribution) Figure 5.1 Integrated Hazard Function of Models D to H in Analysis Period (continued) 120 The estimation results are presented in Tables 5.8 to 5.10 and the following observations are made:  Based the log likelihood values, it is found that the generalized gamma model provides the best model fit. The log likelihood of the generalized gamma model is -3859.1 with data in 2012, which is higher than that of the Weibull (-3907.5), exponential (-4009.1), log-logistic (-4111.1) and log-normal (-4154.5) models. This is also consistent with the estimation result in 2013 (see Table 5.9) and the 2012-2013 analysis period (see Table 5.10).  The signs of coefficients suggest that the modeling results of the loglogistic model and log-normal model are not satisfied. For example, the sign of ship gross tonnage (SGT) coefficient should be positive since container ships with larger internal volume can accommodate more containers which in turn need more time for loading and unloading operations i.e. a longer ship turn-around time. The negative signs of SGT in both log-logistic and log-normal models in Table 5.8 reflect an unsatisfactory model result. Furthermore, the sign of numbers of cranes ( nC ) coefficient should be negative because ports with more cranes can service container ships in a shorter time.  The distribution parameters P and  show that the hazard results in the Weibull model and exponential model are not satisfactory. The hazard in the Weibull model is monotonic decreasing since P  1 (0.79) (as shown in Table 5.8). This means that the longer a container ship is in berth, the less 121 likely it is to leave soon. The hazard in exponential model is constant i.e. P  1 and this means that the hazard is independent of ship turn-around time. These behaviors are contrary to what is expected in actual operations and the two models should not be adopted. Combining these findings, the generalized gamma model is considered to be the most appropriate duration model for modeling ship turn-around time. 5.3.2.2 Parameter Estimation for Generalized Gamma Model The estimation results of the generalized gamma model are shown in Table 5.11, for models developed using 2012, 2013 and the analysis period data. It is found that the number of berths ( nB ), the number of ships ( nS ) and berth occupancy rate (BOR) are the dominant predictors of ship turn-around time. The relationship between ship turn-around time and ship and port characteristics of ships and ports are reflected by the sign of coefficients. Container ships with larger gross tonnage (SGT), higher ship height (SH) and more containers (TPS) result a longer ship turn-around time. A longer ship length (SL) requires more quay-cranes to load and unload containers simultaneously and hence result in a shorter ship turn-around time. Ports with larger area (PA) and more berths ( nB ) show a longer turn-around time while ports with longer berth length (BL) and more cranes ( nC ) result in a shorter ship turn-around time. The shape properties of the generalized gamma hazard function are determined by parameters P and  (Glaser, 1980). The shape of the hazard function is U-shaped when P  1  0 and P  1 . It is shown in Table 5.10 that the generalized gamma 122 hazard is U-shaped since P has a value of 1.63 and  has a value of 0.35 using 2012 data. This indicates that hazard is decreasing to minimum at a point t s (a certain dwell time of ships in the port and determined by P and  ) and then increasing thereafter. It means that for ship turn-around time less than time t s , the longer a container ship stays in berth, the less likely it is to leave soon. For ship turn-around time large than t s , a ship is more likely to leave as the time it stays in the berth increase. The result reflects the probability of a ship’s departure from the berth given the time it stays in a berth. The hazard results of the generalized gamma model using 2012 data are also consistent with the estimation result using 2013 data and the 2012-2013 analysis period’s data as shown in Table 5.11. The generalized gamma hazard function is U-shaped for models using 2013 data and the analysis period’s data. 123 Table 5.8 Duration Models for Ship Turn-around Time using Data in 2012 ( t -statistics in parentheses) Variable Model Generalized Weibull Exponential Log-logistic gamma Constant 7.53 5.97 6.27 4.59 (51.36) (36.79) (60.39) (21.79) Ship gross tonnage 0.48E-05 0.29E-05 0.4E-05 -0.51E-05 (SGT) (in 100 ft3) (2.08) (1.02) (2.13) (-1.39) Ship length (SL) (m) -0.003 -0.002 -0.002 0.002 (-7.10) (-1.97) (-4.55) (1.39) Ship height (SH) (m) 0.006 0.011 0.009 0.015 (1.08) (1.68) (2.08) (2.64) 0.024 0.038 0.033 0.056 Number of berths ( nB ) (6.16) (8.77) (11.28) (13.2) Berth length (BL) (m) -0.58E-04 -0.88E-04 -0.77E-04 -0.14E-03 (-3.67) (-4.83) (-6.31) (-7.19) Port area (PA) (in 1000 0.95E-04 0.001 0.99E-04 0.93E-04 m2 ) (3.49) (3.15) (4.58) (2.48) -0.14E-03 -0.13E-03 -0.13E-03 -0.35E-04 Number of cranes ( nC ) (-1.95) (-1.55) (-2.35) (-0.4) Number of ship arrivals -0.81E-04 -0.11E-03 -0.99E-04 -0.15E-03 (-7.11) (-8.64) (-11.47) (-11.69) ( nS ) (ships/year) TEUs per ship (TPS) 0.11E-03 0.14E-03 0.14E-03 0.12E-03 (TEU/ship) (2.65) (2.68) (4.12) (2.03) Berth occupancy rate 0.014 0.021 0.019 0.032 (BOR) (%) (10.16) (12.94) (17.37) (18.08) Location of port (PL) -0.28 -0.38 -0.35 -0.58 (-3.78) (-3.96) (-5.6) (-4.92) Distribution Parameter 1.63 0.79 1.00 1.06 P 0.39E-03 0.11E-02 0.98E-03 0.002  0.35  Log-likelihood at -3859.1 -3907.5 -4009.1 -4111.1 convergence Number of observations 2109 124 Log-normal 4.55 (19.77) -0.45E-05 (-1.12) 0.002 (1.11) 0.015 (2.31) 0.059 (10.74) -0.19E-03 (-8.37) 0.12E-03 (2.84) 0.99E-04 (1.05) -0.15E-03 (-10.22) 0.64E-04 (0.98) 0.033 (16.24) -0.47 (-3.57) 0.58 0.002 -4154.5 Table 5.9 Duration Models for Ship Turn-around Time using Data in 2013 ( t -statistics in parentheses) Variable Model Generalized Weibull Exponential Log-logistic gamma Constant 6.69 5.57 5.76 4.28 (40.66) (34.45) (49.94) (21.96) Ship gross tonnage 0.21E-05 0.15E-05 0.17E-05 -0.35E-06 (SGT) (in 100 ft3) (0.53) (0.34) (0.51) (-0.08) Ship length (SL) (m) -1.73E-03 -0.81E-03 -1.14E-03 1.94E-03 (-1.53) (-0.65) (-1.23) (1.45) Ship height (SH) (m) 0.013 0.016 0.015 0.020 (2.25) (2.79) (3.39) (4.36) 0.036 0.046 0.043 0.059 Number of berths ( nB ) (9.62) (11.28) (14.2) (13.86) Berth length (BL) (m) -0.74E-04 -0.82E-04 -0.8E-04 -0.69E-04 (-4.75) (-4.38) (-5.92) (-2.96) Port area (PA) (in 1000 0.4E-04 0.34E-04 0.36E-04 0.39E-04 m2 ) (1.71) (1.27) (1.81) (1.16) -1.6E-04 -1.9E-04 -1.8E-04 -2.7E-04 Number of cranes ( nC ) (-1.86) (-1.81) (-2.4) (-2.04) Number of ship arrivals -0.75E-04 -0.1E-03 -0.93E-04 -0.17E-03 (-8.55) (-9.88) (-12.45) (-13.89) ( nS ) (ships/year) TEUs per ship (TPS) 0.14E-03 0.12E-03 0.13E-03 0.53E-04 (TEU/ship) (3.24) (2.39) (3.52) (0.98) Berth occupancy rate 0.017 0.022 0.02 0.031 (BOR) (%) (11.94) (13.58) (17.21) (17.47) Location of port (PL) -0.295 -0.308 -0.305 -0.261 (-4.19) (-3.56) (-4.88) (-2.43) Distribution Parameter 1.39 0.84 1 1.18 P 0.56E-03 1.25E-03 1.14E-03 2.16E-03  0.47  Log-likelihood at -3003 -3019.8 -3062.8 -3150.4 convergence Number of observations 1702 125 Log-normal 4.28 (19.37) 0.25E-05 (0.5) 1.92E-03 (1.21) 0.014 (2.18) 0.057 (11.39) -0.59E-04 (-2.21) 0.54E-04 (1.48) -3.1E-04 (-1.92) -0.17E-03 (-13.14) -0.27E-04 (-0.47) 0.031 (15.42) -0.159 (-1.28) 0.62 2.55E-03 -3216.8 Table 5.10 Duration Models for Ship Turn-around Time using Data in Analysis Period ( t -statistics in parentheses) Variable Model Generalized Weibull Exponential Log-logistic Log-normal gamma Constant 7.07 5.75 6.00 4.41 4.45 (65.96) (52.48) (83.60) (31.66) (28.77) Ship gross tonnage 0.33E-05 0.21E-05 0.27E-05 -0.29E-05 -0.17E-05 (SGT) (in 100 ft3) (1.93) (0.95) (1.88) (-1.1) (-0.58) Ship length (SL) (m) -2.55E-03 -1.65E-03 -2.07E-03 1.83E-03 1.78E-03 (-7.93) (-2.45) (-5.26) (1.94) (1.64) Ship height (SH) (m) 0.014 0.017 0.016 0.018 0.014 (3.64) (4.14) (5.48) (5.18) (3.25) 0.03 0.041 0.038 0.056 0.058 Number of berths ( nB ) (10.75) (13.78) (17.66) (18.71) (15.32) Berth length (BL) (m) -0.69E-04 -0.87E-04 -0.8E-04 -0.10E-03 -0.14E-03 (-6.01) (-6.73) (-8.96) (-7.22) (-8.83) Port area (PA) (in 1000 0.65E-04 0.71E-04 0.68E-04 0.78E-04 0.1E-03 m2 ) (3.65) (3.36) (4.74) (3.11) (3.72) -0.14E-03 -0.16E-03 -0.15E-03 -0.17E-03 -0.14E-04 Number of cranes ( nC ) (-2.54) (-2.42) (-3.4) (-2.43) (-0.18) Number of ship arrivals -0.78E-04 -0.11E-03 -0.95E-04 -0.16E-03 -0.17E-03 (-10.82) (-13.01) (-16.87) (-18.24) (-16.71) ( nS ) (ships/year) TEUs per ship (TPS) 0.12E-03 0.13E-03 0.13E-03 0.91E-04 0.14E-04 (TEU/ship) (4.02) (3.51) (5.38) (2.23) (0.31) Berth occupancy rate 0.016 0.022 0.02 0.032 0.032 (BOR) (%) (16.02) (19.07) (25.29) (25.4) (22.65) Location of port (PL) -0.294 -0.353 -0.333 -0.439 -0.36 (-5.91) (-5.59) (-7.87) (-5.63) (-4.03) Distribution Parameter 1.46 0.81 1 1.11 0.59 P 0.47E-03 1.16E-03 1.04E-03 2.05E-03 2.47E-03  0.42  Log-likelihood at -6886.3 -6944 -7092.7 -7288 -7407.3 convergence Number of observations 3811 Note: Analysis period is from 2012 to 2013. 126 Table 5.11 Estimation Results of Variable Coefficients in Generalized Gamma Model ( t -statistics in parentheses) Variable Generalized Gamma Model 2012 2013 Analysis period Constant 7.53 6.69 7.07 (51.36) (40.66) (65.96) Ship gross tonnage (SGT) (in 0.48E-05 0.21E-05 0.33E-05 3 100 ft ) (2.08) (0.53) (1.93) Ship length (SL) (m) -0.003 -1.73E-03 -2.55E-03 (-7.10) (-1.53) (-7.93) Ship height (SH) (m) 0.006 0.013 0.014 (1.08) (2.25) (3.64) 0.024 0.036 0.03 Number of berths ( nB ) (6.16) (9.62) (10.75) Berth length (BL) (m) -0.58E-04 -0.74E-04 -0.69E-04 (-3.67) (-4.75) (-6.01) Port area (PA) (in 1000 m2) 0.95E-04 0.4E-04 0.65E-04 (3.49) (1.71) (3.65) -0.14E-03 -1.6E-04 -0.14E-03 Number of cranes ( nC ) (-1.95) (-1.86) (-2.54) -0.81E-04 -0.75E-04 -0.78E-04 Number of ship arrivals ( nS ) (-7.11) (-8.55) (-10.82) (ships/year) TEUs per ship (TPS) 0.11E-03 0.14E-03 0.12E-03 (TEU/ship) (2.65) (3.24) (4.02) Berth occupancy rate (BOR) 0.014 0.017 0.016 (%) (10.16) (11.94) (16.02) Location of port (PL) -0.28 -0.295 -0.294 (-3.78) (-4.19) (-5.91) Distribution Parameter 1.63 1.39 1.46 P 0.39E-03 0.56E-03 0.47E-03  0.35 0.47 0.42  Log-likelihood at -3859.1 -3003 -6886.3 convergence Number of observations 2109 1702 3811 Note: Analysis period is from 2012 to 2013. 127 5.3.3 Elasticity Analysis using Generalized Gamma Ship Turn-around Time Model To better understand the influence of ship and port variables on the ship turnaround time, the elasticity of all variables in the generalized gamma model in the analysis period are computed and are presented in Table 5.12. Elasticity is defined as the percentage change in dependent variable due to a 1% change in the independent variable. It is found that the number of berth ( nB ) has the largest elasticity, followed by berth occupancy rate (BOR) and ship height (SH). For example, a 1% increase in the total number of berths of a port results in a 32.75% increase in ship turn-around time and a 1% increase in the height of an arrived container ship results in 14.89% increase in ship turn-around time. This is because these factors are closely related to the terminal loading and unloading operations, which in turn determines the ship turn-around time. When ships arrive at a port, the number of berth and berth occupancy rate of the port affect its capability to serve the shipping demand and thus yield influence on ship turn-around time. Ship height is indicative of the capacity of ships and has an impact on terminal service time. 128 Table 5.12 Elasticity of Variables in Generalized Gamma Model Variables Elasticity Explanation Ship gross 0.004 1% increase in each ship gross tonnage (in 100 ft3) tonnage (SGT) results in 0.004% increase in ship turn-around time (in 100 ft3) (minutes) Ship length (SL) -2.76 1% increase in each ship length (m) results in 2.76% (m) decrease in ship turn-around time (minutes) Ship height (SH) 14.89 1% increase in each ship height (m) results in (m) 14.89% increase in ship turn-around time (minutes) Number of 32.75 1% increase in the total number of a port’s berths results in 32.75% increase in ship turn-around time berths ( nB ) (minutes) Berth length -0.07 1% increase in total berth length (m) of a port results (BL) (m) in 0.07% decrease in ship turn-around time (minutes) Port area (PA) 0.07 1% increase in each port area (in 1000 m3) results in 2 (in 1000 m ) 0.07% increase in ship turn-around time (minutes) Number of -0.15 1% increase in the total number of a port’s cranes results in 0.15% decrease in ship turn-around time cranes ( nC ) (minutes) Number of ship -0.08 1% increase in the total number of a port’s arrived container ships results in 0.08% decrease in ship arrivals ( nS ) turn-around time (minutes) (ships/year) TEUs per ship 0.13 1% increase in TEUs per ship results in 0.13% (TPS) increase in ship turn-around time (minutes) (TEU/ship) Berth occupancy 17.78 1% increase in berth occupancy rate of each port rate (BOR) (%) results in 17.78% increase in ship turn-around time (minutes) 5.3.4 Temporal Stability of Ship Turn-around Time Model The temporal stability of the developed model is next assessed. To test the temporal stability in the generalized gamma model, the likelihood ratio test is performed and the test results are shown in Table 5.13. Table 5.13 shows that the computed  2 value is 48.17 and the  20.05,12 value of 21.03. The null hypothesis (i.e. the model is stable over time) is rejected because  2   20.05,12 , indicating that instability exists in the analysis period. There are a number of possible 129 explanations for this instability. One is that the data of ships in 2013 is not sufficient as only three month data was collected (from Oct to Dec) compared with the five month data collected in 2012 (from Aug to Dec). A more likely explanation of temporal instability is that port’s infrastructure can be constructed and port performance change over time, resulting in a change in the dynamics of port performance. Table 5.13 Temporal Stability Tests for Generalized Gamma Models 2012 2013 Analysis period Log-likelihood at -3859.14 -3003.04 -6886.27 convergence Degrees of freedom 12 12 12 2 2   2[ LL(T )  LL(a )  LL( b )]  48.17 Null hypothesis is  2   20.05,12  21.03 rejected (i.e. instability existed) 5.3.5 Comparison between Poisson Regression Model with Heterogeneity and Generalized Gamma Model In order to compare which model is better for modeling ship turn-around time between the Poisson regression model with heterogeneity and generalized gamma model, the predicted ship turn-around time in the two models is calculated as shown in Table 5.14. Table 5.14 presents the 5th, 25th, 50th, 75th and 95th percentile of ship turn-around time distribution. It was found that the estimated ship turn-around time in the generalized gamma model fits better to the actual observation as compared to the Poisson regression model with heterogeneity. Figure 5.2 illustrates the fit from the Table 5.14 when the cumulative probability (i.e. the probability that the ship turn-around time is 130 less than or equal to a value) is 5%, the estimated ship turn-around time is 552 min and 13 min in the Poisson regression model with normal heterogeneity and generalized gamma model respectively compared against the observed ship turnaround time of 7 min. The results suggest that it is more appropriate to study ship turn-around time using duration models. It is not surprised that the generalized gamma model outperforms the Poisson regression model. This is because the former one is a continuous probabilistic model. Table 5.14 Comparison of Ship Turn-around Time between Probability Models and Observation Cumulative Probability Ship turn-around time (min) 5% 25% 50% 75% Observed Time 7 211 753 1470 Time predicted from Poisson regression with normal 552 698 857 1094 heterogeneity (7786%) (230%) (13%) (25%) (Percentage difference in parentheses) Time predicted from generalized gamma 13 185 609 1392 (Percentage difference in (86%) (12%) (19%) (5%) parentheses) 131 95% 3623 1535 (58%) 3604 (0.5%) Figure 5.2 Comparison of Ship Turn-around Time Obtained from Probability Models and Observed Values 5.5 Summary This chapter explores the potential of using count data and hazard-based duration models to relate ship turn-around time to ship and port factors. Three count data models were developed to study the relationship between ship turn-around time and port and ship characteristics. The models include Poisson regression model, negative binomial regression model and Poisson regression model with normal heterogeneity. It was found that the Poisson regression model with normal heterogeneity is the most appropriate in modeling ship turn-around time. The number of berths ( nB ) and berth occupancy rate (BOR) are considered to be the two most dominant predictors of ship turn-around time. Five hazard-based duration models (namely the exponential, Weibull, log-logistic, log-normal and generalized gamma model) were developed to study the ship turn132 around time. From the analyses presented in the chapter, the generalized gamma model is considered to be the most appropriate. The hazard function of generalized gamma model is U-shaped, suggesting that the hazard of a ship departing a berth is not always monotonic increasing with time. Elasticity of all variables in the generalized gamma model shows that the number of berth ( nB ) has the largest elasticity and the result of temporal stability test implies that instability exists in the analysis period of the generalized gamma model. A comparison between the Poisson regression model with heterogeneity and the generalized gamma model was made to determine the model with the best fit. It was shown that the estimated ship turn-around time by the generalized gamma model is much closer to the observed value as compared to that in the Poisson regression model with heterogeneity. 133 CHAPTER 6 PORT EFFICIENCY ANALYSIS CONSIDERING SHIP TURN-AROUND TIME 6.1 Introduction Container throughput is the most widely accepted container terminal output variable in past port efficiency studies since it relates closely to the need for cargo-related facilities and services (Roll and Hayuth, 1993; Tongzon, 2001; Cullinane et al., 2004; Park and De, 2004). It was noted that most of these studies do not consider the ship turn-around time as an output measure in DEA models for port efficiency analysis. As highlighted in Chapter 5 that it is important to consider the ship turn-around time as the service measure of port efficiency. There is a need to consider this parameter in the DEA models. This chapter aims to study the efficiency of 61 international container ports taking into consideration ship turn-around time. Both DEA-CCR and DEA-BCC models are developed using single-output and multiple-output measures. Port efficiency is evaluated using throughput and ship turn-around time as two separate single output variable in the DEA models, followed by using these two measures in the multiple-output DEA model. 6.2 Concept of Container Port Production 6.2.1 Container Port Production and Operations Understanding the process of production is a key prerequisite to study the efficiency of ports. The operations in a container port mainly include the transfer of containers by quay cranes along the berth, the storage system of containers in 134 yard terminals, gate operations, and so on. Cullinane et al. (2005) emphasized that the quay transfer operation fundamentally decides the efficiency of a port and summarized the elements of efficient production and operations of container terminals or ports:  Utilization of quay cranes and the optimization of berth length  Container ship turn-around time at port  Approaches to optimum containership stowage  Optimum yard operations Ship turn-around time can reflect port efficiency as it is heavily dependent on the quay-crane loading and unloading operations in the terminal. A shorter ship turnaround time in a port is analogous to a better level of service. 