WORLD BANK TECHNICAL PAPER NUMBER 38 WTP38 May 1985 Bulk Shipping and Terminal Logistics SECTORAL LIBRARY !NTERNATIONAL BANK 12 t4? FOR Ernst G. Frankel, John Cooper, RECONSTRUCTION AND DEVELOPMENT Yoo Whan Chang, and George Tharakan J UL 1 1 1985 _1 U -__ C'~ ~ ~~~~~~N ! \ w l | | zE'ILE COP YI \a '1 WORLD BANK TECHNICAL PAPERS No. 1. Increasing Agricultural Productivity No. 2, A Model for the Development of a Self-help Water Supply Program No. 3. Ventilated Improved Pit Latrines: Recent Developments in Zimbabwe No. 4. The African Typanosomiases: Methods and Concepts of Control and Eradication in Relation to Development (No. 5.) Structural Changes in World Industry: A Quantitative Analysis of Recent Developments No. 6. Laboratory Evaluation of Hand-operated Water Pumps for Use in Developing Countries No. 7. Notes on the Design and Operation of Waste Stabilization Ponds in Warm Climates of Developing Countries No. 8. Institution Building for Traffic Management (No. 9.) Meeting the Needs of the Poor for Water Supply and Waste Disposal No. 10. Appraising Poultry Enterprises for Profitability: A Manual for Investors No. 11. Opportunities for Biological Control of Agricultural Pests in Developing Countries No. 12. Water Supply and Sanitation Project Preparation Handbook: Guidelines No. 13. Water Supply and Sanitation Project Preparation Handbook: Case Studies No. 14. Water Supply and Sanitation Project Preparation Handbook: Case Study (No. 15.)Sheep and Goats in Developing Countries: Their Present and Potential Role (No. 16.)Managing Elephant Depredation in Agricultural and Forestry Projects (No. 17.)Energy Efficiency and Fuel Substitution in the Cement Industry with Emphasis on Developing Countries No. 18. Urban Sanitation Planning Manual Based on the Jakarta Case Study No. 19. Laboratory Testing of Handpumps for Developing Countries: Final Technical Report No. 20. Water Quality in Hydroelectric Projects: Considerations for Planning in Tropical Forest Regions ( ) Indicates number assigned after publication. (List continues on the inside back cover) WORLD BANK TECHNICAL PAPER NUMBER 38 Bulk Shipping and Terminal Logistics Ernst G. Frankel, John Cooper, Yoo Whan Chang, and George Tharakan The World Bank Washington, D.C., U.S.A. Copyright (© 1985 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing May 1985 This is a document published informally by the World Bank. In order that the information contained in it can be presented with the least possible delay, the typescript has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. The publication is supplied at a token charge to defray part of the cost of manufacture and distribution. The World Bank does not accept responsibility for the views expressed herein, which are those of the author(s) and should not be attributed to the World Bank or to its affiliated organizations. The findings, interpretations, and conclusions are the results of research supported by the Bank; they do not necessarily represent official policy of the Bank. The designations employed, the presentation of material, and any maps used in this document are solely for the convenience of the reader and do not imply the expression of any opinion whatsoever on the part of the World Bank or its affiliates concerning the legal status of any country, territory, city, area, or of its authorities, or concerning the delimitation of its boundaries or national affiliation. The most recent World Bank publications are described in the annual spring and fall lists; the continuing research program is described in the annual Abstracts of Current Studies. The latest edition of each is available free of charge from the Publications Sales Unit, Department T, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from the European Office of the Bank, 66 avenue d'16na, 75116 Paris, France. Ernst G. Frankel is an advisor on ports, shipping, and aviation, John Cooper and Yoo Whan Chung are marine consultants, and George Tharakan is a shipping consultant for the Transportation Department of the World Bank. Library of Congress Cataloging in Publication Data Main entry under title: Bulk shipping and terminal logistics. (World Bank technical paper, ISSN 0253-7494 ; no. 38) 1. Bulk solids--Transportation. 2. Bulk carrier cargo ships. 3. Terminals (Transportation) I. Frankel, Ernst G. II. International Bank for Reconstruction and Development. III. Series. HE595.B84B85 1985 387.1'53 85-9288 ISBN 0-8213-0531-X - iii - ABSTRACT This paper is intended as a guide to sea port planners in the preliminary assessment of projects involving bulk terminals. An overview of the bulk trades and bulk shipping provides the framework for assessing the market for bulk terminal services. Siting considerations are addressed taking into account the inland transport network, and the relative merits of various transport modes for bulk movements. Terminal equipment and facility layouts are discussed with regard to operational characteristics, and some indication of relative costs are provided. Mathematical models useful in evaluating preliminary design options are presented for various aspects of terminal design such as berth congestion and storage capacity calculations. - iv - CONDENSE Ce document doit aider les planificateurs de ports maritimes a proceder a une evaluation preliminaire de projets de terminaux a vrac. Une vue d'ensemble des echanges et du transport maritime du vrac sert de cadre a 1'6valuation du marche des services de terminaux a vrac. On etudie l'emplacement des terminaux compte tenu du reseau des transports int6rieurs et des merites relatifs de divers modes de transport pour les mouvements du vrac. On examine les caracteristiques operationnelles de 1'6quipement des terminaux et de la disposition des installations en donnant certaines indications des prix relatifs. Des modeles mathematiques utiles pour evaluer les options d'avant-projets sont presentes pour divers aspects de la conception des terminaux comme les calculs de l'encombrement des postes a quai et de la capacite d'entreposage. -v - EXTRACTO Este trabajo tiene como objetivo servir de guia a los planificadores de puertos maritimos en la evaluaci6n preliminar de proyectos relacionados con terminales para carga a granel. Una visi6n general del comercio y el transporte naviero a granel proporciona el marco para evaluar el mercado de los servicios de terminales para este tipo de carga. Se abordan aspectos pertinentes a los emplazamientos, teniendo en cuenta la red de transporte interior y las ventajas relativas de las diversas modalidades de transporte para el movimiento de la carga a granel. Se analiza la disposici6n de los equipos y servicios de los terminales con referencia a las caracteristicas de funcionamiento, y se suministran algunos datos sobre costos relativos. Se presentan modelos matematicos utiles para la evaluaci6n de las opciones preliminares para diversos aspectos del disenio de los terminales, como la congesti6n en los muelles y los calculos de la capacidad de almacenamiento. - vii - TABLE OF CONTENTS I. Overview 1. Description of the Bulk Trades ........................... . 1 2. Bulk Trades of Developing Countries ......... .. ............ 12 3. Interests of Developing Countries in the Bulk Trades ...... 17 4. The Role of Ports and Harbors in the Bulk Trades .......... 24 5. Bulk Shipping ......................... 27 II. Forecasting in the Bulk Trades 1. General Comments .................... 35 2. Predicting Potential Volumes .............................. 38 3. Market Share Estimates .................................... 40 4. Bulk Logistic Forecasting ................................ 45 III. Bulk Shipping Economics 1. Bulk Carriers and the Bulk Carrier Fleet ..... ............. 53 2. Efficiency in the Use of Bulk Carrier Tonnage .... ......... 67 3. Interests of Organizations in Bulk Shipping ................. 72 4. Vessel Chartering .......... ............................... 79 5. Ship Ownership ............................................. 83 IV. Inland Transportation to Alternate Terminal Sites 1. Introduction ...................... ........................ 97 2. Truck Transport ..................... ...................... 98 3. Rail Transport .................... ........................ 101 4. Inland Waterways Transport .................... 103 5. Small Ocean Bulk Carriers ............................... 107 6. Slurry Pipelines ............... ................ 108 7. Conveyors .. ................... 117 8. Aerial Tramways .....................1 117 9. Modal Comparisons ......................................... 120 V. Terminal Siting 1. Introduction .............................................. 125 2. Methodology for Terminal Siting .......... ................ 125 3. Information Requirements .................................. 129 4. Ranking Procedure ..................... , 129 5. Terminal Siting and Regional Development ....... .......... 131 VI. Terminal Equipment Options 1. Introduction ... 133 2. Ship Loaders ........................... 135 3. Ship Unloaders .............................. 141 4. Material Storage .....,,.,.,,.,.151 5. Cargo Movement and Transfer ......................... .... 170 6. Ancillary Equipment ................................... . 181 VII. Terminal Design and lavout 1. General Design Considerations .. 191 2. Mathematical Tools ....................................... 198 3. Process Selection Options ................... 206 4. Terminal Design Optimization ............................. 216 - viii - VIII. Terminal Cost Estimation 1. Introduction . ............................................. 244 2. Civil Engineering Costs ................................... 244 3. Equipment Costs .......... ................................. 250 4. Bid Invitation and Evaluation .............................. 255 APPENDICES A. Regional Development Analysis ....... ....................... 261 B. Techniques for Port Development Analysis ................... 265 C. Queueing Theory .................. 277 D. Inventory Management ........... ............................ 283 E. Methods of Making Pile Capacity Computations ................ 287 - ix - LIST OF FIGURES I.1.1 Seaborne Trade 1981 - Iron Ore ................................... 6 I.1.2 Seaborne Trade 1981 - Grain ...................................... 7 I.1.3 Seaborne Trade 1981 - Coal . ...................................... 8 I.2.1 Value of Selected Bulk Exports of Developing Countries .... ....... 13 I.2.2 Weight of Selected Bulk Exports of Developing Countries .... ...... 14 I.2.3 Weight of Principal Bulk Imports of Developing Countries ...... P.. 16 I.3.1 Distribution of Exports by Value .................................. 21 I.3.2 Vessel Sizes Employed in the Grain Trade .......................... 25 I.5.1 Major Bulk Commodity Trades ...................................... 28 II.3.1 Costs and Benefits from New Technological Innovation .... ........ 43 II.4.1 Elements of Bulk Logistics Forecasting ............ ............. 47 II.4.2 Forecasting Dry Bulk Shipping Decision ModelRequirements ........ 49 III.1.1 Densities of Bulk Cargoes ....................................... 54 III.1.2 Gearless Pure Bulk Carrier ....................................... 55 III.1.3 Typical Dry Cargo Ship .......................................... 56 III.1.4 Car Deck In Medium-Sized Geared Bulk Carrier .................... - - 58 III.1.5 Grain Stowage in Tankers Compamud With Other Ship Types ..... ..... 61 III.1.6 Approximate Composition of the World's Bulk Fleet .... ........... 62 III.1.7 Comparison of Ore Carrier with Normal Bulk Ship .... ............. 62 III.1.8 Self-Discharging Vessel ......................................... 64 III.1.9 Bulk Carrier Dimensions ......................................... 68 III.1.10 Maximum Good Practice Design Dimensions of Bulk Carriers .69 III.2.1 Trading Pattern for Small-Sized Dry Bulk Carriers .... ........... 70 III.4.1 Single Voyage Charter Rates ..................................... 81 III.4.2 Time Charter Rates .............................................. 84 III.4.3 Charterer's Voyage Estimate Form ................................ 85 III.5.1 Historic Trends in Orders for Bulk Carriers ..................... 88 TII.5.2 Scrap Ship Prices ............................................... 90 III.5.3 Trends in Used Ship Prices ...................................... 92 III.5.4 Current Bulk Carrier Prices ..................................... 93 III.5.5 Comparison of Returns from Dry Cargo Vessel Operation ............ 94 IV.4,1 LASH and SEABEE Barge Equipment ................................. 105 IV.6.1 General Process for Slurry System ............................... 109 IV.6.2 Comparison of Relative Freight Rate t ........................... 110 IV.6.3 Slurry Pipeline Transportation Cost ............................. 115 IV.8.1 Two Types of Aerial Tramways~ ...................................... 119 VI.2.1 Layout of Pier for Travelling Ship Loader ....................... 137 VI.2.2 Layout of Radial Ship Loader Installation ....................... 137 VI.2.3 Travelling Ship Loader .......................................... 139 VI.2.4 Comparison of Radial and Linear Ship Loaders .................... 139 VI.3.1 Level Luffing Ship Unloader ..................................... 142 VI.3.2 Gantry Trolley Type Ship Unloader ............................... 142 VI.3.3 Continuous Type Ship Loader ..................................... 144 x VI.3.4 Maximum Ship Size for Unloaders ................................. 145 VI.3.1.1 Barge Mounted Ship Unloader ..................................... 148 VI.3.2.1 Typical Pneumatic Grain Discharge System ........................ 152 VI.4.2.1 Typical Pile Configurations ...................................... 154: VI.4.2.2 Stacking and Blending with Conveyors ............................ 156 VI.4.2.3 Typical Rail Mounted Stacker ..................................... 157 VI.4.2.4 Economic Re,gions of Stockpiling Systems ......................... 160 VI.4.2.5 Radial Stacking and Reclaiming ................................... 161 VI.4.2.6 Segregation of Material In Gravity Stacked Pile ............... ..163 VI.4.2.7 Typical Stacker/Reclaimer ......................................... 166 VI.4.2.8 Typical Stack/Rake ............................................... 166 VI.4.2.9 Travelling Scraper Bridge ....................................... 167 VI.4.3.1 Silos for Denser Materials ....................................... 167 VI.4.3.2 Comparative Costs of Covered Storage ............................ 169 VI.4.3.3 Covered Stockpile Scheme ........................................ 169 VI.5.1.1 Typical Transfer Tower .......................................... 174 VI.5.1.2 Typical Conveyor Installations .................................. 175. VI.5.1.3 Crawler Mounted Conveyor ........................................ 180 VI.5.2.1 Air Slide Conveyor Design Information ........................... 182 VI.5.3.1 Bucket Elevators ................................................ 183 VI.6.1 Coal Loading Station - Key Diagram .185. VI.6.2 Rail Car Loader .186 VI.6.3 Rail Car Unloader .187 VI.6.4 Rail Car Economic Zones .189: VII.1.1 Design Process for a Bulk Handling Facility .192 VII.1.2 Quantitative Factors for Terminal Design .194. VII.2.2 Pipeline Analogy for Transportation Planning .208 VII.3.1 Steps in Initial Preliminary Design .208 VII.3.2 Flow Diagram for Large Scale Mineral Export .213 VII.3.3 Coal Handling Using Railways or Conveyor Transport System . .214 VII.3.4 Sy-stem Facilities Comparison .215 VII.4.1.1 Queueing Analysis of Waiting Times for Ships in Port .. . 218 VII.4.1.2 Queueing Theory in Preliminary Planning ......................... 220. VII.4.2.1 Total Cost vs. Ship Loader Capacity ............................. 222 VII.4.3.1 Storage Options ................................................. 224 VII.4.3.2 Typical Variation in Dry Bulk Cargo Terminal Inventory Level ... 228 VII.4.3.3 Options for Stockyard Layout .................................... 232 VII.4.3.4 Iron Pellet Imports ............................................. 239 VII.4.3.5 Coal Export Terminal ............................................. 240 VIII.l.l Port Kembla Coal Loader Estimates and Tender Responses . ..........242 VIII.2.1 Breakwater Cost EFstijation .................... ................. 246 VIII.2.2 Quay Costs per Linear Meter ................ 247 VIII.3.1 Ship Unloader Costs (1983) ...................................... 251 VIII.3.2 Cost of Rail Mounted Pneumatic Grain Unloaders (1983) ........... 251 VIII.3.3 Price of Crawler Mounted Bucket Wheel Excavators (1983) ......... 252 VIII.3.4 Cost of Crawler Mounted Conveyors (1983) ......................... 252 VIII.3.5 Belt Conveyor Prices (1983) ...................................... 253 - xi - LIST OF TABLES I.1.1 Trade in Primary Commodities ..................................... 4 I.1.2 Annual World Exports of 8 Major Bulk Commodities . ................ 5 I.1.3 Total World Transport Requirements for 8 Major Bulks .... ......... 5 I.1.4 Possible Ways to Ship "Aluminum" from Mine to Consuming Market Area . ........................................................... 9 1.1.5 Percentage of Commodity Flows Moving in Voyage Chartered Ships ... 10 I.1.6 Savings In Kaolin Transport Using Slurry Carriers .... ............ 11 I.2.1 Wheat Imports of Major Developing Countries ...................... 15 1.3.1 Illustrative Foreign Exchange Savings from the Operation of a 60,000-ton Bulk Carrier Built Overseas ........... 17 I.3.2 Comparison of Bulk Exports to Indicators of Country Welfare ...... 22 1.5.1 Bulk Carrier Fleet Development ....................................... 29 II.2.1 Elements of Bulk Logistics Forecasting ........................... 39 III.1.1 Size Distribution of Bulk Carriers as of January 1983. .. 65 III.1.2 Bulk Carrier Dimensions for Standard Classes ..... ................ 66 III.2.1 Trading Pattern for a Small Bulk Cgrrizer .......................... 71 III.3.1 Five Types of Vessel Service Arrangements ........................ 74 III.4.1 Vessel Charter Party Provisions ........ .......................... 80 III.4.2 Single Voyage Charters by Commodity ....... ....................... 80 LII.4.3 Selected Current Charters ........................................ 82 III.5.1 Methods of Vessel Procurement ..................................... 86 III.5.2 The Bulk-Carrier Orderbook ....................................... 88 IV.1.1 Comparison of Alternate Transport Modes .............................. 99 IV.6.1 Comparison of Costs Between Marconaflo and Conventional Trans- portation Systems - A Hypothetical Case ........................ 112 IV.6.2 Comparison of Slurry Pipeline Costs with Railway Costs .... ....... 114 IV.6.3 Selected Commercial Slurry Pipelines ............................. 116 IV.8.1 Investment and Operating Cost of a Ropeway System .... ........... 121 IV.8.2 Selected Examples of Long-Distance Aerial Ropeways .... ........... 122 IV.9.1 Relative Costs for Alternative Modes of Mineral Transport Over Short Distances . ................................................ 123 IV.9.2 Putnam Coal Mine: Alternative Belt Conveyor Systems .... ......... 124 V.2.1 Present Value Analysis ............. .............................. 128 V.5.1 Ranking Procedure Example .......... .............................. 131 VI.1.1 General Equipment Options for Bulk Ports ......................... 134 VI.1.2 Summary of Design Commodity Handling Rates ....................... 135 VI.2.1 Dimensions of Ship Loaders ............. .......................... 141 VI.3.1 Typical Ship Unloader Dimensions and Rates ........................ 146 VI.3.2 Capacity of Continuous Ship Unloaders ........... ................. 147 Vi.3.3 Specifications for a Barge Mounted Ship Unloader ................. 150 VI.4.2.1 Typical Stacker Dimensions .................... . ................... 158 VI.4.2.2 Alternative Reclaim Systems ...................................... 159 VI.4.2.3 Typical Dimensions of Bucket Wheel Reclaimers .................... . 164 VI.4.3.1 Comparison of Characteristics of Silos and Warehouses .... ........ 168 - xii - VI.5.1 Material Transfer Equipment Options ............................. 171 VI.5.2 Cycle Times of Earth-Moving Vehicles ............................ 172 VI.5.1.1 Preferred Types of Conveyors for Bulk Materials .................. 176 VI.5.1.2 Maximum Belt Speed for Various Materials ............... e ........ 178 VI.5.2.1 Typical Materials Moved by Air Float Conveyors .... .............. 179 VI.6.1 Comparison of Rail Car Unloading Equipment .188 VII.2.1 Various Terminal Design Methods and Their Applicable Areas .199 VII.2.2 Mathematical Models Available for Preliminary Design . 200 VII.2.3 Generally Accepted Rules of Thumb About Port Design ., 201 VII.2.4 Equipment Selection Factors .207 VII.3.1 Initial Performance Specification for Coal Logistics System .211 VII.3.2 Ship Performance Specification for Coal Logistics System .211 VII.3.3 Important Physical Properties of Coal .212 VII.3.4 Major Decision Variables for Coal Port Design .216 VII.4.2.1 Example on Estimation of Number of Berths and Optimum Size of Ship Loader ..221 VII.4.3.1 Equipment Capacities as Function of Reclaimer Capacity ... ...... 225 VII.4.3.2 Storage System Costs ............................................ 225 VII.4.3.3 Stockpile Size and Volume ....................................... 227 VII.4.3.4 Safety Stock Calculation ....................................... 230 VII.4.3.5 Geometries of Piles ...............................................232 VII.4.3.6 Geometry of Pile System ......................................... ,234 VII.4.3.7 Comparative Costs for Stockpile Alternatives ..................... ,235 VII.4.3.8 Guide to Belt Selection ......................................... 238 VII.4.3.9 Other Terminal Equipment Installed .............................. 238 VII.4.3.10 Initial Coal Port Estimate ...................... 238 VIII.1.1 Cost Breakdown -- Port 1 ......................................... 243 VIII.1.2 Cost Breakdown -- Port 2 ......................................... 243 VIII.1.3 Cost Breakdown --Port 3 ......................................... 243 VIII.2.1 Unit Construction Costs for Pavement ............................. ,248 VIII.2.2 Bulk Terminal Civil Works Costs ........ ......................... 249 VIII.2.3 Bulk Terminal Storage Construction Costs ..... ................... 249 VIII.3.1 Bulk Terminal Equipment Costs ................................... 250 VIII.3.2 Selected Bulk Equipment Project Cost Data .......................... 254 - xiii - PREFACE Bulk terminals are transportation facilities where vehicles (i.e., ships, barges, rail cars, trucks, etc.) are accommodated and where particular functions necessary to the transportation process are performed. Terminals are required at the ends of a transport chain and at intermediate points. Intermediate facilities are involved with the transshipment of goods between intra-and inter-modal transport. The functions of a bulk terminal include the following: 1. Cargo loading and unloading 2. Cargo consolidation and deconsolidation - this is often beneficial because of economies of scale in ocean trans- portation, feeder transportation, and/or terminal operation. 3. Storage - this means temporary or long-term facilities, as a service to port users or to effect transport economies 4. Classification of cargo 5. Intrasystem or intersystem cargo transfer 6. Vehicle marshalling and cargo stowage (a) Cargo stowing on a ship or other transport vehicle is accomplished usually by using loading equipment, such as shiploaders (i.e., grabs, buckets, chutes, etc.) which must be compatible with the vehicle or ship cargo space. (b) Securing, tie-down, or stowage on or in the transport 7. Vehicle maintenance, servicing, and/or modification 8. Physical form change of cargo (a) Cargo conversion, such as slurrying, bagging, baling, etc., of dry bulk cargo may be done for reasons not associated with the loading function. (b) It may be performed also to facilitate cargo handling or transfer operations. - xiv - 9. Packaging of cargo 10. Safeguarding of cargo 11. Cargo and vehicle information management and documentation. Each of these functions needs to be considered in the design of bulk terminal systems. An analysis of the terminal costs associated with the loading function must include time costs incurred by both cargo and vehicles. Bulk cargo loading involves delays in vehicle time and resultant costs to the vessel owner. Therefore, ship, truck, rail-car, etc., turn-around time costs and cargo inventory holding costs must be included in determining terminal operating costs. Similarly, cargo conversion costs, if accomplished as an integral part of the overall transportation process, must be included as part of the terminal costs. Various descriptive models of the terminal classification and cost processes can be developed and used. It is necessary, here, to recognize that terminals are networks of operational links which often are random in nature with regard to the time factor and the sequence of operations used by cargo in its flow through the terminal. Section 1 provides an overview of the bulk trades and discusses the major commodities moving in the trade. Section 2 describes tech- niques for projecting trade and traffic levels. The economics of bulk shipping is discussed in Section 3. Inland transport for bulk commodities and bulk terminal siting are the topics of Section 4 and 5. Equipment options, facility layout, and terminal cost estimation are dealt with in Sections 6, 7 and 8. Mathematical techniques useful in bulk terminal analysis are presented in Appendices A through E. - xv - This report is a sequel to the report entitled, "Container Logistics and Terminal Design". It presents the results of a study of bulk terminal systems and their essential design considerations. The issues involved in design, from project identification to traffic forecasting, siting, capacity determination, equipment selection, cost determination et al., are discussed to provide an up-to-date basis for decision-making in bulk terminal project development. Information was received from many sources, all of whom are identified and acknowledged at the appropriate places in the text. The figures and the tables developed by the authors are presented without the source. The inclusion of any particular type of equipment does not signify that it is recommended by the World Bank. The World Bank has not verified the information provided by manufacturers or designers as to the performance or any other features of the equipment. - xvi - ACKNOWLEDGEMENTS This paper vas written as part of a research effort into the design of bulk terminals within the Ports Unit of the Transportation Department of the World Bank. The views expressed in the paper are those of the authors and should not be attributed to the World Bank. Helpful comments on the paper were provided by Ezra Bennathan (Economic Advisor), Clell C. Harral (Highway Advisor), and Liviu L. Alston (Railwav Advisor) of the Transportation Department, and Richard Scheiner (Port Engineer) of the Western Africa Projects Department. - 1 - I. OVERVIEW I.1 Description of the Bulk Trades World maritime trade in 1982 amounted to 3.21 billion metric tons. This figure is down from the highest trade level, 3.77 billion metric tons, achieved in 1977. The transportation of petroleum accounted for about one-half of the tonnage in worldwide maritime trade. Major dry bulk commodities, including iron ore, coal and grain, accounted for another one-quarter of the tonnage shipped. The remaining quarter came from the transportation of general cargo items. Traditionally, shipping has been split into two separate markets - liner shipping and bulk shipping. The liner market provides small consign- ment services to many different shippers, as well as full ship loads shipped on "liner terms"; consignments may vary in size and each consignment must have a separate bill of lading. Liner operations ship goods for anyone on request. A single shipload of goods may represent as many as five hundred different consignments. This requires coordination among many shippers, freight forwarders and other parties, who can be reached only with a regu- larly published schedule and standardized commercial arrangements. The shipper may be unable or unwilling to arrange any part of the goods trans- portation (including inland movements), so liner operations frequently offer door-to-door service. Most liner operations provide containerl/ services, but also provide noncontainer capacity for commodities which cannot physically fit in a container. 