6.2.2 Considering Variables Affecting Port Efficiency in DEA Analysis The choice of input and output variables in port efficiency studies has been discussed by many scholars (Estache et al., 2002; Panayides et al., 2009). Most scholars suggested that the inputs of a port are variables related to the land, labor and equipment (Dowd and Leschine, 1990; Tongzon, 2001; Cullinane and Song, 2003) and container throughput is the most widely used output (Roll and Hayuth, 1993; Tongzon, 2001; Park and De, 2004; Cullinane et al., 2004). To study the efficiency of ports considering ship turn-around time, some scholars used variables such as ship calls, ship working rate and average number of containers handled per hour per ship in DEA models (Tongzon, 2001; Barros et al., 2003; Rios and Maçada, 2006). In most cases, those variables are used as 135 output measures in DEA models. For example, Roll and Hayuth (1993) used throughput and ship calls as output variables in DEA models. Tongzon (2001) used throughput and ship working rate as output measures to study port efficiency. A detailed discussion on the variables used in past literature had been earlier listed in Table 2.2. Compared to the other approaches, one advantage of using DEA models is that the performance of a decision making units (DMU) can be measured by using multiple inputs and outputs. This is suitable for port efficiency studies as there are a variety of port-related activities that need to be considered. 6.3 Empirical Setting 6.3.1 Ports and Analysis Period 61 international container ports are studied in this chapter and they are the same as those listed in Table 4.1. Port data was obtained from the Containerization International Yearbooks (2012 and 2013) and the relevant websites of port authorities. The analysis period in this chapter is 2012 to 2013. 6.3.2 Input and Output Variables Similar to Chapter 4, the selection of input variables focuses on the necessary port infrastructure including berths, terminal area, storage capacity and equipment. There are a total of seven input variables considered in the DEA models, namely number of berths, total length of berths, average berth depth, terminal area, storage capacity, number of quayside gantry cranes and number of yard cranes. The list of input variables and their descriptive statistics are listed in Table 6.1. 136 Table 6.1 also shows the four output variables considered in this chapter. The definition of each variable is described as follows:  Annual Throughput (AT): the annual throughput of each container port, measured by the total number of containers loaded and unloaded in 20foot equivalent units (TEUs)  Ship Turn-around Time (STT): the average ship turn-around time of all arrived container ships at a port  Ships ( nS ): the number of arrived container ships of a port in a year  Berth Occupancy Rate (BOR): the ratio of the total time that all ships stayed at a port to the total working time of all berths in a year 6.3.3 Models Considered in Study Chapter 4 had presented the use of output-oriented DEA in port performance studies. In this chapter, a total of six output-oriented DEA models are considered to analyze the efficiency of ports with ship turn-around time taken into account. Table 6.2 shows the various DEA-CCR and DEA-BCC models considered in this study. Method DEA-CCR DEA-BCC Table 6.2 Models Considered in Study Model Description A Throughput (AT) as single output B Ship turn-around time (STT) as single output C Throughput (AT), Ship turn-around time (STT), Ships ( nS ) and Berth occupancy rate (BOR) as outputs D Throughput (AT) as single output E Ship turn-around time (STT) as single output F Throughput (AT), Ship turn-around time (STT), Ships ( nS ) and Berth occupancy rate (BOR) as outputs 137 Table 6.1 Descriptive Statistics of the Input and Output Variables Considered in DEA Models Annual Throughput (TEU) Output Variable Ship Ships Turn(no.) around Time (min) Berth Occupancy Rate (%) Berth (no.) Total Berth Length (m) Average Berth Depth (m) Input Variables Terminal Area (m2) Storage Capacity (TEU) Quayside Gantry Cranes (no.) Yard Cranes and Tractors (no.) 2012 Max 32,529,000 3,672 23,925 90.5 82 21,896 16 9,441,323 2,788,326 204 1943 Min 596,000 114 346 2.4 2 540 7.8 175,000 5,950 4 24 Mean 5,666,455 725 4,077 37.3 20 5,553 12.9 2,397,550 206,691 41 464 Standard Deviation 2013 Max 6,622,854 721 4,968 25.9 17 4,092 1.4 2,035,050 425,577 36 403 33,626,000 1881 24,977 91.5 82 21,896 16 9,441,323 2,788,326 210 2040 Min 638,000 106 361 2.3 3 540 7.8 175,000 5,950 4 25 Mean 5,852,269 613 4,155 34 20 5,553 12.9 2,397,550 206,691 42 482 Standard Deviation 6,802,093 565 4,999 25.8 17 4,092 1.4 2,035,050 425,577 37 422 138 6.4 Results of the Efficiency Analysis and Interpretation 6.4.1 Throughput as Single Output in DEA Models Table 6.3 shows the efficiency of studied ports when throughput is the single output variable. From the results presented in the table, it was found that the ports estimated to be efficient by the DEA-CCR model in 2012 are the same with that in 2013. 6 ports are considered to be CCR-efficient in the analysis period, namely Hong Kong, Lianyungang, Qingdao, Shanghai, Singapore and Xiamen. The results in the DEA-BCC model show that 10 and 12 out of 61 ports were considered to be efficient in 2012 and 2013 respectively. From the results in Table 6.3, it was found that only large ports that have an absolute advantage in throughput are efficient. The six CCR-efficient ports all had throughput above 5,000,000 TEUs during the analysis period. Small ports with low container throughput tend to be inefficient, for example, Duisburg and Dalian. The low average port efficiency in both DEA-CCR (0.4070 in 2012 and 0.41 in 2013) and DEA-BCC (0.4632 in 2012 and 0.4761 in 2013) models is suggestive that most of the ports are inefficient during the analysis period. 139 Table 6.3 Efficiency Estimations in DEA models when Throughput is the Single Output 2012 2013 Port DEA-CCR DEA-BCC Scale DEA-CCR DEA-BCC Scale Efficiency Efficiency Efficiency Efficiency Efficiency Efficiency Shanghai 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 Singapore 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 Hong Kong 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 Xiamen 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 Qingdao 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 Lianyungang 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 Balboa 0.7665 1.0000 0.7665 0.7289 1.0000 0.7289 Duisburg 0.3619 1.0000 0.3619 0.3602 1.0000 0.3602 Haifa 0.2462 1.0000 0.2462 0.2567 1.0000 0.2567 Manila 0.5218 1.0000 0.5218 0.4882 1.0000 0.4882 Fuzhou 0.4177 0.9999 0.4177 0.4255 1.0000 0.4255 Freeport 0.1867 0.2043 0.9139 0.3148 1.0000 0.3148 Algeciras 0.3818 0.3866 0.9876 0.4030 0.4032 0.9994 Antwerp 0.3723 0.3879 0.9598 0.3533 0.3698 0.9555 Barcelona 0.1941 0.2012 0.9647 0.1922 0.1956 0.9826 Bremen 0.4096 0.4370 0.9373 0.3886 0.4013 0.9682 Brisbane 0.1186 0.1319 0.8992 0.1167 0.1295 0.9011 Buenos Aires 0.2065 0.2153 0.9591 0.2071 0.2104 0.9840 Busan 0.7759 0.7760 0.9999 0.7774 0.7791 0.9979 Charleston 0.1603 0.1715 0.9347 0.1640 0.1753 0.9352 Colombo 0.5028 0.5034 0.9988 0.4694 0.4763 0.9855 Constantza 0.0812 0.0928 0.8750 0.0809 0.0903 0.8958 Dalian 0.6557 0.7399 0.8862 0.7855 0.8682 0.9048 Dubai 0.6833 0.6907 0.9893 0.6806 0.6958 0.9781 Felixstowe 0.3371 0.3602 0.9359 0.3286 0.3518 0.9340 Genoa 0.1868 0.1872 0.9979 0.1941 0.1961 0.9900 Gioia Tauro 0.3442 0.3971 0.8668 0.6057 0.6057 0.9999 Guangzhou 0.8435 0.8515 0.9906 0.8508 0.8531 0.9973 Hamburg 0.4137 0.4228 0.9785 0.4146 0.4246 0.9763 Houston 0.2252 0.2254 0.9991 0.2127 0.2132 0.9979 Incheon 0.3744 0.3807 0.9835 0.3790 0.3839 0.9873 Jawaharlal 0.3830 0.3939 0.9723 0.3507 0.3609 0.9716 Nehru Note: efficiency score ‘1.0000’ equates to maximum efficiency and means efficient 140 Table 6.3 Efficiency Estimations in DEA models when Throughput is the Single Output (continued) 2012 2013 DEA-CCR DEA-BCC Scale DEA-CCR DEA-BCC Scale Port Efficiency Efficiency Efficiency Efficiency Efficiency Efficiency Kaohsiung 0.9397 0.9874 0.9517 0.8833 0.9194 0.9607 Karachi 0.2265 0.2367 0.9569 0.2267 0.2369 0.9569 Keelung 0.2262 0.2406 0.9401 0.2072 0.2160 0.9592 Kingston 0.1776 0.1790 0.9922 0.1542 0.1547 0.9967 Kobe 0.2186 0.2190 0.9982 0.2172 0.2178 0.9977 La Spezia 0.1648 0.1857 0.8875 0.1447 0.1625 0.8907 Laem 0.2968 0.3102 0.9568 0.2850 0.2990 0.9529 Chabang Le Havre 0.1798 0.1827 0.9841 0.1751 0.1781 0.9831 Long Beach 0.3767 0.3824 0.9851 0.3764 0.3826 0.9838 Los Angeles 0.4477 0.4541 0.9859 0.4326 0.4394 0.9845 Melbourne 0.2867 0.2982 0.9614 0.2812 0.2993 0.9395 Montreal 0.1688 0.1774 0.9515 0.1570 0.1702 0.9224 Nagoya 0.2455 0.2455 0.9997 0.2308 0.2310 0.9990 New York 0.3495 0.3607 0.9689 0.3354 0.3464 0.9682 Oakland 0.1904 0.1932 0.9855 0.1778 0.1806 0.9843 Osaka 0.1840 0.1891 0.9730 0.1845 0.1895 0.9740 Rotterdam 0.4449 0.449 0.9909 0.4225 0.4276 0.9880 Seattle 0.1407 0.1566 0.8985 0.1242 0.1381 0.8995 Southampton 0.2190 0.2386 0.9179 0.2660 0.2805 0.9481 St Petersburg 0.2252 0.2299 0.9796 0.2287 0.2337 0.9789 Tacoma 0.1481 0.1644 0.9009 0.1593 0.1706 0.9336 Taichung 0.1705 0.1973 0.8642 0.1721 0.1970 0.8736 Tianjin 0.9348 0.9857 0.9484 0.9195 0.9696 0.9483 Tokyo 0.3122 0.3137 0.9952 0.2927 0.2942 0.9949 Valencia 0.4231 0.4654 0.9091 0.5247 0.5549 0.9454 Vancouver 0.2255 0.2511 0.8980 0.2205 0.2472 0.8920 BC Yantai 0.2657 0.2970 0.8946 0.2986 0.3229 0.9246 Yokohama 0.2374 0.2375 0.9996 0.1792 0.1794 0.9987 Zeebrugge 0.2517 0.2725 0.9237 0.2063 0.2219 0.9296 Average 0.4070 0.4632 0.9171 0.4100 0.4761 0.9119 Note: efficiency score ‘1.0000’ equates to maximum efficiency and means efficient 141 6.4.2 Ship Turn-around Time as Single Output in DEA Models Table 6.4 shows the efficiency of studied ports when ship turn-around time is used as single output in the DEA-CCR and DEA-BCC models. It was found that the numbers of efficient ports estimated by both DEA-CCR and DEA-BCC models are less than that shown in Table 6.3. Of the 61 ports, only 3 ports are considered to be CCR-efficient and 4 ports are BCC-efficient in 2012. In 2013, 2 out of 61 ports are estimated as CCR-efficient and 6 ports are BCC-efficient. The few efficient ports in Table 6.4 as compared to that in Table 6.3 suggest that most ports have unsatisfactory ship turn-around time. Most of the estimated efficient ports in Table 6.3 are considered to be inefficient in Table 6.4 and this includes Singapore, Hong Kong and Shanghai. This is because large ports with high container throughput may have longer ship turn-around time. Given the large area, long berth length but unsatisfactory ship turn-around time, these large ports are considered as inefficient use of terminal facilities. It was also found that efficient ports in Table 6.4 have a significant difference in their throughput performance. For example, in 2012, efficient ports include large ports with container throughput more than 10,000,000 TEUs (such as Qingdao) and small ports with throughput less than 1,300,000 TEUs (such as Duisburg). 142 Table 6.4 Efficiency Estimations in DEA models when Ship Turn-around Time is the Single Output 2012 2013 Port DEA-CCR DEA-BCC Scale DEA-CCR DEA-BCC Scale Efficiency Efficiency Efficiency Efficiency Efficiency Efficiency Qingdao 1.0000 1.0000 1.0000 0.6360 0.8966 0.7093 Duisburg 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 Balboa 1.0000 1.0000 1.0000 0.7988 1.0000 0.7988 Lianyungang 0.9039 1.0000 0.9039 0.8668 1.0000 0.8668 Rotterdam 0.5660 0.5680 0.9965 1.0000 1.0000 1.0000 Freeport 0.3188 0.3761 0.8476 0.2698 1.0000 0.2698 Guangzhou 0.5117 0.5133 0.9969 0.8964 1.0000 0.8964 Shanghai 0.2332 0.2340 0.9966 0.2321 0.2543 0.9127 Singapore 0.2018 0.2271 0.8886 0.1544 0.1765 0.8748 Hong Kong 0.2089 0.2093 0.9981 0.1280 0.1528 0.8377 Xiamen 0.7649 0.7655 0.9992 0.8048 0.8396 0.9586 Haifa 0.1136 0.9997 0.1136 0.2082 0.9998 0.2082 Manila 0.1280 0.9997 0.1280 0.1440 0.9997 0.1440 Fuzhou 0.4462 0.9993 0.4465 0.4389 0.9993 0.4392 Algeciras 0.2287 0.2819 0.8113 0.2087 0.2686 0.7770 Antwerp 0.1413 0.1447 0.9765 0.0943 0.1248 0.7556 Barcelona 0.2450 0.2496 0.9816 0.3121 0.3323 0.9392 Bremen 0.1170 0.1259 0.9293 0.1199 0.1550 0.7735 Brisbane 0.0847 0.0848 0.9988 0.1484 0.1484 0.9999 Buenos 0.4754 0.5576 0.8526 0.6373 0.8096 0.7872 Aires Busan 0.1248 0.1250 0.9984 0.0808 0.0978 0.8262 Charleston 0.5700 0.5708 0.9986 0.4033 0.4056 0.9943 Colombo 0.1252 0.1294 0.9675 0.1511 0.1726 0.8754 Constantza 0.1319 0.1333 0.9895 0.1231 0.1256 0.9801 Dalian 0.0868 0.0999 0.8689 0.0819 0.0960 0.8531 Dubai 0.1082 0.1085 0.9972 0.1224 0.1434 0.8536 Felixstowe 0.2124 0.2132 0.9962 0.2076 0.2334 0.8895 Genoa 0.1749 0.1854 0.9434 0.2437 0.2925 0.8332 Gioia Tauro 0.1741 0.1904 0.9144 0.1495 0.2761 0.5415 Note: efficiency score ‘1.0000’ equates to maximum efficiency and means efficient 143 Table 6.4 Efficiency Estimations in DEA models when Ship Turn-around Time is the Single Output (continued) 2012 2013 Port DEA-CCR DEA-BCC Scale DEA-CCR DEA-BCC Scale Efficiency Efficiency Efficiency Efficiency Efficiency Efficiency Hamburg 0.0696 0.0697 0.9986 0.0759 0.0988 0.7682 Houston 0.0396 0.0451 0.8780 0.0880 0.1020 0.8627 Incheon 0.7864 0.8024 0.9801 0.6934 0.7956 0.8715 Jawaharlal 0.1628 0.1629 0.9994 0.1492 0.1741 0.8570 Nehru Kaohsiung 0.1149 0.1363 0.8430 0.1077 0.1437 0.7495 Karachi 0.1385 0.1420 0.9754 0.1295 0.1333 0.9715 Keelung 0.1688 0.1724 0.9791 0.2036 0.2081 0.9784 Kingston 0.1102 0.1103 0.9991 0.1247 0.1359 0.9176 Kobe 0.1843 0.2045 0.9012 0.1589 0.2075 0.7658 La Spezia 0.1930 0.1963 0.9832 0.1646 0.1678 0.9809 Laem 0.1689 0.1804 0.9363 0.1225 0.1711 0.7160 Chabang Le Havre 0.2114 0.2460 0.8593 0.1571 0.2026 0.7754 Long Beach 0.0429 0.0445 0.9640 0.0665 0.0918 0.7244 Los Angeles 0.0543 0.0551 0.9855 0.0708 0.0902 0.7849 Melbourne 0.1026 0.1029 0.9971 0.1545 0.1583 0.9760 Montreal 0.0971 0.0972 0.9990 0.2211 0.2266 0.9757 Nagoya 0.1157 0.1243 0.9308 0.1127 0.1409 0.7999 New York 0.0584 0.0702 0.8319 0.0832 0.1062 0.7834 Oakland 0.2612 0.3063 0.8528 0.0922 0.1186 0.7774 Osaka 0.1802 0.2090 0.8622 0.1556 0.1965 0.7919 Seattle 0.0461 0.0545 0.8459 0.0589 0.0835 0.7054 Southampton 0.1117 0.1171 0.9539 0.1612 0.2232 0.7222 St 0.0959 0.0962 0.9969 0.1093 0.1208 0.9048 Petersburg Tacoma 0.0981 0.1241 0.7905 0.0664 0.0915 0.7257 Taichung 0.1766 0.1786 0.9888 0.1528 0.1548 0.9871 Tianjin 0.1183 0.1331 0.8888 0.1119 0.1586 0.7055 Tokyo 0.1576 0.1619 0.9734 0.1502 0.2008 0.7480 Valencia 0.1497 0.1497 1.0000 0.1701 0.1701 0.9999 Vancouver 0.0470 0.0665 0.7068 0.0565 0.0844 0.6694 BC Yantai 0.1400 0.1402 0.9986 0.1784 0.1787 0.9983 Yokohama 0.2169 0.2184 0.9931 0.4942 0.6302 0.7842 Zeebrugge 0.6237 0.6244 0.9989 0.5873 0.6253 0.9392 Average 0.2564 0.3055 0.9087 0.2638 0.3507 0.8084 Note: efficiency score ‘1.0000’ equates to maximum efficiency and means efficient 144 6.4.3 DEA Models with Multiple Outputs Table 6.5 presents the efficiency of the studied ports when multiple outputs in the DEA-CCR and DEA-BCC models (throughput, ship turn-around time, number of container ships and berth occupancy rate) are used. It was shown that out of the 61 ports, 9 ports were considered to be CCR-efficient and 14 ports were BCCefficient in 2012 compared with 12 ports were estimated as CCR-efficient and 20 ports were BCC-efficient in 2013. Compared to the results presented in Tables 6.3 and 6.4, the results in Table 6.5 suggest that more ports are likely to be estimated as efficient when both throughput and ship turn-around time are considered output variables. Ports estimated as efficient in Table 6.5 have a satisfactory performance on either container throughput or ship turn-around time. For example, large ports with high throughput such as Shanghai, Singapore, Hong Kong were estimated as both CCR-efficient and BCC-efficient in the analysis period. Small ports with short ship turn-around time such as Balboa and Duisburg were estimated as both CCRefficient and BCC-efficient. The results indicate that it may be more appropriate to use multiple outputs in the DEA models when estimating port efficiency. 145 Table 6.5 Port Efficiency using DEA Models with Multiple Outputs 2012 2013 Port DEA-CCR DEA-BCC Scale DEA-CCR DEA-BCC Efficiency Efficiency Efficiency Efficiency Efficiency Shanghai 1.0000 1.0000 1.0000 1.0000 1.0000 Singapore 1.0000 1.0000 1.0000 1.0000 1.0000 Hong Kong 1.0000 1.0000 1.0000 1.0000 1.0000 Xiamen 1.0000 1.0000 1.0000 1.0000 1.0000 Qingdao 1.0000 1.0000 1.0000 1.0000 1.0000 Lianyungang 1.0000 1.0000 1.0000 1.0000 1.0000 Balboa 1.0000 1.0000 1.0000 1.0000 1.0000 Duisburg 1.0000 1.0000 1.0000 1.0000 1.0000 Dalian 0.8543 1.0000 0.8543 0.8900 1.0000 Kaohsiung 1.0000 1.0000 1.0000 1.0000 1.0000 Haifa 0.7812 1.0000 0.7812 0.4817 1.0000 Manila 0.5222 1.0000 0.5222 0.4884 1.0000 Fuzhou 0.6198 1.0000 0.6198 0.6146 1.0000 Tianjin 0.9601 1.0000 0.9601 0.9551 1.0000 Rotterdam 0.6947 0.7594 0.9148 1.0000 1.0000 Gioia Tauro 0.6796 0.9088 0.7478 1.0000 1.0000 Guangzhou 0.9003 0.9471 0.9506 1.0000 1.0000 Freeport 0.4015 0.4628 0.8675 0.7194 1.0000 Algeciras 0.4266 0.4336 0.9839 0.4461 0.4553 Taichung 0.8745 0.9979 0.8763 0.9002 1.0000 Tokyo 0.9666 0.9909 0.9755 0.9916 1.0000 Antwerp 0.4344 0.4484 0.9688 0.4576 0.5423 Barcelona 0.4858 0.5232 0.9285 0.4494 0.4856 Bremen 0.5932 0.6865 0.8641 0.6060 0.6743 Brisbane 0.3017 0.4148 0.7273 0.2703 0.3372 Buenos Aires 0.4929 0.5691 0.8661 0.6592 0.8186 Busan 0.8276 0.8377 0.9879 0.8539 0.8647 Charleston 0.5944 0.6066 0.9799 0.4281 0.4605 Colombo 0.5128 0.5337 0.9608 0.5064 0.5105 Constantza 0.1794 0.2202 0.8147 0.1825 0.2336 Note: efficiency score ‘1.0000’ equates to maximum efficiency and means efficient 146 Scale Efficiency 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.8900 1.0000 0.4817 0.4884 0.6146 0.9551 1.0000 1.0000 1.0000 0.7194 0.9798 0.9002 0.9916 0.8438 0.9255 0.8987 0.8016 0.8053 0.9875 0.9296 0.9920 0.7813 Table 6.5 Port Efficiency using DEA Models with Multiple Outputs (continued) 2012 2013 Port DEA-CCR DEA-BCC Scale DEA-CCR DEA-BCC Scale Efficiency Efficiency Efficiency Efficiency Efficiency Efficiency Dubai 0.6943 0.6973 0.9957 0.6888 0.6965 0.9889 Felixstowe 0.3444 0.3831 0.8990 0.3717 0.3948 0.9415 Genoa 0.2287 0.2300 0.9943 0.3092 0.3253 0.9505 Hamburg 0.7121 0.8440 0.8437 0.6707 0.7503 0.8939 Houston 0.4103 0.4851 0.8458 0.2464 0.2686 0.9173 Incheon 0.9104 0.9597 0.9486 0.9259 0.9656 0.9589 Jawaharlal 0.4004 0.4157 0.9632 0.4096 0.4493 0.9116 Nehru Karachi 0.2826 0.3154 0.8960 0.2974 0.3370 0.8825 Keelung 0.8487 0.8704 0.9751 0.8914 0.9142 0.9751 Kingston 0.2448 0.2532 0.9668 0.2642 0.2655 0.9951 Kobe 0.5200 0.5201 0.9998 0.5330 0.5334 0.9993 La Spezia 0.2623 0.2702 0.9708 0.2480 0.2573 0.9639 Laem 0.4768 0.4976 0.9582 0.3578 0.3994 0.8958 Chabang Le Havre 0.2616 0.2788 0.9383 0.2418 0.2582 0.9365 Long Beach 0.4730 0.6239 0.7581 0.3895 0.4445 0.8763 Los Angeles 0.4993 0.5798 0.8612 0.4378 0.4721 0.9273 Melbourne 0.6130 0.6152 0.9964 0.5035 0.5588 0.9010 Montreal 0.1972 0.2145 0.9193 0.2830 0.2901 0.9755 Nagoya 0.7753 0.8740 0.8871 0.8049 0.8788 0.9159 New York 0.3497 0.3746 0.9335 0.3422 0.3490 0.9805 Oakland 0.3061 0.3334 0.9181 0.2064 0.2170 0.9512 Osaka 0.4982 0.5004 0.9956 0.5062 0.5184 0.9765 Seattle 0.2309 0.3369 0.6854 0.1979 0.2811 0.7040 Southampton 0.4313 0.5301 0.8136 0.3855 0.4042 0.9537 St Petersburg 0.3561 0.3752 0.9491 0.3406 0.3559 0.9570 Tacoma 0.1673 0.2197 0.7615 0.1847 0.2680 0.6892 Valencia 0.5125 0.6691 0.7660 0.5872 0.6211 0.9454 Vancouver 0.2277 0.3190 0.7138 0.2279 0.2936 0.7762 BC Yantai 0.5087 0.6712 0.7579 0.5501 0.5792 0.9498 Yokohama 0.5282 0.5282 1.0000 0.7061 0.7062 0.9999 Zeebrugge 0.6346 0.6399 0.9917 0.6318 0.6433 0.9821 Average 0.5903 0.6919 0.9026 0.6007 0.6570 0.9124 Note: efficiency score ‘1.0000’ equates to maximum efficiency and means efficient 147 6.4.4 Comparison between Single and Multiple Output DEA Models Table 6.6 combines the results from Tables 6.3 to 6.5 and summarizes the efficient ports during the analysis period. The following observations can be made from the table:  When using container throughput as the single output measure in DEA models, large ports with high container throughput are more likely to be estimated as efficient.  