1/ Containers are usually 8 x 8 x 40 or 8 x 8 x 20 foot boxes into which goods are loaded at the shipment point. The containers can be trans- ported door-to-door, which includes the ocean leg, without the goods being removed from the container. This system has many advantages, notably in the reduction of stevedoring costs. In bulk shipping, one load is shipped usually under a single bill of lading. Multiple consignments do occur occasionally, but may not always be possible. In the United States, for example, there is a law requiring ship operators who issue more than three bills of lading per sailing to register as common carriers. There is, how- ever, considerable overlap between liner and bulk commodities, simply because the principal differentiating factor is consignment size. Goods such as timber, plywood, steel, grain, cement, and fertilizer are commonly shipped in either market sector. These so-called "neobulk" commodities could be described as bulk substances, but often move in consignment sizes so small that many, many shipments would be- needed to fill a vessel. Shipment size of bulk traffic is large and can be controlled by the shipper. Considerations of economies of scale play a major role in the design of bulk logistics systems.2/ Because of the control possible, shippers can obtain large savings by carefully integrating shipping activity with other operational aspects. Oil, steel, grain and aluminum companies, for example, use linear pro- gramming models to help manage their logistics planning functions. From these models it is possible to note how economies of scale in ship capacity can be traded off against inventory holding costs and storage capacity at the producing, loading, unloading or distribution centers. 2/ Economies of scale do not play such a large role in the liner trades. The largest decrease in unit costs occurs in goods that are shipped in full container lots (about 10 tons). After this, while economies of scale still exist, seldom are they passed on, except with large volume shippers with sufficient power to nego- tiate a rate reduction for a specific commodity, as the rate charged per full container does not depend on how many are shipped. -3- The greatest market penetration of bulk trades is in the "primary" commodities (Table 1.1.1). However, the market share in these commod- ities varies widely, with some being shipped in small shipment sizes as, for example, hides, fibers (wool and cotton) and beverages (coffee, cocoa and tea), which remain major liner cargoes. One reason they are shipped as liner cargo is that their packaging is not compatible with the bulk cargo handling processes available. Cotton bales, for instance, are more expensive to load on a ship than containers filled with cotton bales, and hogsheads of tobacco are not strong enough to be stacked to fill a hold without the structural support of a container. Sometimes, bulk commodities are shipped on liner ships because they are consigned to di-fferent destinations and cannot be handled economically in bulk. Occasionally large volume manufactured products, such as paper and automobiles, are shipped as bulk cargoes. The development of the modern, small-sized bulk carrier has in- creased the number of commodities shipped in a bulk mode. Improve- ments in commercial warehousing and distribution systems also have facilitated this trend; bulk ocean shipments can often be divided into retail-sized quantities at the warehouse. Fertilizer and rice are good examples, with bulk shipments commonly ranging from 3,000 to 8,000 tons. There are five major bulk commodities carried in maritime commerce-- iron ore, coal, bauxite, phosphate rock and grains. Tables I.1.2 and I.1.3 give historic shipment levels and an estimate of the transport requirements for major bulk commodities. Figures I.1.1-3 show the trade patterns for the three major bulk commodities. The remaining commodities are termed minor bulks; these include sugar, sulfur, steel scrap, timber and fertilizer. -4- Table I1.1. Trade in Primary Commodities (Average 1978-1980) (in millions of U.S. dollars) Developing Commodity Total Value Country Share Bananas 1167 1079 Bauxite 768 648 Beef 7577 1243 Cocoa 3139 2969 Coffee 11983 11063 Copper 8991 5529 Fibers 10551 4059 Fish Meal 990 488 Ground Nut & veg. oils 3676 2710 Hides 3307 407 Iron Ore 6265 2702 Lead 1709 480 Maize 9879 1334 Manganese Ore 410 319 Petroleum 245234 212294 Phosphate Rock 1847 1206 Rice 4193 1811 Rubber 3821 3762 Sugar 10614 3672 Tea 1817 1394 Timber 18458 5515 Tin 2659 2157 Tobacco 3806 1706 Wheat 12789 782 Zinc 2001 473 Source: Commodity Trade and Price Trends The World Bank 1982/83 edition. -5- Table I.1.2 Annual World Exports of Light Major Bulk Commodities (in millions of tons) Commodity 1970 1977 1979 Growth rate [1960(1)-1978(9)I Iron ore 321.7 357.3 390.7 2.1% Coal 102.0 136.8 155.5 3.8% Bauxite 27.2 33.9 - 5.4% Phosphate rock 38.5 48.5 - 5.5% Wheat 50.1 66.8 72.4 3.3% Corn 28.9 55.1 74.7 8.3% Coarse grains 48.2 81.2 103.2 7.2% Rice 8.8 10.8 11.9 9.8% Source: World Bank Commodity Handbooks, Except Coal Coal-Coal Trade Transportation and Handling, CS Publications, Surrey, U.K. 1981. Table I.1.3 Total World Transport Requirements for Eight Major Bulks (1977) Commodity Distance Ton-miles Mean ship Number of No. ships in (n.m.) size (dwt) Voyages trade Iron ore 5,000 1,786 100,000 3752 390 Coal 4,700 - - - - Bauxite 3,750 127 60,000 2118 158 Phosphate 3,500 169 15,000 11320 183 Wheat 5,300 383 28,000 2590 294 Corn 5,300 395 28,000 2670 303 C. grains 5,300 547 28,000 3960 418 Rice 5,300 63 28,000 1190 135 Source: World Bank staff estimates - ~~~~SEABORNE TAEoM@ IRON ORE (TOTAL TRADE 303 MILLION TONNES,\ 1508 MMM (.000 MILLION) TON-MILES. &N~~~~~~~~~~~~~~~~~~~~~~~~~1 0 500 Kilometers ~~~IRON ORE SHIPMENTS IN MILLION METRIC TONNES0 . , - --- 100 49 Totol Trade in Million Tonnes EQUATORIA L SC AL E / 2 0 1491 (000 Million) Ton-Miles Tha map hasha, repaored yTe 1 p adr rhvGrIshsIahexcissisrlyIhe camrr eohhh 01 50 erny main routes are shown. Area figures>w hiune roDoaea reders ed ,,hhshIdrIh usned aIOheaIls IhrldoaAams IIhe nbsIesh'OOO 20 a t \ re towa/s inclu(ding smaller routes not -< r arl.s,r hs panrt rlhe Worid harrI andffarhe hrahosal Fe,rune carporahas. an psprnlw A -.~~ hw sprt/ on he heoal slab,s cl arr Ierhlerv or arTy eedreeoeFhln or acr esIashe al SUCh hesardahes sw eaaey L (~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~After Feorrnlcys. Oslol °O SEABORNE TRADE= 1o9 GRAIN ( TOTAL TRADE 206 MILLION TONNES, ) 1131 MMM ('000 MILLION) TON-MILES. t2 129(791) E A%N~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1214 GRAIN SHIPMENTS IN MILLION METRIC TONNE5° 0 500 Itilometees __ - ---100 49 Total Trade in Million Tonnes I * I~~~~~~~~~~~~~~~~~~910)IRIS EQUATORIAL SCALE ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~FRC 17TRA SCL/L--51491 (000 Million) Ton-Miles> Thes map has Deen p,eharred hy The WoOd hank's staff esseel to, the hoonche at \ |f n/ oinrutesoesowG5 ro{grs>u th. reader aBdn 5n'eXcbyh to, the eto,nrron of Thref World hOkn and thle hnerfnulbnal \¢ __ 20 ore totals including small/er routes not , Teaeoe Co,poahoo The denooatahohs osed and eye eeDanes shnfn on tre rhap rb hee 5 - - e, ot,the pald cThe Vtordd hankand thterwhhbeheau,ance Ca'porahen. any 1udmcnn \*-* s hw seOrteV -4~ en the Ieoat stands l anrr tenrr,toI a, ant' endd,Sentettt o acceptance et nnkh ho,dahaes (fe erly,Ol lAt,FoNDey,AOlN SEABORNE TRADE - 1981 COAL AL A AL'() NLI ON TONNF S 2MMM 00 MION) ON MI [S COAL SHIPMENTS IN MILLION METRIC TONNES 0 500 K,IorNeters / 100 49 Total Trade an M,lhloo Tonnes I I EOUATORIA0 SCALP 49) I'000 M~ll Too- M,IUS I I / L -- - o (~~~~I10049) Totc Mil Tode Ton- Millon _n e 50 rd dSenl;.CDi hWr8}H,,sYaCXslevyo seoveSwco ' On/y mion routes are shown Arec lhgures c| ..... :rraSz Tr tn 1/.: >< nd aaD osel on d,ardoo \' 20 ore tota/s ,nc/odsng smo//er routes notr A'qaAra t1 e lesoa S "d"AS A',AAA,f,,"'~ d,f te reSo y ol dlBY e dOf 3e Cs, c, anea f su'nch separately |After Feo-leys, Oslo) - 9 - Bulk commodities are usually basic or intermediate goods that require additional processing after transport in order to be useful. Grain, for example, requires storage, milling and packaging. Thus, bulk shipping is part of a much larger industrial process. The exact shipping service required depends on the industrial process itself, and on the size and location of processing facilities. For bauxite or nickel, the type of shipping required depends mostly on the relative cost of electricity, since this cost determines where processing plants will be. Depending on the location of processing plants, the movement of "aluminum" from mine to market could be done in a number of ways as shown in Table I.1.4. Table I.1.4 Possible Ways to Ship "Aluminum" from Mine to Consuming Market Area Product Value/ton Tons/ton Cubic feet (1978 $) of final per ton Product Bauxite $ 30.00 4.5 25 Alumina $ 140.00 2.25 28 Aluminum ingots $1,130.00 1.01 -- Plates & shapes -- 1.00 250 Source: The Outlook for Bauxite/Alumina Trade and Shipping, H.P. Drewry, London, 1984. Commonly bulk shipping is tightly integrated into the industrial process. In the steel industry, ore carriers may load cargo at only one mine and unload it at only one steel mill. Such stable trading patterns occur when the demand for a commodity is stable and its supply is from a small number of sources. In this case, with careful optimization of a bulk logistics system, large economies can be realized. - 10- Where demand is not steady or the source of material supply is not well determined, a large investment in dedicated facilities and ships cannot be justified because of the likelihood of low utiliza- tion. To serve such a market, a large sector of general purpose bulk carriage has evolved making it possible for one vessel to carry many different cargoes in its life, the cargoes which are carried depend on the vessel's position and the contracts available. Many grain trades operate this way. Much shipping capacity is then pro- cured on a voyage (or spot) basis. Table I.1.5 gives the percentages of cargo carried by short-term contracts in each of the major bulk trades. Table I.1.5 Percentage of Commodity Flows Moving in Voyage Chartered Ships Commodity Percentage Carried On Voyage Charter Basis Iron ore 3% Coal 4% Bauxite 2% Phosphate rock 3% Grain 34% Source: Dry Bulk Charter Market and Trends, H,P. Drewry, London, l983. Where economies of scale exist in maritime transport, it is im- portant to pay strict attention to the details of bulk logistics and, more important, to increase the degree of organization and integration. The transportation of kaolin (China clay) from the United States to Japan offers an excellent example. Such clay is used to coat high quality papers. Until recently kaolin was transported as a slurry from the mine to a small precessing plant where it was dried - ii - and bagged. The bags then were loaded on pallets into a general cargo ship and taken to Japan. In Japan the dry kaolin would be converted back to a slurry and pumped into storage ponds at the paper mill. Today slurry-carrying ships can be substituted for dry-bulk carriers and much of the intermediate processing is, as a result, eliminated. The kaolin example also illustrates how bulk trading operations need not be of immense scale. Only two vessels are required to carry the total export of kaolin slurry. The savings are indicated in Table I.1.6. Table I.1.6 Savings in Kaolin Transport Using Slurry Carriers Step Eliminated Approximate Savings per Ton* 1. Drying slurry prior to $50.00-energy cost bagging 2. Bagging $15.00-cost of bags $ 5.00-cost of bagging 3. Loading ship $60.00-stevedoring 4. Unloading ship $60.00-stevedoring 5. Bag slitting and $ 8.00-cost of operation reslurrification Total savings $198.00/ton * Assuming typical U.S. and Japanese costs Scurce- World'Bank staff estimates, 1983 - 12 - I.2 Bulk Trades of Developing Countries Bulk trades mainly provide the following functions in the world economy: 1) Raw materials are moved from point of origin to point of need permitting optimal location of industrial sites. 2) Fuel supplies are moved from areas of surplus to deficit areas (both crude and product movements). 3): Food supplies are similarly moved from surplus to deficit areas. In these trades, inexpensive bulk shipping allows movement of low value goods in large volumes. Shifting raw materials accounts for about half of the bulk trade in developing countries. Fuel transported in dry bulk is mainly coal, most of which is supplied by developing countries. Although this trade is still small, the growth of developing countries' coal trades is potentially high. Columbia, Venezuela, Chile, Brazil, Botswana and Zimbabwe all have significant coal reserves and could emerge as big coal suppliers. Korea, for example, is increasing steel production capacity and is also relying on coal-fired plants for power generation instead of oil. Approximately a third of bulk traffic is in food, it might be noted, principally, grains, sugar and oil seeds. Only a few developing countries are bulk exporters of food. These include Argentina (grain), Burma and Thailand (rice), and Brazil (sugar). In any event, both the import and export bulk trades of developing countries have been expanding rapidly. Figures I.2.1-2 indicate the total value and weight of major bulk exports of developing countries. The major growth has been in "Metals" which includes ores. Though agricultural commodities not shipped in bulk (i.e. coffee, tea, cotton, etc.) together with non primary goods, represent 85 percent of developing country exports by value, the concentration of metals, timber and - 13 - FIGURE I.2.1 VALUE OF SELECTED BULK EXPORTS OF DEVELOPING COUNTRIES (Million USD) 40,000 37,500 PHOSPHATE 25,000 .METALS GROUND NUTS & VEGO OILS 12,500 1960 64 68 72 76 1980 Source: Commodity Trade and Price Trends. The World Bank, Washington, D.C. 1982/83 edition. - 14 - F IGURE I.2.2 WEIGHT OF SELECTED BULK EXPORTS OF DEVELOPING COUNTRIES (1000's tons) 4001000 300,000 200,000 X ~~~METALS 100,000 GROUND NUTS &0 I I L S / f~~~~~~~~UGAR / ~~~~~~~GRAIN 1960 64 68 72 76 1980 Source: Commodity Trade and Price Trends. The World Bank, Washington, D.C. 1982/83 edition. -1S grains in certain countries make these commodities especially- important to the economies of those nations. Figure I.2.3 (from World Bank data) presents the total weight of the major bulk imports to developing countries. ¶he pxincipal commodity carried and also the one with the highest growth rate is grain. Because imported grain is used as a means to compensate for shortfalls in the domestic crop, the amount of grain which a particular country imports can fluctuate with domestic harvests. As a result, while world grain shipments may present a pattern of steady gxoith, a given nationts grain imports can vary considerably from year to year (see Table 1.2.1). Table I.2.1 Wheat Imports of Major Developing Countries (in thousands of tons) Importing Nations 1961 1965 1970 1976 1977 1979 India 3,090 6,572 3,586 6,289 852 300 Brazil 1,881 1,876 1,969 3,428 2,624 3,654. Egypt 661 1,230 850 1,930 3,346 3,608 Korea 336 476 1,178 1,787 1,989 1,695 Pakistan 1,079 1,515 228 1,185 497 2,236 Bangladesh 252 241 1,061 1,049 623 1,123 S. Europe 3,248 2,273 2,073 1,723 1,603 2,081 China 2,889 5,626 5,583 2,598 7,577 8,951 Other 4,211 5,137 7,400 13,370 15,255 18,818 Developing Countries (Total) 17,647 24,946 23,928 33,359 34,369 42,466 World (Total) 38,511 49,927 48,785 65,675 64,236 76,039 Source: Grains Handbook, The World Bank, Washington, D.C. 1982. Some countries, particularly India, import less grain today than in 1961 because the performance of the agricultural sector improved. One notable fact emerging from an analysis of the table is that the grain trade is growing more diverse--with the older principal importers buying a - 16 - FIGURE 1.2.3 WEIGHT OF PRINCIPAL BULK IMPORTS OF DEVELOPING COUNTRIES ('1000's tons) 120,OOO PHOSPHATE ROCK METALS GROUND NUTS 90,000 / SUGAR 60,000- 30, 000- GRAIN 1960 64 68 72 76 1980 Source: Commodity Trade and Price Trends, The World Bank, WashlingtQn, D,Cj 1982/83 edltrig. - 17 - smaller proportion of the grain moving in international trade. In 1966, 11 percent of the world's grain trade was shipped to "other developing countries" while, in 1979, the percentage reached 25 percent. The conclusion is inevitable that grain today is being shipped to more diverse destinations. The ability of the trade to support these "other developing countries" by the use of large vessels must be carefully assessed. I.3 Interests of Developing Countries in the Bulk Trades Developing countries have two primary interests in owning and operating bulk shipping. The first is profits and foreign exchange earned from trans- portation. The second is the impact of domestic control. Foreign exchange earnings from bulk shipping operations are the difference between the cost in foreign exchange of using foreign tonnage versus domestic tonnage. These are estimated in Table 1.3.1 Table I.3.1 Illustrative Foreign Exchange Savings from the Operation of a 60,000-ton Bulk Carrier Built Overseas (in USD/YR) Cost Component Payments to Others Developing Country Expenses Wages 1/ 571,000 Subsistence 1/ 90,000 Stores 1/ 91,000 M&R 2/ 457,000 Insurance 2/ 422,000 Port & Canal 508,000 Fuel & Lubes 3,086,000 Capital 3,683,000 Subtotal 7,734,000 1,174,000 Profit (10%) 900,000 Totals 7,734,000 2,074,000 Percentage 79% 1/ Sometimes as much as 40% of wages, and substantial portions of subsistence and stores are paid for in foreign exchange. 2/ This assumes no domestic repair capability and a self-sufficient insurance industry. Source: World Bank staff estimates, 1983. - 18 - for a 60,000-ton bulk carrier acquired overseas and operated by a developing country. The 10 percent profit on gross expenses is an arbitrary figure. As many shipping operations are currently losing money, this is optimistic. Most crew expenditures will be in local currency, although it is common to pay at least 40 percent in foreign exchange. Most vessel operating expenses, such as fuel, maintenance (spare parts and shipyard fees), port and canal tolls, and capital costs, will be paid in foreign currency. The money spent on marine insurance could go either way. If a country does not have a marine insurance industry, the insurance costs are foreign exchange expenditures. If a country does have a marine insurance industry, foreign exchange risk would still remain since the place where a given vessel is to be repaired cannot be predetermined. In any case, the maximum savings in foreign exchange is established at about 20 percent. Minimum savings would result if profits were zero and marine insurance was counted as a foreign exchange expense. Were this the case, foreign exchange savings would only be about 10 percent. Table I.3.1 should be considered an optimistic assessment. The theoretical price of bulk shipping (basically a free enterprise) is the marginal cost of the least efficient ship employed in the trade. Earnings from ship operation are the difference between the total cost of the ship operation and revenues, which are earned at the marginal cost price set by the market. It is not possible to determine if a developing country will make earnings without considering market activity and determining what the marginal ship's costs are. This is a consideration given further attention in the section on ship procure- ment. - 19 - Flags of convenience allow extremely low cost ship operations because of liberal certification, crewing, and tax policies. No country can have a cost advantage over flags of convenience, such as those of Panama and Liberia, - unless capital and insurance costs are less. The remainder of the cost factors-- spare parts, crew, and port charges, will be no more for the vessel registered under a flag of convenience than for one that is registered domestically. Insurance costs depend on the loss history. For established maritime nations, such as India, insurance rates would be comparable to world rates. In any case, a developing country's cost of insurance (or risk) is no lower than current world rates. Subsidized ship financing is available for developing countries that wish to purchase newly-constructed ships. Such subsidized financing is usually about 3 points below the going OECD ship financing interest rates with terms- 50 percent longer and loans of up to 90 percent of a ship's cost. An initial two-year holiday on principal repayment makes secondary financing possible, so little or no cash is required. To the extent that this financing is available, it offers developing countries a large competitive edge in ship acquisition financing.3 There are many who maintain that soft-term financing offered by developed countries for new ship sales to developing countries is largely the result of a lack of shipbuilding orders in developed country shipyards. No good reason exists to subsidize ship sales to developing countries in this manner were the demand for ships higher and the yards working at close to full capacity. Developing countries have major interests in ensuring that a bulk logistics system plan, before it is undertaken, is in accordance with their development goals. A country can improve its economic position, either by improving the continuous stream of real income resulting from trade, 3/This is becoming less so as developed countries cut back support for the shipbuilding industry. - 20 - or by reducing the continuous real expenses. Investments in bulk shipping can achieve both to some degree. However, such investments should be made in areas where the country has the greatest comparative advantage, and this may or may not include bulk shipping. All these factors influence a country's bulk commerce in various ways. When countries with large raw material resources do not possess a comparative advantage in processing, the expansion of bulk export facilities may be the best way to increase income from those resources. This is the case with bauxite, since aluminum smelting is localized in those areas of the world where electricity is inexpensive. Improvements in bulk export logistics can provide increased income which can then be invested in areas where a greater return is available. Otherwise, when the country with raw materials has the comparative advantage in its processing, exporting raw materials would subtract from the potential income. The type of shipping capacity provided must then be of a different technical nature. Developing countries as a group are not heavily dependent on bulk exports for foreign exchange earnings. However, some specific countries are. Figure I.3.1 compares the distribution of bulk versus other exports by value of nine developing countries involved in bulk trades. In addition, Chile, Bolivia, Jamaica, Burma, Fiji, and others are in a similar position. As shown in Table I.3.2, bulk exports can account for between 60 percent of per capita income (Liberia) and 3 percent (Argentina). Even for a large country, such as Argentina, for which bulk export earnings are not vital from a welfare standpoint, bulk exports account for 20 percent of its international trade. - 21 - FIGURE 1.3.1 DISTRIBUTION OF EXPORTS BY VALUE (Millions of USD) Argentina B-ulk 1Other |7421 Guyana -Sulk IOther 326 Guinea Bulk| Other l 3311 Peru Bulk|Other 2577 Liberia Bul l Other 542 Mauritania Bulk |Other 157 Zambia Bulk Othe41216 India Bulk Other 7743 Thailand BulkTOther 5289 Source: Commodity Trade and Price Trends The World Bank, Washington, D,C., 1982/83 Edition. - 22 - Table I.3.2 Comparison of Bulk Exports to Indicators of Country Welfare (1980 figure) Country Population Total % Bulk Value Per Capita Per (million) Value of Bulk Bulk Value Capita Exports (million (USD) Income (million USD) (USD) USD) Argentina 25.7 7,421 19.8 1,469 57 1900 Guyana .8 326 79.8 260 332 620 Guinea 4.5 334 58.3 194 43 150 Peru 16.9 2,577 47.7 1,228 73 700 Liberia 1.8 542 72.6 393 218 360 Mauritania 1.3 157 79.8 125 97 220 Zambia 5.1 1,206 89.5 1,079 212 390 Thailand 43.0 5,289 32.8 1,734 40 358 Source: Per Capita Income Figures from The World in Figures, The Economist, London, 3rd Edition, 1981. When countries are heavily involved in bulk exports, the logistics system can be highly developed and technically efficient. Frequently, limitations in the port facilities at the other end of the trade limits the volume and size of shipments. Improvements in the bulk logistics system regarding that end of the process are outside the area of the developing country's control. One way to improve control over the bulk exports is to deliver the cargo CIF at the port of discharge instead of FOB at the mine's shipment facility. Transportation costs can represent between 30% and almost 100% of the value of bulk commodities. Hence, careful planning in the acquisition of shipping capacity is necessary. As time charter rates are now well below cost in many trades, substituting time charters for a portion of the present shipping requirements presents a possible improvement over acquisition of new tonnage. - 23 - A country that has a competitive advantage in processing may find that bulk facilities are not always suitable for exports of semi-finished materials. There are a number of reasons for this. Total volume of the finished products may exceed that of the raw material, thereby vastly increasing transportation costs, loading process may be different, and the ship itself may need to visit more ports for discharge as it becomes part of a distribution channel for the product. This happens to be the case when a country substitutes the export of plywood, for instance, for the export of logs. A tendency exists to substitute liner vessels (inctluding container ships) for bulk carriers to ship the product. There is a large difference between the cost of using liner and bulk ships. Experience shows that the former can sometimes consume much of the industrialization project's proceeds. A more economical course of action is to perform intermediate distribution where possible with dedicated bulk vessels. This has been done successfully with plywood in ocean-going barges. One area of weakness in developing countries' bulk logistics is in the importation rather than in the exportation of goods. Many opportunities exist throughout the world to reduce the cost of food and raw material imports by improving bulk logistics systems. Grain is a trade where large improvements are possible. There are efficient grain discharge facilities in some developing regions, but the typical port facility is inadequate because of slow discharge rates. Port time could be easily reduced and at reasonably low cost. A recurring problem occurs when the grain facility is just too small to handle the quantity of grain aboard the ship. Even when a slow discharge process is being used, the ship often must wait for the grain ashore to be moved to - 24 - other locations before unloading continues. Grain can be carried most cheaply in the largest of shiDs, and the volume of the trade occasionally jugtifies such shipments. Figure I.3.2 shows the breakdown of vessel sizes used in the grain trade. Significantly, water depths available at many ports used for grain imports are insufficient to allow the ship to enter. Ships must often be lightered with overall operations usually poorly organized. Little thought is given to invest- ment in equipment truly suited to the job. The bulk discharge rates in offshore lightering operations is often as low as 2-3,000 tons a day or even less. With proper equipment, rates of 10-20,000 tons per day can be reached. However, when designing grain import facilities the variability in grain imports must be considered before large in- vestments are undertaken. I.4 The Role of Ports and Harbors in the Bulk Trades Adequate ports and terminals are crucial to the efficient organiza- tion of bulk logistics operations. When terminals are inadequate, only labor-intensive handling techniques are possible (which frequently involve long delays due to periodic cessation of cargo handling when storage ashore is lacking). When vessels are handled without modern port facilities, the most frequent cause of inefficiency is a lack of resources. A shortage of cargo storage ashore, in particular, makes optimal planning and organization of the cargo handling operation impossible. In fact, historic long delays in bulk operations are due practically to a lack of planning and over-estimation of the port resources available. Even at primitive harbor facilities the large expenses of queues cannot be averted if cargo is not removed faster - 25 - FIGURE I..3. 