Most large ports with high container throughput may end up being inefficient when ship turn-around time is used as the single output measure in DEA models. This is because given the large area, long berth length, large ports tend to be estimated as inefficient for unsatisfactory ship turn-around time.  It is more appropriate to use multiple outputs to estimate port efficiency. Container throughput relates closely to the need for cargo-related facilities and services and ship turn-around time directly reflects the loading and unloading terminal operations. 148 Year 2012 Models Table 6.6 Summary of Efficient Ports in DEA Models Applied with Single Output and Multiple Outputs A B C D E F Balboa, Dalian Duisburg, Fuzhou Haifa, Hong Kong Kaohsiung, Lianyungang Manila, Qingdao Shanghai, Singapore Tianjin, Xiamen Efficient Ports Hong Kong Lianyungang Qingdao Shanghai Singapore Xiamen Duisburg Balboa Qingdao Balboa, Duisburg Hong Kong, Kaohsiung Lianyungang, Qingdao Shanghai, Singapore Xiamen Balboa, Duisburg Haifa, Hong Kong Lianyungang, Manila Qingdao, Shanghai Singapore, Xiamen Balboa Duisburg Lianyungang Qingdao Total Number of Efficient Ports 6 3 9 10 4 14 2013 Efficient Ports Total Number of Efficient Ports Hong Kong Lianyungang Qingdao Shanghai Singapore Xiamen 6 Duisburg Rotterdam Balboa, Duisburg Gioia Tauro, Guangzhou Hong Kong, Kaohsiung Lianyungang, Qingdao Rotterdam, Shanghai Singapore, Xiamen Balboa, Duisburg Freeport, Fuzhou Haifa, Hong Kong Lianyungang, Manila Qingdao, Shanghai Singapore, Xiamen Balboa Duisburg Freeport Guangzhou Lianyungang Rotterdam Balboa, Dalian Duisburg, Freeport Fuzhou, Gioia Tauro Guangzhou, Haifa Hong Kong, Kaohsiung Lianyungang, Manila Qingdao, Rotterdam Shanghai, Singapore Taichung, Tianjin Tokyo, Xiamen 2 12 12 6 20 149 6.5 Summary This chapter studies the efficiency of container ports considering ship turn-around time. The efficiency of 61 ports in the analysis period (2012 to 2013) was obtained for different single and multiple output DEA models. The first part of this chapter studies the efficiency of ports using throughput and ship turn-around time as two separate single output variable in the DEA models. The results suggest that when throughput is the output variable, most of the ports considered to be efficient are large ports with high throughput. These ports may lose their competitive edge when turn-around time is used as the single output variable. It is shown that only a few ports were estimated as efficient either by the DEA-CCR (3 efficient ports in 2012) or DEA-BCC model (4 efficient ports in 2012 ) when ship turn-around time is the sole output measure. The second part of this chapter discusses the port efficiency results for multipleoutput in DEA models. It was found that ports estimated to be efficient perform well in either container throughput or ship turn-around time. Large ports with high throughput, small ports with short ship turn-around time and ports with high berth occupancy rate are potentially efficient. The results indicate that it is more appropriate to use multiple outputs to estimate the port efficiency as compared to two separate single-output measures in DEA models. In summary, there is a need to consider both throughput and ship turn-around time in port efficiency studies. Large ports with high throughput and small ports with 150 short ship turn-around time are both likely to be efficient when it comes to port performance. 151 CHAPTER 7 CONCLUSIONS AND RECOMMENDATIONS 7.1 Major Findings of Research In this increasingly competitive landscape of port industry, it is important for port operators to constantly review the performance of their ports so that they can keep their competitive advantage. Within such a competitive environment, it is important to have a reliable measurement of port performance so that useful advice can be drawn to port operators or managers to improve their port efficiency. The objectives of this research are (a) to determine an appropriate method to evaluate port’s efficiency using non-parametric approaches, (b) to develop appropriate probability models to relate ship turn-around time to the characteristics of ships and ports, and (c) to study port efficiency with consideration to the ship turn-around time. A comparative study using the DEA-CCR, DEA-BCC and FDH models to evaluate the efficiency of 61 global container ports during 2001 to 2011 was performed. It was found that FDH lacks the sensitivity to analyze port efficiency compared to the DEA models. A port which is estimated to be efficient by the FDH model is not necessarily better and this can cause issues when using FDH model results in port efficiency studies. DEA, however, is more stringent in determining efficient ports and should be used as a preferred method in port efficiency studies. 152 Three count data models and five duration models are used to model the ship turnaround time. It was found that the Poisson regression model with normal heterogeneity is the most appropriate in modeling ship turn-around time compared with the other two count data models (Poisson regression model, negative binomial regression model). It was also shown that the generalized gamma model provides the best fit compared with the other four duration models (Weibull, exponential, log-normal and log-logistic models). A comparison between the Poisson regression model with normal heterogeneity and the generalized gamma model was made. It was shown that the estimated ship turn-around time in the generalized gamma model provides a much better fit to actual data. The ship turn-around time was employed as an output measure in DEA models. It was shown that most of the ports considered to be efficient when throughput was the single output measure lose their competitive edge when ship turn-around time was used as the single output. It was also found that using throughput as the single output measure may not be fair to small ports with limited container throughput despite having an effective management and operations. The study suggests that it is more appropriate using multiple outputs such as throughput and ship turnaround time to estimate the efficiency of ports. 7.2 Recommendations for Further Research The research study in this thesis has assessed the capabilities of different nonparametric approaches to measure port efficiency, benchmarked global container ports and analyzed port efficiency with consideration to the ship turn-around time. Some recommendations for further work can be made: 153 (a) This study had mainly focused on the application of non-parametric techniques (FDH and DEA) in the measurement of port efficiency. Further work can also involve parametric approaches such as the deterministic or stochastic frontiers analysis. (b) It is difficult to obtain reliable data on input variables such as labor and handling cost. Only parameters related physical facilities are considered as input variables in the DEA models such as terminal area and berth length. For future studies, a complete database on land, labor and equipment can be used as input variables in DEA models. (c) In this thesis, the evaluation of port efficiency have emphasized on the selection of multiple output measures in DEA models and found that it is important to incorporate outputs related to ships, such as ship turn-around time into the measurement of port efficiency. However, it does not address the issue of assigning weight restrictions on output variables. It is therefore suggested that future research could incorporate weight restrictions in DEA models to produce a better evaluation of port performance and efficiency. (d) Another potentially interesting discussion point for the future work could be the analysis of slacks in the DEA models and the meaning of these slacks in the port efficiency context. 154 REFERENCES 1. Afriat, S. N. (1972). Efficiency estimation of production function, International Economic Review, 13(3), 568-598. 2. Aigner, D., Lovell, C. A. and Schmidt, P. (1977). Formulation and estimation of stochastic frontier production function models, Journal of Econometrics, 6(1), 21-37. 3. Aigner, D. J. and Chu, S. F. (1968). On estimating the industry production function, The American Economic Review, 826-839. 4. Ashar, A. (1997). Counting the moves, Port Development International, 25– 29. 5. Asia-Pacific Economic Cooperation (APEC). (1999). Assessing APEC trade liberalisation and facilitation: 1999 update, APEC Economic Committee Publication. 6. Banker, R. D., Charnes, A. and Cooper, W. W. (1984). Some models for estimating technical and scale inefficiencies in data envelopment analysis, Management science, 30(9), 1078-1092. 7. Barros, C. P. (2003). The measurement of efficiency of Portuguese sea port authorities with DEA, International Journal of Transport Economics, 30(3). 155 8. Barros, C. P. (2006). A benchmark analysis of Italian seaports using data envelopment analysis, Maritime Economics & Logistics, 8(4), 347-365. 9. Barros, C. P. and Athanassiou, M. (2004). Efficiency in European seaports with DEA: evidence from Greece and Portugal. Maritime Economics & Logistics, 6(2), 122-140. 10. Bauer, P. W. (1990). Recent developments in the econometric estimation of frontiers, Journal of Econometrics, 46(1), 39-56. 11. Bendall, H. B. and Stent, A. (1987). On measuring cargo handling productivity, Maritime Policy and Management, 14(4), 337-343. 12. Bergantino, A. S. and Musso, E. (2011). The role of external factors versus managerial ability in determining seaports’ relative efficiency: An input-by-input analysis through a multi-step approach on a panel of Southern European ports, Maritime Economics & Logistics, 13(2), 121-141. 13. Bichou, K. (2006). Review of port performance approaches and a supply chain framework to port performance benchmarking, Research in Transportation Economics, 17, 567-598. 14. Bichou, K. (2011). A two-stage supply chain DEA model for measuring container-terminal efficiency, International Journal of Shipping and Transport Logistics, 3(1), 6-26. 156 15. Bichou, K. and Rray, G. (2004). A Logistics and Supply Chain Management Approach to Port Performance Measurement, Maritime Policy & Management, 31(1), 47-67. 16. Bonilla, M., Medal, A., Casaus, T. and Sala, R. (2002). The traffic in Spanish ports: An efficiency analysis. International Journal of Transport Economics, 29(2). 17. Boske, L. B. and Cuttino, J. C. (2001). The Impacts of US-Latin American Trade on the Southwest's Economy and Transportation System: An Assessment of Impact Methodologies: Southwest Region University Transportation Center, Center for Transportation Research, the University of Texas at Austin. 18. Braeutigam, R. R., Daughety, A. F. and Turnquist, M. A. (1984). A firm specific analysis of economies of density in the US railroad industry, Journal of Industrial Economics, 33(1), 3-20. 19. Brooks, M. R. (2006). Issues in measuring port devolution program performance: a managerial perspective. Research in Transportation Economics, 17, 599-629. 20. Buckley, P. (1992). A transportation-oriented interregional computable general equilibrium model of the United States, Annals of Regional Science 26:331-48. 21. Burke, P. (1975). Technical Note—Delays in Single-Server Queues with Batch Input, Operations Research, 23(4), 830-833. 157 22. Cameron, A. C. and Trivedi, P. K. (1990). Regression-based tests for overdispersion in the Poisson model, Journal of Econometrics, 46(3), 347-364. 23. Cameron, A. C. and Trivedi, P. K. (2009). Microeconometrics using stata (Vol. 5): Stata Press College Station, TX. 24. Charnes, A., Cooper, W. W. and Rhodes, E. (1978). Measuring the efficiency of decision making units, European Journal of Operational Research, 2(6), 429-444. 25. Clark, X., Dollar, D. and Micco, A. (2004). Port efficiency, maritime transport costs and bilateral trade, Journal of Development Economics 75:417-50. 26. Coelli, T., Prasada Rao, D. S. and Battese, G. E. (1998). An Introduction to Efficiency and Productivity Analysis, Kluwer Academic Publishers: Boston, Dordrecht and London. 27. Commission, A. P. (1998). International benchmarking of the Australian waterfront. Canberra, Australia: Ausinfo. 28. Cooper, W. W., Seiford, L. M. and Tone, K. (2007). Data Envelopment Analysis, A Comprehensive Text with Models, Applications, References and DEA-Solver Software Second Edition, Springer, ISBN, 387452818, 490. 29. Coto-Millan, P., Banos-Pino, J. and Rodriguez-Alvarez, A. (2000). Economic efficiency in Spanish ports: some empirical evidence, Maritime Policy and Management, 27(2), 169-174. 158 30. Cullinane, K. (2002). The productivity and efficiency of ports and terminals: methods and applications, The handbook of maritime economics and business, 803-831. 31. Cullinane, K., Fei, W. T. and Cullinane, S. (2004). Container Terminal Development in Mainland China and Its Impact on the Competitiveness of the Port of Hong Kong, Transport Reviews, 24(1), 33-56. 32. Cullinane, K., Ji, P. and Wang, T. (2005). The relationship between privatization and DEA estimates of efficiency in the container port industry, Journal of Economics and Business, 57(5), 433-462. 33. Cullinane, K. and Song, D.W. (2003). A stochastic frontier model of the productive efficiency of Korean container terminals, Applied Economics, 35(3), 251-267. 34. Cullinane, K., Song, D.W. and Gray, R. (2002). A stochastic frontier model of the efficiency of major container terminals in Asia: assessing the influence of administrative and ownership structures, Transportation Research Part A: Policy and Practice, 36(8), 743-762. 35. Cullinane, K., Song, D.W., Ji, P. and Wang, T.F. (2004). An application of DEA windows analysis to container port production efficiency, Review of Network Economics, 3(2). 159 36. Cullinane, K., Song, D.W. and Wang, T. (2005). The application of mathematical programming approaches to estimating container port production efficiency, Journal of Productivity Analysis, 24(1), 73-92. 37. Cullinane, K. and Song, D.W. (2006). Estimating the relative efficiency of European container ports: a stochastic frontier analysis, Research in Transportation Economics, 16, 85-115. 38. Cullinane, K. P. and Wang, T.F. (2006). The efficiency of European container ports: a cross-sectional data envelopment analysis, International Journal of Logistics: Research and Applications, 9(1), 19-31. 39. Cullinane, K. and Wang, T.F. (2006). Data envelopment analysis (DEA) and improving container port efficiency, Research in Transportation Economics, 17, 517-566. 40. Cullinane, K. P. B. (2002). The Productivity and efficiency of ports and terminals: Methods and applications, Handbook of maritime economics and business, 803-831. 41. Daganzo, C. F. (1990). The productivity of multipurpose seaport terminals, Transportation Science, 24(3), 205-216. 42. Davis, H. C. (1983). Regional port impact studies: a critique and suggested methodology, Transportation Journal, 61-71. 160 43. De Borger, B. and Kerstens, K. (1996). Cost efficiency of Belgian local governments: A comparative analysis of FDH, DEA, and econometric approaches, Regional Science and Urban Economics, 26(2), 145-170. 44. De Monie, G. (1987). Measuring and evaluating port performance and productivity: United Nations. 45. De, P. and Ghosh, B. (2003). Causality between performance and traffic: an investigation with Indian ports. Maritime Policy & Management, 30(1), 5-27. 46. Demirci, E. (2003a). Simulation modelling and analysis of a port investment. Simulation, 79(2), 94-105. 47. Demirci, E. (2003b). Simulation Modelling and Analysis of a Port Investment, Simulation: Transactions of the Society for Modeling and Simulation, 79(2), 94-105. 48. Devarajan, S. and Rodrik, D. (1991). Pro-competitive effects of trade reform: results from a CGE model of Cameroon. European Economic Review, 35(5), 1157-1184. 49. Dionne, G., Desjardins, D., Laberge, N. C. and Maag, U. (1995). Medical conditions, risk exposure, and truck drivers' accidents: An analysis with count data regression models, Accident Analysis & Prevention, 27(3), 295-305. 50. Dodge, Y. (2003) The Oxford Dictionary of Statistical Terms, OUP for International Statistical Institute. 161 51. Dowd, T. J. and Leschine, T. M. (1990). Container terminal productivity: a perspective, Maritime Policy & Management, 17(2), 107-112. 52. Drewry Shipping Consultants. (1997). World container terminals 1997, London: Drewry Market Report. 53. Dyson, R. G. (2000). Performance measurement and data envelopment analysis – Ranking are rank, OR Insight, 13(4), 3–8. 54. Edmond, E. and Maggs, R. (1976). Container ship turnaround times at UK ports, Maritime Policy and Management, 4(1), 3-19. 55. Eeckaut, P. V., Tulkens, H. and Jamar, M.A. (1993). Cost efficiency in Belgian municipalities, The Measurement of Productive Efficiency–Techniques and Applications, 300-334. 56. Estache, A., González, M. and Trujillo, L. (2002). Efficiency gains from port reform and the potential for yardstick competition: lessons from Mexico. World Development, 30(4), 545-560. 57. Estache., A., Beatriz, T. and Trujillo, L.(2004). Sources of efficiency gains in port reform: a DEA decomposition of a Malmquist TFP index for Mexico, Utilities policy 12, no. 4: 221-230. 58. Fararoui, F. (1989). Queuing theory and berthing optimization, PhD Thesis Priority Berthing in Congested Ports. 59. Farrell, M. J. (1957). The measurement of productive efficiency, Journal of the Royal Statistical Society, Series A (General), 253-290. 162 60. Farrell, S. (2002). The Regulation of Privatised Ports: What Are We Trying To Achieve, In 10th Conference of the International Association of Maritime Economists, Panama City. 61. Fergus, G. (2011). Containerisation International Yearbook 2011, London,UK. 62. Fergus, G. (2012). Containerisation International Yearbook 2012, London,UK. 63. Fergus, G. (2013). Containerisation International Yearbook 2013, London,UK. 64. Fink, C., Mattoo, A. and Neagu, I. C. (2002). Trade in international maritime services: how much does policy matter? The World Bank Economic Review, 16(1), 81-108. 65. Fossey, J. (2007). Containerisation International Year Book 2007, National Magazine Company. 66. Fossey, J. (2009). Containerisation International Year Book 2009, National Magazine Company. 67. Fourgeaud, P. (2000). Measuring port performance, The World Bank. 68. Frankel, E. G. (1991). Port performance and productivity measurement, Ports & Harbors. 163 69. Gaver Jr, D. P. (1959). Imbedded Markov chain analysis of a waiting-line process in continuous time. The Annals of Mathematical Statistics, 698-720. 70. Glaser, R. E. (1980). Bathtub and related failure rate characterizations, Journal of the American Statistical Association, 75(371), 667-672. 71. González, M. and Trujillo, L. (2008). Reforms and infrastructure efficiency in Spain’s container ports, Transportation Research Part A: Policy and Practice, 42(1), 243-257. 72. Goodchild, A. and Daganzo, C. (2004). Reducing ship turn-around time using double-cycling. 73. Gordon, J. R., Lee, P. M. and Lucas Jr, H. C. (2005). A resource-based view of competitive advantage at the Port of Singapore, The Journal of Strategic Information Systems, 14(1), 69-86. 74. Greene, W. (1993). The econometric approach to efficiency analysis, In: H. Fried, C. Lovell & S. 75. Greene W. (2012). LIMDEP Version 10.0, Econometric Software Inc., Plainview, NY. 76. Haddad, E. A., Hewings, G. J., Perobelli, F. S. and Santos, R. C. (2010). Regional effects of port infrastructure: a spatial CGE application to Brazil, International Regional Science Review. 164 77. Hamedi, M., Haghani, A. and Yang, S. (2007). A Genetic Algorithm for Stowage Planning Problem Considering Crane Utilization, Paper presented at the Transportation Research Board 86th Annual Meeting. 78. Hamilton, G. L., Rasmussen, D., Zeng, X. and Arkansas, L. R. (2000). Rural Inland Waterways Economic Impact Kit Users Guide: University of Arkansas, Mack-Blackwell National Rural Transportation Study Center. 79. Harahap, W., Santoso, I. and Suryadi, K. (2005). Selection of private participation model in seaport terminal operation case: port of Tanjung Priok Jakarta Indonesia, In Proceedings of the International Symposium on the Analytic Hierarchy Process. 80. Harold, O. (1993), The measurement of productive efficiency: Techniques and applications, New York: Oxford University Press, pp. 68–119. 81. Hensher, D. A. and Mannering, F. L. (1994). Hazard‐based duration models and their application to transport analysis, Transport Reviews, 14(1), 63-82. 82. Hofmann, J. (2001). Latin American ports: results and determinants of private sector participation, International Journal of Maritime Economics, 3(2), 221-241. 83. Hummels, D. (2001). Time as a Trade Barrier. GTAP Woking Paper, Purdue University, Department of Agricultural Economics, Centre for Global Trade Analysis, West Lafayette. 165 84. Hung, S.W., Lu, W.M. and Wang, T.P. (2010). Benchmarking the operating efficiency of Asia container ports, European Journal of Operational Research, 203(3), 706-713. 85. Imai, A., Sun, X., Nishimura, E. and Papadimitriou, S. (2005). Berth allocation in a container port: using a continuous location space approach, Transportation Research Part B: Methodological, 39(3), 199-221. 86. Jara-Díaz, S. R., Martínez-Budría, E., Cortés, C. E. and Basso, L. (2002). A multioutput cost function for the services of Spanish ports' infrastructure, Transportation, 29(4), 419-437. 87. Jones, B., Janssen, L. and Mannering, F. (1991). Analysis of the frequency and duration of freeway accidents in Seattle, Accident Analysis and Prevention 23. 88. John, P. M. and Gonzalez, J. J. (2008). Critical Infrastructure At U.S. West Coast Intermodal Terminals: Are Canadian or Mexican Terminals an Alternative? TRB. 89. Cullinane, K., Song, D.W. and Richard, G. (2002). A stochastic frontier model of the efficiency of major container terminals in Asia: assessing the influence of administrative and ownership structures, Transportation Research Part A: Policy and Practice, 36 (8), 743–762. 90. Khadaroo, J. and Seetanah, B. (2008). The role of transport infrastructure in international tourism development: A gravity model approach, Tourism management, 29(5), 831-840. 166 91. Kia, M., Shayan, E. and Ghotb, F. (2000). The importance of information technology in port terminal operations, International Journal of Physical Distribution & Logistics Management, 30(3/4), 331-344. 92. Kia, M., Shayan, E. and Ghotb, F. (2002). Investigation of port capacity under a new approach by computer simulation, Computers & Industrial Engineering, 42(2), 533-540. 93. Kiefer, N. M. (1988). Economic duration data and hazard functions, Journal of economic literature, 26(2), 646-679. 94. Kim, E. and Hewings, G. J. D. (2003). An application of integrated transport networkmultiregional CGE model II: Calibration of network effects of highway, Discussion Paper, REAL 03-T-24, Regional Economic Applications Laboratory, University of Illinois at Urbana-Champaign. 95. Kim, K. H. and Bae, J. W. (2004). A look-ahead dispatching method for automated guided vehicles in automated port container terminals, Transportation Science, 38(2), 224-234. 96. Kim, K. H. and Kim, H. B. (2002). The optimal sizing of the storage space and handling facilities for import containers, Transportation Research Part B: Methodological, 36(9), 821-835. 97. Kim, M. and Sachish, A. (1986). The structure of production, technical change and productivity in a port, The Journal of Industrial Economics, 209-223. 167 98. Kozan, E. (1997). Comparison of analytical and simulation planning models of seaport container terminals, Transportation Planning and Technology, 20(3), 235-248. 99. Lai, K. and Leung, J. (2000). Analysis of gate house operations in a container terminal, International journal of modelling & simulation, 20(1), 89-94. 100. Langen, P. D. (2002). Clustering and performance: the case of maritime clustering in The Netherlands. Maritime Policy & Management, 29(3), 209-221. 101. Lawrence, D. and Richards, A. (2004). Distributing the gains from waterfront productivity improvements. Economic Record, 80(s1), S43-S52. 102. Lee, T.W., Park, N.K. and Lee, D.W. (2003). A simulation study for the logistics planning of a container terminal in view of SCM, Maritime Policy & Management, 30(3), 243-254. 103. Le Havre Port. (2000). Impact socio-economique du port 2000, Online, available at www.havre-port.net/pahweb.html 104. Levinson, Marc. (2013). How the Shipping Container Made the World Smaller and the World Economy Bigger, Princeton University Press. 105. Liu, C.I., Jula, H., Vukadinovic, K. and Ioannou, P. (2004). Automated guided vehicle system for two container yard layouts, Transportation Research Part C: Emerging Technologies, 12(5), 349-368. 106. Liu, C. I., Jula, H., Vukadinovic, K. and Ioannou, P. A. (2000). Comparing different technologies for containers movement in marine container 168 terminals, Paper presented at the Intelligent Transportation Systems, 2000. Proceedings. 2000 IEEE. 107. Liu, Z. (1995). The comparative performance of public and private enterprises: the case of British ports, Journal of Transport Economics and Policy, 263-274. 108. Lloyd, M. I. U. and Fossey, J. (2001). Containerisation International Yearbook 2001. London: Lloyd's MIU. 109. Lloyd, M. I. U. and Fossey, J. (2003). Containerisation International Yearbook 2003. London: Lloyd's MIU. 110. Lloyd, M. I. U. and Fossey, J. (2005). Containerisation International Yearbook 2005. London: Lloyd's MIU. 111. Lovell, C. K. and Eeckaut, P. V. (1993). Frontier tales: DEA and FDH: Springer. 112. Maher, M. J. and Summersgill, I. (1996). A comprehensive methodology for the fitting of predictive accident models, Accident Analysis & Prevention, 28(3), 281-296. 113. Mannering, F., Murakami, E. and Kim, S.G. (1994). Temporal stability of travelers' activity choice and home-stay duration: Some empirical evidence, Transportation, 21(4), 371-392. 114. Mannering, F. L. (1993). Male/female driver characteristics and accident risk: some new evidence, Accident Analysis & Prevention, 25(1), 77-84. 169 115. MARAD. (1978). Economic Impact of the U.S. Port Industry: An Input- Output Analysis, Port Authority of New York/New Jersey Planning and Development Department, August 1978, vol.I. 116. MARAD. (2003). Public port finance survey for the financial year 2001, Washington, DC: US Maritime Administration. 117. Marchand, M. G., Pestieau, P. and Tulkens, H. (1984). The performance of public enterprises: concepts and measurement (Vol. 33): Elsevier Science Ltd. 118. Marine Traffic Services. (2012).Available on line, www.marinetraffic.com/ais/cn/, cited from 2012 August to December and 2013 October to December. 119. Martinez-Budria, E., Diaz-Armas, R., Navarro-Ibanez, M. and Ravelo- Mesa, T. (1999). A study of the efficiency of Spanish port authorities using data envelopment analysis, International Journal of Transport Economics, 26(2). 120. Meeusen, W. and Van den Broeck, J. (1977). Efficiency estimation from Cobb-Douglas production functions with composed error, International Economic Review, 435-444. 121. Micco, A. and Pérez, N. (2001). Maritime transport costs and port efficiency, Paper presented at the Seminar" Towards Competitiveness: The Institutional Path". Santiago, Chile: IDB. 170 122. Min, H. and Park, B.I. (2005). Evaluating the inter-temporal efficiency trends of international container terminals using data envelopment analysis, International Journal of Integrated Supply Management, 1(3), 258-277. 123. Moloney, R. and Sjostrom, W. (2000). The economic value of the Port of Cork to Ireland in 1999: An input–output study, Report to the Irish government, Cork: National University of Ireland. 124. Montgomery, D. C. and Runger, G. C. (2010). Applied statistics and probability for engineers: John Wiley & Sons. 125. Nam, D. and Mannering, F. (2000). An exploratory hazard-based analysis of highway incident duration, Transportation Research Part A: Policy and Practice, 34(2), 85-102. 126. Langen, P.W. (2004). Shipping and Ports in the Twenty-first Century, Routledge, London. 127. Panayides, P. M., Maxoulis, C. N., Wang, T. F. and Ng, K. Y. A. (2009). A critical analysis of DEA applications to seaport economic efficiency measurement, Transport Reviews, 29(2), 183-206. 128. Park, R.K. and De, P. (2004). An alternative approach to efficiency measurement of seaports, Maritime Economics & Logistics, 6(1), 53-69. 129. Paselk, T. A. and Mannering, F. L. (1994). Use of duration models for predicting vehicular delay at a US/Canadian border crossing, Transportation, 21(3), 249-270. 171 130. Peter de Langen, M. N. and Martijn, H. (2007). New Indicators to Measure Port Performance, Journal of Maritime Research, IV(No.1), 23-36. 131. Poitras, G., Tongzon, J. and Li, H. (1996). Measuring port efficiency: an application of data envelopment analysis, Singapur, Department of Economics and Statistics, National University of Singapore, inédito. 132. Quddus, M. A. (2008). Time series count data models: an empirical application to traffic accidents, Accident Analysis & Prevention, 40(5), 17321741. 133. Ramani, K. (1996). An interactive simulation model for the logistics planning of container operations in seaports, Simulation, 66(5), 291-300. 134. Rios, L. R. and Maçada, A. C. G. (2006). Analysing the relative efficiency of container terminals of Mercosur using DEA, Maritime Economics & Logistics, 8(4), 331-346. 135. Rodrigue, J.P., Comtois, C. and Slack, B. (1997). Transportation and spatial cycles: evidence from maritime systems, Journal of Transport Geography, 5(2), 87-98. 136. Roll, Y. and Hayuth, Y. (1993). Port performance comparison applying data envelopment analysis (DEA), Maritime Policy and Management, 20(2), 153161. 137. Sánchez, R. J., Hoffmann, J., Micco, A., Pizzolitto, G. V., Sgut, M. and Wilmsmeier, G. (2003). Port efficiency and international trade: port efficiency as 172 a determinant of maritime transport costs, Maritime Economics & Logistics, 5(2), 199-218. 138. Sachish, A. (1996). Productivity functions as a managerial tool in Israeli ports. Maritime Policy and Management, 23(4), 341-369. 139. Talley, W. K. (1994). Performance indicators and port performance evaluation, Logistics and Transportation Review, 30(4). 140. Thanassoulis, E. (2001). Introduction to the theory and application of data envelopment analysis: Springer. 141. Tinbergen, J. (1962). Sharing the World Economy: Suggestions for an International Economic Policy, New York: Twentieth Century Fund. 142. Tiwari, P. and Itoh, H. (2001). A computable general equilibrium analysis of efficiency improvements at Japanese ports, Review of Urban & Regional Development Studies, 13(3), 187-206. 143. Tongzon, J. L. (1995). Determinants of port performance and efficiency. Transportation Research Part A: Policy and Practice, 29(3), 245-252. 144. Tongzon, J. (2001). Efficiency measurement of selected Australian and other international ports using data envelopment analysis, Transportation Research Part A: Policy and Practice, 35(2), 107-122. 145. Tongzon, J. and Heng, W. (2005). Port privatization, efficiency and competitiveness: Some empirical evidence from container ports (terminals), Transportation Research Part A: Policy and Practice, 39(5), 405-424. 173 146. Trujillo, L. and Jara-Díaz, S. (2003). Production and cost functions and their application to the port sector: a literature survey: World Bank Publications, Vol. 3123. 147. Tulkens, H. (1993). On FDH Efficiency Analysis: Some Methodological Issues and Applications to Retail Banking, Courts and Urban Transit, Journal of Productivity Analysis 4, 183–210. 148. Turner, H., Windle, R. and Dresner, M. (2004). North American containerport productivity: 1984–1997, Transportation Research Part E: Logistics and Transportation Review, 40(4), 339-356. 149. UNCTAD. (1976). Port Performance Indicators, Paper presented at the United nations Conference on Trade and Development New York, US. 150. Upton, G. and Cook, I. (2008). Oxford Dictionary of Statistics. 151. Valentine, V. and Gray, R. (2001). The measurement of port efficiency using data envelopment analysis, Paper presented at the Proceedings of the 9th world conference on transport research. 152. Wang, J. J. (1998). A container load center with a developing hinterland: a case study of Hong Kong, Journal of Transport Geography, 6(3), 187-201. 153. Wang, T.F. and Cullinane, K. (2006). The efficiency of European container terminals and implications for supply chain management, Maritime Economics & Logistics, 8(1), 82-99. 174 154. Wang, T.F., Cullinane, K. and Song, D.W. (2003). Container port production efficiency: a comparative study of DEA and FDH approaches, Journal of the Eastern Asia Society for Transportation Studies, 5, 698-701. 155. Wang, T.F., Song, D.W. and Cullinane, K. (2002). The applicability of data envelopment analysis to efficiency measurement of container ports, Paper presented at the Proceedings of the international association of maritime economists conference. 156. Wanke, P. F. (2013). Physical infrastructure and shipment consolidation efficiency drivers in Brazilian ports: A two-stage network-DEA approach, Transport Policy, 29, 145-153. 157. Wanke, P. F., Barbastefano, R. G. and Hijjar, M. F. (2011). Determinants of efficiency at major Brazilian port terminals, Transport Reviews, 31(5), 653-677. 158. Washington, S. P., Karlaftis, M. G. and Mannering, F. L. (2010). Statistical and econometric methods for transportation data analysis: CRC press. 159. Waters, R. C. (1977). Port economic impact studies: practice and assessment, Transportation Journal, 14-18. 160. Weisberg H.F. (1992). Central Tendency and Variability, Sage University Paper Series on Quantitative Applications in the Social Sciences, P2-3. 161. Wilson, J. S., Mann, C. L. and Otsuki, T. (2003). Trade facilitation and economic development: A new approach to quantifying the impact, The World Bank Economic Review, 17(3), 367-389. 175 162. Woo, Y. P. and John, S. W. (2000). Cutting through Red Tape: New Directions for APEC’s Trade Facilitation Agenda, Vancouver: Asia Pacific Foundation of Canada. 163. Yochum, G. R. and Agarwal, V. B. (1987). Economic impact of a port on a regional economy: note. Growth and Change, 18(3), 74-87. 164. Yun, L., Ji, C., Li, Q. and Li, D. (2011). Comprehensive efficiency measurement of port logistics—study based on DEA two-stage relative evaluation, Journal of System and Management Sciences, 1(4), 1-18. 176 APPENDIX I Port Infrastructure Dataset (2001 to 2011) Port Information in 2001 Port Name Singapore Shanghai Hong Kong Busan Dubai Guangzhou Rotterdam Qingdao Hamburg Kaohsiung Antwerp Tianjin Los Angeles Long Beach Bremen New York Laem Chabang Xiamen Dalian Tokyo Jawaharlal Nehru Colombo Valencia Yokohama Gioia Tauro Algeciras Felixstowe Manila Lianyungang Nagoya Barcelona Kobe Vancouver BC Le Havre Yantai Osaka Oakland Zeebrugge Balboa Melbourne Keelung St Petersburg Kingston Tacoma Houston Buenos Aires Genoa Southampton Seattle Incheon Freeport Montreal Constantza Charleston Haifa Karachi La Spezia Taichung Fuzhou Duisburg Brisbane 2001 Throughput (TEU) 15,520,000 6,340,000 17,900,000 8,072,814 3,501,820 1,730,000 6,102,000 2,640,000 4,688,669 7,540,000 4,218,176 2,010,000 5,183,520 4,462,971 2,896,381 3,316,275 2,312,439 1,290,000 1,210,000 2,535,841 1,573,677 1,726,605 1,506,805 2,303,780 2,488,332 2,151,770 2,800,000 2,296,151 502,300 1,872,272 1,411,054 2,010,343 1,146,577 1,525,000 290,000 1,502,989 1,643,585 875,926 448,565 1,276,476 1,815,854 360,899 888,941 1,320,274 983,451 1,011,748 1,526,526 1,160,976 1,315,109 610,000 860,000 989,427 206,449 1,528,034 839,000 455,000 974,646 1,100,000 418,000 340,000 460,844 No. of Berth 37 3 19 15 14 6 20 8 29 21 30 8 16 25 15 33 6 3 7 11 5 11 7 22 2 16 15 82 2 14 14 33 19 17 1 13 19 27 3 16 14 8 13 8 4 17 20 4 16 7 4 15 4 13 10 5 12 8 6 4 7 Berth Length (m) 5265 2281 6059 4547 3786 1299 10605 200 7993 5727 12901 2450 5862 6992 3840 7098 2000 970 918 3764 1280 2616 3482 5690 3155 4109 3756 6705 540 3755 4506 9355 4258 4750 180 3835 4415 7887 891 3524 3192 1556 2540 2053 1220 4504 5341 1350 4361 13597 914 3570 888 3103 1960 600 1297 2437 1050 930 1650 Berth Depth (m) 13.16 10.80 13.92 14.06 12.45 12.00 12.24 13.25 13.76 12.72 12.19 13.57 13.78 13.01 12.45 11.70 14.00 12.75 12.83 13.22 12.75 11.48 12.40 13.31 13.67 11.50 9.09 8.28 11.00 12.19 10.77 12.78 14.36 12.83 9.20 12.15 12.26 12.71 11.67 10.33 12.00 10.89 10.00 14.96 12.19 9.88 19.31 13.90 13.78 12.50 15.50 10.23 9.23 12.30 10.76 10.10 13.50 14.00 10.73 10.00 12.67 177 Terminal Area (m2) 3556000 882108 2259720 2784891 1898860 264000 5356100 1134200 4223859 2060532 4703342 1604400 2959576 1948300 2295858 5863612 196000 641080 590000 1165494 714000 343430 1916000 1823325 950000 556000 2291670 2242107 175000 1219200 1038000 2268940 1397312 1787000 31000 933393 1736065 2249100 181000 1547453 368000 252500 791983 1786584 784110 1306200 1357877 772375 1821814 385000 37000 800051 197900 3673000 110000 159220 289000 414200 388000 177000 560470 Storage Capacity (TEU) 65614 60800 175148 155445 101876 8750 44055 96200 123554 76286 99437 22100 51974 86700 53500 64150 4200 15000 31626 56265 55960 44200 50080 67668 24000 9425 66968 46900 7000 53537 790 113537 36100 36000 4000 31391 32570 10900 5950 15899 9457 13800 21650 3510 21500 38755 26051 12100 52150 7200 15000 39000 8700 26800 16800 43567 20500 26352 30000 21046 16971 No. of Quayside Gantries 118 15 66 40 25 8 159 14 57 24 62 10 44 38 40 42 3 10 11 25 14 11 18 40 14 18 27 17 3 24 37 52 13 22 2 20 24 17 3 15 25 14 10 19 1 15 22 12 25 5 7 14 10 20 20 3 1 1 7 3 9 No. of Yard Cranes and Tractors 385 151 916 848 70 189 1798 102 747 243 553 263 219 222 261 992 14 138 246 408 317 433 83 276 94 210 1032 926 68 90 196 667 506 805 130 215 353 869 61 211 72 82 154 284 213 281 212 159 409 1010 25 325 35 241 16 74 114 345 21 10 96 Port Information in 2003 Port Name Singapore Shanghai Hong Kong Busan Dubai Guangzhou Rotterdam Qingdao Hamburg Kaohsiung Antwerp Tianjin Los Angeles Long Beach Bremen New York Laem Chabang Xiamen Dalian Tokyo Jawaharlal Nehru Colombo Valencia Yokohama Gioia Tauro Algeciras Felixstowe Manila Lianyungang Nagoya Barcelona Kobe Vancouver BC Le Havre Yantai Osaka Oakland Zeebrugge Balboa Melbourne Keelung St Petersburg Kingston Tacoma Houston Buenos Aires Genoa Southampton Seattle Incheon Freeport Montreal Constantza Charleston Haifa Karachi La Spezia Taichung Fuzhou Duisburg Brisbane 2003 Throughput (TEU) 18,100,000 11,280,000 20,449,000 10,407,809 5,151,958 2,761,700 7,106,779 4,239,000 6,138,000 8,840,000 5,445,436 3,015,000 7,178,940 4,658,124 3,189,853 4,067,812 3,181,050 2,331,000 1,670,000 3,313,647 2,268,989 1,959,354 1,992,903 2,504,628 3,148,662 2,515,908 2,500,000 2,552,187 502,300 2,073,995 1,652,366 2,045,714 1,539,058 1,977,000 290,000 1,609,631 1,923,136 1,012,674 448,565 1,721,067 2,000,707 649,812 1,137,798 1,738,068 1,243,866 590,677 1,605,946 1,377,775 1,486,465 821,071 1,057,879 1,108,837 206,449 1,690,846 1,069,000 455,000 1,006,641 1,246,027 590,000 500,000 639,570 No. of Berth 37 10 23 62 14 6 18 8 30 19 44 8 21 7 16 42 12 3 7 11 5 11 7 22 2 16 15 79 2 12 14 31 17 20 3 14 25 28 3 17 14 9 16 8 6 15 17 4 15 7 4 16 5 13 10 5 12 6 6 4 7 Berth Length (m) 5280 2281 7259 11040 3786 1299 11520 3367 8223 5122 10014 2450 7388 2118 4040 8569 3600 1110 918 3686 1280 2976 4019 5440 3155 3804 3756 8278 540 3355 4506 8755 4031 6075 753 4085 7134 7235 891 3724 3192 1696 2911 2057 1525 4619 5141 1350 4055 13597 914 3570 1126 3102 2518 600 1297 1800 1050 930 1610 Berth Depth (m) 13.16 10.80 14.33 13.82 12.45 12.00 11.54 