2 VESSEL SIZES EMPLOYED IN THE GRAIN TRADE Percentage 100 - 90 60 10| . . . . . . . . . . . . . ...... .. ,X\\\\\\\\\\ 70 ////-. . . . . . . . . . . . . . . .*v 60 ......60. 40 .... ........ ................ . ......................... ..... ..................................... Sorc: or evlomet, U~Ae................. . , , 1966 .... ............... 196i 196 1999 197 8 Vessel Sizes~~~........ 60,000 DwI~~~~~~~~~~~~~~~~~~~~~~............... El and above B 18/25,000 ~~~~................................. 0 40/60,000 DWr 3 below 18,000 DN~.......................... 30 .................. Sorc : otDvlpnn,U0A ertra 10~ ~ ~~~Ute &R:nJN) ok 98 - 26 - than it is discharged. Modern harbor facilities with large-scale, fixed ship loaders and unloaders are at least ten times more productive than undeveloped sites. While loading and unloading methods vary, one common method is to fill bags in the hold of the ship and then discharge them using the ship's gear. This is, of course, an expensive operation considering ship turn- around and berth occupancy costs. Generally cargo handling gear mounted on bulk carriers is much less efficient than shore-based gear. Cargo gear mounted on a ship gets in the way of fixed ship loaders and unloaders. Consequently it is hardly desirable on a vessel which is frequently handled at modern facilities. Modern high capacity self-unloading ships, however, are an exception. Unfortunately most of the large bulk carriers are not fitted with any cargo handling equipment. Trades where modern equipment is not available must be conducted in small, more expensive vessels even though the draft in the harbor allows the use of a larger vessel. This can be very inefficient indeed. The scale of bulk facilities must be roughly that of the ship itself. Except for proper organization of existing facilities and an increase of storage ashore, there is not an "entry level" facility in the bulk trades. Small facilities are feasible in container trades because mobile cranes make efficient, small-scale operations possible even with the largest container ships. Due to the scale of these bulk facilities, their cost is largely a fixed investment. Unit costs are higher if the facility is under- utilized. The location and demand for bulk facilities is very important. In situations where the utilization of bulk facilities is not certain - 27 - such facilities present a risky investment and that risk should be carefully evaluated. This means distribution systems development for bulk cargo must be integrated with large-scale port investments. I.5 Bulk Shipping Total bulk shipping trade in international commerce grew from a level of 512 million tons in 1970 to over 801 million tons in 1980 indicating an average annual growth rate of 4.3%. Since 1970 the major bulk trades have passed through two major cycles with sharp in- creases in volumes between 1972-74 and, again, between 1978-80. The only major commodity which remained nearly static in volume and, therefore, declined proportionally (between 1970-82) was iron ore as indicated in Figure I.5.1. Coal and grain trades have had the largest and most consistent growth. Generally it is assumed that the volumes of major bulk commodities traded are functions of the following: a) Political Stability b) World Economic Growth Rates c) Relative Energy Demand d) Confidence in the World Economy e) Degree of Freedom of International Trade As a result of its dependence on these conditions bulk trades are usually fragile and subject to major changes in the volume and direction of trade. Past long-term bulk trading relationships no longer define world bulk trades. Now most bulk commodity importers are continuously seeking more attractive sources of supply with lower delivered costs, which, in most cases, include significant ocean transportation costs. - 28 - FIGURE I.5.1 MAJOR BULK COMMODITY TRADES 1000 E Phosphate Rock 900 E Bauxite Alumina Grain 800 - Coal EIron Ore 700 - 600 z500 300 200 100 1971 1973 1976 1978 1980 19826 Source: World Bank, - 29 - Considering the medium range future of major bulk trades, total volume of world trade in iron ore is expected to grow from 304 million tons in 1981 to 356 million tons in 1990 for an increase of nearly 16.5% over a 10 year period. The coal trade is expected to grow from 204 million tons in 1981 to 322 million tons in 1990 for an increase of nearly 55%. Grain, bauxite/aluminum and phosphate rock are all expected to grow only marginally by about 20% over the same period with total world dry/bulk shipping trade growing from about 800 million tons in 1981 to 1,040 million tons in 1990 - a growth of under 30% over the ten year period. Table I.5.1 Bulk Carrier Fleet Development (in thousands of dwt) Vessels over 10000 dwt Figuiresin number of ships and '000 dwt Tankers Combined Bulk Carrier Total Year Mth. Carrier No dwt No dwt No dwt NO dwt 1973 1.7 3223 201419 327 33023 2672 84358 6222 318800 1974 1.1 3293 215574 355 37415 2781 89393 6429 342382 1.7 3383 234162 373 40221 2868 93132 6624 367515 1975 1.1 3406 254327 386 42081 2992 97812 6784 394220 1.7 3398 272879 392 43443 3094 101873 6884 418195 1976 1.1 3439 290891 398 44208 3197 105749 7034 440848 1.7 3398 306627 405 45445 3311 110212 7114 462284 1977 1.1 3384 320531 414 46808 3454 116586 7252 483925 1.7 3339 327339 419 47545 3662 123735 7420 498619 1978 1.1 3301 331940 419 48273 3826 129629 7546 509842 1.7 3184 329886 417 48722 3930 133516 7531 512124 1979 1.1 3129 329657 418 48589 3960 134931 7507 513377 1.7 3096 327603 415 48779 4017 136856 7528 513238 1980 1.1 3071 326836 410 48179 4020 137657 7501 512672 1.7 3079 326785 411 48376 4056 138875 7546 514036 1981 1.1 3081 324706 401 47266 4116 142058 7598 514030 1.7 3085 322387 404 47393 4198 147234 7687 517014 1982 1.1 3084 320158 385 45250 4316 154713 7785 520121 1.7 3011 310689 380 45078 4438 161796 7829 517563 1983 1.1 2944 300923 362 43145 4545 169231 7851 513299 1.7 2861 289768 363 43248 4630 174331 7854 507347 Source: World Bulk Fleet, Fearnleys, Oslo, August 1983. - 30 - To discuss bulk shipping capacity or supply in relation to bulk shipping demand, ton-mile demand is of especial interest. Total ton- mile demand increased from 10.6 billion in 1970 to 16.78 billion in 1980, but fell shortly after to only 14.19 billion ton-miles in 1982. Ton-miles of crude oil and product movement remained almost constant between 1970 and 1982. Over the same period, however, ton- mile demand of coal and grain more than doubled. The world bulk fleet (both tankers and drybulk carriers) increased from 6,222 vessels in 1973 to 7,854 vessels in 1983, or 26%, while total carrying capacity increased by 58% over the same 10 year period as noted in Table I.5.1. However, it should be noted that the fleet capacity increased to 510 million tons (dwt) by 1978 and has since remained at a fairly constant size. Tanker capacity has declined by over 40 million dwt since 1978 while dry bulk carrier capacity increased by about 40 million dwt over the same 1978- 83 period. According to the size of the dry bulk carriers the largest increase is noted in the 60-80,000 dwt class. In 1965 only seven ships with a combined capacity of less than 500,000 dwt were available in this size class. This has now grown to over 430 vessels with a capacity of 28.7 million dwt. There is a similar, though more recent increase in the number of 100-150,000 dwt bulk carriers of which there are now 188 with a total capacity of 23.4 million dwt. The largest component of the fleet, however, remains the handy-sized 25-40,000 dwt carrier with 1,727 vessels with a carrying capacity of 53.6 million dwt or 43% of the total. The average age of bulk carriers is 9.3 years with carriers of sizes less than 40,000 dwt having an - 31 - average age of 10.98 years and larger carriers having an average age of 7.9 years. Considering ownership of the bulk fleet, it is noted that 57% of the world bulk fleet by number and 56.2% by dwt is registered under the flags of Liberia, Greece, Japan and Panama; 28% of world bulk carrier capacity is registered under the flags of Liberia and Panama. Japan has only 8.3% of the bulk carriers under its flag, but controls over 11.5% of bulk carrier deadweight, as the average size of Japanese- registered bulkers is significantly larger than the world average size. Considering tonnage on order in 1983, 396 out of 562 bulk carriers were ordered from yards in Japan (344) and Korea (52). The average size of bulker on order in 1983 was 43,263 dwt. With regard to future fleet developments, it is expected that total bulk carrier fleet capacity will increase from 174.3 million dwt in 1983 to 190 million dwt in 1986, 210 million dwt by 1990 and 252 million dwt by 1995. An increase in the number and fleet capacity of vessels of 25-40,000 dwt is expected over that period with a lesser increase in the 60-80,000 dwt class. Important considerations in organizing bulk shipping which must be studied before decisions are made on ship acquisition, chartering and various aspects of bulk shipping operations can be summarized as follows: 1. Strength of Demand for Transportation * Political and strategic considerations * Seasonal consideration * Demand for shlipping of other commodities * Demand for shipping of same commodity in other regions - 32 - * Long-term changes in production, consumption and method of distribution of commodities 2. Supply of Vessels * World order book * Utilization of the current fleet * Long-term changes in the composition of the world fleet 3. Ship Market Conditions * Current spot and term charter rates * Aggregate of past market behavior * Volume of market activity * Volume of scrapping and laid-up vessels 4. Expectations and Forecasts * Forecasts of market level, operating costs and shipbuilding prices * Availability of backhaul arrangements 5. Nature of Commodity Using Transportation * Type, grade and quality * Value per ton * Seasonal or nonseasonal 6. Control over Commodity Source * Degree of outright control * Ability to schedule production and shipment 7. Control over Commodity Use * Nature of use (continuous or intermittent) * Ability to schedule use * Volume of turnover * Size of consignment acceptable 8. Availability of Alternate Forms of Transport - 33 - 9. Cost Related Variables 10. Impact of Transportation Costs on Specific Parties * Percent of transportation cost of total operating expenses * Percent of transportation costs in raw materials costs 11. Expected Cost of Product Shortages * Cost of slowdown or shutdown * Loss of sales or goodwill * Probability of shortage under various arrangements 12. Costs of Raw Material Shortage * Cost of physical shortage * Cost emergency procurement * Inventory holding cost * Risk of spoilage or obsolescence - 35 - II. FORECASTING IN THE BULK TRADES II.1 General Comments Forecasting is the evaluation of future prospects based on past experiences and expectations for future occurrences. While producing a "perfect" forecast is desirable, many well-designed, useful forecasts have been "wrong" because their basic assumptions were later found to be incorrect. Examples are changes in the price of oil, the closing and opening of the Suez Canal, and the sizes of grain harvests. Frequently the accuracy of a forecast is immaterial, as the goal is to make the best possible decision with the information available. A useful by-product of forecasting is the research performed in areas where information is lacking. It is not uncommon to find that the in- creased understanding which results from this research is more valuable than the forecast itself. A common assumption in forecasting is that the best indicators of the future are the present and the immediate past. While in many cases this is a sound assumption, this need not be true in every case. Often "trends" can be started by improvements in statistical collection procedures. In developing countries this is frequently the case when automated data processing is implemented for the first time. It is also possible that a goal of a project is to change the structural organization of a particular sector or industry. Examples of this are investments in export-oriented industry, transshipment ports or implementation of large industrial projects well beyond the historical scale of past development. In situations such as these history has little place in the "forecast" and a more market-research- oriented approach should be attempted. - 36 - Five fundamental issues always arise when establishing the basis for a forecast. These include the following: 1. How much should the present and immediate past count in predicting the future? 2. How accurately does available information describe current and past situations? 3. What mechanisms are at work undermining the relationship between the past and the future? 4. What specific plans exist to alter the structural variables affecting the forecasted quantity? 5. What are the uncertainties present in the basic assumptions? The treatment of uncertainty is important. For much maritime information, such as charter rates, the random portion of the quantity is so large that, even if correctly estimated, the trend portion may have little utility in planning. Here, establishing upper and lower bounds on the variable may provide sufficient planning information. Monte Carlo simu- lation can also be of assistance; the IBRD Staff occasional paper number 11, "Risk Analysis and Project Appraisal," by Louis Pouliquen, provides a lucid treatment of this approach. In cases of poor project performance, many problems have their origin in inadequate forecasts. A frequent occurrence is for forecasts to overestimate the revenues of a project. A World Bank study found that the most important reasons for this were the following: * Slower than expected growth of the basic economy or particular sector * Incorrect assumptions based on too static a view of the relation- ship between transportation requirement and production * Insufficient allowance for competition from other modes * Poor operating performance of the operating authority Another factor causing an overestimation of the returns from projects - 37 - is that, once sufficient resources are available for a large scale project, the goal of the forecasting effort becomes more a matter of justifying the project than of evaluating it. Such forecasts are not unbiased assessments of the future. Still another difficulty with forecasts is that, generally, they have both an engineering and an economic purpose. From an engineering standpoint the maximum or potential use will probably establish design criteria for the facility. This is because the savings from constructing facilities too undersized to meet maximum demand may be outweighed by the costs of congested facilities. From an economic standpoint the forecast must predict the average revenues generated by the project and hence determine its economic viability. This requires an estimate of the expected market size of the project and its risk. For some projects the market size cannot be determined by the present situation. It is the rate of market share growth of the project which determines its basic viability. Here, market research plays a crucial role in project evaluation. To prepare such a market share forecast a marketing plan must be prepared to estimate the volume of revenue generated by the facility. This should include measures which should be undertaken to market the services of the facility and which should consider the design elements that will expedite the marketing effort. When a dedicated marine facility such as an ore or coal terminal is under consideration, the marketing plan should be for the commodity itself as well as the trans- portation demand derived from it. A final problem affecting forecast accuracy is the late completion of the project or, even worse, of a single vital element of the project when everything else is on time. - 38 - II.2 Predicting Potential Volumes The potential volume of cargo to flow in a bulk logistics system is important because it helps to set the optimal design capacity for the system and places an upper limit on the revenue generated from the use of the facility. Potential volumes are generally determined from forces exogenous to the bulk logistics project. These can be the countries' Gross National Product, population growth, or developmental projects. The exception to this is the cost of transportation of the commodity which, of course, exerts a substantial effect on the situation. Efforts to forecast cargo flows tend to fall into three categories: trend extrapolation, model building, and policy (or opinion) eapture. There are many technical methods available for each group. These are fully summarized in Table II.2.1. Trend extrapolation, pattern identification and probabilistic forecasting are all based on series of historical data that are analyzed in various statistical ways to arrive at forecasts of the future. In general, these techniques are the most commonly used and are the most comprehensible of forecasting methods. The second group includes dynamic models, cross-impact analysis, KSIM, input-output analysis and policy capture. These methods are based on models or simulations of the phenomena to be forecasted. Be- cause they demonstrate the interactions of the separate elements of a system as well as their combined overall effect, they are called structural models. These models are helpful in attaining a broad per- spective and better grasp of the totality of a problem, in foreseeing effects that might otherwise be overlooked and in anticipating public reaction to alternative problem solutions. Table II.2.1 Elements of Bulk Logistics Forecasting Technique Input Output Span Example Trend historical data array of time (No specific span) population estimates Extrapolation series forecast Pattern historical data array of time short medium commodity demand Identification series forecast Probabilistic historical data various sets of probability (No specific span) risk analysis Forecasting matrices decision tree, etc. Dynamic historical data results of various alternative long water resource planning Model growth and decline of the system variables Cross Impact events and their table showing interaction (No specific span) estimates of impacts on the Analysis effects among items project from occurrence of lower population rate KSIM variable and its numerical forecasts long impacts on deep water value ports from new policy Input-Output sectors and amount of input-output table showing short medium estimates of output increased Analysis transaction interrelationship among by additional demand sectors Policy preference of graphics illustrating CNo specific span) measurement gf the relatiye Capture participants weights and functions preference of individuals for competing issues Scenario data and information formal documents (definition, long scenarios for future base assumption, data, findings, etc.) growth pattern Expert Opinion opinions of experts varies (No specific span) future of American water Method resource Alternative historical data varies medium long alternative future of inland Futures waterway traffic Values survey data document (generalization (No specific span) forecasts of changes in Forecasting and prediction about behavior) peoples' lifestyles Source: Handbook of Forecasting Techniques, Stanford Research Institute, 1975 - 40 - The last group includes scenario, expert opinion method, alternative futures, and values forecasting. This kind of forecasting tends to be more global, more qualitative, and "softer" than more conventional approaches. In general, these are the least developed of the forecasting techniques. Policy capture methods are based on the presumption that policies and plans tend to come true and that determining what they are and what their goals are provides a sound base for a forecast. A variant of this is the expert opinion method. This method assumes that experts in the field can base their opinions on extensive knowledge of the field and hence make good forecasts. Modelling can be extremely productive because it encourages thinking about the causes of events. This can be invaluable, especially if it can be communicated outside the model which is being created. One way to do this would be to provide a complete statement of the model's assump- tions in simple language. Models frequently rule out relationships. One might suppose, for example, that orders for new ships are related to the average age of the fleet. The idea being here is that new ships are ordered to replace worn out ones. All attempts to model the shipbuilding market have shown that this is not the case; rather, new orders are closely related to current revenues. II.3 Market Share Estimates A product's market share is the percentage of the total market serviced by that product. Major markets are generally defined by the type of product (such as fuel oil) and submarkets by areas of differing -elasticity of demand. Thus, the ship fuel market and the home heating - 41 - oil market are different submarkets because of the different elasticity of demand for the product, even though the product itself is the same. Depending on the circumstances, a bulk logistics system's market share may be based on the total market or on a submarket. In discussions of market share it is important that the market under consideration be rigorously defined because of a possible confusion of market share growth with a total expansion of the market. Confusion between growth of market share and expansion of demand is common, especially in Marine Transportation. For example, people who view the container trades as a self-contained market would tend to see good prospects for expansion based on the historic rapid increase in container traffic. While those people who view the fundamental market as general cargo would see that the volume of general cargo in international trade is growing slowly. Also they would note immediately that the limits of expansion of the container trades were being rapidly approached despite the high growth rate. Failure to see this point clearly has caused much overinvestment in ships of nearly every type and in the shipyards where they are built. Market shares change for a variety of reasons. A major reason is that of diverted demand. Examples of diverted demand are the sub- stitution of container vessels for general cargo ships and the penetra- tion of bulk carriers into the grain trades, again at the expense of general cargo ships. In many cases, this diverted demand results from dramatic technical innovation of ship type presenting vastly different requirements for port capacity. In the case of the introduction of the container ship, VLCC, and large bulk carrier, older port facilities became unusable and totally new ones were required. Estimating the rate at which this technical progress occurs is of some importance in making - 42 - the proper decision about the timing of the building of new port facilities and ensuring that surplus facilities are not built because of an over- estimate of the demand for the new facility. One of the factors which differentiates bulk shipping from liner shipping is that technical innovation occurs at a faster rate than in the liner trades. For instance, it required fifteen years for containers to become commonplace after proving the concept in the trades between the United States and Hawaii and Puerto Rico. By way of contrast, the concept of the slurry carriage of ore was deployed in less than five years and the expansion of maximum tanker sizes from 50,000 tons to 250,000 tons in less than seven years. The reasons for this are not entirely clear, but it is likely that the structure of the market for the services of these ships plays a major role. The liner trades are highly cartelized while the bulk trades more closely resemble a free market. This distinction is not always clear-cut as trades in many bulk commodities such as grain and gypsum are highly concentrated (even though the sources of the ships themselves may not be). Consider Figure 11.3.1. Here the cost of installing new technology is given by a cost curve as a function of time. The cost of earlier implementation is higher because of