13.18 13.87 12.56 14.35 13.57 13.85 13.40 12.58 11.98 14.00 12.75 12.83 13.22 12.50 11.83 11.86 13.08 13.67 10.75 9.09 7.92 11.00 12.19 10.77 13.03 14.08 13.06 11.60 12.36 12.94 11.95 11.67 10.37 12.00 10.65 10.24 15.28 12.19 9.85 20.00 13.90 14.14 12.50 15.50 10.23 9.72 13.00 10.75 10.10 13.50 14.00 10.73 10.00 13.00 178 Terminal Area (m2) 3556000 849034 2567316 3325495 1949260 264000 5063100 1484200 4453859 1493906 5075170 1604400 3296800 1236780 2525000 6292100 723450 718200 590000 1057355 714000 439730 1026500 1779204 1300000 556000 2291670 2240307 175000 1131240 1038000 2096700 1673712 2087000 471000 1166193 3252958 2796000 181000 1547453 368000 336825 1107870 1790775 784110 1296700 1361877 797375 2116814 385000 37000 800051 255280 3369200 110000 159220 401000 993402 388000 134000 538530 Storage Capacity (TEU) 65614 60000 202050 180466 102176 8750 59197 146200 129554 69511 159876 22100 98574 36487 61880 56536 42373 15000 31626 50965 55960 52200 50080 87717 55038 12718 66968 55290 7000 50174 790 105309 30100 46500 20632 37142 60770 5400 5950 15899 9457 19000 21388 6860 23400 36226 26051 25100 47150 7200 15000 39000 11629 27011 16800 43567 20500 55106 30000 21046 22737 No. of Quayside Gantries 118 20 87 67 28 8 67 22 63 25 74 10 56 18 32 53 19 12 11 25 15 26 22 40 18 17 28 17 3 23 39 52 15 28 6 24 38 14 3 16 25 16 14 20 1 15 22 11 25 5 7 14 8 21 21 3 11 13 7 4 9 No. of Yard Cranes and Tractors 385 201 990 1015 76 189 1482 147 793 327 554 263 306 157 244 1150 415 154 246 407 340 561 153 350 132 254 998 811 68 91 196 617 531 823 140 218 465 825 61 235 72 83 197 273 217 324 206 89 366 1010 25 325 53 242 16 74 120 622 21 10 96 Port Information in 2005 Port Name Singapore Shanghai Hong Kong Busan Dubai Guangzhou Rotterdam Qingdao Hamburg Kaohsiung Antwerp Tianjin Los Angeles Long Beach Bremen New York Laem Chabang Xiamen Dalian Tokyo Jawaharlal Nehru Colombo Valencia Yokohama Gioia Tauro Algeciras Felixstowe Manila Lianyungang Nagoya Barcelona Kobe Vancouver BC Le Havre Yantai Osaka Oakland Zeebrugge Balboa Melbourne Keelung St Petersburg Kingston Tacoma Houston Buenos Aires Genoa Southampton Seattle Incheon Freeport Montreal Constantza Charleston Haifa Karachi La Spezia Taichung Fuzhou Duisburg Brisbane 2005 Throughput (TEU) 23,192,200 18,084,000 22,427,000 11,843,151 7,619,222 4,685,000 9,300,000 6,307,000 8,087,545 9,471,056 6,482,061 4,801,000 7,484,624 6,709,818 3,735,574 4,792,922 3,765,967 3,342,300 2,655,000 3,593,071 2,666,703 2,455,297 2,409,821 2,873,277 3,160,981 3,179,614 2,700,000 2,665,015 1,005,300 2,491,198 2,071,481 2,262,066 1,767,379 2,118,509 550,462 1,802,309 2,273,990 1,407,933 663,762 1,862,993 2,091,458 1,119,346 1,670,820 2,066,447 1,582,081 1,075,173 1,624,964 1,375,000 2,087,929 1,153,465 1,211,500 1,254,560 771,126 1,986,586 1,122,580 850,000 1,024,455 1,228,915 1,177,200 712,000 766,275 No. of Berth 63 25 26 65 13 8 17 13 35 22 38 8 47 35 15 44 22 5 9 14 8 12 10 23 2 18 15 79 3 12 14 34 14 17 3 15 25 28 5 17 14 12 16 13 5 16 23 4 15 5 3 17 9 13 10 8 12 6 6 6 6 Berth Length (m) 10835 5892 7999 12090 1350 3119 11420 5100 9248 6711 14414 2450 9002 6736 4040 9037 8600 1490 1759 4016 1725 3176 5088 5830 3155 4264 4026 8382 830 3355 5571 9595 4019 5475 1073 4435 7134 7488 1511 3794 3192 2313 3277 3478 1525 5793 9219 1350 4055 4076 381 3565 1966 3102 2518 1200 1899 1800 1050 1230 1427 Berth Depth (m) 12.67 11.34 14.50 13.96 13.39 13.25 12.21 15.00 13.33 13.60 14.54 13.73 13.85 13.98 12.58 12.25 14.78 12.10 13.00 13.22 11.67 11.50 11.27 13.20 13.67 10.75 9.09 7.92 10.07 12.19 11.04 13.02 15.43 12.93 14.00 12.60 12.94 11.81 12.75 10.37 12.00 10.69 10.24 14.96 12.19 9.88 16.46 13.90 14.14 12.50 9.10 9.82 12.35 13.00 10.75 11.40 14.17 14.00 10.67 10.00 14.00 179 Terminal Area (m2) 3589438 5215999 2841316 4135725 1826150 2084000 4997100 1484200 5718859 1471481 7641815 1604400 6194200 4849780 2525000 6292100 3550080 483200 1534000 1184555 903400 494730 1603750 2158738 1300000 625184 2291670 2046406 175000 1131240 1433745 1965174 1673712 2157000 476500 1341193 3258665 2913400 182000 1621453 368000 486825 1107870 2594641 784110 1336700 4645411 797375 2116814 515000 14000 840215 480280 3369200 500000 295220 351000 978402 671000 277196 339000 Storage Capacity (TEU) 65614 192084 260548 259813 102176 94646 54697 145072 150550 72335 200127 22100 98574 138015 61880 60073 52660 15000 31626 55465 62550 63320 67459 110243 60000 12902 66968 55290 7000 52218 13071 93157 35832 52500 20632 55725 60770 12450 5950 19272 9457 30000 21388 28860 23400 363155 27811 25100 47150 7200 250 40000 22229 27011 16800 48367 13000 55106 30000 29046 15976 No. of Quayside Gantries 131 63 95 76 48 27 65 43 71 21 96 10 59 46 32 71 56 8 18 27 15 26 22 42 18 21 31 17 4 23 38 52 17 30 8 26 38 8 9 16 25 27 14 25 1 17 16 11 25 8 1 14 13 22 21 6 13 13 7 8 10 No. of Yard Cranes and Tractors 428 496 919 1067 1177 311 1636 196 852 279 771 306 333 237 248 1156 480 61 246 441 296 597 201 375 237 261 998 797 68 83 255 459 539 823 139 210 528 847 69 270 72 103 198 184 217 344 266 88 366 940 2 369 158 242 16 176 98 594 21 19 77 Port Information in 2007 Port Name Singapore Shanghai Hong Kong Busan Dubai Guangzhou Rotterdam Qingdao Hamburg Kaohsiung Antwerp Tianjin Los Angeles Long Beach Bremen New York Laem Chabang Xiamen Dalian Tokyo Jawaharlal Nehru Colombo Valencia Yokohama Gioia Tauro Algeciras Felixstowe Manila Lianyungang Nagoya Barcelona Kobe Vancouver BC Le Havre Yantai Osaka Oakland Zeebrugge Balboa Melbourne Keelung St Petersburg Kingston Tacoma Houston Buenos Aires Genoa Southampton Seattle Incheon Freeport Montreal Constantza Charleston Haifa Karachi La Spezia Taichung Fuzhou Duisburg Brisbane 2007 Throughput (TEU) 27,932,000 26,150,000 23,998,449 13,270,000 10,653,026 9,200,000 10,790,604 9,462,000 9,900,000 10,256,829 8,175,952 7,103,000 8,355,039 7,312,465 4,892,239 5,299,105 4,641,914 4,627,000 4,574,192 4,123,920 4,059,843 3,381,693 3,042,665 3,428,112 3,445,337 3,414,345 3,300,000 2,869,447 2,001,000 2,896,221 2,610,099 2,472,808 2,307,289 2,600,000 2,214,631 2,309,820 2,387,911 2,020,723 1,833,778 2,206,567 2,215,484 1,697,720 2,016,792 1,924,934 1,768,687 1,710,896 1,855,026 1,900,000 1,973,504 1,663,800 1,634,000 1,363,021 1,411,414 1,750,000 1,148,628 1,219,724 1,187,040 1,250,000 1,177,200 901,000 1,000,066 No. of Berth 72 32 75 55 11 19 27 14 35 22 39 12 28 35 18 36 30 5 15 16 8 11 10 23 2 19 15 69 2 13 14 24 14 20 7 15 25 28 5 16 14 11 12 10 12 16 23 5 10 8 3 25 9 13 10 8 12 6 8 6 9 Berth Length (m) 18346 9142 10999 12610 1350 5219 10505 5356 9248 6714 12010 3472 9278 7902 4470 7615 10300 1490 3536 4669 1725 2954 5028 5680 3155 4944 4026 7252 540 3755 5571 6985 4019 6075 1883 4435 6869 8490 1511 3574 3192 2203 4129 2959 3525 5793 9646 1500 3423 14772 381 4305 1966 3102 2518 1200 1899 1800 1658 1700 2124 Berth Depth (m) 12.81 12.24 14.70 13.75 12.52 12.60 11.99 14.50 13.33 13.75 14.39 14.30 13.57 14.49 13.00 12.89 14.60 12.10 13.90 14.00 11.33 11.89 11.27 13.43 13.67 11.17 9.09 7.82 10.20 12.61 11.04 13.31 15.43 13.06 14.00 12.60 12.82 13.65 12.75 10.30 12.00 10.98 11.16 15.40 12.19 9.88 16.50 12.30 15.00 12.50 9.10 10.32 12.52 13.00 10.75 11.23 14.17 14.00 12.00 10.00 14.00 180 Terminal Area (m2) 4360000 8569837 3519061 4752455 1826150 4774000 5635900 1484200 5708859 1471481 6835601 1382400 6510336 4903227 2961000 6585100 4399280 483200 2078579 1755276 903400 494730 2173750 2142140 1300000 805184 1803670 2094814 175000 1364879 1370345 1783522 1635342 2157000 843600 1341780 3101641 3249900 182000 1584453 368000 668500 1307870 2442642 1145410 1336700 4680411 894375 2030000 515000 14000 1053100 526600 3369200 500000 369000 451000 980402 1328000 401000 830090 Storage Capacity (TEU) 33438 350084 258934 296741 102176 295068 61000 145072 155550 72335 197627 78350 104006 142515 86500 51923 121047 15000 36528 278362 52500 63320 87569 108729 60000 18302 66968 54111 7000 62994 13071 82655 35832 52500 54972 55725 57570 36351 5950 813700 9457 37200 88355 397260 36084 363155 39656 35600 42150 7200 250 45500 37629 27011 16800 48367 20500 55106 60000 31046 21840 No. of Quayside Gantries 159 113 133 80 51 37 87 45 79 21 65 37 67 60 47 65 85 8 42 35 16 26 26 38 18 23 31 17 4 27 37 39 19 30 16 26 38 15 9 18 25 11 17 24 16 17 30 11 25 8 1 18 13 25 21 8 13 13 12 9 13 No. of Yard Cranes and Tractors 41 992 1041 1431 1206 370 1577 208 906 279 556 311 352 237 325 1107 748 61 345 412 611 536 475 311 287 274 998 800 58 45 230 603 685 823 88 204 507 881 69 253 72 139 301 143 216 345 416 101 421 940 2 411 212 258 19 211 119 578 34 23 158 Port Information in 2009 Port Name Singapore Shanghai Hong Kong Busan Dubai Guangzhou Rotterdam Qingdao Hamburg Kaohsiung Antwerp Tianjin Los Angeles Long Beach Bremen New York Laem Chabang Xiamen Dalian Tokyo Jawaharlal Nehru Colombo Valencia Yokohama Gioia Tauro Algeciras Felixstowe Manila Lianyungang Nagoya Barcelona Kobe Vancouver BC Le Havre Yantai Osaka Oakland Zeebrugge Balboa Melbourne Keelung St Petersburg Kingston Tacoma Houston Buenos Aires Genoa Southampton Seattle Incheon Freeport Montreal Constantza Charleston Haifa Karachi La Spezia Taichung Fuzhou Duisburg Brisbane 2009 Throughput (TEU) 25,866,600 25,002,000 21,040,096 11,954,861 11,124,082 11,190,000 9,743,290 10,280,000 7,007,704 8,581,273 7,309,639 8,700,000 6,748,994 5,067,597 4,578,642 4,561,528 4,537,833 4,680,355 4,552,000 3,810,769 4,061,343 3,464,297 3,653,890 2,555,000 2,857,438 3,043,268 3,100,000 2,815,004 3,020,800 2,112,738 1,800,214 2,247,024 2,492,107 2,240,714 2,342,262 1,843,067 2,045,211 2,328,198 2,011,778 2,236,633 1,577,824 1,341,850 1,692,811 1,545,853 1,797,198 1,412,462 1,533,627 1,400,000 1,584,596 1,559,425 1,702,000 1,247,690 607,483 1,181,353 1,140,000 1,307,000 1,046,063 1,193,943 1,177,000 1,006,000 918,998 No. of Berth 72 32 73 55 56 19 31 14 35 22 40 12 29 34 18 36 29 9 15 13 10 12 6 20 2 19 15 66 2 14 13 23 14 24 7 15 20 29 5 16 15 11 12 10 14 16 27 5 10 8 3 16 9 13 10 8 12 6 8 7 9 Berth Length (m) 18346 9142 11109 12610 13820 5219 16125 5449 9248 6714 13120 3472 9369 7902 4470 7615 9800 2483 3536 4114 2437 3154 3983 5150 3155 4944 4026 7252 540 4105 5091 6635 3974 7005 1883 4435 6194 8235 1511 3574 3516 2203 4129 2959 3525 4908 9720 1500 3423 14772 1033 4305 1966 3102 2518 1209.7 1899 1800 1658 2050 2124 Berth Depth (m) 12.81 12.24 12.75 14.00 14.08 12.60 12.31 14.70 13.33 13.75 14.81 14.30 13.59 14.49 13.00 12.89 14.44 12.66 13.90 14.31 11.88 11.55 12.43 13.54 13.67 11.06 9.09 7.82 10.20 13.19 10.54 13.26 15.39 13.00 14.00 12.60 14.24 12.95 12.75 10.30 12.30 10.98 11.16 15.40 12.19 9.81 12.23 12.30 15.00 12.50 15.50 10.32 12.52 13.00 10.75 12.00 14.17 14.00 12.00 10.00 14.36 181 Terminal Area (m2) 4593000 8569837 3518871 4516720 1826150 4720000 6950900 1671000 5708859 1471481 7600273 1382400 6510336 4903227 3051500 6585100 4399280 483200 2078579 1727299 1423400 494730 1638550 2140873 1300000 805184 1803670 2088126 175000 1511475 1105820 1783522 1635312 2712000 945600 1341193 3001023 3079900 182000 1607453 508000 1014500 1307870 2442642 1340620 1341350 4340293 894375 2030000 515000 490000 1053100 526600 3369200 500000 453000 451000 980402 1328000 438500 830090 Storage Capacity (TEU) 34038 350084 308935 296105 102176 295068 81000 145072 155550 72335 197527 78350 104006 142515 90500 51923 121047 15000 36528 278362 97500 63320 88126 107350 60000 18302 66968 54111 7000 72922 12384 82655 46832 62900 55925 55725 38928 17700 5950 813700 1209800 37200 88355 397260 36084 369605 42656 35600 42150 7200 28327 45500 39129 27011 16800 58567 20500 55106 60000 31046 21840 No. of Quayside Gantries 159 113 129 80 51 37 112 45 79 21 69 37 67 60 54 66 75 10 50 35 24 26 32 38 18 23 31 17 4 54 32 39 19 38 16 26 34 15 9 17 29 15 19 24 16 17 54 11 26 8 1 18 13 25 21 10 13 13 12 10 13 No. of Yard Cranes and Tractors 41 992 1004 1420 1206 370 2086 208 906 279 626 304 479 237 406 1142 767 66 343 412 827 536 396 284 297 264 998 819 68 220 200 603 685 854 143 183 309 866 69 245 81 139 302 186 243 341 411 101 420 940 77 411 213 258 19 247 119 578 34 24 166 Port Information in 2011 Port Name Singapore Shanghai Hong Kong Busan Dubai Guangzhou Rotterdam Qingdao Hamburg Kaohsiung Antwerp Tianjin Los Angeles Long Beach Bremen New York Laem Chabang Xiamen Dalian Tokyo Jawaharlal Nehru Colombo Valencia Yokohama Gioia Tauro Algeciras Felixstowe Manila Lianyungang Nagoya Barcelona Kobe Vancouver BC Le Havre Yantai Osaka Oakland Zeebrugge Balboa Melbourne Keelung St Petersburg Kingston Tacoma Houston Buenos Aires Genoa Southampton Seattle Incheon Freeport Montreal Constantza Charleston Haifa Karachi La Spezia Taichung Fuzhou Duisburg Brisbane 2011 Throughput (TEU) 29,937,700 31,739,000 24,384,000 16,163,842 12,617,595 14,260,400 11,876,920 13,020,100 9,014,165 9,636,289 8,664,243 11,587,600 7,940,511 6,061,091 5,915,487 5,503,485 5,731,063 6,454,200 6,400,300 4,416,119 4,307,622 3,651,963 4,327,371 2,992,517 2,264,798 3,608,301 3,248,592 3,342,200 3,870,000 2,471,821 2,033,747 2,725,304 2,507,032 2,215,262 2,342,262 2,172,797 2,342,504 2,207,257 3,232,265 2,467,967 1,749,388 2,365,174 1,724,928 1,485,617 1,866,450 1,851,701 1,847,648 1,324,581 2,033,535 1,924,644 1,116,272 1,362,975 556,694 1,381,352 1,238,000 1,545,434 1,307,274 1,383,578 1,318,958 1,181,000 1,004,983 No. of Berth 82 32 73 58 56 20 29 14 34 19 35 16 29 23 17 36 30 9 17 16 10 12 7 21 2 20 12 69 2 12 13 26 18 26 7 22 24 28 5 13 30 14 17 10 13 13 23 5 12 6 3 16 8 13 13 7 11 6 8 7 10 Berth Length (m) 21896 9142 11409 14610 13820 5370 16125 5449 9148 5898 15130 4674 9381 7323 5259 7615 10300 2483 4253 4669 3749 3154 4793 5390 3155 6036 4062 7252 540 3670 5878 7275 5504 7065 2013 4785 6804 8485 1511 2995 3920 2927 3954 2959 3220 4908 5813 1500 4231 2335 1036 4305 1722 3102 3468 1563 1698 1800 1658 2050 2497 Berth Depth (m) 12.94 12.24 13.09 13.62 12.88 12.50 12.29 14.70 13.66 13.63 15.05 14.74 13.20 14.48 13.30 12.89 14.60 12.66 14.26 14.00 12.00 11.55 13.57 13.24 13.67 12.96 11.61 7.82 10.20 12.80 11.38 13.39 15.46 13.10 14.00 12.79 13.05 12.99 12.75 10.48 12.67 11.25 11.10 15.40 12.19 13.00 12.22 12.57 15.05 12.50 16.00 10.32 13.58 13.00 11.94 11.90 13.50 14.00 12.00 10.00 14.30 182 Terminal Area (m2) 6233000 9441323 3518871 4617786 3536905 4689600 6978400 1671000 6103550 1460376 7555573 1949400 6510368 4466227 4589000 5810100 4399900 483200 2078579 1603246 2355000 742730 1853750 2140873 1600000 1167459 1586350 1198234 175000 1405549 1085320 1912293 1835312 3038000 945600 2241860 3250719 3134900 182000 1317954 409900 2634000 1607870 2442642 1340620 1341350 1611298 894375 2294016 515000 477428 1053100 615000 2704200 710000 499324 451000 965402 1328000 435500 1055200 Storage Capacity (TEU) 35600 350084 308935 345263 346196 283784 1253000 145072 132725 72051 2788326 107362 104006 127432 88046 48323 119737 15000 36528 202362 118266 67256 104989 107350 75000 43272 109000 55136 7000 77430 12084 80959 46822 61300 52925 55725 58102 12100 5950 813700 1209800 68750 88350 398820 36084 375385 68156 35600 1032150 7200 28327 45500 20477 27011 16800 69999 28500 55106 60000 31046 30868 No. of Quayside Gantries 204 113 125 80 78 70 124 45 86 21 72 33 69 55 108 70 75 10 42 41 26 26 36 38 25 35 38 17 4 27 32 44 28 40 22 41 37 15 9 22 29 42 19 26 16 18 31 11 30 8 20 19 8 24 30 13 10 13 12 10 15 No. of Yard Cranes and Tractors 43 980 977 1657 1508 380 1943 208 1041 259 594 279 567 238 299 1093 790 66 353 408 846 586 558 299 278 323 497 926 68 276 193 622 700 810 147 206 426 878 69 250 88 307 222 186 243 363 461 116 469 940 84 432 207 181 31 380 103 534 34 24 250 APPENDIX II Ship Turn-around Time Dataset (2012 to 2013) Part Ship Information in 2012 Vessel Name WIND JOLLY CRISTALLO HORAI BRIDGE ANAKENA OCEAN LOHAS WLADYSLAM ORKAN NONA AN NING JIANG BLUE LEAF KONG QUE SONG AN LONG JIANG CHENG SHAN WEI ZHONG CHANG 28 DORIC VALOUR SC LOTTA QUAN CHENG MARCLIFF TAI HANG3 JIN HAI YU PACIFIC ENDEAVOR ZHE HAI 362 MOSKVA JING HAI BAO HONG 10 STADT ROSTOCK JIN SHA LING ST UNION BAO AN J STAR ZHONG CHANG 28 GUANG PING MAGNA STX KNIGHT KING FORTUNE JU JIN YUAN DE XIN HAI ELPIS BAO MAY DA YANG BAI LI MAXIMUS TAI HANG 1 DONG HAN ZHONG CHANG68 LING GANG 9 IRON LINDREW LI DIAN 5 LI DIAN 6 ROXANNE D ROXANNE D XIAO JIANG TAI CANG HE KRASZEWSKI ZIM CHICAGO BAHAMIAN EXPRESS SANUKI DANU BHUM DANU BHUM ACX MARGHERITE MITRABHUM MOL EMINENCE MOSEL TRADER NYK MARIA OOCL SHANGHAI Ship Turn-around Time (min) 237 962 1005 231 1781 2419 1005 1779 287 2579 4617 1516 348 278 2626 221 501 275 291 2602 200 288 263 188 1782 1778 1657 227 303 193 194 590 310 2577 228 242 300 198 387 2705 297 1064 208 2527 247 2 2 7 182 220 1119 5 1833 1313 453 1495 1515 1408 1314 1637 1196 962 2324 Gross Tonnage (in 100 ft3) 4860 56000 17211 28148 9340 24167 27915 11505 5308 20609 11495 6980 12539 32351 88367 6577 9610 38520 26922 24139 22382 2360 3995 5217 27971 16725 1500 6524 5556 12539 5275 5403 10549 17101 23525 40892 2238 91385 7092 93196 38639 1446 18121 37663 43158 32879 32879 32450 32450 4713 4879 24221 91558 21018 13448 9675 9675 18602 9917 54940 28048 27051 66289 183 Ship Length (m) 111 240 172 185 131 199 215 149 112 180 149 112 160 190 288 128 143 225 190 184 180 108 101 115 222 169 75 126 98 160 117 115 150 188 194 225 80 290 110 292 225 76 195 225 229 190 190 189 189 116 113 200 334 180 159 146 146 244 148 294 215 210 277 Ship Height (m) 18 38 28 32 20 26 30 23 19 27 23 19 22 32 45 19 23 32 30 30 29 14 16 18 30 27 12 19 18 22 20 19 22 23 28 32 14 45 20 45 32 12 23 32 32 32 32 32 32 16 19 28 43 28 25 22 22 42 23 32 30 30 40 Arrived Port Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Vessel Name PDZ MASYHUR CMA CGM KAILAS PAC BANDA PAC BANDA PAC BANDA DANU BHUM APL SOKHNA SEASPAN NINGBO OOCL BEIJING HANJIN SAO PAULO OSSIAN SINAR SUMBA SINAR SUMBA KMTC SINGAPORE MCP LARNACA SINAR BIMA THANA BHUM CAPE FULMAR KOTA DELIMA SINAR BITUNG HYUNDAI BRIDGE ANAN BHUM KAPITAN MASLOV WANHAI 502 SINAR BANTEN NORTHERN PRACTISE SINAR BIAK KOTA WISATA NILEDUTCH LUANDA NILEDUTCH SHANGHAI CARLA RICKMERS APL BANGKOK CARLA RICKMERS KOTA DAMAI APL SPINEL DALIAN EXPRESS HANSA CASTELLA IKARUGA PAOLA OOCL CANADA NILEDUTCH SHANGHAI HANJIN SAO PAULO SINAR BITUNG PAC BANDA COSCO TAICANG MOL EMINENCE KOTA DAMAI PDZ MASYHUR NANTA BHUM NYK ALTAIR MOL EMISSARY MOL DISTINCTION AUGUSTE SCHULTE APL AUSTRIA HYUNDAI PROGRESS ANL WARRINGA SINAR BANDUNG NYK ARGUS SANUKI CAP BLANCHE SINAR SANGIR NYK VEGA MOL DIRECTION YOSSA BHUM SINAR BROMO LILAC Ship Turn-around Time (min) 464 209 762 532 905 1503 1620 4 1470 1935 1029 848 788 897 607 1322 1129 981 673 1132 796 1201 1193 554 1644 596 1606 1834 698 2 2 1853 2 2 1579 2485 2 2 1244 1 1 1843 3 2 2 1963 2 2 1378 2 3 5 5 2 2 2 1347 2 892 1 1 1447 3 2 5 2 Gross Tonnage (in 100 ft3) 5914 21971 3085 3085 3085 9675 35975 39941 91563 16472 12514 18321 18321 16659 5315 9957 21932 15995 6245 13596 21611 9675 16575 42579 12598 47855 15184 17125 34642 25630 14278 35991 14278 6245 53519 88493 16915 18619 9931 91563 25630 16472 13596 3085 115933 54940 6245 4007 11079 105644 54940 42110 27093 71867 21611 39906 12584 75484 13448 28372 17515 97825 42110 11788 12545 28927 184 Ship Length (m) 115 196 90 90 90 146 231 260 335 172 158 174 174 171 117 148 197 170 116 162 182 145 184 269 147 264 167 182 221 207 159 231 159 116 293 320 169 193 148 335 207 172 162 90 349 294 116 108 146 333 294 260 210 295 182 260 147 300 159 222 172 338 260 145 147 222 Ship Height (m) 21 28 20 20 20 22 32 32 43 27 24 27 27 27 20 24 28 25 21 26 35 23 25 32 25 32 27 28 32 30 25 32 25 21 32 42 27 28 23 43 30 27 26 20 46 32 21 16 25 46 32 32 30 40 30 32 25 40 25 28 28 46 32 25 25 30 Arrived Port Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Vessel Name MOL DISTINCTION WADI ALRAYAN WILHELME MOL DIRECTION OOCL GUANGZHOU MERKUR BRIDGE SINAR BANTEN HANJIN NORFOLK SINAR BROMO EURO MAX MOHEGAN MCP AMSTERDAM THORSTREAM KING ALFRED MOL CREATION RACHA BHUM JARU BHUM KOTA HASIL YM PINE WAN HAI601 BUTTERFLY CSAV ROMERAL APL CHONGQING ZIM DALIAN XIN QIN HUANG DAO OOCL CHICAGO OOCL AUSTRALIA MOL DISTINCTION APL IRELAND APL PUSAN ALEX MAERSK MOL GRANDEUR CSCL URANUS VIRA BHUM KOTA PURI CMA CGM MARGRIT HANJIN MANZANILLO YM PINE BARENTS STRAIT APL KENNEDY WARNOW MASTER WAN HAI 502 COSCO YANTIAN MOL EMPIRE MOL DEVOTION EVER LAMBENT BIEN DONG STAR TIM S KATHARINA SCHEPERS MOL EMPIRE KUO CHIA ZIM LIVORNO MAERSK LA PAZ AMBASSADOR BRIDGE NILEDUTCH GUANGZHOU APL ZEEBRUGGE OOCL LUXEMBOURG WES JANINE OOCL CHARLESTON SARA PONA CAPE FAWLEY HANSA COBURG PONA YM INTERACTION HANJIN SHENZHEN Ship Turn-around Time (min) 5 655 978 2 1069 1 5 1137 5 1103 2 1164 824 2 2255 1046 3 2 39 27 771 817 1960 21 25 24 1089 917 943 30 897 927 30 789 257 25 560 31 1461 33 406 25 986 24 29 890 1872 28 1 916 897 23 1559 597 41 20 912 636 1077 588 47 29 1289 48 1811 36 Gross Tonnage (in 100 ft3) 42110 34083 14844 42110 40168 9597 12598 40487 12545 32284 6158 5316 16803 28007 86692 32060 8571 13272 64005 66199 111249 36097 113735 40030 66452 66677 41479 42110 66462 25305 93496 59307 135000 24955 27104 141635 27061 64005 18102 61926 17068 42579 109149 54940 39906 98882 6899 35581 10318 54940 15095 39906 88237 40839 27100 86679 89097 10350 40168 9590 27968 15995 18327 27968 16488 74962 185 Ship Length (m) 260 214 161 260 260 150 147 261 147 211 106 117 184 222 316 211 137 158 275 276 350 220 349 260 279 277 263 260 280 207 352 274 366 194 199 366 199 275 183 275 180 269 351 294 261 335 121 223 151 294 169 260 300 261 210 316 323 151 260 143 222 170 176 222 172 304 Ship Height (m) 32 32 25 32 32 22 25 32 25 32 20 20 25 30 46 32 21 25 40 40 43 30 46 35 40 40 32 32 40 30 42 40 52 32 32 48 32 40 26 39 25 32 43 32 32 46 20 32 23 32 28 32 46 32 30 46 43 23 32 23 30 25 28 30 27 40 Arrived Port Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Vessel Name NAVARINO CMA CGM TARPON XIN QUAN ZHOU WAN HAI 313 TIM S KATHARINA SCHEPERS MOL EMPIRE KUO CHIA ZIM LIVORNO MAERSK LA PAZ AMBASSADOR BRIDGE NILEDUTCH GUANGZHOU APL ZEEBRUGGE OOCL LUXEMBOURG WES JANINE OOCL CHARLESTON SARA CMA CGM AZURE HYUNDAI SPLENDOR MSC TORONTO HYUNDAI SUPPERME APL SPINEL HYUNDAI HIGHNESS YM INAUGURATION BOX VOYAGER LANTAU BREEZE NORTHERN PRACTISE HYUNDAI SPLENDOR NYK FURANO KUO WEI OOCL NINGBO WAN HAI231 WAN HAI 101 AMERICA EXPRESS CAPE FAWLEY HANJIN SHENZHEN WAN HAI 603 WAN HAI 603 NORTHERN PRACTISE SITC INCHON AN CHUN AN CHUN WINCHESTER STRAIT BONANZA EXPRESS YM INTELLIGENT OOCL TAICHUNG MOBILANA UNI CONCERT CHINA STEEL INVESTOR CAPE FRANKLIN WAN HAI 312 KING ADRIAN EVER UNICORN YM INITIATIVE YM UNISON WAN HAI 275 KUO CHANG ULYSSES SZCZECIN TRADER APL TEXAS YM MANDATE OOCL ZHOUSHAN SPAARNE TRADER OOCL NINGBO CAPE FAWLEY LAST TYCOON UNI CONCERT Ship Turn-around Time (min) 1165 1109 945 834 28 1 916 897 23 1559 597 41 20 912 636 1077 588 1138 22 35 33 945 808 913 22 1 209 22 45 877 43 26 564 1099 29 36 733 33 32 7 927 927 868 1729 628 2361 4762 705 4163 402 915 519 693 632 1246 895 411 593 367 1670 1326 1039 672 766 454 3165 853 Gross Tonnage (in 100 ft3) 91354 54309 41482 27800 35581 10318 54940 15095 39906 88237 40839 27100 86679 89097 10350 40168 9590 39906 94511 89954 52581 53519 64054 16488 36087 9610 47855 94511 44925 15095 89097 17751 9834 42894 15995 74962 66199 66199 47855 13267 10383 10383 18358 6352 16488 16705 29688 12405 82112 15995 27800 27915 69246 16488 90389 16776 15095 27061 16803 75582 73675 41479 17068 89097 15995 22549 12405 186 Ship Length (m) 335 294 263 213 223 151 294 169 260 300 261 210 316 323 151 260 143 260 340 325 294 293 274 172 228 142 264 340 267 168 323 191 144 269 170 304 277 277 264 162 151 151 175 111 172 183 195 153 289 170 213 215 285 173 335 172 168 211 184 304 299 263 180 323 170 180 153 Ship Height (m) 43 32 32 32 32 23 32 28 32 46 32 30 46 43 23 32 23 32 46 43 32 32 40 27 32 23 32 46 36 27 43 28 23 32 25 40 40 40 32 25 24 24 27 19 27 28 32 25 45 25 32 29 40 28 43 27 27 30 24 40 40 32 25 43 25 29 25 Arrived Port Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Vessel Name YM KEELUNG CSAV LANALHUE WAN HAI 301 ITAL ONORE XIN LIANG YM PLUM MEDCORAL EVER DECENT OOCL TEXAS TIAN LI 26 YM GREEN IAL 001 DERYOUNG SPRING TY LOTUS EVER ALLY OOCL TIANJIN HENG SHUN HAI MAPLE LEAF 25 DONG TENG ZENITH BUSAN VULKAN WARNOW BOATSWAIN CHENG LU 15 YM UBIQUITY AN CHUN EVER UNITY TRANSFORMER OL XIN FU XING FAR EAST GRACE APL ENGLAND APOLLO LYNUX DONG FANG FU STADT LAUENBURG BERLIN EXPRESS UNI ARDENT FEN JIN 2 KUO CHIA LAUREL ISLAND HANJIN OSAKA EVER CHIVALRY MARE THRACIUM OCEAN MATE MAERSK DRURY HYUNDAI OAKLAND MASOVIA SAGITTARIUS ACX SATSUMA ACX SATSUMA CSCL CHIWAN APL PUSAN YM INTERACTION YUSHO ANGEL II XIN YANG ZHOU YM HAWK UNI ARDENT UNI ARDENT HYUNDAI NEW YORK MASOVIA TIMBER TRADER XI TONG CHENG 602 EVER ALLY OCEAN MELODY MAERSK DANBURY APL ILLINOIS GLORY DILIGENCE YM INTELLIGENT BUXMELODY Ship Turn-around Time (min) 711 423 657 634 550 805 548 1087 859 3650 857 1466 3669 2197 1001 706 3271 1142 2070 650 416 622 1007 777 997 558 1863 385 796 898 3154 590 387 1091 1727 1448 442 939 676 1260 1178 1998 661 1142 574 737 291 200 587 758 671 2350 515 985 1650 1301 731 487 3692 351 1042 2141 444 950 725 625 666 Gross Tonnage (in 100 ft3) 40030 40541 26681 32968 3580 64254 17068 52090 40168 4049 64254 11810 6278 2216 14807 89097 2930 6249 1413 4713 16800 17068 8461 90532 10383 69246 17018 2772 6253 65792 6290 13199 9610 88493 14807 2561 15095 16980 51754 90449 29383 22009 53481 71786 17285 16803 6773 6773 39941 25305 16488 1512 41482 15167 14807 14807 71786 17825 5542 2981 14807 17979 54271 75582 5394 16488 28050 187 Ship Length (m) 260 260 200 212 93 274 186 294 260 115 275 146 100 82 165 323 97 125 79 102 184 180 140 305 151 285 169 101 125 276 100 162 142 321 163 93 169 170 289 335 195 188 294 293 175 185 123 123 260 207 172 74 263 169 163 163 293 175 98 89 165 171 295 304 97 172 216 Ship Height (m) 32 32 32 32 16 40 25 32 32 17 40 25 20 14 27 43 16 20 12 17 25 25 20 56 24 20 27 31 20 40 19 26 23 42 27 14 28 26 32 42 32 28 32 40 26 26 28 18 32 30 27 12 33 27 27 27 40 26 18 21 27 27 32 40 19 27 30 Arrived Port Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Vessel Name SHUNHENG TS CHINA HO MAO LIANG GU DERYOUNG SHINESTONE WAN HAI 162 SALLY MAERSK THORSWAVE SINOP PROSRICH WANLI 8 HOUSTON BRIDGE OOCL TAICHUNG UNI ACCORD CAPE FAWLEY WAN HAI 275 FPMC CONTAINER 9 OOCL ZHOUSHAN STX TOKYO CMA CGM CALLISTO STAR EVVIVA KATRINA COSCO NEW YORK SONG YUN HE HANSA CENTURION PUFFIN ARROW PELICANA PANDORA BBC QUEBEC TIAN REN CMA CGM CALLISTO MINERAL SINES SPRUCE ARROW OCEAN PIONEER SEA LAND LIGHTNING TIAN REN CMA CGM CALLISTO MINERAL SINES DA XIN HUA YAN TAI CMA CGM VIVALDI COSCO NEW YORK ANTIGONI PAN HE INES TRIDENT XIANGTONG9 OMSKIY115 TAKEKO SAVANNAH EXPRESS FLECHA ZHONG WAI YUN QUAN ZHOU HENG YU XIANG LIAN CMA CGM VIVALDI OSG ALPHA MAERSK WARSAW NORTHERN VIVACITY XIN HAI HONG LIAO YUAN11 BLUE WAVE JUNG GANG 5 APL KOREA NEW GOLDEN BRIDGE V BEI LUN 1 COSCO EXCELLENCE LOWLANDS ERICA Ship Turn-around Time (min) 249 409 2189 547 3023 556 660 1280 2781 335 1422 1093 1985 1847 477 825 540 1126 21 16 3 20 16 6 6 7 23 19 25 15 9 3 9 15 15 9 8 3 19 20 3 17 24 1 6 7 6 28 4 5 1 Gross Tonnage (in 100 ft3) 1720 17515 14599 3580 6283 13246 91560 29022 35760 7589 2612 96801 16705 14807 15995 16776 9909 41479 8306 131332 24479 7170 54778 16737 16915 36925 39258 6701 9611 2358 131332 87495 32458 29970 49985 2358 131332 87495 34231 91038 54778 9610 9951 9602 9549 2797 2463 8957 94483 87440 18487 Ship Length (m) 81 172 157 93 100 159 347 196 200 127 86 335 183 165 170 172 147 263 143 363 179 133 294 184 167 200 213 133 138 82 363 289 190 190 292 82 363 289 216 331 294 143 148 149 145 95 108 138 332 289 190 Ship Height (m) 13 18 26 16 20 25 42 32 32 20 14 46 28 27 25 27 25 32 21 46 30 20 32 28 28 32 32 19 21 14 46 45 30 32 32 14 46 45 32 43 32 23 22 22 22 14 14 22 44 45 28 Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao 9 8 3 1 22 18 23 38 22 14 3 19 7 13 14 7589 4966 91038 7167 18123 27437 4090 2838 47984 1081 64502 29554 24111 141823 89603 127 122 331 134 172 222 107 98 229 66 277 196 185 366 289 21 18 43 20 27 30 17 14 37 12 40 30 29 48 45 Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao 188 Arrived Port Vessel Name OCEAN LOHAS ANNA LISA VEGA DOLOMIT COSCO KIKU YU JIN MOL GARLAND STX SINGAPORE SINOTRANS NAGOYA DELIA OCEAN PIONEER EVER UNICORN HARRIER EVER REWARD FPMC CONTAINER 8 PACIFIC ENDURANCE GOOD FAITH SKY QUEEN LOWLANDS ERICA SITC TIANJIN CHANG RONG HC RUBINA SAFMARINE MULANJE FRIESEDIJK GUO SHUN XIN JIA HE SALAMANCA TIAN XIU HE PACIFIC BLESS YM HAWK XIN YANG PU EQUINOX DAWN BUSAN STAR ASIAN JOY SUN STAR GONG YIN 1 FU WEN SHAN BING HE XIN ZHENG ZHOU OOCL JAKARTA CMA CGM BAUDELAIRE HE DE SUNNY PINE YUAN DA NO.9 FENG KANG SHAN JIN MA TIAN JIN HE HUAXIANG MOON BRIGHT SW OUTSAILING 8 KING DIAMOND KARMEN PENG FA STAR APEX COSCO OCEANIA DARYA JAMUNA CS CHAMP KOTA NAGA HONG CHANG HARRIER EASLINE TIANJIN CHATTANOOGA JOSCO TAIZHOU MOKPO STAR MOKPO STAR FAVOR SAILING FAVOR SAILING FAVOR SAILING Ship Turn-around Time (min) 18 21 3 17 31 5 3 20 12 21 16 17 20 28 6 7 11 3 8 14 14 14 21 16 3064 3187 1347 5887 674 876 3136 2766 1126 2059 1080 5238 663 1813 1084 1053 1018 862 3415 4248 2322 1053 1026 5145 27 4042 1323 1276 827 1484 5886 3122 558 35 1081 1316 714 6215 4 4 7 62 45 Gross Tonnage (in 100 ft3) 9340 7464 7209 8917 2974 59307 16731 9443 9983 29970 69246 9971 53103 9954 92752 41101 1955 89603 9531 1457 8821 50686 9983 1998 6084 26084 54005 32300 15167 42000 30049 33308 5578 4137 32488 13823 23542 47815 40168 73172 19987 3986 2497 13367 2921 54005 1983 21718 2926 15350 37209 24550 9522 115776 22000 32987 20902 1433 9971 10649 9744 30988 45026 45026 1479 1479 1479 189 Ship Length (m) 131 129 133 138 92 274 185 140 140 190 285 148 294 148 292 225 89 289 145 75 126 292 140 88 119 183 294 190 169 263 190 190 104 97 190 162 201 255 260 300 178 107 91 156 92 294 79 176 84 152 236 190 143 349 186 190 180 75 148 156 143 190 229 229 74 74 74 Ship Height (m) 20 20 19 22 14 40 27 23 23 32 40 23 32 23 45 32 14 45 22 12 20 32 22 14 18 31 32 32 27 32 32 32 17 28 32 23 28 37 32 40 28 18 15 24 15 32 14 28 15 27 32 32 22 46 28 32 28 12 23 23 23 32 32 32 12 12 12 Arrived Port Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Vessel Name FAVOR SAILING K.OPAL SINO 3 TENG YUN HE SINOKOR QINGDAO INTREPID INTREPID UNI PACIFIC HANJIN SEMARANG YONGDINGHE YM INTERACTION PONA WAN HAI 231 CHATTANOOGA MOL DAWN HIMAWARI NO.2 EVER REWARD NYK DANIELLA SHIN OH MARU CAMELLIA MARU SUNFLOWER HAKATA OHRYU MARU YM SKY WAN HAI 221 MOL CELEBRATION HALCYON IKUTA SHINSEN MARU JIN PING OOCL CHARLESTON KARIYUSHI ITAL UNIVERSO GLORY FORTUNE NYK MARIA TAI CANG HE PANCON GLORY SHIN ZUI MARU KOUN MARU NO.18 TAIKO MARU OGASAWARA MARU HUTUOHE KIYOHAMA MARU IKUTA TAISHO MARU HANJIN MONTEVIDEO MOSEL TRADER LANTAU BAY SHINSEN MARU NAKAHARU MARU SHIN YU MARU SHIN OH MARU J.PIONEER CAPE FRIO NYK LODESTAR BAY BRIDGE SUMISE MARU NO.2 CAMELLIA MARU KATSUEI MARU NO.18 KATSUEI MARU NO.19 SHIN ZUI MARU LANTAU BAY BERMUDIAN EXPRESS APL PUSAN NYK TERRA MEDAEGEAN KYORIKI MING ZHOU 77 Ship Turn-around Time (min) 37 5520 47 860 843 36 49 779 1046 762 906 731 696 1159 635 545 304 560 336 248 988 373 675 745 2064 627 211 599 30 644 2123 679 670 579 281 557 450 2 335 1215 872 206 211 1954 532 961 1117 524 399 4 8 229 1045 620 1314 444 328 560 395 426 679 699 721 724 693 2389 2052 Gross Tonnage (in 100 ft3) 1479 31532 17061 20569 9030 30046 30046 17887 9592 9471 16488 27968 17751 9744 42894 7323 53103 27051 11790 3837 10507 5195 17153 16911 86692 9971 749 13089 1997 40168 9943 68888 9928 27051 4879 11000 13097 497 5389 6700 9471 499 749 3215 40542 28048 9610 13089 4944 499 11790 4879 14308 75201 44234 5468 3837 997 997 13097 9610 16850 25305 76928 9946 498 8282 190 Ship Length (m) 74 190 177 180 136 190 190 182 145 145 172 222 191 143 269 161 294 210 161 103 167 115 172 162 316 148 96 160 80 260 154 285 148 210 113 145 160 76 118 131 144 70 96 93 261 215 143 160 111 74 161 113 154 299 267 118 103 86 86 160 143 168 207 304 139 76 126 Ship Height (m) 12 33 26 27 25 32 32 28 23 23 27 30 28 23 32 24 32 30 24 15 27 18 28 31 46 23 14 27 14 32 23 40 23 30 19 22 26 12 18 17 22 12 14 16 32 30 23 27 18 12 24 19 24 40 34 12 15 15 15 26 23 26 30 40 23 12 23 Arrived Port Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Vessel Name SITC MOJI SHIN OH MARU SUMISE MARU NO.2 SUNFLOWER TOKYO MUSASHI MARU MUSASHI MARU HANGCHENG OOCL CHARLESTON SITC HOCHIMINH EVER ETHIC POSEN SHIN ZUI MARU SHIN ZUI MARU SHINMEI MARU XING YUAN JI XIANG SONG JOYOUS WORLD TT STAR LIAN MENG 9 CHANG MING SHENG SHANG CHENG XIN HAI 68 HC RUBINA WEN CHENG JIN SHUN HAI SKY QUEEN SKY QUEEN HAI WANG XING NEW VENTURE RI XIN CHANG XING LONG GUANGYUAN AN PING 1 DONG AN SUN JUNE RICKMERS NEW ORLEANS SPRING HOPE ASIAN INFINITY ORIENT SUNRISE LIAO YUAN 11 CHIPOLBROK GALAXY TIAN LONG XING TONG DA CHANG WANG LONG LAN HAI LI LIANG EMERALD CORAL LEO MONO DA JIA HANSA CENTURION WAN HAI 263 DONG FANG XING SITC OSAKA CSCL AFRICA CAIYUNHE APL TEXAS MIN TAI YI HAO APL ROTTERDAM CAPE NORVIEGA APL ROTTERDAM DERYOUNG SPRING FORMOSA CONTAINER NO.4 SONG HE CSCL ZEEBRUGGE BALTIC STRAIT HUA RONG SHAN MAGSENGER 11 Ship Turn-around Time (min) 649 326 592 956 457 953 2440 585 1184 563 566 426 278 337 4102 258 6256 2653 2932 1665 1379 6015 3075 4121 119 1590 33 5 1373 2814 680 566 275 300 20 458 415 516 238 7 4610 1590 4023 371 587 40 901 49 619 542 1500 384 812 383 844 644 911 573 911 35 1564 Gross Tonnage (in 100 ft3) 9734 11790 5363 10503 13927 13927 1427 40168 9744 76067 27968 13097 13097 13091 3015 20684 35879 5675 182 35874 9683 4732 8821 2423 2972 1955 1955 24964 32505 1444 19370 2470 24489 5552 4233 23119 9665 5577 17431 2838 24142 24964 1970 19370 32964 9910 24724 10625 16915 18872 4879 9000 90645 16738 75582 1495 71786 17609 71786 6278 9280 Ship Length (m) 143 161 117 167 166 166 70 260 143 298 222 160 160 160 95 180 225 122 84 224 147 96 126 92 97 89 89 187 190 76 140 89 195 99 110 193 114 104 170 98 200 187 80 141 189 128 200 153 167 198 113 141 334 184 304 74 293 182 293 100 138 Ship Height (m) 23 24 19 26 27 27 11 32 23 43 30 26 26 26 15 27 32 18 13 32 22 18 20 14 16 14 14 29 32 12 24 13 28 18 16 28 23 17 27 14 28 29 14 24 32 20 31 22 28 28 19 23 43 28 40 13 40 28 40 20 23 Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Lianyungang 2389 1843 2354 3496 3819 24438 108069 18102 26835 64110 199 337 183 195 254 28 46 26 32 43 Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang 191 Arrived Port Vessel Name CAPE CANARY YM HARMONY STX TOKYO YM HEIGHTS WIN LONG XIN HAI KOU MIN HE SKY PRIDE KATE XIN BIN CHENG SUN NEW TUO FU 3 TENG YUN HE VECTIS CASTLE OCEAN HAWK FAREAST HONESTY MIN HE LIAN XING SHEBELLE ADMIRALENGRACHT BEI LUN 6 SHABDIS SITC TIANJIN CENTRANS SUN SAMOS LEGEND PACIFIC TREASURE FORMOSA CONTAINER NO.4 GOLDEN TRADER UMEKO DINTEL TRADER SITC LIANYUNGANG TAKEKO CHENG SHAN WEI STX YOKOHAMA JOSCO TAIZHOU HANJIN QINGDAO GRAIG CARDIFF STX TOKYO WAN HAI 233 KANG FU PACIFIC ISLAND ANANGEL DESTINY TAI HE KANG PING JIN PU JI PENG KAI DA 358 DE QIN 57 XING HANG HAI 998 SHUN HANG 98 GUO LIANG 69 RUI ZHOU 69 XIN LONG 996 XIANG JING 6 LIAN HE 10 BAO MA 227 MIN JIE 586 YONG LONG 102 MING FEN ZHI YUAN SHUN 303 ZHONG XING 1 DONG SHENG 7 SHENG AN DA 19 DONG FANG FU WAN HAI 231 SHIN CHUN Ship Turn-around Time (min) 1460 1640 1781 1814 3180 2101 1428 1571 4151 1506 2068 2874 1814 5881 6667 6668 1951 6874 6964 6549 5256 7 892 3367 4925 4388 1660 Gross Tonnage (in 100 ft3) 93235 15167 8306 15167 1535 41482 37143 9520 91374 9683 18870 39666 20569 7168 22662 33042 37143 4430 20471 7949 25766 74175 9531 94710 36615 92752 9280 Ship Length (m) 292 168 143 169 79 263 235 142 292 147 177 228 180 123 186 190 235 98 178 129 186 299 145 295 225 292 138 Ship Height (m) 46 27 21 27 12 32 32 22 45 22 28 32 27 20 28 32 32 17 27 19 29 40 22 46 32 45 23 Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang 4363 2760 1866 1792 1279 3723 3312 3285 1372 4399 1371 1268 6124 5953 5855 2340 782 2125 429 6 94 12 31 17 598 29 465 1419 117 23 2213 760 25 721 20 115 11 393 885 28420 8957 6701 9734 8957 6980 8306 30988 27104 23444 8306 17751 28613 23300 87523 35963 5275 15786 1984 648 6892 614 578 624 2378 647 2321 2134 621 638 6214 1587 602 2654 5412 1598 13199 17751 9965 192 137 132 143 138 112 144 190 199 180 143 191 189 180 289 236 117 167 127 57 130 54 52 54 98 53 97 97 53 54 135 88 53 97 123 79 162 191 152 32 22 19 23 22 19 21 32 32 30 21 28 32 30 45 32 20 26 20 9 20 9 9 9 16 9 16 16 9 9 19 13 9 16 17 13 26 28 24 Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Taichung Taichung Taichung 192 Arrived Port Vessel Name UNI ARISE TS CHINA TE HO SEALORD ORIENTAL CORE MAPLE LEAF 25 HON CHUN KUO WEI KANG PING HUA HANG 3 YI YUN BLUE STAR BAO CHANG QI MEN HUA HANG 1 WAN HAI 305 TS HONGKONG NEW DYNAMIC MAENAM 1 UNI CHART AN CHUN WAN HAI 203 KEN HO TAIPOWER PROSPERITY VI XIAO JIANG WAN HAI 205 GLORY WISDOM KUO CHIA AS SAVONIA HAMMONIA ADRIATICUM WU XING 5 POS TOPAS BLUE OCEAN WAN HAI 267 JUI HO MING CHUN DERYOUNG SPRING NESRIN AKSOY NA HA SITC PYEONGTAEK YI YUN DA SHEN TENG YUN HE BO DUN 2 DONG FANG FU WAN HAI 233 ASIA CEMENT NO.2 NEW LEADER STADT LAUENBURG DS ABILITY UNI ARISE KAI PING TONG MAO 7 KM NAGOYA HUA HANG 3 KUO WEI HON CHUN BANOWATI NOTO III SEA BAILO CITADEL HERMINA STAR COMET OPDR LISBOA THARSIS MORSUM PERSEUS Ship Turn-around Time (min) 12 448 5706 5613 5445 545 366 479 375 322 371 7 4603 387 400 814 304 2144 495 371 510 951 2146 3727 342 882 778 441 518 223 2552 5854 495 542 3936 20 787 4810 1467 386 184 2062 384 2859 345 360 52 3780 258 619 818 235 2908 2849 2 489 479 1638 481 1721 17 14 18 19 25 20 34 Gross Tonnage (in 100 ft3) 14796 17515 41372 7345 12840 6249 9965 15095 5275 6392 4822 7111 50697 6650 6387 26681 15487 21059 11810 12405 10383 17117 13465 50236 4713 17134 5394 15095 16850 9581 32965 51195 9949 18872 14233 10383 6278 27011 992 9530 4822 17107 20569 2385 13199 17751 8165 37623 9610 9966 14796 5275 3770 50625 6392 15095 9965 35119 9243 17172 3990 91374 6277 7545 1801 9983 1392 193 Ship Length (m) 165 172 225 127 149 125 152 168 117 130 113 124 229 124 128 200 165 176 146 157 151 174 159 235 116 174 97 169 167 150 190 229 151 198 153 151 100 189 99 150 113 173 180 81 162 191 130 223 142 148 165 117 100 235 130 168 152 224 130 172 111 292 133 129 88 140 72 Ship Height (m) 27 28 32 19 24 20 24 27 20 18 19 21 39 21 18 32 24 28 25 25 24 27 25 38 16 27 18 28 28 22 32 38 23 28 26 24 20 30 13 22 19 27 27 14 26 28 20 32 23 24 27 20 16 38 18 27 24 32 20 26 14 45 19 20 12 20 11 Arrived Port Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Vessel Name HS SCHUBERT SUECIA SEAWAYS SOPHIA FRANCISCA LAZURITE CEG COSMOS WILSON HOOK STAS FINNLANDIA BG ROTTERDAM BG ROTTERDAM BG ROTTERDAM BG ROTTERDAM HAMMONIA GALICIA PACHUCA CMA CGM ANNE SIBUM RANTUM AURORA STARA PLANINA ANTWERP CLAVIGO KORSIKA MSC STELLA TRAVEBERG LARGONA BASLE EXPRESS ANTWERP FESCO VORONEZH GRUMANT SWE BULK TRANS DANIA COUNTESS ANNA HELGALAND HS SCHUBERT GRETE SIBUM ATHENS CAROLINA IDA LYSVIK SEAWAYS HYUNDAI GLOBAL KLENODEN AURORA MOL CHARISMA HAMMONIA GALICIA ANDROMEDA J LAPPLAND ELM K GRANDE AFRICA REINBEK Q IOANARI ANDREA GALAN EM ITHAKI MATHILED MAERSK MOL MODERN SOFIA SCHULTE APL ENGLAND YM WEST APL PHILIPPINES EVER ELITE SEA LAND INTREPID PORT HAINAN CMA CGM DOLPHIN MAERSK WINNIPEG BLOBAL MIRAI COSCO CHINA Ship Turn-around Time (min) 25 30 5 20 22 34 21 24 16 20 15 567 24 3 24 3116 544 718 1291 3533 2794 8002 1043 4367 1610 1482 4518 2794 555 1209 706 516 3263 1654 1015 743 2596 1092 3443 893 3874 1499 1340 3759 1230 1405 1203 4634 2186 2061 3225 4820 5040 1143 3866 3726 900 1622 3109 2977 3228 2913 2247 898 1657 6780 3544 Gross Tonnage (in 100 ft3) 18480 24196 7464 4015 3505 1139 2993 2765 9981 8273 8273 8273 8273 42069 6901 152991 10585 1984 9981 25327 2451 2446 2997 73819 1939 866 141700 2451 16803 15868 2480 5167 1589 7519 18480 10585 91373 6362 1616 7409 94511 3828 9981 86692 42609 8273 5056 20992 56642 16324 44625 9981 2473 25497 98268 78316 28616 65792 46697 64502 76067 49985 33036 54309 18123 32376 91649 194 Ship Length (m) 177 196 130 100 90 63 89 103 134 139 139 139 139 269 139 366 151 90 134 186 88 88 100 304 88 63 366 88 183 181 87 113 180 