increased technical risk, crash programs, the cost of scrapping obsolete equipment, etc. The cost far in the future rises primarily beause of the cost of maintaining a capability in unused technology. Assuming that investment decisions are made to maximize the surplus of revenues over cost, the timing of the deployment of the technology depends primarily on the benefits. Three benefits curves are postulated. - 43 - FIGURE II.3.1 COSTS AND BENEFITS FROM NEW TECHNOLOGICAL INNOVATION 0 N E Y V Benefits of Monopoly or cartel V Benefit for few competitors C1 Cost of implementation of new technology V3 Benefit with many v 3 competitors < V INNOVATION TIME DEPLOYED - 44 - The top V1 is the base and represents gains accruing to a monopoly or tightly controlled cartel. The time when the new technology is deployed is the point where the difference between the two curves is at its greatest. The curve is downward sloping because of the increased time available to reap benefits if the technology is deployed earlier and possibly because of the improved competitive position of the firm. When more firms are involved in the trade, the benefits curves will be shifted down because of the decreased market share and will have a steeper slope because earlier deployment increased a firm's market share at the expense of its competitors. This benefits curve is denoted V2. Note that with an increase in the number of firms technical innovation occurs earlier. As the number of firms increases, the combined shifts will result in a benefits curve V3 for which the introduction of new technology results in losses. In this case firms will abandon the trade if the losses are large enough or deploy the new technology even earlier to minimize losses. It is likely that the liner industry is represented by curve VI, the bulk shipping industry by either curve V2 or V3, and the tanker sector by curve V3. From this standpoint it is probable that innova- tion in the bulk shipping industry will begin earlier and be completed more rapidly than in the liner sector. This indicates that basing forecasts on the historic rate of adoption of a new idea in the bulk trades is risky. The actual popular- ity of the idea may be misjudged by the short-term rate at which it is deployed. In other words, one should be cautious about adopti,ng it too quickly. This note of caution applies especially to decisions tegarding the dredging of ports for extremely large ships. - 45 - II.4 Bulk Logistics Forecasting The goals of bulk logistics forecasting can be divided into sector planning, bulk shipping planning, and bulk port planning as follows: A. Sector Planning (a) Provide estimate of commodity sales or purchases (b) Provide estimate of income or expense projections (c) Provide basis for investment in industrial facilities (d) Provide estimate of transportation requirements by mode B. Bulk Shipping Planning (a) Provide projection of shipping requirements (b) Determine vessel types and sizes required (c) Provide information to plan procurement of vessel * Purchase new * Purchase used - Time charter * Spot charter C Contract of affreightment (d) Estimate cost of providing marine transport C. Bulk Port Planning (a) Depth and geometry of dredging (b) Determine areas in port devoted to specific activities (c) Determine number and type of berths required (d) Determine number, type and size of cargo handling equipment procured (e) Determine port interface and support requirements (e.g., generating capacity installed in the port) (f) Determine the market serviced by the port Demand mechanisms in the bulk trades fall into three general categories: * Normal - determined by existing mechanisms and growth rates * Diverted - determined by competition * Generated - determined by industrial development - 46 - These three mechanisms produce different growth patterns. Normal growth assumes that general demographic factors such as GNP or population are good indicators of traffic and cargo flow. This could apply to auto- mobiles, building materials, etc. Normal growth assumes that trade growth is independent of investments in the logistics systems. Figure II.4.1 offers a general method to prepare a forecast for a bulk port facility. The goal of this procedure is to determine, first, what the potential market for the bulk trade is. Then, to evaluate the actual commercial prospects of servicing the trade and, finally, to use this information to estimate actual sales. In step one a naive forecast is made using the general techniques described in the section on predicting potential volumes. In step two the existing production capacity is inventorized and evaluated. A similar exercise is performed in step three for planned industrial development. Projections are then made for future consumption and production and subsequently checked for plausibility by determining if sales or purchases of the difference are possible on the target markets. This, then, is used to make a raw forecast which is used as a base for the marketing analysis. Marketing targets, next, are established and a formal marketing plan is developed to ensure that the marketing targets are reasonable. This marketing plan is translated into a revenue estimate which can be used for economic project evaluation. The pro- jected maximum flow achieved when marketing efforts no longer yield increases in sales is used as a basis for the engiiteering design of the facility. In essence the method consists of extrapolating past data and - 47 - Figure II-4.1 Elements of Bulk Logistics Forecasting Analyze recent trends and produce simple extrapolation for the period in question | ssess internal production potentiall Identify development and investment plans FProject future production Project future consumption or demand |Calc-ulate surplus (deficiency)l jDetermine if it can be marketed (or bought)l Make forecast Establish Marketing Targets . Investigate feasibility of meeting marketing targets Revise forecast Make revenue estimate - 48 - checking the plausibility of the extrapolation. There are a number of methodologies that produce valid extrapolations when the problem meets certain assumptions. Not all of them give the same answer, so the selection of the method used is highly important. Figure II.4.2 is a schematic representation of the process used in developing fleet projections for bulk shipping. II.5 Bulk Traffic Forecasting in Ports In any port analysis determining the actual as well as the potential port hinterland is essential. Port hinterland or area of influence studieS are therefore required to determine the geographic area dependent on such port facilities. Part of a hinterland study is the evaluation of developments of socio-economic activities likely to affect the demand for port commerce (and, in turn, be affected by it). Port tributary or hinterland area determination is often performed on the basis of factor such as the following: - Equal Transport Cost - Equal Distance - Equal Time - Interport Analysis - Competitive Factors The required data for such an analysis usually consists of the following: - Origin - Destination Records - Imports and Exports - Transport Costs/Time/Special Requirements - Port Costs (Direct - Indirect) - Competing Ports FIGURE II.4.2 FORECASTING DRY BULK SHIPPING DECISION MODEL REQUIREMENTS |world supply of | worl demand for | existing owned| commodity m cmodity fleet jrelative cos i mcocnmc| advantage 0 factors l expected market l demand fluctuations forecasting shipping demand currently char share of commodity , _ 1 error forecast tered vessels t cost of pro- iduction 0 __^ lactual pro- | shortfall of available _ uctiosh _ shipping poution sstock of, cpcty commodity IL I a distant Ihppurchase| immediat chartering capacity ll long shortnew used | em|er m rm||tonnage tonnage - 50 - - Entrepot Trade - Port Hinterland Changes - Demographic Factors - Employment - Per Captta Ixncome (PCI) - PCI Growth Rate Employment Projections - Feeder Transport - Rail / Road / Barge Traffic - Cost / Capacity of Systems - Technology Introduction - Ship Arrivals - Routes Served and Rates Forecasts - Transshipment Traffic - Inter atd Intra-port Transport A regional or hinterland economic analysis normally yields the best projection with substantial supporting evidence and, as a result, develops high user confidence. Such an approach examines all the major factors determining the cargo and traffic flow and estimates how the current patterns will be projected into the future. An in-depth analysis of a single commodity should attempt to identify the principal causal relationships and determine the key factors that contribute to movements of goods through the port. Among the causal factors that should be considered are the following: - Principal sources of the commodity together with information on the present and future capacities of these sources con- ditioned on various developments (such as port expansion). - 51 - - Principal components of major domestic and foreign demands with respect to the nature of the demand, the product end-use, the product form and the elasticity of the demand. - Constraints on cargo flow imposed by regulation or other constraints to the free flow of goods. - Available modes of transportation, both complementary and competitive, their capacity and relative appeal. - Impact of technology, both current and anticipated with regard to the production, distribution and consumption of the cargoes under study. Only after a clear understanding of the commodity flow pattern has been achieved can one prepare models that relate levels of domestic and world-wide demand and supply of potential interest to a port. I - 53 - IIl. BULK SHIPPING ECONOMICS III.1 Bulk Carriers and the Bulk Carrier Fltet Bulk cargoes are generally powders, granules or lumps stowed in the vessel without packaging. Typically, they are dry and should be kept dry while they are being handled. They are often shipped in grades, and the lot size is usually large enough so that a single grade will fill the hold. However, grades can also be separated with tarpaulins. The materials are strong enough so that they can fill an entire hold without support. Although the cargoes are loaded in a simple way, the loading technique is quite sophisticated because bulk cargoes tend to move with the ship when it rolls. Recently, considerable research has been done to find ways to prevent this movement from damaging the ship. The most important difference between bulk cargoes is their stowage factor (called bulk density in literature about mining and materials handling). The stowage factor ranges from 14 cubic feet per ton for iron ore to over 60 cubic feet per ton for wood chips. Figure III.l.1 lists the specific quantities of common bulk cargoes. Bulk cargoes also differ in the lot sizes in which they are shipped, the hazards (such as fire) to which they are prone, and the type of handling equipment that should be used. However, the shipment size and stowage factors are the most important details to consider in designing bulk carrying ships and other portions of the bulk logistics system. Basically there are three types of bulk carrying ships.- Pure bulk carriers are designed to carry only bulk (see Figure III.1.2), small general purpose ships are designed to carry a variety of cargoes besides bulk (see Figure III.1.3), and combination carriers are designed to carry - 54 - FIGURE III.1.] DENSITIES OF BULK CA1RGOES RA-NE Iron Ore -1.91-3.49 Iron Pellets . 1.84-3.49 Manganese Ore . 1.79-3.15 Chrome Ore 2.09-2.72 Bauxite 1 .03-1.38 Salt . 1.20 Phosphate - 1.1 Coal _ 0.75-0.85 Raw Sugar 1 0.78 Petroleum Crude _ 0.78-0.92 Residual Fuel Oil - 0.94-1.00 Distillate Fuel Oil - 0.84-0.94 Gasoline 0.74 Wheat 0.64-0.77 Corn - 0.59-0.72 Rye - 0.63-0.72 Grain Sor'ghums 0 0.63-0.71 Barley 0.45-0.61 Linseed 0.61-0.63 Oats - 0.35-0.49 .3 .4 .6 .8 1.0 .1.5 2.0 3.0 4.0 Specific Gravity Source; World Bank staff. - 55 - FIGURE 111.1.2 - GEARLESS PURE BULK CARRIER Length Overall ................................ 186.5 m (611.8 ft) Length Between Perpendiculars ................. 178.0 m (584.0 ft) Beam, Molded .................................. 28.4 m (93.2 ft) Depth, Molded ................................. 15.3 m (50.2 ft) Draft, Molded Designed ........................ 9.8 m (32.0 ft) Draft, Molded Scantling ....................... 10.7 m (35.2 ft) Sea Speed, Knots .............................. . 16.9 Deadweight At Design Draft .................... 32,100 Tons Gross Tonnage U.S. (Approx) ................... 23,500 Tons Net Tonnage Panama Canal (Approx) ............. 19,000 Tons Net-Tonnage Suez Canal (Approx) ............... 21,000 Tons Cargo Hold Capacity (Grain) ................... 45,417 m3(1,603,880 ft3) Water Ballast Tank (Full) ......... ............ 19,763 m3(697,920 ft3) Fuel Oil Tank .............................. 2,010 m3(70,990 ft3) Fresh Water Tank .............................. 230 m3(8,120 ft3) Diesel Oil Tank ............................... 190 m3(6,710 ft3) SHP, ABS Max .............. 15,288 Crew Accommodations .. 26 Total Accommodations .. 34 Propeller (1), Blades ....... 5 Machinery, Twin diesel engines Source: World Bank staff. FIGURE III.1.3 TYPICAL DRY CARGO SHIP - 4 - 4~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- DIMENSIONS AND TONNAGES MACHINERY DETAILS CARGO DETAILS CARGO GEAR: 2 Cranes G.R.T: 13200 MAKE: Burmeister & Wain GRAIN: 26900 cu.m. 2 x 10.5 T D.W.T: 20000 TYPE: 7K67GF BALE: 25000 cu.m. 2 Derrick lOT L.O.A: 161.5m ENG. BLDR: Hitachi Zosen SADDLE TANKS: 2600 cu.m. ric s L.B.P: 152m OUTPUT: 13100bhp @ 145 rpm CONTAINERS: 305 HATCH DIM: (1) 13.6m x BREADTH: 22.8m FUEL CAPACITY: 1610 cu.m. NO. OF HOLDS: 5 9.94m DEPTH: 13.6m CONSUMPTION: 46.5 t/day NO. OF HATCHES: 5 (2)-(5) 13.6m x DRAFT: 9.8m RANGE: 12000mls TYPE OF HATCH: Single 11.6m GENERATORS: 3x 400 kW SPEED: 16.25 OTHER DETAILS REMARKS No. of Crew: 33 Conforms to St. Lawrence Seaway Regulations Classification: ABS 1R5 cnntRiners car be stowel in the hold, 120 on deck Alternative Classification - LR, NV. Source: World Bank staff. - 57 - oil as well as bulk. Even the pure bulk carrier is, to some extent, flexible because it can be cheaply converted to carry containers. Although such conversions do not use either the volume or the deadweight of the ship well, they can be competitive on short routes when charter rates for bulk carriers are low. Pure bulk carriers are sometimes fitted with portable car decks, allowing automobiles to be carried. Ships of this type are somewhat out of favor now because the hold ventilation system cannot prevent con- densation from damaging the cars. These carriers are rarely fitted with cargo gear, as it interferes with more efficient, shore-based loading and unloading equipment. The smaller, general purpose ships are usually fitted with derrick- type cargo gear. The cargo gear allows the ship to work cargo where bulk facilities do not exist and to perform as a general cargo carrier. Much ingenuity has been put into the design of these vessels to allow bulk cargoes to be loaded without trimming. (Trimming means filling up small voids left in the cargo hold because the ship's structure casts shadows in the stream of cargo being loaded, hence preventing large ship loaders from filling the hold completely). Figure III.1.4 shows this type of a ship fitted with car decks. Because the structure and machinery of tankers and pure bulk carriers is similar, certain aspects of the two were combined and the combination carrier evolved. Building this dual-purpose ship costs about ten percent more than a single-purpose ship. Having a dual- purpose ship allows the owner to operate in whichever trade offers higher freight. Results with this type of ship have been mixed because, desoite their flexibility, these ships generally operate in one trafe for #~~~~~-gq -qu To aao ~~~~~~~~~~~....... .e .w @ .@ . -.-. . _ , , ,, ~~~~~~~~..... ...... . . . . . . . . .. . .. . . .. . _n *_ s1^ s... ... ... .... ... ... m X B 1 @~~~............. .. . .. . . . ., ... ... .. . ........... ......... ..o.o.ol .. . . . . _ . . .. . . . . ESIlIXTd::).xrIn ............ s-r...a. .. ...a ... xv............. nOI - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ . . . . . . - 85 - ~ ......... - 59 - extended periods of time, and the reliability of the equipment intended for the other trade is reduced from lack of use. To ready the ship for a change in trade often requires a shipyard overhaul. Such an overhaul may increase the cost of the trade to the point where the new trade may no longer be justified by the difference in freights. Because bulk materials vary in density, they require different amounts of space for the same material weights. Fifty thousand tons of iron ore fills about 700,000 cubic feet, 50,000 tons of grain 2,250,000 cubic feet and 50,000 tons of wood chips over 3,000,000 cubic feet. As a result, bulk carriers tend to be specialized and carry cargoes that fall within only a small range of stowage factors. However, a few designs exist for "universal" ships which are suitable for a wide range of cargo densities. Bulk cargoes can be carried in general cargo ships (so-called hybrid liner/bulk), but special arrangements to trim the cargo and to restrain it mechanically (with "shifting boards") make their use costly. The LASH (lighter aboard ship) vessel is ideal for carrying small consignments of bulk cargoes. As the name implies, these ships carry barges loaded with various cargoes. However, the general trend has- been to convert LASH vessels to container ships, which indicates that the technical merits of this ship do not compensate for its high cost and the cost of supporting barges. Bulk carrying ships usually are slow vessels that have service speeds between 14 and 16 knots. The low value of bulk commodities and the organized trades, in which they move, do not provide incentives for speed. Because fuel costs are high, new bulk ships tend to have even lower speeds in order to conserve fuel. - 60 - Grain has been carried in great volumes in tankers. The grain is loaded and discharged through manholes in the deck. This is not as efficient a process as is possible with a shore-based unloading facility for normal bulk carriers. Figure III.1.5 compares grain carriage in tankers with carriage in other vessel types. Moving grain in tankers does create an extra cost--that of cleaning the vessel before loading. This cleaning process involves removing loose, rusted steel, as well as oil residues, and it may require up to two weeks at berth. Cleaning becomes especially burdensome after several grain cargoes, as the grain draws the water from the rust scale, which loosens it from the base metal. Figure III.1.6 breaks down different types of bulk carriers and their numbers. There are a few special types of bulk carriers, which can only carry one type of cargo. Lumber carriers, otherwise, provide large hold spaces for a given deadweight. Ore carriers are specially designed to provide small spaces in the cargo holds while having sufficient displacement to support the ship and cargo. The ore is carried high in the ship to minimize ship motions and to keep the cargo from shifting in the holds. Figure III.1.7 shows the difference between the mid-ship section of an ore carrier and that of a general purpose bulk carrier. - 61 - FIGURE III.1.5 GRAIN STOWAGE IN TANKERS COMPARED WITH OTHRER SHIP TYPES .:.: .:.::.,: . ..-. .:--.-..- *::::::::.-.. -.-. .---- -. .......-:::.-.-....... ;L:::Zr---- GENERAL CARGO SHIP > ng *: g - -.-. . . .............- stows cargo :::............ in this area TANKSHIP BULK CARRIER Source: World Bank staff. - 62 - FIGURE III.1.6 APPROXIMATE COMPOSITION OF THE WORLD IS BULK FLEET Bulk Car a Grades hybrid linehybrd ol/bulk bulk vessels (1,000) Dry Bulk Carrier Fleet Specializ td co mmdity General purpose oriented vessels *Lumber calriers (370) *Wood chip carriers (76) 25-40 000 40-..0,000 00+ *Ore carriers (260) (1483) (165) *Cement carriers (30) *Continuous self (34) dischargers *Slurry carriers ( 15) 7!CUUg 111.1..? Comparison of Ore Carrier with Normal Bulk Ship OREi Length 244.39M (801.79 ft.) Beam 32.31M (106.00 ft.) L Lg CAR41ER Depth. 19.76m (64.83 ft.1 *0. 0 i _ ^ Draft 13.48M (44.21 ft.1 Length 281.,g4M C925.00 ft,J Beam 44.20M (1)45.00 ft.) NORMAL BULKER Depth 214.99m (82.00 ft.) Draft 18.46M (60.56 ft.) Source: -World Bank staff. - 63 - Continuous self-discharging ships have internal conveyors and discharge booms to unload cargo to the harbor facility. They are common on the Great Lakes and can achieve high unloading rates. There are about 34 in ocean service and these are optimally used when a single ship delivers cargo routinely to many ports, for example, coal shipments to power plants or fertilizer distribution. The largest self-discharging ship is a 150,000 ton salt carrier. Figure III.1.8 provides additional information about self-discharging ships. A sixth special purpose bulk carrier is the slurry carrier. These ships handle cargo that is a mixture of finely ground powder and water. Iron ore, coal and kaolin clay are sometimes handled as slurry. There are some disadvantages to this process. First, it is never possible to remove all the water from the slurry once the ship is loaded, so only part of the cargo carried is the actual commodity; the rest is water. Since a slurry is specialized, this vessel type is currently limited to few trading opportunities. General purpose bulk carriers come in three general sizes - those able to cross the St. Lawrence Seaway, those which are just able to pass through the Panama Canal, and those that are even larger. Only a few vessels are so large that they cannot pass through the Suez Canal. Panamax bulk carriers are of two kinds--those which can manage the Panama Canal loaded and those which can only manage it in ballast. Table III.1.1 lists bulk and combined carrier fleets by size. The majority of the bulk carrier fleet is under 60,000 tons deadweight, with 71 percent of the capacity beneath the 60,000 dwt mark. Approximately 94 percent of the ships in service are less than 80,000 tons deadweight. To be competitite as tankers, combined carriers must have larger FIGURE III.1.8 SELF-DISCHARGING VESSEL OPTONAL \ el 7-~~~~~~~~~~~~ * - LOA, MLD. 634'-0" CARGO DWT, CDAL 29,998 LT. (33,597 ST.) LBP 621'-6" CARGO HOLD VOLUME 1,276,700 CU FT. BEAM, MLD. 78'-0" UNLOADING RATE 3,500 ST./JR. DEPTH, MLD. 56'-0" UNLOADING BOOM 250'-O" LOADED DRAFT COAL 31'-4" SHP 8,500 DISPLACEMENT 36,430 LTSW. SPEED LOADED 14 KNOTS BALLAST DRAFT 15'-6"FWD., 20'-10" AFT. BOW THRUSTER 1000 HP. OPTIONAL TONS PER INCH 112 LTSW. Source: World Bank staff, NOTE: COAL CAPACITY EIGURED Al 38 CU.FT./ST. - 65 - Table III.1.1 Size Distribution of Bulk Carriers, January 1983 (Figures in number of ships and 1,000 dwt) Size Group Tankers Combined Bulk Total in dwt Carriers Carriers 10 - 18000 250 3612 3 45 744 11252 997 14909 18 - 25000 312 6565 5 114 912 19481 1229 26160 25 - 40000 626 20227 3 93 1680 52041 2309 72361 40 - 50000 93 4094 9 432 281 12353 383 18879 50 - 60000 163 8789 12 669 260 14202 434 23660 60 - 80000 252 17347 69 5096 400 26790 721 49233 80 -100000 311 27502 37 3294 52 4504 400 35300 100-150000 247 31047 126 15122 184 22804 557 68973 150-200000 80 13066 70 11404 25 4211 175 28681 200-250000 201 46169 18 4157 6 1325 225 51651 250-300000 294 79235 10 2719 1 268 305 82222 300-400000 83 28325 - - - - 83 28325 400000 33 14945 - - - - 33 14945 Total 2944 300923 362 43145 4545 169231 7851 513299 Source: World Bulk Fleet, Fearnleys, Oslo, August 1983. capacities than pure bulk carriers.- In fact, 60 percent of the combined earrier fleet has a mean deadweight between 100,000 and 200,000 tons. Bulk ports should be designed to handle both "random" arrivals and regularly scheduled large bulk carriers in the quantities in which they arrive. When considering the possibility that a port will be required to handle exceedingly large ships, their availability in the marketplace and economic desirability should be noted. Large vessels usually have long-term charter parties for specific work, at least when they are new, because a firm bankable charter is required as collateral for the vessel's financing. Table III.1.2 gives the dimensions for common, series built bulk - 66 - Table III.1.2 - Bulk Carrier Dimensions for Standard Classes DWT VOLUME LENGTH BEAM DRAFT SPEEP BHP (f?/ton) (ft) (ft) (ft) (kts) a and p sd-14 14861 55.5 451.1 66.9 29.2 15.1 8600 bv liberty 14960 55.6 433.0 68.9 30.1 16.0 8400 ihi freedom mk ii 15353 57.0 440.9 68.9 29.5 14.5 6850 sasebo (mp) 16 15800 52.3 479.0 75.8 31.5 16.6 11400 hyundai 18b 17716 47.2 465.9 74.1 30.1 14.6 8000 mitsui concord 18 18208 51.9 458.3 65.1 30.5 15.2 8300 boelwerf 19 18700 56.7 504.5 75.0 33.7 15.5 12400 nippon kokan kk 20 19192 56.5 478.0 75.0 30.6 15.2 9000 hitachi zosen ut-20 20000 56.7 498.7 74.8 32.0 16.2 13100 nippon kokan kk 21 21349 47.1 478.0 75.0 32.5 15.0 9000 sumitomo 22 21500 50.1 506.9 74.8 31.5 15.3 11400 horten verft 22 21602 47.2 496.0 75.0 31.9 15.2 10000 ihi friendship 21751 53.7 509.8 75.1 30.8 15.0 7800 mitsui 22 22000 46.1 518.3 75.0 31.1 15.0 9400 ihi fortune 22000 52.9 510.0 75.0 32.3 15.0 8000 hyundai 24b(11) 24113 51.2 550.0 74.8 33.3 14.0 9400 sumitomo 25 24500 55.2 551.1 75.3 31.9 15.1 11400 harland & wolf 35 34445 45.4 590.6 91.8 34.4 15.2 11400 aesa 34447 44.2 606.9 79.4 36.4 15.2 11540 emaq-brazil 35 34447 51.6 600.4 90.5 33.9 15.0 02000 nippon kokan 35 34666 43.5 547.9 91.2 36.6 15.0 12000 sasebo 35 35400 49.3 577.4 91.2 35.5 14.7 12000 hyundai 35b 35431 47.9 550.7 105.8 35.7 15.0 11200 mitsubishi 35 35500 49.3 577.4 91.2 35.5 15.1 12000 helenic shipyards 37 36415 47.4 610.2 86.9 37.3 15.3 12000 korea shipbuilding 37000 48.1 570.0 105.8 35.5 15.4 11200 sanoyasu 40 40386 48.8 567.6 90.5 39.7 15.0 14000 swan hunter 40 40443 46.0 590.5 97.7 37.1 14.9 14000 stocznia paryskiej 53147 53.6 674.2 105.6 40.7 16.0 17400 mitsui 56930 50.7 688.9 105.6 40.0 15.7 16800 sumitomo 59 58100 48.1 715.2 105.6 40.0 15.1 14404 hitachi hi-bulk 59775 46.2 705.4 105.6 40.8 14.8 12200 burmeister&wain 60 59904 50.1 698.8 105.7 41.3 15.6 16650 nippon kokan kk 60 60035 49.1 721.8 105.6 41.0 16.0 17400 hyundai 60b 60528 45.9 705.3 105.6 40.9 16.5 16500 mirauviahi 63 62000 49.6 693.2 104.3 43.7 14.6 14000 harland & wolff 64 63675 45.6 721.5 105.8 42.5 15.2 16800 astano 69848 52.2 787.4 105.6 43.5 16.5 18400 gotabverken 72 70859 46.0 748.0 106.0 43.9 15.9 18500 sunderland 72 70918 47.0 715.2 105.7 46.0 15.0 20000 boelwerf 75 73812 42.8 761.1 105.8 45.6 15.0 19200 hyundai 76b 65261 43.8 775.5 105.6 45.2 15.6 20100 italcantieri 80 79800 44.5 813.6 105.8 45.9 16.0 20300 Source: U.S. Maritime Administration - 67 - carriers up to the PANAMAX size. Figure III.l.9 present much of the same information on a graph plotted with typical larger vessels. Typical good pTactice design dimensions are shown in Figure III.1.10. Virtually every dimension of those ships using shore-based cargo gear play a part in the design of harbor facilities. The beam of the ship is important because it determines the outreach (and cost) of ship loaders and unloaders. The hold's span length determines the length of crane rails (and support piers) for movable ship loaders and unloaders. It also determines what the dimensions (or number) of quadrant type ship loaders should be. The ship's draft determines harbor dredging require- ments. III.2 Efficiency in the Use of Bulk Carrier Tonnage In liner shipping, because itineraries can be adjusted to meet existing demand, vessel utilization on outbound and inbound voyages can be high. In most tanker trades no return cargoes are available, and one leg of the voyage must be in ballast. The bulk trades fall in a middle category because cargo is frequently available for loading at ports not far from the vessel's discharge port. Because bulk cargo may not be available for prompt loading or may not have destinations appropriate for the vessels to resume their primary trade, vessels may pursue almost random trades to minimize steaming in ballast. A typical trading pattern for a small bulk carrier is shown in Figure III.21.and Table III.2.1. It is necessary to be able to arrange such cargoes in order to obtain the lowest cost operation of bulk carriers. An alternative to seeking return cargoes is to use larger vessels, as the costs of larger ships with a ballast leg will resemble those of a smaller ship -68 - FIGURE III.1, BULK CAARIER DIMENSIONS 150 125 _ @........ .. 9. 0 100 _ BEAM F- .. E @9 . E X X T x x x ?x x x 75.. ***i** Xy x X X YK X X xK x LENGTH/ 10 Xx x xx 50 x44, KEX iX XYXX xxx 4+ t x X+444+ 4 4 4444* 4 4 44 + 4 4++4 DRAFT 25 _ _ _ _, _,_,_ - __ 10,000 30,000 50,000 70,000 90,000 110,000 130,000 Vessel Deadweight Sourcet World Bank staff. FIGUlRE !Uf.I. 10 MAXIMUM GOOD PRACTICE DESIGN DIMENSIONS OF BULK CARRIERS L 1T , s 12-1 k~c= 5. == t __ DIM1ENSION SUIF SIZIE. ( DEADW rEIGM TON _ 16,000 20.00 000000 0000 W)0O 12C A6,000 000 2S,000 S2,000 ,0- 500 . LENGTH L 140 151 163 190 184 OS 218 235 245 Z5S 70 LVI. FORECASTLE R 14 15 16 18 19 2o Z2 24 25 26 e7 Sh" OF HOLDS S l01 5oS 118 130 40 143 isi 169 174 191 191 LGTS.OF HousE T 25 27 29 32 3S 37 39 42 44 48 49 BEAm a 19.0 20.2 21.8 23.2 25.0 Z7.0 29.2 31.7 34.5 373 403 WArcH WDZ74 bf 9.0 9,4 9.8 ,0.2 10.6 11.4 12.2 13. . 14.8 16, 8* , bz 9.0 9.0 9.0 9,0 9.0 9.0 93. 9.0 . 9.0 9.0 9.0 DMEPH :ro a/EF6Cc 0 tot 11.3 12.2 13.5 14.0 (55 1(6 56.0 183 19.6 0A.4 CAMBER E 1.2 1.4 1.4 1.6 1.6 I.- 1.6 1.8 1. 2.0 2.0 DEPW TO 4ATC$ " 12.0 12., 13.6 15.1 16 s 11.1 18.5 19.8 20.1 '14 21.4 LOADED dRAPT C 7.7 8. 8.5 52 9.71 10.4 11.2 12.0 12 13.2 4.0 LI6ltT D^t dL. j 2.61 2.7 2.8 .3.1 3.2 3.5 3.1? 4-0 4.2 4. 4. 1: ORE/8L44K/OIL SOURCE: MITSIBISHI HEAVY IMDUSTRIES ADNGJ PATTE RN FOR $SMALL OXE D DIRY B1UL CAROMERS CEMENT, GRAIN, AND FERTILIZERS I." '~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~' 0 500 Kilometers I *I EOUATORIAL SCALE TRADE ROUTES Thn map has btee brepared by The vbed ranks stalfl fc -1-eet IDr th. -snvesece ot c Pt re der and . eu chdvefr lor the hters5t use or rhe T udd Ba * and the W -rterrah /> C Fmnanon C orp orah on The d e t -rnn s used and 1ie b - da sh u et Dh t h5 mraP d D nu l hpiy. on the art of The WhedS Butnk and the U nternattutat F rrne r Curp-rssun. any j ginent < cnthIV.,a statiof. my t-em . -y or f am--- pi- .1 -rBonementDch b-d-e . (After: Ferrnloys, Oslo) ._ . . . . .. o~~~~~~~~~~~~~~~~~~~~~~~- - 71 - Table III.2.1. Trading Pattern for a Small Bulk Carrier Loading Discharge Rate Rate Port Commodity Draft (tons/day) (tons/day) Monrovia Cement 45 12,000 -- Fle Kkefjord - 12,000 -- Halifax 42 -- 12,000 New Orleans Grain 38 12,000 -- Valencia 30 -- 3,000 Seville 24 __ 3,000 Toledo Grain 26 10,000 -- Glasgow - 3,000 New Orleans Grain 38 12,000 -- Acajutla 38 __ 3,000 San Francisco Fertilizer 35 12,000 -- Chittagong 1,000 Source: Fearnley's, Oslo, 1984. with most sailings fully laden. This course possesses much less risk for the ship user. The market risk and the inconvenience and scheduling problems associated with spot chartering are eliminated from the opera- tion. It is of course true that, if the larger ship can be chartered for the ballast passages, the freight rate is reduced. However, the spot market for large ships is small and may not fulfill expectations for backhaul cargoes for a large ship. A second consideration is that the stowage factor of the cargo booked must closely match the ship's design stowage factor. If the cargo's stowage factor is larger than that of the vessel, the vessel will be unable to lift its full deadweight because there will not be sufficient space in which to put the cargo. If the cargo's stowage factor is too small, full advantage will not be taken of the vessel's capacity and large expenses may be incurred for stevedoring measures required to prevent the cargo from shifting. The ship's intrinsic design will only prevent the cargo from shifting if the holds are full. - 72 - A final general consideration is that of vessel size. While large economies of scale exist as the size of hulk carriers grow, it is not possible to evaluate the usefulness of large ships without noting the diseconomies of scale which exist. Considering only the positive aspects creates a false statement of worth for the larger vessel, and it biases decisions toward the use of large ships. Large ships need onshore storage for cargo being discharged, and there may not be adequate storage in existing or planned port facilities. It is not uncommon for a very large crude carrier to have to go to four ports in Europe to discharge a full cargo from the Persian Gulf. Multiple port calls forced by lack of storage ashore reduce the efficiency of large ships. E-xpenses of larger ships based simply on extrapolation from smaller vessels can be grossly underestimated. This is also true of a vessel's out of service time. Repairs that can be made on smaller vessels may require expensive drydocking. The implications of shipwreck are much more serious with larger vessels than with smaller ones, as are scheduling problems. These and other difficulties have depressed the resale and charter value of large bulk carriers more than smaller ships. II1.3 Interests of Organizations in Bulk Shipping The bulk shipping industry produces an intermediate product, which is usually a component of a large scale distribution or production system. The transportation segment, including vessel loading and dis- charge costs, can be the largest single cost and typically the most variable. A ship user should try to obtain transportation at the lowest cost commensurate with the risk level acceptable. The marine industry has produced a variety of market channels to - 73 - procure shipping at varying degrees of involvement, risk and cost. Typically, there are five ways to offer the vessel's services. They are listed in Table 111.3.1. The table also indicates which services are included as part of the basic contract and which must be provided by the ship user. A spot (or voyage) charter is a contract to carry a specific cargo between two generally defined ports for one voyage. Occasionally, several are arranged in succession, creating a consecutive voyage charter. Voyage charters generally provide for all operating costs. Furthermore, voyage charters always specify the amount of time that the charterer may use to load the vessel and establishes a penalty (called demurrage) for each additional hour. They may or may not provide steve- doring services. Usually these are arranged for by the charterer. Generally, charter payment is made in advance and is considered earned wehen the cargo is on board. A time charter makes provisions for the vessel to be available for a specific period of time, generally between one month and five years. Fuel is included in this type of charter account, as are other expenses connected with the ship's trade route, such as canal tolls. Short time charters are common. Long time charters occur only when the price expecta- tions of the owners and charterers coincide. The variability of time charter rates depends on their period; the shorter ones change more rapidly with trends in the demand for shipping then longer ones. While time charters do reduce some of the organizational requirements for shin use, the arrangements that the charterer must take, including fuel oil, tug boats, berths, pilots, etc., consume a lot of organizational time. Buying a vessel is the third way to procure tonnage. Owning a ship - 74 - Table III.3.1 Five Types of Vessel Service Arrangements Ship Ship Arrangement Cost Component owner user 1. Spot charter vessel financial cost X operating expenses X insurance X fuel X port costs X stevedoring costs * * demurrage X 2. Time charter vessel financial cost X operating expenses X insurance X fuel X port costs X stevedoring costs X demurrage X 3. Ownership vessel financial cost X (New) operating expenses X (Second hand) insurance X fuel X port costs X stevedoring costs X 4. Bareboat charter vessel financial cost X operating expenses X insurance X fuel X port costs x stevedoring costs X 5. Contract of vessel financial cost X Affreightment operating expenses X insurance X fuel X port costs * * stevedoring costs * * * - Indicates that the responsibility could be either the ship owner's or user's Source: World Bank staff. - 75 - may require no more organizational resources than the use of chartered tonnage, as it is common for one owner to contract for the management of a vessel. It is possible to buy both new and secondhand ships. While similar in organizational framework, these options differ greatly in cost, risk and potential return. Conceivably, a company that owns ships may never use them, but, instead, rely entirely on charters to meet its shipping needs. In that case, vessel ownership is seen as financial risk management rather than operational technique. Bareboat charters are a cross between ownership and a normal charter. The charterer has most of the owner's responsibilities to man and repair the ship, but, like normal charter, these responsibilities end at a particular time, and the risk involving the value of the vessel is borne by its owner. While obligations are placed on the charterer to maintain the ship, they are not well-defined and it is common to defer maintenance until the ship reverts to the owner. Bareboat charters are frequently used as purely financial instruments. When both parties wish the ship to change hands, but cannot agree on the value or do not wish to renegotiate the vessel's financing with the holder of the mortgage, this type of charter can be employed. If a shipping company goes bankrupt, it may leave many bareboat charters in its wake. A contract of affreightment is a common way to procure tonnage. It is an agreement to provide transportation for goods in defined quantities and between specified points. No specific vessels are named in the con- tract. It is not necessary to own vessels to enter into a contract of affreightment which operates as a futures market for shipping. An important difference between a contract of affreightment and the voyage charter, and the other methods of obtaining ships, is that it is - 76 - the responsibility of the shipowner to arrange for backhaul cargoes. For a large scale organization, although all of these methods are viable ways to procure tonnage, they differ with respect to the following: (1) Cost (2) Organizational resources required (3) Ability to service variable demand (4) Market risk (5) Risk of insolvency of contractual partners. It is difficult to rank each market channel in the abstract, because so much depends on specific circumstances. A contract of affreightment, for example, is a potentially risky contract, because both sides may make important assumptions about the movement of spot and voyage rates and may be unable to complete the contract if their judgement is incorrect. The extent to which this matters depends on the resources and business ethics of the parties involved. For intermittent ship users the spot market usually is the best way to procure shipping, as this is its basic function in the market- place. Continuous users of bulk shipping will probably use a variety of procurement strategies. The general goals of arriving at the best mix are the following: 1. To obtain transportation at as close to marginal cost as possible at an acceptable level of risk. 2. To organize and finance the operation with an acceptable commitment of funds. (Because shipping can be highly lever- aged, the commitment of funds may mean iittle or possibly zero commitment of own funds). 3. To minimize the taxes paid (if privately owned) or paid to others (if governmental). 4. To retain sufficient control to allow overall coordination. - 77 - 5. To minimize risks, including foreign exchange, political casualty losses, and insolvency of business associates. The matter of risk management has unfortunately become an area of vital concern in today's shipping market, as it is possible that all parties involved are in marginal financial condition - ship owner, ship charterer, ship yard, and financing bank. An accurate assessment of risks and subsequent steps to minimize exposure to them is an important feature of the agreement. Ship users and suppliers have different attitudes toward risk. Large organizations, those which use marine transportation, tend to put premium on risk reduction because negotiating contracts of that magni- tude requires long lead times. In addition, they would like to reduce the cost of ocean transportation, as it is the cost element with the largest variability. Studies of the charter markets made by the Institute of Shipping Research in Norway34indicate that the price of risk is very high in these markets. Hence, shipowners find it profitable to go for an agressive chartering policy. Ship operators can assume a role of almost an insurance company, agreeing to bear some of the long-term market risks in exchange for long-term charter payments considerably above costs of delivering shipping services. There are three areas of risk in the bulk transportation field: (1) Foreign exchange and financing risk (2) Operational or "catastrophe" risk (3) Business or "market"risk Of the three, the foreign exchange and financial risk is the most difficult to manage because interest rates and currency values fluctuate. 4/ Market Strategies in Bulk Shipping, Victor D. Norman, Studies in Shipping Economics, 1982, ledriftsokomomens Forlag Als, Oslo. - 78 - Somehow the risk must be diversified since it cannot be eliminated. Actual problems with foreign exchange losses have not been common in the marine field (other than a few ships financed in Deutsche Marks). Operational risks involve ship loss or damage. There are two types of risk - financial responsibility for the loss and consequential damage to the operations of the ship user. Under maritime law, the ship cargo owner and ship owner are considered to have a common venture and share many maritime risks in proportion to the value of the ship and cargo (called general average). The potential risk is especially large for the cargo owners. Many companies use general average claims to move profits from highly taxed areas to untaxed ones (i.e., shipping), so they obtain general average claims at every opportunity. Specialized bulk handling operations, using slurry ships or continuous self-dischargers, are especially prone to operational risk. What to do should the ship sink must be addressed specifically in the project plan. The cost of replacing a vessel depends on the way the shipping is procured. If a voyage charter strategy is pursued, there is no additional risk because of the short contract. If the boat is owned, hull and machinery insurance will provide funds to buy another ship. Long time charters, however, pose the greatest risk because the charter terminates when the ship is lost and, unless the charterer had insured the ship for his own account, there is little recourse. Market risk depends on uncertainty about the availability and cost of shipping and on the current shipping rate levels and expectations about the direction of future movements. For many reasons no formal options market has developed to institutionalize the risk of shipping rates. While the charter market for shipping meets many of the criteria for a - 79 - commodity suitable for options trading, the high value of individual transactions and the availability of capital from other sources prevent the development of an options market for tramp or bulk shipping. It is not uncommon, however, for a ship user to make a long-term contract of affreightment with concerns that do not own any vessels but who fulfill the contract with chartered tonnage (both spot and term charters). It is also not uncommon for charters to be arranged so the vessel will be delivered up to eight months after the agreement is reached. Ififormal arrangements of this nature are the principal means of handling market uncertainties. To the extent that international marine transportation makes up a significant part of bulk cost, the options market for the cargo may absotb some of the transportation risk, since most bulk commodities are traded on various exchanges. 111.4 Vessel Chartering When a vessel is to be obtained for a fixed period, the usual instrument is a charter. While there exist many standard form charter parties for different trades, they all cover the general items outlined in Table III.4.1. Of these, the place where the ship will be delivered to the owner at the end of the charter is a potentially costly and easily overlooked issue. The two most common charters are spot (or voyage) and time charters. Most statistics regarding spot charter costs relate to particular major trade routes; this is based on the assumption that other trades run in parallel. Specific spot charter costs may reveal little about the true cost of shipping, as they may 6nly be revenue marginal to the ship - 80 - Table III.4.1 Vessel Charter Party Provisions 1. Vessel description and performance (including warranties of fuel consumption at various speeds) 2. Period and trading limits 3. Place and date of delivery of vessel 4. Place and date of redelivery of vessel 5. Vessel condition 6. Payment of hire 7. Off hire 8. Responsibility for fuel 9. Drydocking (including vessel cleaning and steaming to shipyard) 10. Pollution 11. Vessel modifications 12. Layup 13. Pollution 14. Settlement of disputes 15. Attachments owner. A low rate may be quoted only to obtain the fuel to reposition the ship for a more lucrative charter. Averages of many charters may provids the best estimate of short-term costs of shipping. Table III.4.2 Table III.4.2 Single Voyage Charters by Commodity (April 83) Ship size (1,OOOdwt)grain agr forest iron mins coal ferts steel scrap other 4-10 9 12 - - - - 1 1 - 3 10-20 25 20 - 2 - 14 1 - 2 20-30 64 8 2 - - 2 4 1 3 1 30-40 23 - - - - 1 - - - - 40-50 13 - - 1 - 1 - - - 1 50-70 19 - - 2 - 8 - - - - 70-100 4 - - 3 - 5 - 100+ 1 - - 9 - 3 - - - - Soaitree Dry Bulk Charter Markets: Developments and Trends, H-,P.' Dr'eQt, Lonidoin; IMS-.- - 81 - FIGURE III.4,1 SINGLE VOYAGE ChIARTER RATES $/Tonne Grain (20/40,000 DWCT: U.S. Gulf to Japan) ------ Coal (50/60,000 DWCT: Hampton Roads to Japan Iron Ore (80/100,000 DWCT: Brazil to Continent) 40 35 30 o5_~ I I 25 - . O~~~I _I . 20 15 0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ----19* -1981 -198 - - 6-1983 Source: Cargo Vessels Voyage Rates 1980-1983, H.P. Drewry, London, 1984. - 82 - gives a general idea of the size of vessels working in the spot market of various trades in April of 1983. Historic spot charter rates are given in Figure III.4.1 with selected current charters given in Table III.4.3. For large vessels the spot market can be thin and may not represent the true, short-term value of the vessel. Time charter rates are generally quoted in dollars/dwt per month and are segregated by ship size. As there is considerable varia- tion in different ships' fuel consumption, some of the apparent varia- bility of time charter rates is due to differences in ships' machinery. This is particularly true of larger ships in which a significant proportion of the fleet is steam-driven and consumes more fuel oil than diesel ships Table III.4. 3 Selected Current Charters Annual rate of return, measured as revenue less operating costs under Norwegian flag as a percentage of building price, for an 80,000 dwt tanker, built 1966-1967.- Alternative A Alternative B Alternative C Year spot market 1-3 year time 3-7 year time chartering charters charters 1967 46.7 13.0 12.7 1968 35.5 15.2 9.6 1969 18.8 20.1 13.3 1970 48.8 14.4 8.8 1971 36.3 34.9 9.8 1972 11.3 40.7 17.0 1973 77.2 18.1 18.4 1974 7.5 40.2 7.8 1975 -14.2 22.2 20.7 1976 - 1.6 - 6.0 25.6 Average rate of return 26.6 21.3 14.4 Standard deviation 25.5 13.0 5.0 Source: World Bank staff. - 83 - of the same size. The difference is compensated for somewhat imperfectly by differences in charter rates. When market rates are low, the difference between steam and diesel ships may be such that the steam vessel cannot be chartered at all. Figure III.4.2 presents historic charter rates for bulk carriers of different sizes. Figure III.4.3 is a typical charterer's voyage estimate form developed by Fairplay. III.5 Ship Ownership Ships can be purchased new, nearly new, used and ready for demolition. Each method of procuring tonnage has advantages and disadvantages. These are partially summarized in Table III.5.1. Ship ownership is a common way to procure tonnage. As the costs of ownership do not change with the market level, ownership produces predictable ship usage costs. Most large organizations, those which do bulk shipping, use ship ownership as the way to manage risk exposure in the charter market, as owned tonnage-is a "riskless" source of transportation against which a portfolio of charters can be built to provide the rate of return and risk levels desired. Even though the use of chartered tonnage cannot be avoided for operational reasons, owning tonnage equal to the capacity required and chartering it out eliminates the market risk from vessel chartering. The cost of this is the brokerage commissions. Hence, many bulk shippers own ships, even though their operating costs may be a little higher due to higher operation standards. The most obvious way to obtain a ship is to buy a new one. In periods of medium to high shipping activity, this can be the least expensive approach. This is mainly because modern diesel engines have been improved and are able to burn low quality fuels. Because subsidized financing can always be a part of a ship purchase agreement, much smaller - 84 - FIGURE III.4.2 TIME CHARTER RATES ($/DWT-MTH) 15 7.5 0 Jan.. Apr., Jul.. Oct. Jan. 1978 1979 1980 1981 1983 Bulk carrier 50-69,999 DWT Bulk carrier 35-49,999 DWT Bulk carrier 24-34,999 DWT Multidecker 10-13,000 DWT Source: Dry Bulk Charter Markets: Developments and Trends, H.P. Drewry, London, 1983. - 85 - FIGURE III.4.3 CHARTERER''S VOYAGE ESTIMATE FORM Date: Cargo Detail5 Daily Bunker Consumption At Ses In Port FO DO Idle Working Vessel_ Speed L Miles L _ B _ Daily B_ VOYAGE LEGS Miles Davs FO DO Bunkering Port: Canal Transit: Port Time. Loading- Discharging: CARGO CALCULATIONS TOTALS- Zone Load:- DRAFT AND DEADWEIGHT CALCUILATIONS Dwt: +5 Less: Hunkers C Weights Cargo,- VOYAGE EXPENSES HUNKFERS tons in - S- PFO t tonsin in S S _______________ tonsin in S -Y I ___________ _ {ltonsin i n* DO90 tons in C, S- ________________ tons in _ _ _ S _ S = g OTHER COSTS: Loading port dishursements - S Discharging port disbursements - S Bunkering porl disbursements - S Canal Trans,t Expenses * Insurance Premiums - S Ste-edoring Charges - S Tlme Charter Hire - _S Vo age Freight Deadfreight - S DemurraRe: - GROSS VOYAGE EXPENSES - S Gross Voyage Expenses Add Comm: Nett Voxage Expenses Cargo Rate Per Ton S S _ S S $ $ S Copyrighto 1 978 by Fa.rplay Publcations Ltd. Additional copies nsy be obaisned from Faiplaxy Publication. Minaterilouse. Arthur Streel London itC4R 9AX. Tel. 01-623 1211 - 86 - Table III.5.1 Methods of Vessel Procurement NEW SHIPS Advantages *Lower fuel consumption *Cheaper fuel grade *Easier private financing 1. Lower equity 2. Longer loan period 3. Subsidized interest rate *Common spares if multiple ship procurement *Exact design requirements can be met *Known risk Disadvantages *Break-in maintenance *Currently-higher cost *Delivery of ship not at first loading port *Unknown performance of custom features 1. crude oil burning 2. Ro/ro ramp NEARLY NEW SHIPS Advantages *Competitive fuel consumption *Currently-lower cost *Break-in maintenance accomplished *Technical performance known *Delivery of vessel more flexible *Possible to charter first to try out Disadvantages *Shorter period before overhaul *Ship condition marginally riskier *Non-common spares *More difficult financing OLD SHIP Advantages *Lower cost *Delivery of ship flexible *Possible to charter first to try out Disadvantages *Higher fuel consumption very high if steam *High maintenance especially if diesel *Poor financing terms *Unstable vessel value *Technically risky *Short useful life *Higher insurance *Higher crew cost *Probable imminent repair expense *Higher skills required of management - 87 - equity investments are required to buy new ships than otherwise would be required. Even under OECD finanicing rules, it is possible to obtain nearly 100 percent financing on ships obtaining an engine financed with export credit from country A and using it as a down payment for a ship built in country B. As new ships need not be built, new ship prices have a floor, which is equal to the cost of building the ship minus the loss the shipyard is willing to take and any direct subsidy it is able to obtain. Because governments may choose to subsidize ship sales through indirect means (loans, guarantees, lower interest ratesf etc.), the evaluation of competing offers must consider discounted cash flows and never be based on price alone. The terms are as important as the price in determining the viability of a ship sale. Historic trends in orders for bulk carriers are shown in Figure II.5.1. One advantage for new construction is that properly engineered, new ships have much lower fuel costs than older vessels. Table III.5.2 projects bulk carrier deliveries based on shipyard order books. Purchasing a nearly new vessel can be a good investment because their value fluctuates more than that of a new ship; the price floor usually is the outstanding value of the ship's mortgage. It is difficult to buy a nearly new vessel without retiring the existing financing. Nearly new ships may be more reliable than newly built ones because the initial owner may have solved most of the technical problems. The ship's performance is also a matter of record rather than specification. The life of the nearly new ship may be as long as newly built ones, as a con- tinued program of good maintenance will keep the ship in good condition. - 88 - FIGURE I11.5.1 HISTORIC TRENDS IN ORDERS FOR BULK CARRIERS Million DWr 14 REPORTED ORDERS FOR BULK VESSELS AT WORLD SHIPYARDS 12 10 8 6 S . ~~~~~~~~~~~~.............. . S...:..:.. . -. ''.... '..' .'.:.:... :, :..:::', _ S '','''. . . . . . . '. ':':.,.