136 177 150 292 122 82 129 339 104 134 316 269 139 118 186 214 169 229 134 114 194 367 302 222 276 276 276 299 292 190 300 175 190 334 Ship Height (m) 25 26 21 17 14 11 15 12 22 22 22 22 22 33 20 52 22 13 22 30 12 13 13 40 11 11 48 12 26 23 13 18 20 20 25 22 45 18 11 18 46 16 22 46 33 22 18 28 32 27 32 23 13 32 42 44 30 40 32 40 42 38 32 32 28 32 43 Arrived Port Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Long Beach Long Beach Vessel Name HANJIN GDYNIA KOTA JATI CSL TRAILBLAZER ZIM BEIJING GLOBAL MIRAI HORNCAP MSC ORNELLA MSC ORNELLA MSC SOLA HANJIN BALTIMORE MAIZURU BENTEN FEDERAL SKEENA FEDERAL SKEENA BAY BRIDGE CMA CGM IVANHOE KOTA WARIS HORIZON DISCOVERY OOCL NETHERLANDS MSC RITA MOL DESTINY CSCL BRISBANE HUDSON BAY ZIM YOKOHAMA GREY SHARK CAPUCINE CAPUCINE CAPUCINE SUECIA SEAWAYS MSC SILVANA ENERGIZER FLANDRIA FLANDRIA ANDROMEDA J BRITANNIA SEAWAYS DOLFIJN MARTIN MAERSK SINGAPORE BERNHARD PRUDENCE EST ARKLOW RESOLVE PAULA C APL INDONESIA SEVERINE MARNEDIJK AASVIK SCI CHENNAI SEA KESTREL SEVEN STAR APL BALBOA GELIUS 3 TOLAGA ASPEN MUKARNAS LUKA ZIM KINGSTON CHARLOTTE BORCHARD SORMOVSKIY 121 NOVOROSSIYSK STAR REBECCA BORCHARD ZUIDERDIEP ZENA A MSC JOANNA IREM KALKAVAN ZIM ALABAMA ADA A ZIM IBERIA Ship Turn-around Time (min) 40 2692 2486 53 6780 667 69 786 5103 5089 2881 10 9 4315 4385 2901 2274 2286 3006 1685 2833 3470 1858 529 8 5 8 255 4 1065 1322 206 2327 4212 5 10 4070 2 8 5 7 5 810 5 365 9 1582 11 5208 1453 5514 3369 1192 5459 5411 3053 43 4527 4585 2074 4259 3735 1306 3303 3258 3016 2347 Gross Tonnage (in 100 ft3) 40487 18502 18241 54626 32376 12887 54304 54304 131771 82794 48034 24196 24196 44234 111249 16772 18888 66086 89954 39906 39941 17986 39906 4688 16342 16342 16342 24196 94489 7642 13073 13073 8273 24196 1987 794 93511 10318 1556 920 2999 2990 40541 16342 7545 3088 43679 1382 1939 8203 2678 6362 8289 14430 13066 40030 9962 2466 20624 7702 5638 25503 107894 10308 40542 4919 41507 195 Ship Length (m) 261 194 178 294 190 153 294 294 364 300 229 190 190 267 350 185 212 276 324 260 260 170 260 110 152 152 152 196 332 134 142 142 139 198 81 58 334 150 82 66 90 90 261 152 130 94 262 78 88 129 89 122 144 150 153 260 134 114 180 142 121 185 337 148 261 113 254 Ship Height (m) 32 28 26 32 32 22 32 32 46 42 38 28 28 34 42 28 28 40 43 32 32 27 32 19 22 22 22 26 43 22 23 23 22 26 12 10 42 23 11 11 14 14 32 22 21 15 32 11 11 23 15 18 20 26 24 32 23 13 28 20 16 30 46 22 32 16 32 Arrived Port Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach New York New York New York New York New York New York New York New York Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Haifa Haifa Haifa Haifa Haifa Haifa Haifa Haifa Haifa Haifa Haifa Haifa Haifa Haifa Haifa Haifa Haifa Haifa Haifa Vessel Name URANUS HELENA SIBUM MSC MIA SUMMER MEDONTARIO ZIM LIVORNO WARNOW PORPOISE MSC KYOTO MSC SOCOTRA MSC SARISKA TROPIC EXPRESS MSC TOKYO MSC BANU HARMEN OLDENDORFF MSC BARCELONA MSC ROSARIA MSC ALESSIA MAERSK WAKAMATSU MSC RITA MSC PILAR YUE SHAOGUANHUO 2833 HENGXIANSIGUAN8866 XIJIANG11HAO YUE HE YUAN HUO 2220 SHAOGUANHUO 0966 YUESHAOGUANHUO 2631 XIN HANG HAI 168 MINGZHU 863 GUI XIANG JIANG 2012 GUIPING NANHUO 6789 FO_HANG_868 HUI HONG 838 PING NAN YONG JIAO18 JIAN CHENG SHUI 11 YUE. WEI. HANG.003 YUE AN SHUN 618 ZENG CHI HANG 168 YUE DU CHENG HUO 8986 HE SHUN 168 YUESHAOGUANHUO0911 GUI NAN 8228 GANG SHUN 3688 GUI GUI PING HUO0809 TENGXIANXIANGQ 1208 SUI DE YANG 998 YUE HAI LONG 003 YUE_SHAOGUAN_HUN2602 JIAN_GONG_628 HUI YUE 666 YUE GANG 6 HAO HAIFUHUA 1 HAO Ship Turn-around Time (min) 1606 2294 1381 1486 225 490 1222 2110 1383 541 2032 1856 1606 2354 806 935 299 627 798 39 41 27 34 57 341 26 42 14 251 34 407 29 319 24 23 31 60 46 40 49 49 37 43 1159 3176 1323 2258 2985 2734 864 Gross Tonnage (in 100 ft3) 23722 6701 25219 15334 39906 15334 43325 60117 52181 3744 65483 35954 42033 61870 50963 75590 17280 89954 52181 874 747 687 547 765 987 874 879 825 849 1002 985 824 579 1102 1321 489 748 697 579 747 698 630 479 690 827 790 870 960 1021 987 196 Ship Length (m) 195 132 216 166 260 166 270 300 294 107 275 231 225 270 274 301 172 324 194 67 50 49 38 51 67 53 57 51 49 66 49 55 42 64 66 48 50 49 50 50 52 49 32 49 63 54 50 50 60 50 Ship Height (m) 28 19 27 25 32 26 32 38 32 21 40 32 32 40 32 40 28 43 32 13 11 11 7 12 13 9 13 12 13 15 16 13 8 13 13 10 11 13 11 10 12 12 8 12 12 12 15 16 12 16 Arrived Port Haifa Haifa Haifa Haifa Haifa Haifa Freeport Freeport Freeport Freeport Freeport Freeport Freeport Freeport Freeport Freeport Freeport Freeport Freeport Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Part Ship Information in 2013 Vessel Name DK IRENE NING AN 10 JIN HAI OU XIN HUI ZHOU MARITEC SHEN HUA 801 TONG DA HUA0028 QUEEN YAN MING HENG 1 XIN SHI JI 198 YONG HENG 1 HAO WEI LUN 198 SHENG FENG LU JI NING HUO 3239 CHANG YUE HAI DING SHENG 11 HUA LUN 9HAO HUA LUN 9HAO WANHUAI YUAN HUO CHANG TIAN HAI HUALONG98 HAI LI 5 WANSHOUXIANHUO LUZAOZHUANGHUO LUZAOZHUANGHUO JIN TAI 618 WAN SHOU XIAN CHANG JIN HAI WANG DA 269 YU HONGXIANG CHONGLUNS5001 CHONGLUNS5002 NIN LIAN HAI 618 JIN SHA YUAN 998 YI CHUN 2 NITHI BHUM APL TENNESSEE HANJIN ARGENTINA BERLIN EXPRESS X PRESS HOOGLY MOL MAXIM SINAR TANJUNG YANGON STAR CHANA BHUM OOCL SOUTHAMPTON SAVANNAH EXPRESS MOL COMPETENCE CHICAGO EXPRESS IBN AL ABBAR PAC LOMBOK PAC LOMBOK THANA BHUM TORRES STRAIT MCP ROTTERDAM NORTHERN PRECISION HAMMONIA GALLICUM LILA BHUM EURO MAX JITRA BHUM BRUNO SCHULTE SINAR BANDUNG DONGJIANG IONIC STORM MOL DEVOTION MOL EMISSARY Ship Turn-around Time (min) 5587 26392 43774 47815 20763 41500 1238 6980 5153 32351 2238 5403 7092 1500 5475 9340 4860 4660 1332 21382 1760 8524 786 988 899 13211 586 13242 4980 1020 6980 6986 6058 856 9989 9661 75582 37200 88493 8971 78316 17613 9606 9675 89097 94483 86692 93811 16705 5272 5272 21932 18123 5536 47855 29383 8443 32284 15533 40542 12584 5250 31263 39906 54940 Gross Tonnage (in 100 ft3) 210 839 281 791 759 94 722 267 304 246 1061 1139 245 58 1321 464 5 10 15 216 264 5763 1 63 6 18 1 7 318 1 13 8 1 2 214 1592 936 672 1564 1088 1001 1678 670 1516 1858 1498 1602 2350 1137 649 1018 885 1747 513 303 1156 1583 747 667 291 254 214 934 1103 966 197 Ship Length (m) 93.9 185 190 230 180 200 43 112.2 119.8 199 68 108 122 63 119 133 108 108 60 178 86 133 50 58 58 160 46 160 98 53 110 110 104 48 146 137 304.2 231 316 154.9 302 182.8 149.6 148 323 332.4 316 336.2 172 117 117 196.9 244 117 264.4 195.7 135.9 210.9 170 261.1 147 112 190 260.7 294.1 Ship Height (m) 18 32 30.5 32 29 25 8 19 16.8 32 14 18 18 12 19 20 18 18 11 27 13 22 10 12 12 24 9 24 16 10 20 20 17 10 21 25 40 32 46 21.5 43.4 28 22.6 22 42.8 43.3 45.8 43 27 19.7 19.7 27.8 42 20 32 32.3 22.3 32.3 25 32.3 25 20 32.3 32.3 32.7 Arrived Port Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Singapore Vessel Name YANTRA BHUM XETHA BHUM KOTA RAJIN NEW BLESSING FO SHAN 12 HAO DONGGANGYUN062 JIN LONG 85 SIRI BHUM OOCL CANADA OOCL CANADA GUANGBOYUN1003 FEI FAN 88 HANJIN NHAVA SHEVA YUE FU LONG 318 JIA HANG 838 SHI TAI 28 MAERSK SEOUL ZHONGHANG 918 ZHONGHANG 918 GLORY CHALLENGER GLORY CHALLENGER ZHUCHUAN 513 ZHAO HANG 828 XIN YUNTONG 288 XIN YUNTONG 288 SHI TAI 18 HAO SHI TAI 18 HAO OOCL HAMBURG MEDBOTHNIA ZHEN DONG 833 HUI HAI LONG 188 HUI HAI LONG 188 XI BOHE HAN JIN BOSTON XINHONG 69 ZHONGQILUN 128 ZHONGQILUN 138 HAI BANG DA 8 YM MASCULINITY YUN XUN 288 ZHONG HANG 911 BO YUN 588 OOCL CANADA WAN HAI 301 HANJIN MANILA MELL SATUMU YM MASCULINITY IAL 001 ZHONG HE APL LOS ANGELES ITAL ONORE YM UNITY YM INSTRUCTION DONG PENG FPMC B 107 HAN ZHI OOCL CHARLESTON SHUN HENG SITC PYEONGTAEK SAN GEORGIO EVER DEVOTE FEI HE OOCL CANADA SFL HUNTER ITAL OTTIMA WAN HAI 301 UNI CORONA Ship Turn-around Time (min) 11086 11086 9678 8203 786 867 896 9757 91563 91563 684 849 17280 765 788 1020 94483 896 896 8650 8650 689 857 980 980 1002 1002 89097 9946 751 886 886 36772 93542 640 688 688 754 76787 580 890 878 91563 26681 17225 9610 76787 11810 48311 43071 32968 90389 16488 1064 50695 5799 40168 1720 9530 20239 52090 48311 91563 28592 27779 26681 12405 Gross Tonnage (in 100 ft3) 920 988 1211 807 1614 696 2360 1471 32 967 3556 493 380 764 3241 1776 527 185 2177 38 26 2518 2919 107 180 280 1747 30 635 2644 7 10 30 839 3035 1235 2161 2025 908 467 2875 1712 967 949 920 35 1224 1016 664 271 743 1903 750 3546 3505 1103 753 464 361 509 975 2085 711 840 1101 838 1927 198 Ship Length (m) 145 145 145.9 128.9 50 55 50 143.9 335 335 49 50 188 49 49 73 332 50 50 138 138 49 50 50 50 62 62 323 139.1 47 50 50 223 300 49 49 48 50 310 49 50 50 335 200 172 142.7 310 145.7 275.1 267.2 212.8 335 173 69.7 292 107.4 260 81.6 149.7 196 294.1 265 335 207 222.1 200 152.1 Ship Height (m) 22 22 22.6 23 15 11 16 22 42.8 42.8 13 15 22 12 16 15 43.2 16 16 22 22 10 13 15 15 13 13 42.8 22.6 16 10 10 32 42.8 11 10 10 14 40 13 16 16 42.8 32.2 27.6 22.6 40 25 32.2 32.2 30.1 43 28 12 30 19.7 32.3 13 22.6 26 32.2 32 42.8 32 30 32.2 25.7 Arrived Port Singapore Singapore Singapore Singapore Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Hong Kong Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Vessel Name CORAL SW FPMC CONTAINER 9 WAN HAI 303 SPRING RETRIEVER WAN HAI 261 KEUM YANG PRIME MEDPEARL CSE CLIPPER EXPRESS QI MEN DONG GANG SHUN OOCL ASIA YM INTELLIGENT LANTAU BRIDE SITC PYEONGTAEK SANYA SONG YUN HE ELLY HONG DA XIN 29 HONG DA XIN 29 CANES CANES SHI TAI 298 SHI TAI 298 SHI TAI 298 JIANG XIN 8 JIANG XIN 8 CAPE NEMO REVERENCE WAN HAI 216 WAN HAI 216 TENKO MARU HANJIN ALGECIRAS HANJIN ALGECIRAS XIN HAI WANG XIN HAI WANG XIN HAI WANG APL CHARLESTON HAI LAN ZHONG GU 3 RONG NA 6 UNI CROWN DE XIANG THAI LAKER WAN HAI 235 BAO DONG 198 XIN YUAN 68 XIN YUAN 68 YI CHUN 2 MAERSK DELMONT MAERSK DELMONT SEAGLASS II ZHONG WAI YUN QUAN ZHOU MAERSK DELMONT VENTURE PEARL KWK PROVIDENCE CHANGHONG2128 CEDAR ARROW CEDAR ARROW OMSKIY 122 NEW HISTORY CHIPOLBROK GALAXY HEBEI XINGTAI CHANG AN 103 GLOBAL DREAM SEAGLASS II SITC SHENZHEN CHENG GONG 77 Ship Turn-around Time (min) 5471 9909 26681 4724 18872 4713 17068 16962 6650 6094 89097 16488 9610 9530 16705 16737 6701 5860 5860 4562 4562 540 540 540 10322 10322 18257 9990 17138 17138 32415 35595 35595 1248 1248 1248 105000 17688 456 12404 1882 5471 17751 1866 1989 1989 9989 50350 50350 18499 18487 Gross Tonnage (in 100 ft3) 596 364 773 2147 463 450 592 1985 331 942 1312 754 500 361 451 7 8 21 8 24 979 711 847 531 23 6 24 31 25 8 21 1 5 15 7 19 14 20 2298 16 14 16 1 20 17 8 36 1 8 4 808 Ship Length (m) 98.2 148 200 99.9 198 109 192 169.3 116 115.8 323 160 142.7 149.7 183.2 182.8 132.6 114 114 109.5 109.5 49 49 49 155 155 175.5 146 174 174 220 248 248 97 97 97 350 180 44 152.1 80 98.2 191.5 88 88 88 146 292.2 292.2 170 193 Ship Height (m) 18 23 32.2 19.2 28 17 26 27.2 22 18.2 42.8 28 22.6 22.6 27.6 27.6 19.2 16 16 16.6 16.6 13 13 13 22 22 27.4 24 28 28 28 30 30 16 16 16 50 28 9 25.7 12 18 28 13 13.2 13.2 21 32.2 32.2 27 28 Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Kaohsiung Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Qingdao Tianjin 50350 32672 88856 366 32458 32458 2463 21796 24142 52709 4049 20395 18499 9734 9810 1078 41 3446 8 34 27 1350 1787 3856 2956 1033 2441 8 799 108 292.2 212 289 30 189.9 189.9 108.4 175.5 180 294 114 153.8 170 143.2 133 32.2 32 45.1 6 31 31 15 29 30 32 16 26 27 23 20 Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin 199 Arrived Port Vessel Name YOU GUO 188 PING XIANG 16 HUA SHENG 158 EDWIN YM HARMONY CMA CGM NERVAL SITC SHIMIZU HAI EN 311 JIN HUA 8 JIANGYUJI 3938 TIAN SHENG HE XIN HAI XIN XINGLONGZHOU 518 GUO MAO 3 SERDOLIK REN JIAN 7 LIBAIDA 16 KAI HANG XING 3 JIN FENG CHENG HUA XIANG 999 HUA XIANG 1000 SHANGDIAN XIANG AN YUAN SHUN HANG AN QUAN ZHOU 88 ZHOU GONG 6005 TIAN CHENG 19 XIN HONG XIANG 57 XIN HONG XIANG 58 APL CAIRO MATHIS STX TOKYO RYUKAKU KOUYUMARU KOTOHIRA MARU NO.8 TAISEI MARU NO.37 MAIKO MAIKO MAIKO DAISHINMARU NO.5 DAISHINMARU NO.5 WAKA MARU HIMAWARI 3 NICHTOKUMARU NO.10 WAIHAI 317 OGASAWARA MARU KUROSHIO MARU GREEN COSTA RICA CSCL YOKOHAMA EVER POWER APL PUSAN WAN HAI 275 MOL EMISSARY SHENG JIE 1 STX YOKOHAMA STX YOKOHAMA SALVIA MARU SALVIA MARU SALVIA MARU CAMELLIA MARU CAMELLIA MARU OHYU MARU FLUVIAL JIN TAI 7 JIN LONG 28 JIN LONG 28 XINGGUANG2 ANJI 1 Ship Turn-around Time (min) 1002 8480 5899 91792 15167 72884 9744 899 5220 1022 54005 6250 3336 3340 3505 17156 2450 5021 3002 8461 8461 50082 2584 1899 879 19784 9687 9687 25305 9983 8306 894 786 842 842 1228 1228 1228 989 989 446 8668 1002 32642 6700 4824 7743 9850 17887 25305 16776 54940 4696 8306 8306 4992 4992 4992 3837 3837 9841 478 6225 2002 2002 1221 4002 Gross Tonnage (in 100 ft3) 2676 1107 606 2716 647 1235 817 3459 2800 1072 1013 1085 361 587 2001 2890 9 881 390 1240 1057 1123 881 409 1049 2213 972 1101 749 611 685 1220 3548 390 1332 481 175 98 467 534 204 49 544 1367 2637 937 1425 757 708 738 630 702 4 177 654 204 177 186 256 195 651 114 3742 1077 2062 1213 217 200 Ship Length (m) 73 149 126 291.8 168.8 300 147 77 100 70 295 130 98 99 89 186.1 88 128 95 140 140 200 98 88 68 188 140 140 207.4 140.7 143.3 74 77 73 73 85 85 85 78 78 50 133 84 213 131 113 131.3 145 181.8 207.4 184 294.1 101 143.3 143.3 120.5 120.5 120.5 102.9 102.9 167.7 38 126 96 96 89 100 Ship Height (m) 13 21 18 45.1 27 40 22 16 16 13 30 19 16 16 18 27.6 13 18 15 20 20 32 16 13 16 29 20 20 29.9 23.2 20.5 12 11 15 14 14 14 14 17 17 13 22 14 32 17.2 17.8 19.6 22.5 28 29.8 25 32.7 19 20.5 20.5 15.2 15.2 15.2 15 15 24 5 20 14 14 13 16 Arrived Port Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tianjin Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Dalian Dalian Dalian Dalian Dalian Dalian Vessel Name ANJI 1 XIAN TAN NEW VENTURE NJ XIN CHENG LONG HUA FU 108 JIN HAI SHUN 1 ZHOU SHENG BAOTONGHAI 3 XIAN TONG FU HAI 6 JIAN GONG 9 HENGSHENG688 ZHONG HONG 1 HAI YUN SHENG 808 JIN LONG 28 JIN LONG 28 JIN LONG 28 JIN ZHOU 26 YU LIN AN JI 4 JIN LONG 28 JIN LONG 28 JIN LONG 28 AN JI 2 TONGXING36 JINJIANGHE LONG SHUN HUI YE 2 JIN CHUAN 9 DONG FANG YONG HENG ZHONG HE LI HUA LI HUA YI YUN IDEAL BULKER TAI SHENG 2 TAI SHENG 2 GUAN HUA WAN SHUN WAN SHUN XIN HENG 97 LONG LUN 103 LONG LUN 103 LONG LUN 103 LONG LUN 103 XIN FU TAI 6 AO TONG 1 AO TONG 1 AO TONG 1 ZHONG HONG 8 HAI RUN 589 HAI RUN 589 HAI RUN 589 HAI RUN 589 HAI RUI 16 HAI RUN 668 HAI RUN 668 HAI RUN 668 TIAN XIANG 26 HAI RUN 967 HAI RUN 967 HAI RUN 987 HAI RUN 987 HAI RUN 987 HAI RUN 987 HAI RUN 987 HAI RUN 987 Ship Turn-around Time (min) 4002 432 21932 2112 2456 9675 4007 985 654 1121 8571 2121 2006 987 2115 2115 2115 2654 2325 687 1846 1846 1846 988 789 12210 6971 1321 1101 4368 48311 15189 15189 4822 16721 1486 1486 32642 688 688 868 1002 1002 1002 1002 6158 6316 6316 6316 4262 426 426 426 426 768 788 788 788 3085 846 846 688 688 688 688 688 688 Gross Tonnage (in 100 ft3) 21 3260 27 1877 2191 1068 235 1395 1125 1045 37 503 65 4 1077 11 3929 5753 864 811 3929 1455 1805 342 3469 4248 3173 2051 2594 2964 492 30 34 272 37 921 816 89 1280 1512 49 373 258 644 317 43 140 606 684 1242 24 40 216 261 49 210 56 37 1004 454 191 32 643 641 33 573 586 201 Ship Length (m) 100 53 190 99 97 140 109 74 52 88 137 96 98 82 96 96 96 99 97 74 96 96 96 81 78 149 125 97 82 113 275.1 175.9 175.9 113 169 98 98 230 50 50 62 75 75 75 75 106 122 122 122 115 53 53 53 53 63 64 64 64 98 76 76 63 63 63 63 63 63 Ship Height (m) 16 9 32 15 16 20 16 12 9 13 20 14 16 19 14 14 14 16 16 13 14 14 14 13 12 21 18 16 12 16 32.2 25.4 25.4 19 27.2 16 16 32 8 8 12 10 10 10 10 16 17 17 17 16 10 10 10 10 13 14 14 14 16 16 16 13 13 13 13 13 13 Arrived Port Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Dalian Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Vessel Name HAI RUN 987 HAI RUN 987 HANG SHUN DA 96 HANG SHUN DA 96 HANG SHUN DA 96 HANG SHUN DA 96 HANG SHUN DA 96 HANG SHUN DA 96 HANG SHUN DA 96 HANG SHUN DA 96 HAI RUN 567 HAI RUN 567 HAI RUN 567 XIN WAN LONG CONTI BILBAO HUI JIN QIAO 09 HUI JIN QIAO 09 MING FEN MING FEN MARCLIFF CHANG PING XIN YI HAI 3 HONG RUN 2 HUIYE 3 WEN FENG 18 HUIHONG1 HUIHONG1 LIDA 3 YUAN SHENG 7 YUAN SHENG 7 LU HAI SHUN HENG XU 168 CHANG YONG HAI YONGFENG8 XIN HAI YUE TONGXING19 TENG FENG 30 ZHOU GANG HAI 8 CHANG PING CHANG PING GUAN XIN 508 HENG XU 168 HENG XU 168 JIANG HAI YANG JI HAI ZHI HONG HUIHONG1 HUIHONG1 YUAN DA ZHI HUI HONG XIN 6 JIN ZHAO 1 CHANG LING HAI HUA MIN 18 FENG SHUN 28 XIN LU SHENG 6 MINGYING 56 JIE HAI 5 MIN JIANG 809 XING HANG 288 SHEN ZHOU 9 YONG LONG 102 JIN MA 59 SHENG FENG 2 WAN YUAN 19 ZHONG XING 1 XIN YUN SHENG 18 MIN JIANG 805 DONG SHENG 7 Ship Turn-around Time (min) 688 688 768 768 768 768 768 768 768 768 688 688 688 826 25630 4020 4020 1248 1248 8931 1328 14228 466 1468 18616 1624 1624 12584 1228 1228 1228 1228 13448 1428 4914 3268 1020 22611 1248 1248 7170 1248 1248 3468 9030 1411 1411 1411 9020 3433 8021 5412 4007 1021 6874 512 587 601 589 6214 497 5316 10547 2654 542 563 5412 Gross Tonnage (in 100 ft3) 633 56 60 56 56 61 59 60 48 46 148 214 213 1405 830 926 657 985 659 297 6657 6414 1299 5723 2663 3395 1233 1397 1747 1183 123 158 5977 4563 1398 5235 1761 4817 2582 153 4296 4142 158 557 2083 3096 1233 1736 5754 13 4045 890 740 45 1712 37 34 18 14 2213 63 1191 846 721 21 24 20 202 Ship Length (m) 63 63 64 64 64 64 64 64 64 64 63 63 63 85 208 92 92 88 88 143 97 159 55 97 175 98 98 149 96 96 96 96 160 98 116 100 96 200 97 97 133 96 96 83 140 98 98 96 140 86 159 115 106 84 132 51 53 55 54 135 53 122 147 97 52 53 123 Ship Height (m) 13 13 12 12 12 12 12 12 12 12 13 13 13 14 30 16 16 13 13 23 16 24 13 16 25 16 16 21 16 16 14 14 24 14 16 16 16 32 16 16 19 14 14 14 20 16 16 16 20 13 24 16 16 13 18 8 9 9 9 19 8 18 21 16 9 9 17 Arrived Port Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Xiamen Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Lianyungang Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Vessel Name KAI WANG XING JIN PU LIAN HE 10 BAO MA 227 MIN JIE 586 XIN WAN LONG MING FEN ZHI YUAN SHUN 303 CHANG JIN 26 SAN GEORGIO UNI CORONA FPMC CONTAINER 9 WAN HAI 303 WAN HAI 261 QI MEN WAN HAI 221 BLUE OCEAN DONG FANG FU WAN HAI 233 FPMC CONTAINER 6 LANTAU BRIDE YI YUN YI YUN HANJIN MANILA KANWAY GLOBAL TAIPOWER PROSPERITY VII IDEAL BULKER JIUH CHI WU XING 6 SITC PYEONGTAEK TA HWA ASIA CEMENT NO.6 CEMTEX CREATION UNI ACCORD WAN HAI 206 TROPICAL STAR FPMC B 202 WAN HAI 105 WAN FU TS TAIPEI CHEN CHANG CHIEN HORNG VEERHAVEN V JURA MER-BLUE MER-BLUE PROMINENT PROMINENT PROMINENT PROMINENT VERITAS H VERITAS H VERITAS H KARMEL KARMEL CELTIC CRUSADER DUBLIN EXPRESS ENSEMBLE ENSEMBLE ENSEMBLE QUATTRO MER-BLUE MER-BLUE CLIPPER LOTUS CLIPPER LOTUS BORUSSIA FENNY I Ship Turn-around Time (min) 7214 15272 2134 621 638 648 1587 602 4897 20239 12405 9909 26681 18872 6650 16911 9949 13199 17751 8766 9610 4822 4822 17225 18502 50236 16721 1321 32965 9530 1428 7614 44720 14807 17136 22361 22852 9834 16605 15487 1002 868 22196 7480 1020 1020 1020 1020 1020 1020 3452 3452 3452 3183 3183 2450 46009 3850 3850 3850 3713 1184 1184 17019 17019 6378 3088 Gross Tonnage (in 100 ft3) 412 60 1419 117 23 320 760 25 879 432 434 518 792 330 2 423 4 5 506 341 320 438 272 499 682 3742 1834 805 2955 412 2801 2477 1344 673 1 3507 3501 894 5237 839 1958 1741 240 23 17 27 10 17 271 34 18 21 21 36 27 1567 26 24 489 52 35 39 301 29 26 18 7 203 Ship Length (m) 134 167 97 53 54 85 88 53 113 196 152.1 148 200 198 116 172.2 150.4 161.9 191.5 139.2 142.7 113 113 172 193 234.8 169 95 212 149.7 99 125.5 265 165 174.6 179.4 188 147 170 168 74 71 193 135 86 86 86 86 86 86 100 100 100 90 90 88.4 281 110 110 110 105 86 86 169.4 169.4 121.4 110 Ship Height (m) 19 26 16 9 9 4 13 9 16 26 25.7 23 32.2 28 22 27 22.6 25.6 28 19.6 22.6 19 19 27.6 28 38 27.2 16 32 22.6 14 20.2 35 27.3 27 29 30 25 26 22 13 12 23 15 10 10 10 10 10 10 16 16 16 15.2 15.2 12.8 32.2 11 11 11 9 10 10 27.2 27.2 18.5 12 Arrived Port Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Fuzhou Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Taichung Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Vessel Name APOLLON OBBOLA HELENA THARSIS MADEGRO SR DB LIBRA BOW2 GOTTARDO COLORADO ORION HENJOR HENJOR TANNENBERG 1 SOPHIA SORAYA SOPHIA SORAYA NIEDERSACHSEN 2 JONNI RITSCHER HOEGH TRIDENT TORONTO EXPRESS NEDLLOYD ADRIANA SOPHIA SORAYA CERES LORE PRAHM AKACIA NORDIC STANI SAN ANTONIO AJUG II LYSBRIS LEONIE P JESSICA B BUXLINK BALTIC SKIPPER DUBLIN EXPRESS DONAU HANJIN ASIA FRI STREAM EVER LOGIC TINA HS BEETHOVEN SILVER SKY WINDSTAR STEFAN SIBUM HEKIA BARMBEK NYK HERCULES PAGE AKIA JORK ROVER APL CANADA CAP CLEVELAND APL NINGBO CAP PASLEY APL SCOTLAND NYK TERRA SC TIANJIN OAKLAND EXPRESS HORIZON RELIANCE EVER SIGMA YM ORCHID SANTA RAFAELA HANJIN CONSTANTZA CAMELLIA STAR KVARVEN MOL MODERN DALLAS EXPRESS MAERSK WINNIPEG LT CORTESIA OOCL ITALY Ship Turn-around Time (min) 38871 20186 6788 1801 899 7100 847 7458 899 2985 466 466 866 1322 1322 10021 17360 56164 55994 26833 1322 9983 1156 11662 10318 22914 989 7409 9991 6326 25375 2208 46009 3995 141754 2051 98882 7519 50243 39043 2237 10585 2281 16324 141003 7519 7852 65792 42789 86679 22914 65792 76928 39941 54437 34077 75246 64254 45803 35595 28927 37158 78316 54437 18123 90449 66462 Gross Tonnage (in 100 ft3) 34 23 31 31 29 155 97 30 16 19 26 33 3373 990 1133 548 1104 587 1943 1470 609 1326 330 1461 1699 1657 2400 418 2675 2326 1847 825 2479 2324 2242 745 2163 1532 983 947 828 1255 1660 1439 1429 1254 918 1821 2130 3546 1524 2396 2306 1811 1403 1379 823 789 796 754 746 718 2856 1163 1357 2681 2908 204 Ship Length (m) 225 170.4 135 88 70 122 75 110 85 98.4 65 65 74 96 96 162 178.6 200 294 210.1 96 140.7 99 149.1 151.7 186.3 87 129 139.1 133 207.4 97 281 111.4 350 89.9 338 137.5 282.1 182.5 82.7 146 106 169 350 137.5 140.6 277 269 316 186 277.3 295 259.8 294.1 272.3 300 274.7 281 222.5 222.2 208.8 302 295 175.1 334 280 Ship Height (m) 32.3 23.9 10 11.4 7 23 7 12 9 16.9 8 8 10 14 14 9 27.6 32.3 32.3 30.2 14 23.2 12 22.5 23.8 27.6 8 18 24.2 18.9 29.8 15 32.2 13.4 50 12.7 46 21.3 32.2 29.6 12.6 22 12 27.2 48 21.7 22 40 32 45.6 27.6 40 42 32.4 32.3 30.5 42.9 40 32.2 32.2 30 32.3 43.4 33 27.9 42.8 40 Arrived Port Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Rotterdam Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Vessel Name HAMMONIA ROMA GJERTRUD MAERSK YM EMINENCE APL INDIA XIN FEI ZHOU NYK PEGASUS APL MALAYSIA BRUSSELS BRIDGE OSAKA TOWER COSCO TAICANG KNOSSOS WAVE SAGARJEET SHENGKING PANAMAX TRADER AEGEAN LEADER MAHIMAHI MSC BILBAO OOCL MEMPHIS MOL EXPLORER SALOME METIS LEADER COSCO KOREA SEVILLA CARRIER SUNBELT SPIRIT CSAV BRASILIA APOLLON LEADER R.J.PFEIFFER ISS SPIRIT LOS ANGELES COSCO QINGDAO KOTA WIRAWAN CMA CGM NORMA PYXIS LEADER NYK JOANNA APL EGYPT YM VANCOUVER CCNI ANTOFAGASTA CAFER DEDE COSCO VENICE MOL DESTINY OOCL KUALA LUMPUR APL MELBOURNE ZIM SAVANNAH ATLANTIC CONVEYOR MSC MAEVA TOKYO EXPRESS CLIPPER TENACIOUS HORIZON DISCOVERY OOCL NETHERLANDS MSC RITA GREY SHARK MAERSK LOWA CSCL BRISBANE HUDSON BAY HOEGH TROOPER ZIM YOKOHAMA EEMS SPACE CAPUCINE SCOT EXPLORER FLANDRIA SEAWAYS FLANDRIA SEAWAYS FLANDRIA SEAWAYS SEAGO FELIXSTOWE MSC RAFAELA MSC CORDOBA SUECIA SEAWAYS SUECIA SEAWAYS Ship Turn-around Time (min) 26435 97933 42741 65792 90757 76199 54415 44234 39941 115933 47984 32343 24700 35890 48319 41036 89941 91563 54098 75251 59650 91051 5994 60587 52726 60213 32664 20927 106847 65140 16731 107711 62195 27051 54415 40030 35881 21092 41500 39906 66462 40541 54626 58438 89954 54465 19918 18888 66086 89954 4688 50686 39941 17986 55164 39906 1862 16342 1882 13073 13073 13073 48853 42307 50963 24196 24196 Gross Tonnage (in 100 ft3) 1516 2255 2216 1577 11 1479 780 3736 3171 3528 3441 3422 1907 2239 746 1555 1811 1374 1355 587 748 676 102 1196 1673 633 1425 787 722 838 98 1485 1460 2747 1244 1556 1161 873 1437 1685 1855 1267 1528 712 2360 1946 55 2274 2286 3006 529 827 2833 3470 1734 1858 8 2 5 200 220 219 1784 1303 1524 253 276 205 Ship Length (m) 208.9 367 268.8 277 335.1 300 294 266.7 260.1 348.5 229 190 189 225 180 262.1 335 335 294.1 304 297 334 134 212.1 292 297 217.5 172 342 280 184.5 341 199.9 210 294 260 224 182.9 261.1 260 280 261 294.1 291.9 324.8 294.2 178.7 212 276 324 110 292.1 260 170 200 260 83 170 81.7 142.5 142.5 142.5 294.1 244.2 275 197 197 Ship Height (m) 29.8 42.8 32.3 40 42.8 40 32.2 35.4 32.3 45.6 36.8 32 30 32.2 32.3 32.2 42.8 42.8 32.2 40 32 42.8 20.2 32.3 31 34 32.3 28 44 39.9 27.6 43 32 32.2 32.2 32.3 32 28 32.3 32 40 32.3 32.2 32.4 42.8 32.2 28.6 28 40 43 19 32.4 32 27 32.3 32 14 25 12.5 23.2 23.2 23.2 32.3 32.3 32.3 26 26 Arrived Port Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Los Angeles Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach Long Beach New York New York New York New York New York New York New York New York New York New York New York New York New York New York New York New York New York New York New York New York New York New York New York New York Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Vessel Name BG ROTTERDAM SEVERINE PERU SELANDIA SEAWAYS SEVERINE HANSE VISION DORIS T JANA MAERSK OHIO SVENDBORG MAERSK ESTRADEN BUXCOAST VICTORIA C DOUWENT MAULE MSC MATILDE NATALI MAERSK BENTONVILLE CONMAR AVENUE MARGRETHE MAERSK FO SHAN 12 HAO FO SHAN 12 HAO ZHONGQILUN 138 ZHONGQILUN 138 BO YUN 588 YUNFENG666 JIN LONG 688 U Q N Y ANHYO8912 U Q N Y ANHYO8912 JIAN_GONG_168 GUIPING FEIDA 379 GUIPING FEIDA 379 YUEYINGDEHUO3212 HUI WAN 305 LONG AN DONG YI 688 GUIGANGSISI698 GUANGHONGYUE113 GUANGHONGYUE113 GUI GANG SI SI 578 SUI DE YANG 108 SUI DE YANG 108 GUIGANGTIANYOU2368 YUE HE YUAN HUO 2043 GUIPINGNANHUO 3318 PINGNANYOUJIA 3233 YUE SI HUI HUO 3333 YUEYINGDEHUO 8068 HONG XIANG 308 YUE HE YUAN HUO 2111 YUESHAOGUANHUO2396 YUESHAOGUANHUO2396 YUESHAOGUANHUO2396 YUESHAOGUANHUO2396 YUESHAOGUANHUO2396 Ship Turn-around Time (min) 8273 16342 2993 24196 16342 7713 1973 8273 50686 91560 18205 72760 2990 1311 75752 53208 2837 48853 10585 98268 786 786 688 688 878 899 988 1002 1002 589 864 864 668 764 688 648 802 802 423 668 668 402 412 456 502 308 302 488 589 688 688 688 688 688 Gross Tonnage (in 100 ft3) 3456 1 6 262 1 1390 8 1065 1035 1587 2 599 8 1 1228 728 5 58 554 1622 33 52 25 30 40 166 2380 42 31 50 55 45 31 778 39 42 31 25 59 30 31 39 29 33 34 34 39 33 34 27 37 29 63 29 206 Ship Length (m) 139.6 180 85 197 168 141.6 79 139.6 292.1 347 180 300 87 79.7 305.6 294.1 91 294.1 148 366.9 50 50 48 48 50 50 50 64 64 49 56 56 40 50 44 49 50 50 40 50 50 49 44 49 49 39 38 56 40 67 67 67 67 67 Ship Height (m) 22.2 22 16 25.9 25 20.9 12.4 22.2 32.4 42.8 27 40 18 11.2 40 32.2 13.5 32.3 23 42.9 15 15 10 10 16 11 16 15 15 15 13 13 9 11 10 11 11 11 9 15 15 11 8 11 10 8 8 12 10 13 13 13 13 13 Arrived Port Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Felixstowe Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou Guangzhou APPENDIX III Port Infrastructure Dataset (2012 to 2013) Port Information in 2012 Port Name Singapore Shanghai Hong Kong Busan Dubai Guangzhou Rotterdam Qingdao Hamburg Kaohsiung Antwerp Tianjin Los Angeles Long Beach Bremen New York Laem Chabang Xiamen Dalian Tokyo Jawaharlal Nehru Colombo Valencia Yokohama Gioia Tauro Algeciras Felixstowe Manila Lianyungang Nagoya 2012 Throughput (TEU) 31,649,000 32,529,000 23,117,000 17,040,567 13,280,000 14,547,000 11,870,000 14,503,000 8,860,000 9,780,000 8,640,000 12,303,000 8,077,714 6,045,662 6,120,000 5,529,913 5,930,000 7,200,000 8,060,000 4,165,000 4,260,000 4,260,000 4,470,000 3,172,000 2,721,749 4,070,000 3,950,000 3,710,690 5,020,000 2,655,230 Ship Turn-around Time (min) 730 587 613 986 1203 262 241 114 1763 1077 788 964 2345 2700 1081 2243 681 454 1654 761 861 1166 1129 585 1699 570 680 1679 828 1137 No. of Ships 18567 23925 21171 14193 9388 10749 7544 9814 7569 8228 7276 7734 3243 2475 4894 2307 6299 4866 4824 7818 1847 2271 1896 5097 992 1581 1424 2078 1118 4294 Berth Occupancy Rate (%) 31.88 84.65 34.69 46.54 38.90 27.16 12.08 15.42 75.72 89.97 31.59 89.90 50.59 56.05 60.05 27.73 27.58 47.35 90.52 71.73 30.68 42.56 58.98 27.39 81.27 8.69 15.56 9.76 89.25 78.47 No. of Berth 82 32 73 58 56 20 29 14 34 19 35 16 29 23 17 36 30 9 17 16 10 12 7 21 4 20 12 69 2 12 207 Berth Length (m) 21896 9142 11409 14610 13820 5370 16125 5449 9148 5898 15130 4674 9381 7323 5259 7615 10300 2483 4253 4669 3749 3154 4793 5390 3155 6036 4062 7252 540 3670 Berth Depth (m) 13 12 13 14 13 13 12 15 14 14 15 15 13 14 13 13 15 13 14 14 12 12 14 13 14 13 12 8 10 13 Terminal Area (m2) 6233000 9441323 3518871 4617786 3536905 4689600 6978400 1671000 6103550 1460376 7555573 1949400 6510368 4466227 4589000 5810100 4399900 483200 2078579 1603246 2355000 742730 1853750 2140873 1600000 1167459 1586350 1198234 175000 1405549 Storage Capacity (TEU) 35600 350084 308935 345263 346196 283784 1253000 145072 132725 72051 2788326 107362 104006 127432 88046 48323 119737 15000 36528 202362 118266 67256 104989 107350 75000 43272 109000 55136 7000 77430 No. of Quayside Gantries 204 113 125 80 78 70 124 45 86 21 72 33 69 55 108 70 75 10 42 41 26 26 36 38 25 35 38 17 4 27 No. of Yard Cranes and Tractors 41 980 977 1657 1508 380 1943 208 1041 259 594 279 567 238 299 1093 790 66 353 408 846 586 558 299 278 323 497 926 68 276 Port Name Barcelona Kobe Vancouver BC Le Havre Yantai Osaka Oakland Zeebrugge Balboa Melbourne Keelung St Petersburg Kingston Tacoma Houston Buenos Aires Genoa Southampton Seattle Incheon Freeport Montreal Constantza Charleston Haifa Karachi La Spezia Taichung Fuzhou Duisburg Brisbane 2012 Throughput (TEU) 1,758,647 2,896,000 2,664,000 2,306,000 2,490,000 2,312,000 2,344,424 1,953,000 3,300,000 2,600,000 1,607,566 2,520,000 1,607,566 1,711,134 1,922,529 1,974,000 2,065,000 1,651,000 1,869,492 1,981,821 1,193,000 1,375,327 596,000 1,514,585 1,377,162 1,661,000 1,247,000 1,488,000 1,790,000 1,264,000 1,077,000 Ship Turn-around Time (min) 1757 682 2382 609 1230 732 495 688 860 1565 1064 1563 1377 1206 3672 273 788 2118 2429 911 1237 1684 1940 340 2743 1601 1057 1090 468 169 2004 No. of Ships 2110 4825 1098 971 1765 3846 957 967 1416 2351 3503 1471 1073 607 799 1151 809 580 769 1667 489 597 346 1257 770 584 573 3386 994 518 957 Berth Occupancy Rate (%) 23.24 24.42 28.03 4.39 59.83 24.68 3.81 4.58 46.98 54.59 23.96 31.68 16.77 14.11 43.52 4.66 5.35 47.43 30.02 48.84 38.87 12.12 16.17 6.34 31.33 25.77 10.62 88.99 11.22 2.41 37.00 No. of Berth 13 26 18 26 7 22 24 28 5 13 30 14 17 10 13 13 23 5 12 6 3 16 8 13 13 7 11 8 7 7 10 208 Berth Length (m) 5878 7275 5504 7065 2013 4785 6804 8485 1511 2995 3920 2927 3954 2959 3220 4908 5813 1500 4231 2335 1036 4305 1722 3102 3468 1563 1698 1800 1658 2050 2497 Berth Depth (m) 11 13 15 13 14 13 13 13 13 10 13 11 11 15 12 13 12 13 15 13 16 10 14 13 12 12 14 14 12 10 14 Terminal Area (m2) 1085320 1912293 1835312 3038000 945600 2241860 3250719 3134900 182000 1317954 409900 2634000 1607870 2442642 1340620 1341350 1611298 894375 2294016 515000 477428 1053100 615000 2704200 710000 499324 451000 965402 1328000 435500 1055200 Storage Capacity (TEU) 12084 80959 46822 61300 52925 55725 58102 12100 5950 813700 1209800 68750 88350 398820 36084 375385 68156 35600 1032150 7200 28327 45500 20477 27011 16800 69999 28500 55106 60000 31046 30868 No. of Quayside Gantries 32 44 28 40 22 41 37 15 9 22 29 42 19 26 16 18 31 11 30 8 20 19 8 24 30 13 10 13 12 10 15 No. of Yard Cranes and Tractors 193 622 700 810 147 206 426 878 69 250 88 307 222 186 243 363 461 116 469 940 84 432 207 181 31 380 103 534 34 24 250 Port Information in 2013 Port Name Singapore Shanghai Hong Kong Busan Dubai Guangzhou Rotterdam Qingdao Hamburg Kaohsiung Antwerp Tianjin Los Angeles Long Beach Bremen New York Laem Chabang Xiamen Dalian Tokyo Jawaharlal Nehru Colombo Valencia Yokohama Gioia Tauro Algeciras Felixstowe Manila Lianyungang Nagoya 2013 Throughput (TEU) 32,600,000 33,626,000 22,352,000 17,740,000 13,620,000 15,300,000 11,650,000 15,550,000 9,300,000 9,898,000 8,610,000 13,000,000 8,220,000 6,430,000 5,820,000 5,556,000 5,996,000 8,032,000 9,884,000 4,179,000 4,015,000 4,269,000 4,617,000 2,560,000 3,271,000 4,591,000 4,003,000 3,716,000 5,400,000 2,687,000 Ship Turn-around Time (min) 863 514 880 1335 930 132 106 161 1455 1037 1033 920 1621 1571 952 1423 846 393 1585 712 843 871 898 231 1788 564 626 1347 781 1054 No. of Ships 19290 24977 19018 14471 9801 11221 7843 10246 7901 8378 7596 8118 3386 2584 5109 2389 4442 5079 5036 8162 1928 2370 1979 5321 1035 1650 1486 2170 1121 4531 Berth Occupancy Rate (%) 39.15 77.45 53.45 64.27 31.40 14.31 5.52 22.79 65.24 88.20 43.23 80.01 36.50 34.04 55.18 18.22 24.17 42.76 90.60 70.09 31.36 33.20 48.97 11.31 89.28 8.98 14.97 8.17 84.48 76.73 No. of Berth 82 33 73 58 56 22 29 16 34 20 35 18 29 23 17 36 30 9 17 16 10 12 7 21 4 20 12 69 4 12 209 Berth Length (m) 21896 9142 11409 14610 13820 5370 16125 5449 9148 5898 15130 4674 9381 7323 5259 7615 10300 2483 4253 4669 3749 3154 4793 5390 3155 6036 4062 7252 540 3670 Berth Depth (m) 12.94 12.24 13.09 13.62 12.88 12.50 12.29 14.70 13.66 13.63 15.05 14.74 13.20 14.48 13.30 12.89 14.60 12.66 14.26 14.00 12.00 11.55 13.57 13.24 13.67 12.96 11.61 7.82 10.20 12.80 Terminal Area (m2) 6233000 9441323 3518871 4617786 3536905 4689600 6978400 1671000 6103550 1460376 7555573 1949400 6510368 4466227 4589000 5810100 4399900 483200 2078579 1603246 2355000 742730 1853750 2140873 1600000 1167459 1586350 1198234 175000 1405549 Storage Capacity (TEU) 35600 350084 308935 345263 346196 283784 1253000 145072 132725 72051 2788326 107362 104006 127432 88046 48323 119737 15000 36528 202362 118266 67256 104989 107350 75000 43272 109000 55136 7000 77430 No. of Quayside Gantries 210 116 129 82 80 72 128 46 89 22 74 34 71 57 111 72 77 10 43 42 27 27 37 39 26 36 39 17 4 28 No. of Yard Cranes and Tractors 43 1029 1026 1740 1583 399 2040 218 1093 272 624 293 595 250 314 1148 830 69 371 420 871 604 575 308 286 333 512 954 70 284 Port Name Barcelona Kobe Vancouver BC Le Havre Yantai Osaka Oakland Zeebrugge Balboa Melbourne Keelung St Petersburg Kingston Tacoma Houston Buenos Aires Genoa Southampton Seattle Incheon Freeport Montreal Constantza Charleston Haifa Karachi La Spezia Taichung Fuzhou Duisburg Brisbane 2013 Throughput (TEU) 1,720,383 3,077,000 2,831,000 2,400,000 2,647,000 2,460,000 2,346,000 1,728,000 3,369,000 2,739,000 1,600,000 2,685,000 1,477,000 1,837,000 1,980,000 2,104,000 2,308,000 1,788,000 1,719,000 2,146,000 1,275,000 1,388,000 638,000 1,661,000 1,532,000 1,785,000 1,189,000 1,600,000 1,960,000 1,353,000 1,154,000 Ship Turn-around Time (min) 1257 714 1791 740 872 765 1267 667 998 931 797 1236 1094 1611 1493 183 510 1326 1707 950 1321 668 1881 435 1356 1547 1120 1139 430 153 1035 No. of Ships 1978 5037 1146 1013 1843 4015 999 1009 1478 2454 3841 1536 1120 633 827 1202 845 606 796 1740 510 623 361 1302 804 610 598 3335 1037 540 999 Berth Occupancy Rate (%) 17.36 26.69 22.01 5.56 44.28 26.95 10.17 4.64 56.89 33.91 19.68 26.15 13.91 19.67 18.32 3.26 3.62 31.01 21.84 53.13 43.33 5.02 16.37 8.41 16.16 26.00 11.74 91.56 10.76 2.27 19.96 No. of Berth 13 26 18 26 7 22 24 28 5 13 30 14 17 10 13 13 23 5 12 6 3 16 8 13 13 7 11 8 8 7 10 210 Berth Length (m) 5878 7275 5504 7065 2013 4785 6804 8485 1511 2995 3920 2927 3954 2959 3220 4908 5813 1500 4231 2335 1036 4305 1722 3102 3468 1563 1698 1800 1658 2050 2497 Berth Depth (m) 11.38 13.39 15.46 13.10 14.00 12.79 13.05 12.99 12.75 10.48 12.67 11.25 11.10 15.40 12.19 13.00 12.22 12.57 15.05 12.50 16.00 10.32 13.58 13.00 11.94 11.90 13.50 14.00 12.00 10.00 14.30 Terminal Area (m2) 1085320 1912293 1835312 3038000 945600 2241860 3250719 3134900 182000 1317954 409900 2634000 1607870 2442642 1340620 1341350 1611298 894375 2294016 515000 477428 1053100 615000 2704200 710000 499324 451000 965402 1328000 435500 1055200 Storage Capacity (TEU) 12084 80959 46822 61300 52925 55725 58102 12100 5950 813700 1209800 68750 88350 398820 36084 375385 68156 35600 1032150 7200 28327 45500 20477 27011 16800 69999 28500 55106 60000 31046 30868 No. of Quayside Gantries 33 45 29 41 23 42 38 15 9 23 30 43 19 27 16 18 32 11 31 8 21 19 8 25 31 13 10 13 12 10 15 No. of Yard Cranes and Tractors 199 641 721 834 151 212 439 904 71 258 91 316 229 192 250 374 475 119 483 968 87 445 213 186 32 391 106 550 35 25 258 [...]... Also, port operators can use the information from performance analysis to improve their port planning and operations 1.2 Difficulties in Port Performance Measurement and Benchmarking In the literature, there have been extensive studies that focus on port performance measurement and benchmarking (Ashar, 1997; Cullinane, 2002; Bichou and Gray, 2004) Topics such as individual performance metrics, performance. .. direction and will cause unintended negative consequences 5 The performance of a port can influence the economic growth of a region greatly because ports connect the sea transport and inland transport modes They are also crucial providers for the activities of vessels, cargo and inland transport A port with good performance provides satisfactory service for ships and efficient cargo operations and contributes... to identify port performance indicators relevant to the activities of vessels, cargo and terminals Through the analysis of ports efficiency using identified indicators, insights on port performance benchmarking on an international scale can be obtained 1.4 Objectives In this thesis, performance benchmarking of global container ports using efficiency analysis is studied There are three important objectives... between container ports results in the interest of port operators to improve their efficiency Port efficiency, which measures the utilization of port resources, is of importance to contribute a nation's international competitiveness (Wang et al., 2002) The analysis of port efficiency allows port operators to compare performance of different ports This allows them to enhance operations and produce as much... suitable for benchmarking port performance This is because each port- country has its own economic structure and a separate inter-sectoral configuration In addition, the data relevant to the port economic impact studies by input-output models and gravity models (such as the profit, price and cost of cargo, transport and labor cost) are limited 2.2.2 Port Trade Efficiency Studies Port trade efficiency. .. importance to researchers due to the growing importance of understanding the role of ports in trade facilitation Better trade facilitation allows improved efficiency in administration and procedures as well as enhanced logistics at ports and customs (Wilson et al., 2003) In most port trade efficiency studies, port efficiency is often studied in conjunction with evaluated in relation to transport and. .. performance and port traffic It was found that higher efficiency induces higher traffic at most of India ports and suggested that government should give priority to improve port performance by enhance facilities Tongzon and Heng (2005) used PCA to investigate the quantitative relationship between port ownership structure and port efficiency and found that private sector participation in the port industry... study port benchmarking performance are introduced: (1) performance metrics and index methods, (2) port impact studies and (3) frontier approaches (Bichou, 2006) Research studies on ship turn-around time in port industry are then presented, including the applications of ship turn-around time in port classical operation strategies and port logistic process simulations 2.1 Performance Metrics and Index... between port trade and the regional economic impacts Port impact studies literature typically involve: port economic impacts and port trade efficiency studies (Bichou, 2006) 2.2.1 Port Economic Impact Study Port economic impact study is an important aspect of determining the regional economic influence of a port It is useful in determining the capital and operating budgets for publicly-owned port facilities... as the single performance indicator to evaluate port performance No consensus on a single framework for port performance benchmarking has been established to date Bichou (2006) reviewed the most practical and theoretical approaches to port performance measurement benchmarking over the last three decades and summarized the core differences in these studies(Roll and Hayuth,1993; Christmann and Taylor,