- 1979 1980 1981 1982 1983 z Tanker Orders jzJ Combined Carrier Orders =:* Bulk Carrier Orders Source: World Bank staff. TABLE III.5.2 The Bulk Carrier Order Book FUTUKR SCIIEDULED DELIVEK7ES SUIP SIZE CUKKENT IST HALF 2hD HALF 1ST IALF 2ND HLF 1ST HALF 2ND HALF TOTAL (Ht.BW) FLEET 1983 1983 1984 1984 1985 1985+ ON LkDEt N0. H, DWT UO. M.IWT NO. tt.DWT NO. H.DWT NO. M. DWT NO. h.DwT No. M.DOT NO. n. M.? 10-20 1,044 16,633 II £181 3 53 1 18 - - - - - - IS 252 20-30 1.467 37.592 43 1,125 20 490 23 567 16 432 6 147 3 6U 111 2,826 30-40 833 29179 64 2.254 54 1,878 19 699 16 582 9 333 4 160 166 5.906 *6-50 295 12.949 28 1,179 17 725 32 1.325 8 351 - - 4 160 89 3.740 5-80 644 39,603 70 4.424 53 3,271 20 1,265 12 792 2 140 2 130 159 IO.u22 0-lU0 57 4,927 - - 2 169 - - - - - - - - 2 169 HAD-150 170 20,750 6 833 8 1.152 1 130 3 392 - l a _ 18 2.5us 15U-2U0 23 3,D43 4 672 2 332 3 522 - - _ _ _ _ 9 1,526 2IUD 6 1,3ST 2 421 - - - - _ _ _ _ _ - 2 421 Tilal 4,539 166.828 228 11,092 159 8.072 99 4.526 55 2,547 17 620 13 510 571 27.367 ftaumeld at ea mrrlera Source; Dry Bulk Charter Markets: Developments and Trends, H.P. btewiy, London, 1983. - 89 - A drawback to nearly new ship purchase is that it is difficult to find lenders who are willing to extend more than 50 percent of a vessel's sales price as a loan. Thus, projects involving used ships are initially less liquid and have negative cash flows in the first few years. As no one has an interest in subsidizing the financing of used ships, the interest rate will also be higher. If a ship owner has access to low cost financing from sources other than those typically available to ship buyers, many of the reasons to buy a new ship are eliminated, and the older ship becomes the better investment from every point of view. In the purchase of used tonnage, there is less choice of ship con- figuration, and many types of specialized ships may not be available. One alternative here is to convert an existing ship to a new use. Tankers, for example, can be converted to bulk carriers. While this option can provide a satisfactory product at reasonable cost, the real advantage of conversions is that the ship is available for use sooner than new con- struction is. However, it is not always cheaper. For instance, to retro- fit an inert gas system on a 30,000 ton tanker costs $4,000,000. An existing ship with a similar system installed in the last 10 years could be purchased secondhand for less than this. The final way to acquire an equity interest in ships is to buy them in the last stages of their lives and operate them for a short time before scrapping them. The lowest value of such ships is determined by their value as scrap minus the cost of the voyage to the scrap-yard. The value of scrap is given in dollars/ton of ship (not deadweight tons) and is shown in Figure III.5.2. As the weight of the ship itself is about 20 percent of .a vessel's deadweight, it can be seen from the figure that, in 1981, it was possible to buy a 60,000 ton bulk carrier for as little as - 90 - FIGURE III.5.2 SCRAP SHIP PRICES (USD per-Lt. wt. ton) 225 MAXIMUM REPORTED SCRAP PRICES 200 175 . 100 LX., 125 50 25- 1981 1982 1983 Far East Indian Sub-continent Europe Source: Ship Scrapping, H.P. Drewry, London, 1983. - 91 - $600,000. The actual value of old ships is the profits from the few voyages that can be sailed before scrapping it. Well-selected, old ships can stay in service sometimes for a few years before required repairs force scrapping. As in 1981, one month's time charter for the above vessel was $450,000, so a large fraction of a vessel's scrap value can be earned on the short-term. Many owners make most of their profit from specula- tion in older ships rather than operation. Ship prices vary considerably (see Figure III.5.3), and the greatest risk of ship ownership is capital loss should the ship no longer be required. Figure III.5.4 shows prices for new building and two and five year old bulk carriers. New building prices are based on construc- tion in Japan or Korea. The graph shows two year old vessels selling for about 70 percent of the cost of new vessels, and five year old vessels for about 50 percent of new ones. The 27,000 ton vessel (which is the largest which can pass through the St. Lawrence Seaway) has a higher resale value than the other sizes; the resale value is 90 percent of the new price for the two year old ship and 63 percent of the new price for the five year old one. A collparative analysis of various options was carried out for a prospective shipowner and is printed here as an example. The break-even charter rate for various options for ship procurement was calculated and is presented against the one year time charter rate in Figure III.5.5. The ship, the one ready for demolition, was assumed to be purchased for cash while, for the sake of simplicity, the remaining ships were presumed to be 100 percent financed. The new ship, sold with OECD terms, was assumed to have an eight year, 10 percent mortgage and the FIGURE III.5.3 TRF.NDS IN USED SHIP PRICES INDEX 360 - ARAB/ ISRAELI CONFLICT 320 - -\ 3280 U.S. ~ ~ ~ ~ ~ ~ ~ I RECESSION 280 e , II, - U.S. RECESSION /INRECESSIO 40~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. 1970 '1971 1972 19373 1974 1975 1976 1977 1978 1979 100,000 (1972=100) 50,000 IYT(1967=100) 25,000 D*?T (1966=100) Tranm Time ._ *- Charter Index Source: World Bank staff. (198 1/t) - 93 - FIGURE III.5.4 CURRENT BULK CARRIER PRICES 30 Million USD Neiv Ships 2-year Old Ships 20 5-year Old Ships 10 + I 25,000 50,000 75,000 100,000 125,000 150,000 DWT Source: Naess Hallman, Inc., - Shipbrokers, Oslo, 1984. - 94 - FIGURE III.5.5 COMIPARISON OF RETURNS FROM DRY CARGO VESSEL OPERATION Time Charter Rate $/Dfr/Mo ith 10 -9 Break even _ _ _ _ _ new 70, 000 ton bulk carrier 7 OECD terms 6- Break even new 70,000 ton - -\- - - -- bulk carrier 5\ soft terms - - - - - - _. - -Break even 5-year old 70,000 4 - ton bulk carrier Break even 3- short term operation of vessel scheduled for demolition 2- 1- 1__-1980 1^ 98,1 ._-1982 - ^ 198 3 Note: Based on one year period with fairly prompt delivery. (Bulk carriers 50/69,999 DWT) Source: World Bank staff. - 95 - same ship sold on "soft" terms was assumed to have a.twelve year, 8 percent mortgage, but the five year old ship has a five year, 12 percent mortgage. The comparison does not treat the different lifetimes of the vessels adequately, since it ignores the decreased costs when the mort- gages of newer vessels are paid off. These effects, however, are of second order because, if the analysis were done using discounted cash flows, the future cash flows would be heavily discounted. From the graph it appears that the older the ship, the less it costs to use it because of the lower financial cost. This will be true if the older ship is only slightly less efficient than the new one and if its off-hire time can be kept to comparable levels. In general, vessels less than 8 years old have mechanical performance that is as good as that of a new ship and they do not have higher off-hire times. The operation of demolition bound ships (those over 15 years old) is a different matter and inexperienced maintenance on these ships is often more detrimental than beneficial because it creates more off-hire time and higher repair costs. Maintaining older ships is a very specialized skill, as only machinery required should be kept in good condition, some machinery should be temporarily repaired, and some items, particularly the hull steel and paint, should not be repaired. One must remember that all repair work will be scrapped in a few years with the ship anyway. It is, therefore, tempting not to repair even the items that are required for the safety of the ship, if the ship is able to move the cargo and the required repairs are expensive. Even if a ship user intentionally avoids the buying of old ships, lack of space and problems related to coordinating his chartering policy may leave him with no option but to charter old ships without inspecting - 96 - them. Chartering old ships can be just as risky as owning them. A general average claim, for example, would be almost all for the cargo, as the value of the cargo would easily be ten times that of the ship carrying it. One point that can be drawn from the graph is that owners of newly built ships in the 70,000 dwt range lost money over the last three years. The purchaser of the older vessel at least broke even. It is clear that investment in new ships is today advantageous only under special circumstances. - 97 - IV. INLAND TR NSPORTATION TQ ALT TI E TE, TES IV.l Introduction Usually there exist several possible sites on which to build a bulk facility. The options available may include expanding an existing port or constructing a new port(s) at one or more alternative locations. New sites may have the advantage of having lower site preparation costs, as the sites can be selected where the water is deep and there are small filling costs. Expanding an existing port is advantageous because much of the infrastruc- ture development work is complete and, as a result, development costs will be decreased. Experience demonstrates that one of the major differences between alter- native sites is the cost of inland transportation to and from the port. When no inland transportation facilities exist, their cost may equal the difference between development costs at the least expensive and the most expensive alternative port sites. A large scale iron ore exporting port, for example, may cost approximately $75 million (U.S.). Of this, work done at the port site may cost about $20 million (U.S.). Of this total it is unlikely that different sites vary by more than 50 percent or $10 million. If the cost of railway construction is about half a million dollars per mile, the cost of the inland transportation will exceed the variable element in the port's cost should the line haul distance exceed twenty miles. In such a situation, selecting the least expensive port facility will require optimizing the combined cost of development at the site and the inland transportation facil- ities. There can be advantages to developing existing port facilities: for instance, using the existing road and rail transportation system. It is important, in the feasibility study, to determine early whether or not - 98 - these facilities are in reasonable condition and their design is adequate. Table IV.l.l lists the transport alternatives available for inland movement for bulk goods. This information is presented with a qualitative evaluation of the different transport modes. The second portion of the table indicates the magnitude of the relative costs of right of way, vehicle acquisition and operating costs. These are for typical situations and have no universal application. If a railroad right of way already exists, for example, it may be much cheaper to lay new track than to construct a road. There are many situations in which the intermittent foundations required for a pipeline are less expensive to construct than the continuous civil engineering works required for both roads and railroads. IV.2 Truck Transport Where roads already exist, trucks provide flexible transport that can be installed in stages and removed if the demand decreases. Often, existing roads may be adequate to handle additional traffic, and the only additional investment required is for the vehicle fleet. Expanding the system is easier and cheaper than most of the other options available. Additionally, since trucks (and rail cars) have recoverable salvage values, truck transportation can be used as an interim step, while construction of the final distribution system is progressing. Trucks can also be used to defer installation of the final distribution system until available traffic justifies the higher costs of the other modes. While not always the case, road developments required for the road alternative may also be beneficial for other areas of commerce. - 99 - Table IV.1.1 Comparison of Alternate Transport Modes Inland Aerial Belt Modes Highway Train Water Pipelines Tramways Conveyor Route med low low low high low Flexibility Terminal high med med low low low Flexibility Speed med-high low-med low low low low Operating Range (miles) high high high low low low Expandability high high high low low low Independence med low -- high high med Climate very very Independence med high med high high high Flexibility high high med low med med Capacity low-med med-high low-high -- low -- Impact on Environment med-high med low low med-high med Cost Elements Carrying Right-of-way Vehicle Operating Highway Medium Low High Train High Medium Low-medium Ocean Ships Low High Low Inland Water Low Low Medium-high Pipelines High None Low Belt Conveyors High None Low Aerial Tramways Medium Low Medium Source: World Bank. - 100 - Truck transport delivers many small loads frequently by mechanically independent means. Barring major accidents on the road, truck-centered transportation systems are less susceptible to complete breakdowns than other systems. This means that maintenance problems must be large before the overall system performance is degraded. On the other hand, conveyor systems must be almost completely operational to ensure that the system is available for use as needed. Trucks also provide a more diverse distribution system than other means of transport and are particularly good if the source of supply or delivery is not a single point. For this reason truck or rail transport are the only viable options for many bulk commodities, such as grain. Truck transport can be the most expensive of any transport system if it is operated only during the normal work week (one or two shifts, five or six days a week). Additionally, truck schedules are difficult to maintain, as variations in the time between arrival for trucks is generally larger than for other systems. Both of these factors, in essence, mean that the system needs more storage. The first factor requires that operational reserve storage be installed at both the truck loading and unloading facilities. If a pure conveyor system were installed, investments in storage facilities for these factors would not be necessary. When port facilities are not used often and are served by trucks, there must be enough bulk storage capacity so that stevedoring work on the ship is not interrupted by any lack of truck capacity. If this is done, it is not necessary to invest in the truck capacity that would be required to meet the peak demands when a ship is being stevedored. Road haulage costs are difficult to estimate because costs depend to a large extent on the quality and design of the roads. Poor roads - 101 - raise vehicle maintenance costs, tire wear, and fuel consumption, and lower the average speed. This increases the number of trucks that will be needed. The cost of building roads depends largely on the terrain, climate and soil conditions. Additionally, much domestic labor is used in road building, so that portion of the cost depends on local wage rates. As an estimate, the cost of non-urban, two-lane paved roads in flat-rolling terrain may vary between $150,000 to $300,000 per kilometer in 1982 prices; costs in urban areas or more hilly terrain may be 2 to 4 times as much. IV.3 Rail Transport Rail transport plays an important role in bulk transportation and is generally found to be the least'costly transport mode when distances exceed 20 to 30 miles from the port and the quantity to be transported is in excess of 500,000 tons per year. As a result, rail connection and transfer facilities are usually incorporated into bulk terminals. Rail interface facilities usually consist of marshalling, classification and transfer yards, l6a'ding/unloading tracks, car dumpers (unloaders/ loaders), car cleaning facilities, run out and transfer tracks and related facilities. In-terminal railway services including transfer, consoli- dation, and movement of railcars is usually performed by the bulk terminal, which as a result, requires shunting locomotives, tracking winches, and various communication and safety equipment to perform these in-terminal rail operations. - 102 - Railway service to and from bulk terminals is usually by unit or block trains consisting of 50-100 cars. Such trains are broken up in the mar- shalling or transfer yard into smaller train sections for effective transfer by shunting locomotive to particular port locations. Cars can be unloaded (using car dumpers, unloaders, or gravity gates on the car) onto conveyors which either transfer the cargo to a stacker and hence, to a stockpile or which transfer the cargo directly to ship/barge loaders. Facilities usually provide for both direct and indirect transfer of bulk cargo between railcar and ship. Similarly, direct and indirect service can usually be performed in the transfer of cargo from ship to railcars. In the import mode, railcar sections are after loading, assembled into unit or block trains in the mar- shalling and classification yards. Rail networks in bulk terminals usually have zero grade and are laid on well compacted leveled soil. Because the speed of operations in the terminal or port is quite low (only shunting and transfer operations), used rail is sometimes utilized for laying out of the track at substantial savings in investment. The most important consideration in the design of a rail network in a bulk terminal is assurance that unloaded (or loaded) cars can be readily returned to the marshalling yard and outside rail network without affecting terminal operations. In other words, effective switches, branchings, return tracks, turnouts, turntable crossings and other facilities which assure the means for efficient railcar turnaround under any foreseeable operating conditions, must be provided. Railcars should not be parked or marshalled in the terminal and should achieve an in-terminal turnaround of a few hours. It can be shown that rail transport using new rail can be cheaper than road transport. Existing rail can be cheaper than even barge transport for all but the very large tows found on the Mississippi River, over short distances. - 103 - However, before a decision to use an existing rail link is taken, the con- dition of the link should be carefully assessed. In general, for bulk traffic, track conditions are not critical to efficient operations, and if the line is presently in operation, it will probably prove adequate. Use of railway presents a number of advantages over truck use for bulk movements. Railcars sometimes present advantages in blending different grades of bulk materials. Blending can be accomplished relatively easily by selectively dumping railcars carrying various grades. The need for blending should, however, be carefully assessed. In the case of coal, the need for blending at the port is much greater when the port serves a number of small producers as opposed to a larger mine where blending can be carried out prior to loading the coal into railcars. In general, for in-terminal rail operations a number of factors must be considered: o need to construct additional rail spurs o the transport distance, diversity of traffic and shipment sizes o ease of access to the main rail links o availability of yard facilities at the port The investment cost for track can vary considerably depending on terrain, soils, labor costs, availability of materials and other factors. ]V.4 Inland Waterways Transport Inland waterways provide the least expensive transportation available in many regions of the world. The system relies on barges with 1,500 to 3,000 ton capacities. These can be combined into multiple "tows," which consist of many barges attached together. The size of waterways in Europe - 104 - limits the size of tows to four barges, while in the United States, tow sizes of up to forty barges are common on the lower Mississippi River. Many different types of cargo can be carried in the same tow, and it is possible to realize large economies of scale, even though each shipment is relatively small. Where this system is in use, industrial facilities are located on the water and "door to door" delivery is common. River barges are simply and inexpensively constructed. Consequently they allow for low investment costs per ton of capacity. Barges are available in many types and those typical of American practice are shown in Figure IV.4.1. River transportation provides high transport momentum because of the large tow size, even though the speed of the tow is low. The typical speed for river transport is eight miles per hour. The low speed and high transport momentum provide lower transport costs and low fuel costs. While the barges can carry a variety of cargo, the system presents the lowest costs when the barge load is carried between single shippers and single receivers located on the river front. Tows are made up of a group of barges usually pushed by a tow boat or less frequently pulled by a tug boat. The tow boats come in a variety of sizes and range from small units (100 horsepower - 36 feet long by 12 feet wide) to large units (12,000 horsepower - 230 feet long). A six thousand horsepower tug can move a tow of about 15 barges with a cargo capacity of 40 to 50,000 tons. A similarly powered diesel locomotive can pull about 120 cars with a payload of 6,000 tons. The inland waterway's transport system was originally the only way to move large volumes and, as a result, most population centers grew up where there was easy access to waterways. For this reason a large percentage of the world's industry can be served by inland waterway - 105 - FIGURE IV.4.1 LASH ANI SEABEE B3ARGE 'EQUIPMENT watertight hatch cover LASH UNITS T n S , , i, ~~~~~~~~BHOt1 LL L ----~~~------------ ------ ----------- 18.28' 1 18.745 Deck 15.849- Porticulors J.. Yf L*a. 18.745 m T r boao. 9.50 m '---r Clear h opening 13.411 | 0 ot side 3.96 m L 2 d mox l.w. 2.66 m Il ___ ____ d mox I.w. 2.74 m |L Bat. 19.500 cu. ft. !I (563m') . i i .| _ _ _ _ _ _ _ _ _ _ _ _ ____..______h SEABEE UNITS (Commercial Service) TOP 9715 iFT. - lD 1~~ - 1FTH ~~~SIDE 32.75T. 35FT. 16.FT. | 17.5FT. I - 90 FT. | ,-32.75 FT. 2 END IN INTEGRATED TOW I 35FT.-- ___ ZII.11F711 1 y I CAPACITY- 40,000 CU.FT. 850 L/T WEIGHIT - ABT. 150 L/T DRAFT- WITH 850 L/T-10'8" SEAU3EE DRAFT - WIT. G50 t./T - 8'6" UN ITS DOUV3LE SKIN CONS71RUCTION - 106 - transport. Barges are easy to integrate with ship loading and unloading. The most obvious way is to load the barge itself on a ship. The LASH and SEABEE systems both function in this manner. Figure. -,TA.4 . showed both LASH and SEABEE barges. LASH barges are lifted onto the ship with a crane, and SEABEE barges are loaded onto an elevator. Both barges have a relatively small capacity and are suited to "neo-bulk" cargoes, such as construction grade marble slabs, graded chemical clays, etc. SEABEE barges are easily pushed in floats, while LASH barges must be towed in strings, which limit the number that can be towed with a single tug. Historically much of the world's stevedoring has been done by lightering barges to ships anchored in a river without port facilities. As port facilities - and the road and rail transport systems required to support the land-based port - were erected, lightering operations usually stopped. However, the lightering of large grain cargoes is still common. Timber and rubber cargoes are usually loaded from barges or directly from the water. In order to exploit the cost advantage of inland water transport, one must eliminate the double handling of cargo (i.e., discharge from the barge to a shore facility with subsequent shiploading). Recently much progress has been made in this regard and the direct transshipment rates from ships to barges - or the reverse - have risen to approach handling rates possible with shore-based facilities. Costs are con- siderably lower, because rehandling of the cargo has been eliminated and so have the requirements for shore storage. Equipment to transship directly between ships and barges can be - 107 7 barge-, ship-, or shore-mounted, as in the case of shore-based pneumatic grain handling systems, barge-mounted ship unloaders/loaders and more. IV.5 Small Ocean Bulk Carriers There is a growing market for small ocean-going bulk carriers to carry bulk cargo in areas where the weather and sea conditions do not permit the operation of barges which have small freeboards. Smaller ocean-going bulk carriers can be either integrated tug barges_/ or small ships. The feeder ship, in the bulk trades, was created to reduce the time larger vessels spend at ports by having smaller ships make the actual pickups and deliveries and transship the cargo, in turn, to a large vessel for mainline ocean transport. In addition to their feeder function, these ships commonly support a shorter mainline trade. Bulk feeder services are organized on an ad hoc basis in the grain trades to allow large shipments of grain (over 100,000 tons) to be discharged to lighters. These go to four or five different ports, none of which has the capacity to receive the entire shipment. While international trade in steam coal is small, it is likely that a coal feeder trade will evolve. This will make it possible to transport coal in large vessels and, then, to have it distributed to individual powerplants on small, 5/ An integrated tug barge is a barge and tug boat fastened together semi-permanently. The ship's operating profile does not call for their separation, except for repairs. This results in cheaper transportation, because crew costs are less due to differing work rules for tugs and ships and also due to construction rules for tug barges which result in a cheaper vessel than a ship. - 108 - feeder colliers. Small bulk vessels can be most economical as feeders when they are used to distribute intermediate bulk manufactures, such as fertilizers and cement. These commodities are shipped in bulk from the point of manufacture to intermediate distribution points for storage, packaging, and shipment to retail markets. The PUSRI fertilizer company in Indonesia has installed such a system; it uses seven 7,000 ton self-discharging vessels to distribute urea and other fertilizers throughout Indonesia. Overall it does not seem that there are advantages to using ships smaller than 7,000 tons to distribute cargo, because the delivery cost per unit of cargo rises very quickly when the size of the ship is smaller than 7,000 tons. TV.6 Slurry Pipelines Slurry pipelines are used to ship bulk material as a finely ground powder in a moving stream of water. The general process for the slurry system is shown in Figure IV.6.1. In areas of rugged or difficult terrain the slurry pipeline process may be the only way to transport cargo, as with the famous Savage River project in Australia., One of the major costs, created by using a slurry transportation system, is removal of the water from the transported solid. As a rule decanting (i.e., allowing the solids to settle and pumping the water off the top) leaves a mixture that is still 10 percent water. If the material is to be dry, the rest of the water can usually be removed from the mixture by heating it. Coal transported as slurry, for example, has a lower heating value than usual because of the extra water content. Thus, slurry transporta- - 109 - FIGURE IV.6.1 GENERAL PROCESS FOR SLURRY SYSTEMI Primary feed p r | e p a -Grinding _ torage r a t Thckning C'unical Treatment J ......t... n urry Pipe lne Punp U Ihidceni Station t ___, ~~~~~~~~~~~~~~~~i ~~~~~~~~~~~~~~~~~~1 Storage rl ewtrng Eduse z a r ~~~~~~t | Dhryl4; | Cenrifgel | Fi tern | Cclones | i IF ~~~~~~~~~~0 n Source: World Bank staff. - 110 - FIGURE IV.6.2 COMPARISON OF RELATIVE FREIGHT RATE" D R y U L K S 100%0- L U \\X R X\\ R % \ Y %' y F W n R F E R I E G I H G T H T Source: World Bank staff. - 111 - tion is most economical when used for commodities whose processing is in slurry form or where a small water content is not detrimental. Slurry transportation in ships is technically well-developed. Indeed several slurry ships have been in service for many years. The process is used in international shipment by the Marcona Mining Corpora- tion and by transporters of kaolin clay. When the slurry carrying ship was being designed, optimistic reports were developed about the economics of slurry carriage. Figure IV.6.2 gives comparisons of the relative freight rates (for slurry and conventional freight) and Table IV.6.1 shows the total system cost of slurrying with conventional shipment. This information is the original data used in 1967 as a basis for the slurry system's development. There is little evidence that the economics of the system in service differ greatly from the estimates used in the system's design. In practice, slurrying is reported to have worked well. Despite the obvious technical advantages and success of prototype operations over the course of many years, no slurry carrying ships are currently on order and no more than six were built or have been con- verted to carry slurry. A possible explanation for this is that slurry ships are extremely specialized in nature which means that there is no margin for error in the plans. The system depends on the good faith of those parties involved with its use. This is perhaps more than the business climate in the minerals industry can provide. The initial slurry system developed uses a finely ground slurry and was intended to be used for the transport of iron ore. With this process the carrying water settles and is decanted in the ship. The resulting solid is mud-like - "sticky" an4 not free flowing. Table IV.6.1 Comparison of Costs between Marconaflo and Conventional Transportation Systems: A Hypothetical Case (Dollars per ton, except as otherwise indicated) Land transport Loading port Ocean transport Receiving port Total Harconaflo (100 pipeline miles) (5000 ocean miles) (slurry tanks) Capital expenditures 15 million 5 million 60 million 5 million 85 million (total cost) Capital charge a/ 0.45 0.15 1.80 0.15 2.55 Total 0.65 0.20 2.80 0.20 3.85 Conventional (130 railway miles) (stockpiles, (5000 ocean miles) (stockpiles, loading unloading equipment) equipment) Capital expenditures 40 million 15 million 70 million 20 million 145 million Transport cost Direct cost 0.80 0.30. 2.00 0.40 3.50 Capital chargea/ 1.20 0.45 2.10 0.60 4.35 Total 2.00 0.75 4.10 1.00 7.85 Source: Engineering and Mining Journal, September 1970, p. 72. a/ Based on annual rate of 15 percent of total capital cost including interest, insurance and depreciaton - 113 - When discharged, the slurry can form steep walled piles as high as forty feet. This makes it difficult to use conventional ship unloading equipment. Thus, supplying slurry to a port which uses this process is possible only where there exists an integrated system of slurry carrying ships and special purpose slurry discharge terminals. A new slurry process is being developed for coal. Here the basic slurry element is small lumps. This is called Integrated Pipeline Transportation and Coal Separation System (IPTACSS). Large pieces of coal are crushed, then the small pieces are made into briquettes to raise their particle size to a set minimum. A hydrocarbon, such as diesel fuel or gas oil, is mixed with the coal to adjust its viscosity when the coal is mixed with the transport water. The mixture of coal, hydrocarbon and transporting water can then be pumped. The IPTACSS process has many advantages for coal transport. The slurry preparation process can be integrated in the coal washing process. Low grade coal, with as much as 70 percent ash content, can be improved until it has as little as 15 percent ash content. Furthermore, the transport water can be removed from the slurry simply by draining the mixture over a screen and the resulting product can be loaded on the ship with a conventional ship loader. If the coal is loaded on the ship as a slurry, the water can be decanted aboard the ship and the resulting cargo can be discharged with conventional unloaders. This ensures that the product has a wide market. Also, the ship can be fitted to discharge the coal using the slurry system when facilities exist to receive slurry. This system is the most economical. The hydrocarbon can be removed at many stages of the process, or it can be burnt with the coal. - 114 - Figure IV.6.3 indicates what the cost of slurry transport and competing modes were in 1982. The breakeven distances with rail for various volumes is given in Table IV.6.2. Although this information is not current, the relative costs presented are probably still valid. It is important to note that slurry in low volumes is the least expensive means of transport. Table IV.6.2 COMPARISON OF SLURRY PIPELINE COSTS WITH RAILWAY COSTS Tons per year Distance over which competitive with railway 0.5 million All distances 1 million Over 70 miles 2 million Over 150 miles 5 million Over 250 miles 10 million About equal over 300 miles Source: J. H. D. Sturgess, and J. D. Weldon. "The economic prospects for solids pipelines in lieu of new rail branch line construction." Presented before eighth annual meeting of Transportation Research Forum, Montreal, Canada, September 1967. There are over eighty slurry pipelines in use throughout the world These carry coal, iron ore, sulfur, potash, kaolin, phosphate, limestone, and other materials. More information about many of the projects is given in Table IV.6.3. - 115 - FIGURE IV.6.3 SLURRY PIPELINE TRANSPORTATION COST - IRON CONCENTRATE, COPPER CONCENTRATE, AND LIMESTONE 8.0 6.0 HIGHWAY _____ _ E-I 0 0_2.0- __ _ 4f) C 0.1 0.2 0.4 0.6 0.8 1.0 2.0 4.0 e,o 8.0 1.0.0 20. Annual Throghput - Million Tons Sourcet Transportation & Traffic Engineering Handbook, Ed. Wolfgang S. Homburger, Institute of Transportation Engineers, Prentice-Hall, Inc., New Jersey, r.~~192 - 116 - Table IV.6.3 Selected Commercial Slurry Pipelines Length Pipe Size Capacity Operational (miles) (in) (tons/yrxlO6) (year) Coal Consolidation 108 10 1.3 1957 Black Mesa 273 18 4.8 1970 ETSI 1378 38 25.0 198- Alton 180 24 10.0 198- Iron Concentrate Savage river 53 9 2.25 1967 Waipipi (Iron Sands) 6 8 and 12 1.0 1971 Pena Colorado 28 8 1.8 1974 Las Truchas 17 10 1.5 1976 Sierra Grande 20 8 2.1 1978 Samarco 253 20 12.0 1977 Copper Concentrate Bougainville 17 6 1.0 1972 West Irian 69 4 0.3 1972 Pinto valley 11 4 0.4 1974 Limestone Rugby 57 10 1.7 1964 Calaveras 17 7 1.5 1971 Phosphate Concentrate Valip 80 8 2.0 1979 Source: World Bank staff. - 117 - IV.7 Conveyors In many ways conveyors are the best suited device to carry bulk materials. Transfer equipment, required to load and discharge material from the conveyor, is simple and inexpensive. Conveying systems usually have the smallest requirements for internal storage of any transport system. Environmental considerations are easily solved, as are require- ments about protecting the cargo from the weather. Furthermore, for distances less than five miles, conveyors also tend to be the cheapest transportation mode for bulk materials. For distances over five miles, trucks will have lower costs. Thus, most long distance conveyors are less than five miles long and very few are longer than ten miles. Besides the cost over vast distances, the most significant disad- vantage of a conveying system compared to the use of trucks is that a breakdown usually causes the entire operation to stop. Truck and rail services often can continue to operate despite a breakdown, although at reduced capacity. Shorter conveying systems used in material stock- yards usually have redundant capacity, but this is seldom available on long distance conveyors. In some situations the long distance conveyor may be the best transport option. This was the case in the Spanish Sahara, where a 62 mile long conveyor is used to move phosphate rock. This choice of trans- port system was made in part because the terrain is difficult and there is not enough water available to use the slurry approach. IV.8 Aerial Tramways Aerial tramways consist of a loop of wire rope that is supported by towers spaced about every 500 meters. The rope is moved by a motor driven spool Xocated at one end. Buckets are suspended from the cable - 118 - and moved in the manner illustrated in Figure IV.8.1. The buckets, which carry the commodity, may each have a capacity of up to 150 tons. Cable means are well-adapted to difficult terrain. For transporting small quantities their capital cost is low and they are competitive with conveyors when there is only a small amount of cargo, even on the ground. It is possible, although unusual, to mount one end of the tramway directly on the ship - this makes it necessary to assemble the device each time a vessel is loaded. This was a common way to load lumber schooners on the American Pacific Coast up to 1920. The tramway line was attached to the mast and the logs were slid down the line to the vessel. Modern aerial tramways have either one cable or two. Cable ways using the mono-cable principle, employ the same wire rope to support and move the load (see Figure IV.8.1). The mono-cable is the less expensive arrangement because the tramway construction is simple. In the bi- cable tramway the weight of the moving buckets is supported by a stationary cable, while motive power for the buckets (mounted on wheels) is supplied by a second cable. The bi-cable arrangement makes sense because of the properties of wire rope. Rope, sufficiently flexible to be used in continuous movement over sheaves, must have a hemp core and is not as strong as solid steel wire rope that can be used where continuous movement is not required. The bi-cable ropeway separates the supporting and moving functions in a way appropriate to the types of wire rope available. Over level terrain tramways are less costly than conveyors when less than 300 tons per hour is moved. It is approximately equal in cost to move any amount between 300 and 400 tons per hour. On difficult - 119 - FIGURE IV.8.1i TWO TYPES OF AERIA TRAMWAYS Arrangement of a Bicable Aerial Tramway 1 - Track Cables 6 - Driving Sheave 2 - Traction Cable 7 - Shunt Rails 3 - Tension Weight for 8 - Carriers Track Cables 9 - Locking Frame 4 - Tension Weight for 10 - Unlocking of Carriers Traction Cable 5 - Track Cable Anchors Sources E. Franke1 Article. Arrangement of a Monocable Aerial Tramway 1 - Track/Traction Cable 2 - Tension Counterweight 4 - Return Sheave 3 -7 Driving Sheave 5 - Carriers Source; E. prankel Article. - 120 - terrain comparison is meaningless and the tramway may be the only feasible way to move bulk cargo. Properly designed tramways, incidental- ly, are very resistant to earthquakes, as much of the current technology was developed for ship to ship transfer of goods when each vessel is moving. Tramways can have very low total horsepower requirements, as the only energy losses in the system are from internal friction in the wire ropes and wheel of the trolley buckets in the bi-cable system. The total power consumed is a function of the change in the loading and discharging elevation. Systems where the cargo is delivered at a lower point than where it is loaded from are sometimes fitted with generators to make use of the energy of the descending cargo. Table IV.8.1 lists the estimated construction costs of a German aerial tramway in 1982. For convenience these costs have also been converted into dollars. Table IV.8.2 provides further information about some existing aerial tramways. IV.9 Modal Comparisons In selecting a suitable bulk transport system, the choice is usually restricted to some subset of the modes presented in this section. No general rule clearly can exist that indicates the dominance of one choice over others based on a few simple parameters. Each application can have characteristics which rule out, or put at a special disadvantage, certain systems. This section presents some comparisons that have been carried out in specific cases. These can serve as a guide for making comparisons among modes. In many instances rail, truck, ropeway, and tramway are direct competitors. Table IV.9.1 presents a comparison of alternatives - 121 - Table IV.8.1 Investment and Operating Costsof a Ropeway System Characteristics Length (Conveying distance), km : approx. 10 Drop in conveying direction, m : approx. 700 Material conveyed : limestone Conveying performance, t/h : 300 System : 2-cable continuous ropeway Hauling rope speed, m/s : 4 Load capacity per ropeway car, t : 1.82 Interval between cars, s : approx. 22 Distance between cars, m : 88 Drive rating : none Brake power, kW : 450 xDM 1,000 x$1,000 approx. Investment cost Machinery parts, steel structures and electrical equipment free to construction site 9,500 3,808 Ropes, free to construction site 2,000 802 Installation 3,000 1,203 Construction work 2,500 1,002 Other expenses 500 200 17,500 7,016 Operating and transport costs 7% interest and invested capital 1,225 491 Amortization of system in 20 years, excluding ropes (2,44% of DM 15,500,000) 380 152 Amortization of the ropes in 5 years (17.4% of DM 2,000,000) 350 140 Personnel costs (two shifts day) for a total of: 1 foreman DM 65,000 1 specialist DM 45,000 6 trained workers 390 156 (+ 1 stand-by) DM 280,000 Electricity costs not applicable, since the system is generator braked, feeding electricity into the network Maintenance, lubricants, etc. 155 62 Operating and transport costs per year 2,500 1,002 Operating and transport costs per tonne DM 1,90 $ 0.76 Source: Process Economics International, Vol III, Nos. 1&2 (1982) ($1 = DM 2,4941, May 1983) Table IV.8.2 Selected Examples of Long-Distance Aerial Ropeways Capital cost Capacity Total Per mile Length (thousands of (millions (thousands Year Country Commodity (miles) tons per hour) of dollars) of dollars) Type commissioned Remarks Brazil Limestone 18.5 100 0.75 40 Mono 1957 Portland cement,Belo Horizonte Gabon Manganese 47 150 8.0 170 Mono 1962 COMILOG, ore (now Moanda to 250 M'binda (Kinshasa) Germany, Limestone 1.4 300 0.95 680 Bi 1971 Portland Federal cement, Republic of Dotternhausen India Sand 2.4 200 6 330 Mono 1965 Sand-atoring 15.5 450 Bi plant, Jharia, Bihar India Sand 6.2 200 12 430 Mono 1967 Sand-6toring 21.6 450 Bi plant, Jambad, West Bengal India Limestone 1.2 400 0.6 580 Bi 1971 United Provinces Cement Works, Marrapur Sweden Ore 60 50 4.0 67 Bi 1943 Krictenberg to concentrate (now 70) Boliden Source: The Application of Modern Transport Technology of Mineral Development in Developing Countries, United Natitn, N,Y., 1976. - 123 - TABLE IV.9.1 Relatiye Costs for Alternstive Modes of Mineral Transport Over Short Distances Parameter Conveyor Railwa Ropewa Truck Relative Capital Cost a/ 1.00 1.30 0.81 0.97 Relative Operating Cost 1.00 1.26 2.29 2.16 a/ Annual throughput of 3 .million tons. Source: World Bank staff. for the transport of bauxite. The table uses indices with a base of 1.0 for the conveyor option. When the waterway option is available, this can be extremely competitive with rail. The cost of waterway transportation may be about half of the railway cost. For vessels of 9 feet draught the cost in 1976 was 1.75-7.0 mills per ton-mile for waterway transport (1 mill = 0.1 cents). This compares with 5.8 mills per ton-mile for rail, and road haulage costs of 4.6 cents or more per ton-mile. Again, within a given option (i.e., belt convelrorsN, many configurations may be examined. Table IV.9.2 shows comparisons among belt conveyor options. We note that these are financial costs, the actual economic cost may be less depending on the effect of taxes and other factors. - 124 - TABLE IV.9.2 Putnam Coal Mine: Alternative Belt Conveyor Systems (1969) Multiflight Single-flight Single-flight conventional steel-cored cable belt Item conveyor conveyor conveyor Capacity (tons per hour) 200 200 200 Length (feet) 27x 500 27x 500 27x 500 Lift (feet) 130 130 130 Speed (feet per minute) 650 800 650 Width (inches) 42 36 42 Power required (hp) 2x 600 2x 400 lx 200 Number of transfer points Terminals Terminals Terminals plus 5 only only Reliability Ranking 2 3 1 Capital cost comparison factor 1.15 1.35 1.00 Projected operating cost Total per year (dollars) 260,000 155,000 230,000 per ton (cents) 7.4 4.4 6.5 Source; World Bank staff. - 125 - V. TERMINAL SITING V.1 Introduction In this chapter choosing a port site from among several alternatives will be discussed. In general the following factors are of importance: a. Sufficient demand must exist to sustain the port's operation. b. In order to attract sufficient demand, the port's natural conditions, such as depth, should impose as few restrictions as possible on prospective ships that would call on the port, and it should be eisily accessible from trade routes of interest. c. Over 55 percent of all port-related costs are the result of delays in ships' turnaround. Therefore, efficient material handling equip- ment is necessary for loading/unloading of ships and for transferring cargo to storage or to the inland distribution network. d. efficient inland distribution networks should be available to minimize shipper's inland transportation costs. Keeping the above factors in mind, a general methodology for bulk terminal logistics can be developed. V.2 Methodology for Terminal Siting The steps involved in selecting a bulk terminal site may be structured as follows: Step 1 Description of the proposed bulk port projects. This will include a statement of objectives describing the potential benefits and beneficiaries and a description of the alternative port configurations that should be considered. - 126 - Step 2 Forecast of bulk traffic flow. The result of this forecast will be the basis for determining port capacity, bulk handling equipment requirements, and the choice of the inland transportation mode. Step 3 Assessment of costs and benefits. The cost items in a port development can be categorized as follows: (1) Port facility costs. Given the projected bulk handling requirement, a simple queuing model analysis can determine the number of berths needed to provide a satis- factory level of service to port users. The level of service is measured according to the average waiting time and the average number of ships waiting to be loaded or unloaded. (2) Material handling equipment cost. These costs are affected by the type of bulk material to be transferred and also by the layout of the port. (3) Inland transportation costs (4) Benefits. The benefits that can be readily recognized from developing a port are savings in shipping costs due to the economies of scale in vessel size, reduction in ships' wait- ing time, savings in investment costs, etc. Other benefits, such as regional development and improved competitiveness in international markets, are less easily quantified. Step 4 Alternatives for timing of investments. In determining the port capacity, one has to consider the possibility of later expansion to accommodate demand growth. The objective will be to determine a port expansion strategy which results in the minimum cost, while - 127 - providing a desired level of service at all times. Possible strategies are to build enough capacity initially so that no further expansion is necessary over the Vlanning horizon, or to expand capacity as needed, or to build something in-between. Step 5 Economic feasibility analysis. Present value and internal rate of return calculations can now be carried out. From the present value analysis, one can choose the combination of a port site and inland transportation alternative that gives the largest net present value of total costs and benefits. Table V.2.1 pre- sents a format for calculating present values. Step 6 Sensitivity analysis. After the best port site and the inland transportation mode from that site have been selected in Step 5, a sensitivity analysis is performed to understand how the selected site and the inland transport mode will be affected as traffic volume or capital investment costs change. For example, if investment costs for road facilities increase proportionally with traffic volume, the preferred inland transportation mode may be rail, instead of road. It is useful to develop breakeven distance/volume relationships among alternate inland transpor- tation modes. Figure V.2.1 shows such a relationship. - 128 - TABLE V.2.1 Present Value Analysis (Step 5) Port Development Inland Alternative Coata Cost t_ Road Rail r=es Alternate Year Capital Operating Total p.v. Cap. Oper. p.v. Cap. Oper. p.v. Cap. Oper. p.v. coat cost coat (dis- cost cost cost cost cost cost count) 2 3 A 4 econ. life - 2--- - -. -. -_-- - - I . __~~~~ 2 3 4 - - _ - . - _- - -_ -. _,- C 2 FIGURE V. 2.1 Break Even Distance/Volume Relationship between Road and Rail (Single Route) Rail preferred Distance Road preferred Tons - 129 - V.3 Information Requirements The relevant data for terminal site selection analysis can be summarized as follows: 1. Characteristics of alternate terminal sites 2. Audit of current conditions a. Identifying demand centers in a region that is to be served by the bulk port. b. Accurate amount of bulk trade (import and export) generated by each demand center. c. Inventory of available resources including (i) the existing port, its facilities, and handling capacity, and (ii) the existing inland transportation network. 3. Information for economic analysis a. Inland transportation network distances from each alternate terminal site to each demand center. b. New investment costs for each alternate inland transportation mode and its actual operating cost. c. Financial resources available d. Revenue schedule e. Appropriate rate of return f. Foreign exchange earning g. Construction period and economic life of the project, equipment, etc. V.4 Ranking Procedure From the described general methodology for bulk terminal logistics, a site can be chosen based on a pure cost/benefit analysis. But, as stated earlier, Non-quantifiable factors also have important effects on terminal siting decisions and, thus, they should be taken into account. The following procedure will enable port planners to incorporate non-quantifiable factors into the terminal siting decision. First, a list of relevant non-quantifiable factors along with the results of the economic analysis is developed. The relevant factors may - 130 - include the following: a. Availability of labor b. Skill level of labor c. Available financial resources d. Port expandability e. Existing infrastructure around prospective port site f. Regional development considerations g. Attractiveness of location to ship operators - a port's natural conditions, such as depth and the accessibility of the port to ship operators h. Economic analysis result CNPV calculation) i. Sensitivity analysis result Note that, in performing the economic analysist such factors as sufficient demand availability, efficient material handling system, transportation investment and operating costs have already been accounted for. The next step is to assign weights to each factor, which would represent the relative importance of the factor with respect to others in planning for port development. Each factor is then ranked on a scale of 1 to 5 - 1 representing poor and 5 representing excellent - for each alternate site. The total weighted score for each alternate site is calculated by summing up the product of the assigned weight of each factor and the given rank. The example in Table V. .1 illustrates the procedure. From it one can see that, even though the alternate site A has the highest advantage in terms of the economic analysis result, alternate site C has the highest weighted score when all other factors are considered. One difficulty with this procedure is that human judgements are required in assigning weights to each factor and ranking each alternative. - 131 - Table V.5.1 Ranking Procedure Example Rank Weighted Score Alternatives Alternatives Factors Weight A B C A B C Labor availability 0.1 2 2 4 0.2 0.2 0.4 Labor skill 0.05 3 4 3 0.15 0.2 0.15 Financial resource 0.2 2 3 5 0.4 0.6 1.0 Port expandability 0.05 4 3 3 0.2 0.15 0.15 Infrastructure 0.05 4 5 3 0.2 0.25 0.15 Regional development 0.15 2 5 3 0.3 0.75 0.45 Attractiveness 0.15 4 2 4 0.6 0.3 0.6 Cost analysis 0.2 5 4 3 1.0 0.8 0.6 Sensivity analysis 0.05 5 4 3 0.25 0.2 0.15 1.00 3.3 3.45 3.65 Rank I = Severe disadvantage 2 = Mild disadvantage 3 = Equal standing 4 = Mild advantage 5 = Great advantage Nevertheless, the procedure provides a means for taking non-quantifiable factors into account in terminal siting decisions and through careful analysis reasonable judgements can be made on weight assignments and ranking alternatives. V.5 Terminal Siting and Regional Development In selecting,the lo.cation.of a bulk terminal, -the.costs,- which include investment and operating costs for the - 132 - port and for inland distribution, were used as the main criteria. However, appraising it from the national standpoint, balanced regional development of the country is equally important. It would be in the national interest to increase the income of depressed regions in order to decrease the differences in incomes among regions. This is difficult to include in the economic analysis presented earlier. Per capita income of each region indicates the development status of that region in relation to tohers. The effect of this project on per capita income may be included within the ranking procedure described earlier if regional development is one of the goals of port development. The development of a port in a region can increase the income of that region through new employment and development of infrastructure faeilities. Appendix A describes the regional development analysis performed by Paul E. Smith of the University of Missouri. This methodology can be used to calculate the equilibrium income in relation to other regions, given the marginal and average propensity to spend, and the effect of new investment in one region on its own income and on the income of other regions. In Appendix B a number of computer-based techniques useful in port development analysis are presented. - 133 - VI. TERMINAL EQUIPMENT OPTIONS VI.1 Introduction The functions performed by bulk-handling equipment in a port can be classified as loading and unloading ships, storing and reclaiming material, transporting material, transferring material between equip- ment, and, occasionally, cleaning and weighing the material. Except for ship loaders and unloaders, most of the equipment available can be used in other industries, such as the mining industry. The materials handling equipment field is dynamic and offers a rich variety of equip- ment, which can only be summarized here in brief. Table VI.l.l lists the equipment available for port use for each of the major materials handling functions. Much of this equipment is produced in small numbers with a high unit cost for design and engineering. If the available equipment designs are not satisfactory, it is common to custom-design equipment to meet exacting specifications. While subsequent sections of this paper describe general characteristics of the equipment, there is a lot of variability in the precise dimensions available. This is particularly true of the capacity of continuously acting devices using conveyors, such as stackers and ship loaders. Often capacity is poorly related to equipment size and weight. The physical dimensions of the machines and the weight of the materials being trans- ported by the equipment determine the machines' weight and cost. Typically a machine's weight - and cost - varies with the square of its size and the first power of the supported weights. With conveyor-type equipment, it is possible to increase capacity without significantly increasing either its size or weight. This can be done merely by increasing the speed at which the conveyor operates. It is, therefore, - 134 - Table VI.l.1 General Equipment Options for Bulk Ports *Loading Fixed loaders Travelling loaders Quadrant loaders radial linear *Unloading Level luffing grab cranes Trolley type grab cranes Continuous unloaders Pneumatic unloaders Ship mounted equipment grab self discharging ships *Storing/reclaiming Piles gravity reclaim bucket wheel excavator reclaim dragline reclaim Silos and warehouses *Transporting Mechanical vehilcles trains conveyors chutes and elevators *Transferring Chutes Hoppers Gates Feeders *Cleaning *Weighing and grading - 135 - important to think about the equipment in terms of its physical size rather than its rated capacity. Equipment tends to be rated optimistically. Only a few types of equipment can be selected on the basis of their rated capacity. Determining when to ignore manufacturer's claims without being arbitrary is difficult, as the capacity which can be achieved depends to a large extent on how the equipment is used. Table VI.1.2 presents ship loading and unloading equipment capabilities. Care must be administered in applying these figures, as equipment vary as to the extent its actual performance differs from the nominal rate. For example, the average performance of continuous unloaders is closer to rated performance than the average performance of grab unloaders. Hence, a continuous unloader with the same rating is more productive than a grab unloader. Table VI.1.2 Summary of Design Commodity Handling Rates (tons/hr) Typical present Design handling handling rate (s) rate Co;smmodity load unload load unload Grain-small 400 300 700 400 -large 1200 1200 2000 2000 Ores:bauxite 3500 2000 5000 3000 Coal-small 300 200 500 400 -large 2000 2500 5000 4000 Fertilizer 1800 1900 2000 2000 Sugar 250-400 400 Source: Future requirements for mid-America Inland rivers port system - A. T. Roselli, Mississippi Vallev Coal Exporters Council, Niew Orleans, 1982. VI.2 Ship Loaders Ship loading involves two steps - base loading and trimming. Base - 136 - loading is the filling of the ship to the maximum extent possible with the ship loader, and trimming is filling the remaining voids using auxiliary equipment. Most trimming is required because a typical ship's structure casts shadows. Often this prevents the ship loader from filling all the spaces in the hold. When a ship loader is too small for a ship, or it is otherwise geometrically incompatible with the ship, the ship may require a lot of trimming before it can sail. Since trimming is expensive, most of the design developments for ships and ship loaders have attempted to eliminate the need for manual trimming. General cargo ships have been the most troublesome, because portions of the cargo may require bagging before they are manually stowed in the voids of the ship. Because the trimming costs are high general-cargo ships are not, in most cases, competitive in the bulk trades. Ships must always be trimmed, as specified in their loading manuals, to keep the cargo from shifting and damaging the ship. LASH ships are particularly susceptible to this kind of damage, as the loading of bulk into barges may not be supervised. Most bulk carrying ships no longer are required to trim manually, because the shape of their holds ensure that no voids remain when they are full. All bulk loading (other than trimming) can be accomplished using conveyors, which raise the material, and a ship loader, which lets the material fall into specific areas of the hold. Using this two-step process, high loading rates are possible. Current problems have to do with the conveyor's capacity to supply the ship loader. Modern mechanical ship loaders fall into two categories, travelling and quadrant. Travelling ship loaders are illustrated in Figures VI.2.1 and VI.2.3. Quadrant loaders are illustrated in Figures VI.2.2 and - 137 - FIGURE VI.2.1 LAYOUT OF PIER FOR TRAVELLING SHIP LOADER Source: IHI Heavy Industries. FIGURE VI.2.2 LAYOUT OF RADIAL SHIP LOADER INSTALLATION I_~~~~~T 77 rail Loading area Source: IHI Heavy Industries. - 138 - A third type of loader, the fixed loader, uses a spout or chute for each hold. These work well and are common on the Great Lakes. The dimensions of the ships there are standardized to fit under the loaders; no such standardization exists for ocean-going ships generally. The travelling loader moves on crane rails along a pier and, thus, can access the ship's entire hold. The quadrant type loader moves in an arc (see Figure VI.2.2) and, depending on the dimensions of the ship, two loaders may be used to reduce the number of times the ship must be moved during the loading. Quadrant loaders are either of the radial type, with a fixed pivot point, or of the linear type, with a movable pivot point (see Figure VI.2.4). Quadrant loaders were developed to be installed offshore, and using them can reduce the amount of dredging and pile work that would be required to construct the port facility. They are also inexpensive. Their principal disadvantage is that they cannot be used to load all bulk ships for the following reasons: 1. They sweep out an arc over the deck, which means that a geared bulk carrier can be loaded only with much difficulty. 2. They have a fixed geometry, which means that problems may be encountered with gearless ships having dimensions different from those contemplated in the loader's design. 3. When they are installed offshore, they place the vessel in a fixed orientation to sea and, in severe weather conditions, the ship might have to leave the berth. 4. They are not able to trim the ship and can only partially load the ship if trimming is required. Quadrant loaders can be assembled quickly at the site once their foundations are in place with the help of a floating crane. Travelling loaders are more expensive than quadrant loaders and need a pier on which to mount the rails. However, they are more flexible - 139 - FItGU1W Yl,2.3 Travelling Ship Loader Boom in raid poition Oif F i n j , / w . \ / d$t flow -n~~~~~~~~dwo-af -norwwyo for Source: Port Development, UNCTAD, U.N., New York, 1978. FIGURE VI.2.4 COMPARISON OF RADIAL AND LINEAR SHIP LOADERS RADIAL LOADER Q N*?____ ~~~~~~~ojjwr ~ ~ ~ ~ ~ 22 M?IATCH COVERAGE_4 TRAVELING LOADER moa- i/s_ v -2 HATCHI COVERA E q0 L leeLINEAR LOADER eonorATCHC1BAL. LINEAR LOADER Source: Planning Layout and Design of Bulk Terminals, N.J. Ferguson, XXV International Navigation Conference, 10-16 May 1981, Edinburgh, Scotland, U.K., Permanent International Association of Navigation Congresses. - 140 - in the type of ship they service and can often load geared ships, although this cannot be guaranteed. Furthermore, they can be fitted to trim bulk carriers mechanitally and this results in reduced loading costs. There are two types of travelling ship loaders: luffing types and slewing types. With a combination of revolving and longitudinal movement (see Figure VI.2.3), a slewing loader can vary the transverse position of material to be discharged. This arrangement can interfere with the loading of geared bulk carriers, but another type of loader is available that can vary the transverse position of cargo discharge by raising the loader's boom. A loading facility designer must decide to what extent the loading facility should be required to move the ships during the loading. It is relatively easy to move a ship at berth, using its mooring winches, and, if the berth is properly designed, it should not take more than an hour and a half for a move. Turning the ship around is another matter. The ship must get underway, leave the berth and return. It requires the use of the engine and tug assistance. This operation takes about four hours. Table VI.2.1 gives typical ship loader dimensions. The loader's boom outreach determines the maximum size of a ship that can be loaded without recirculation. The largest ship, however, that can be loaded is larger than that given in the table, because the cargo can be moved in the-hold with bulldozers. - 141 - Table VI.2.1 Dimensions of Ship Loaders Travelling Quadrant luff ing slewing Radial Outreach from 30 35 45 55 70 reference (m) Maximum ship size (dwt) 100,000 100,000 200,000 250,000 300,000 without cargo recirculation Outreach from 26 25 34 41.5 60 pier edge (m) Quadrant radius (m) -- -- -- 50 60 Span of crane rails (m) 10 12 12 -- -- Minimum pier width (approx.) (m) 18 12 12 -- -- Speed data crane movement m/sec 30 20 30 boom shuttle sec 10 20 20 10 10 (in&out) slew (rotate) sec -- 15 20 7.5 9 Source: World Bank staff. VI.3 S Unloaders Ship unloaders are larger than ship loaders for the same load. The unloader must access as much of the hold as possible to minimize cleaning expenses. Cleaning (i.e., the removal of the final portions of the cargo) is facilitated by bulldozers. These push cargo from areas that are in- accessible to the unloader to areas which are accessible. The smaller the unloader, in relation to the size of the ship, the greater is the cleaning work required. Not all bulk vessels are intended for in-hold bulldozer use, because the bulldozer blades can puncture bulkheads in unsuitable vessels. Each ship should be verified for suitability. Ship unloaders use grabs, continuous bucket chains (or digging wheels), or pneumatic means to lift the cargo out of the ship's hold. Most ship unloaders are categorized as the grab or pneumatic type. Figures VI.3.1 and VI.3.2 illustrate two types of ship unloaders--the - 142 - FIGURE VI.3.1 LEVEL LUFFING SHIP UNLOADER . 1 Source: Mitsubishi Heavy Industries. FIGURE VI.3.2 GANTRY TROLLEY TYPE SHIP UNLOADER Sore Mi Ha Source: kvlitsubishi Heavy Industries. - 143 - level luffing type and the trolley type. Both these are non- continuous unloaders. The continous unloader consists of a digging device which feeds a bucket elevator. This bucket elevator, in turn, raises the cargo to a discharge conveyor. Sometimes a digging chain (see Figure VI.3.3) or a digging wheel can be used instead of a bucket elevator. A disadvantage of the continuous unloader is that it is im- practical to change the digging device and bucket elevator to suit the density of the cargo, and this unloader's production is reduced with less dense cargoes. Because all three of these unloaders have large overhanging weights, which must be supported over the hold of the ship, their costs grow more with the size of the ship to be handled than with the discharge rate. As a result, the maximum ship size that can be handled is related to the discharge rate; faster machines being geared to larger ships. Figure VI.3.4 illustrates this relationship. It is possible, however, to unload larger ships than those shown. One can move cargo in the hold with bull- dozers or turn the ship around at berth. Table VI.3.1 gives typical specifications for level luffing and trolley-type unloaders. Table VI.3.2 gives specifications for the capacity of continuous ship unloaders. VI.3.1 Portable Ports and Transshipment Terminals In many situations site development costs for the use of land mounted unloaders are high, or cargo can be directly transshipped to barges or small ships. In situations such as this a ship unloader (or combination ship loader/unloader) mounted on a catamaran barge can be used instead of a shore port. Figure VI.3.1.1 illustrates this device. Such devices are now in use on the lower Mississippi River, where they - 144 - FIGURE VI.3.3 CONTINUOUS TYPE SHIP UNLOADER Source: Ship Design and Construction, The Society of Naval Architects and Marine Engineers, New York, 1970. - 145 - FIGURE VI.3.4 MAXISMU4 SHIP SIZE FOR UNLOADERS 200- 4-) 100 4t~~~~~ a a 0 375 750 1125 1500 Nominal Capacity (Tons/Hour) + = grab type unloaders + = level luffing unloaders Source: Mitsubishi Heavy Industries. - 146 - Table VI.3.1 Typical Ship Unloader Dimensions and Rates Grab trolley type Level luffing Capacity (t/hr) 500 640 1000 1500 2500 250 400 700 Hoisting load (tons) 16 20 30 40 42 6.3 10 25 Grab capacity (m3) 8.5 11 16 22 21 3.2 5 13 Lift above crane (m) 18 20 20 22 23 12.5 15 18 rail Lift below crane 15 16 18 20 19 11.5 13 15 rail(m) Outreach from (m) 26 25 30 30 30 19.5 22.5 26.5 reference Span of rails (m) 20 25 30 30 30 14 14 20 Speeds (m/min) hoisting 90 100 100 110 140 100 100 100 lowering 100 120 120 140 140 100 110 110 & grabbing luffing -- -- -- 80 80 100 trolley movement 160 160 180 210 190 -- -- -- crane movement 20 20 20 20 20 20 20 20 Slewing (rad/min) -- -- -- -- -- 1 1 .8 Source: Mitsubishi Heavy Industries Table VI.3.2 Capacity of Continuous Ship Unloaders Rated Capacity-Free Digging Free Flowing Material-tons/hr ~ 801 1201 1602 2002 2403 2803 Kg/m3 Kg/m3 Kg/m3 Kg/m3 Kg/m3 Kg/m3 Bucket Bucket Line (50 (75 (100 (125 (150 (175 Unit Diameter Speed lb/ft3) lb/ft3) lb/ft3) lb/ft3) lb/ft3) lb/ft3) * Capacity 254 mm 183 mpm 159 239 318 397 476 556 MT 0.0082 m3 3.05 mps 175 263 350 438 525 613 ST (10 in.) (600 fpm) 156 235 313 391 469 547 LT (0.29 ft3) 381 mm 183 mpm 340 511 680 851 1021 1191 MT 0.030 m3 3.05 mps 375 563 750 938 1125 1313 ST (15 in.) (600 fpm) 335 503 670 838 1005 1172 LT (1.05 ft3) 457 mm 183 mpm 499 748 998 1247 1497 1746 M1T 0.052 m3 3.05 mps 550 825 1100 1375 1650 1925 ST (18 in.) (600 fpm) 491 737 982 1228 1473 1719 LT (1.82 ft3) 610 mm 183 mpm 817 1225 1633 2041 2449 2858 MT 0.114 m3 3.05 mps 900 1350 1800 2250 2700 3150 ST (24 in.) (600 fpm) 804 1205 1607 2009 2411 2813 LT (4.03 ft3) 762 mm 152 mpm 1179 1769 2359 2948 3538 4128 MT 0.248 m3 2.53 mps 1300 1950 2600 3250 3900 4550 ST (30 in.) (500 fpm) 1161 1741 2321 2902 3482 4063 LT (8.75 ft3) 914 mm 146 mpm 1588 2381 3175 3969 4763 5557 MT 0.411 m3 2.43 mps 1750 2625 3500 4375 5250 6125 ST (36 in.) (480 fpm) 1563 2344 3125 3906 4688 5469 LT (14.5 ft3 *Mt, metric tons; ST, short tons; LT, long tons. Source: World Bank staff. FIGURE VI.3.1. 1 BARGE MOUNTED SHIP UNLOADER Source: Amhoist. - 149 - transship grain and coal from river barges to large bulk carriers for export. Its specifications are shown in Table VI.3.3. Barge mounted units cost between 8 and 13 million dollars (U.S.), depending on purchase specification (American Chain and Hoist Mechanical Excavators Division). In offshore discharge of grain to lighters, a similar device could be mounted on a very large crude carrying tanker with the internal volume of the ship serving as storage capacity for the discharge system. A varia- tion would be to mount loading or discharge equipment on a newly constructed barge, allowing the entire ship handling facility to be constructed in a shipyard and towed to the site. This approach has many potential advantages. First, while the cost of the barge, housing and power generation equipment mounted on the unit will be more than the land mounted unit's cost at the factory, the cost delivered may be less for the floating unit, because of cheaper shipping costs. Secondly, less site preparation is required at the port. Thirdly, because of the over- capacity in the shipbuilding industry, a very favorable price could be obtained. Depending on the device's usefulness in other situations, it could also prove possible to finance it thlrough normal ship financing channels at subsidized interest rates. VI.3.2 Pneumatic Unloading Systems Pneumatic unloading equipment is used to handle commodities with low density, such as grain, which also have little adhesion and are granular or powdery. These unloaders are continous and are made in sizes of up to 800 tons/hr. of grain with a single unit. As the commodity density increases, pneumatic unloader capacity usually decreases. Table VI.3.3 Specifications for a Barge Mounted Ship Unloader Dimensions Performance Specifications Overall length _ 240 feet Hoisting capacity 140,000 lbs. at 100 feet from side Overall width 125 feet of catamaran barge. Overall height above waterline 155 feet Hoisting speed 220 feet per minute Each catamaran barge_ 25 feet wide Traverse speed 550 feet per minute Lateral space between catamaran barge 75 feet Fore-to-aft-speed of HH50 along Front apron length from side of ship 20 feet per minute catamaran barge _ 100 feet Theoretical cycle time barge to Clearances 65 feet under bucket hopper 38 seconds 100 feet clear of catamaran barge Production capacity in excess of 3000 tons per hour free digging. Power: DC electric from shore power via submarine cable, transformed to working voltages on board HH50 Front apron supports conveyors discharge spout and clam-shell bucket. Apron can be raised up to 75-degree elevation to clear ship superstructure. 0 Discharge spout can be remotely positioned anywhere along front apron and ir equippped with dust suppression equipment. Operator's cab swings for optimum view of barges or ship holds, depending on mode of operation. Machinery cab contains hoists and electrical control panels. Fully-enclosed dual conveyor belts. Erie Strayer 50 cubic-yard clamshell bucket. Fully-enclosed dual vertical conveyor belts. Blending hoppers, complete with dust suppression equipment. Collar frames stabilize catamaran barges. Center portions of collar frames contain facilities for crew lounge, control center, office, shop and storage space. Covers provide weather protection when unloading moisture-sensitive material. Shallow-draft catamaran barges permit shoreside as well as midstream operations Catamaran design provides a protected ship for cargo barges. Source: Amhoist - 151 - Similarly, the output is a function of grain size. Pneumatic unloaders are less expensive to procure than mechanical unloaders, but they require substantially more power to operate per unit thcoughput. Pneumatic unloaders (see Figure VI.3.2.1) have a vertical suction pipe with an attached suction nozzle. This is lowered into the hold of the ship by a level luffer or gantry crane, which also supports a horizontal suction pipe. A blower is used to generate the vacuum using a venturi device. The grain or other commodity discharged is sucked into a receiver tank which usually has a dust collector mounted on it. Transport from the unloader to a silo or loader can be by conveyor (as shown) or by pneumatic pipe trapsport. VI.4 Material Storage It is necessary to determine the best way to store a commodity in order to grade, blend, classify and protect it adequately. Material that does not require protection from the environment can be simply stored outside in a pile (or pond, if it is a slurry). Materials, requiring protection, can be stored in bins, silos, or warehouses (i.e., covered piles). Covered facilities are also used for material with low bulk densities that would, otherwise, blow away. While the materials handling equipment available for warehouses resembles that available for open piles, it may be less expensive to install equipment in a ware- house, because the structure of the building can be used to support the equipment. Storage in a bulk handling system is either operational or reserve storage. Operational storage may be either long-term buffer storage, FIGURE VI.3.2.1 TYPICAL PNEUMATIC GRAIN DISCHARGE SYSTEM Bucket Conveyer Silo Conveyer Horizontal sunction pipe Tower structure Boom Dust separatori.I i e Z ~~~~~Receiver tanklll Vertical suction pipe urn head-out La |Driver's UBlowe \r 5 e ~~~Chute Suction ozz e Source:Mitsubhsh Hevy /dustPier ship Source: Mitsubishi Ileavy Industries. - 153 - required because of delays between shipments or deliveries, or surge storage, which is necessary to remove material and allow efficient equip- ment operations. Surge storage is frequently required where continuous devices, such as conveyors, operate with non-continuous devices, such as rail car dumpers. Live storage refers to the portion of the material that a materials handling device can reach by itself. Dead storage is stored material that cannot be reached and additional equipment, such as a bulldozer, is needed to move it to a position where the matetials handling device can reach it. Depending on the probability of use, reserve storage can be dead storage. WE.4.2 Pile Storage The configuration of pile storage depends on the equipment used to stack (i.e., build the pile) and to reclaim the material. There are six typical pile configurations and they are shown in Figure VI.4.2.1. When material must be covered, ramped or radial piles should not be used. These piles are expensive to build as they require larger than usual space in a building. Every conventional method of stacking material involves conveyors or clamshell grabs. The simplest stacker is an inclined belt. If the frame supporting the inclined belt rotates, it is a radial stacker. Radial stackers make crescent-shaped piles (see Figure VI.4.2.1). Conveyors can be used to make wind row (i.e., straight) piles. In such cases the conveyor is suspended over the pile and a "tripper" distributes the material along the length of the pile or creates a series of conical piles. A reversing shuttle (i.e., movable) conveyor Figure VI.4.2.1 Typical Pile Configurations A B C PE F RAfPE . . AADjAL ,. /b . e w _ 'D pobv OL OCA Type ' I Method 11h~__ F f j __ _ - - - - -N Open 0 Covered C 0 O&C 0 O&C O&C OLC Pile Height M 2.5 5 7.5 2.5 5.0 7.5 2.5 5.0 7.5 2.5 5.0 7.5 2.5 5.0 7.5 2.S 5.0 7.5 Base Width 2x 2x 2X in M 6 12 18 6 12 18 6 12 18 6 12 18 6 12 18 11 22 33 Vo 9e in 10 m 0.85 0.53 0.85 0.3 0.53 0.85 1.9 3.9 5.8 3.5 7.0 10.5 7 14 21 15 30 46 Area Utilized % 65 65 65 75 75 75 90 90 90 80 80 80 90 90 90 80 80 80 M2/ton 0.60 0.45 0.3 0.6 0.45 0.3 0.300.230.1 0.460.300.19 0.360.270.1 0.200.150.1 '100 lb/ft3 or 1633 kg/mn material Source: Design Considerations for Storage & Reclaim Systems, H. Colijn, Bulk Materials Handling, Vol. 1, Universitv of Pittsburgh, 1971 - 155 - can also be used to build two wind row stacks (see Figure VI.4.2.2). Stacking with conveyors makes it possible to blend materials, either when the material is stored or when it is reclaimed. Blending on reclaim is useful when blending proportions are not known in advance. For large volumes fixed conveyor or radial stacker storage is expensive, as the area stacked by one device is limited by the height of material discharge, combined with the angle of repose of the material. Overhead conveyors also have a disadvantage; they require a structure over the entire length of the pile. To increase the area stacked per unit investment, the mobile stacker was introduced. This machine mounts the stacking conveyor on a boom which can be rotated as well as elevated. The entire assembly is supported by a chassis, which can either move on crawler treads or rails. This device increases the area that can be stacked by allowing wider piles to be constructed on both sides of the machine. Since no above ground support is required, the foundation for the rail mounted version is less expensive. Crawler mounted stackers require little, if any, site preparation. Rail mounted stackers (see Figure VI.4.2.3) build wind row stacks parallel to the tracks, while the crawler mounted stacker can build virtually any size or shape pile. Table VI.4.2.1 lists typical stacker dimensions. The boom structure and the need to keep the stacker stable at its maximum outreach are the most important contributors to cost. Therefore, the cost of the stacker is closely related to the square of the outreach. Increasing the belt speed is the most common way of increasing a stacker's capacity and larger capacity stackers, those having the same frame as smaller ones, do not cost proportionately more. The foundation costs of the high and low - 156 - FIGURE VI.4.2-2 STACKING AND BLENDING WITH CONVEYORS Tripper forms long extended pile A: < < S4NrRe S~~~/ - Source: Belt Conveyor Transfer Points, Colijn and Conners Bulk Materials Handling, Vol.2, p. 69. University of Pittsburgh, 1973 - 175 - FIGURE VI.5.1.2 TYPICAL CONVEYOR INSTALLATIONS Double-beaded overlapping flights apron conveyor DP apron conveyor Leak-proof apron conveyor. Source: World Bank. - 176 - TABLE VI.5.1.1 Preferred Types of Conveyors for Bulk Materials Meterlal Phms1caL A, W'oolna i1aitL.ono Pr.ferred -referred Comment Cond' tIbn Conveyor ConvwyorS' E1-vtort lb/ft' kgt/u' Acid phosphat '.p 90 1,44u Adheres d. b Sticky Alum Cronular 60-os 960-1,040 Abhrsls a,b.C.e g.b Aluminum oxide Pul,. 60 960 Abrasive a.a I A ntiu nitrete oIul 62 E90 Mygroacopi b,c, g,b Explosive A_onums nitrate aeep 65+ 1,040+ Adheres CSe S.b Sticky Arsenic salts PIu. 100 1,600 Heivy C.e g,b Poisonous Asbes dry Granular 35-40 56u-640 Abra.Ive d,f b Nustr _et Sticky 45-50 720-800 Abrasive f b, Corroslve Btoe Mal PuIv. 55-60 880-960 a,b.c.d.e g.b.c Borax Pulv. 50-70 800-1,120 Abrasin A,b.c.d,e g,b Bran Granular 16-iO 260-320 a.b.c.d.e g,b Sometiees sticky 8revers grain$. hot Granular 55 8O CorroslI c.e. g.b Carbon black (pellets) Granular 40 640 *., g,b Fregile Cemtnt, dry Pulv. 90-l1S 1,440-1,890 a.c,d., g,b Packs Clays Pulv. 35-60 560-960 Adheres a.b..e, g.b Sluggish Coal: anthracite Lumpy 50-54 d00-860 a.b.c,e g,b *tea. sizes Granular 50-60 800-960 a.b,c,d,e g,b.c bituminous. -ap Lu-py 50-60 800-960 s,b.0 b bitualnous. slack Granular 50-60 800-960 a.b,c.d,e g.b,c Chalk Pulv. 70-75 1,120-1,200 *.b,c,d.e g.b.c Sluggish Coffee beans Granular *0-45 640-720 &.c.a g,b fragile Copra, ground Pull. 40 640 Kay be abraaive alb,c.c g,b Sticky Cork, ground Pu.lv. 5-15 S0-240 S, b,c,d,e 5.b Sluggish Cots. sblled Cranular 45 720 Abrsiv shell a,c.s g.b,c Cottonseed Granular 35-40 560-640 Sometimes sticky a,b,c,d,* g,b Cullet Granular S0-100 1,280-1,600 Abrxld s*b,e Sgb Corrosive Flaxted Granular 45 720 Sh3ll abrasive a.b.c.d.e g.b.c Fres-floving Flue dirt Pul,. 100 1,600 Abrseio b.d.a.f g.b Fly meb, clean Pul. 35-45 560-720 Yild abrasive A,b.c.d.e S.b.c *ree-flowing Class batch Grasular 80+ 1,280+ Abrasive *,b.e S.b aGes Granular 45 720 a,b., g.b.c Keep cool Graphite (flour) fulm. 40 640 Lubricant *.b.c,d,e g,b.c Gravel Granular 95-135 1,S20-2, 160 Abrasive A,e.f ,.b Gyps kPui. 60 960 a,b,c. g,b sea" ores Lumpy 100+ 1,600+ a,b,f g.b May bh tougb Us fuel Strina 15-30 240-480 May jam a.b,d.e Least glta ulv. 60-150 960-2.400 Sluggish &,b.c.e g.b Poisoous Lim, pebble Granular 55-80 880-1.280 a.b.c., g.b LiUmstone duet Puiv. 85-95 1,360-1,520 Abrasiv * .b.e Halt Dry 63 720 May be sticky *.b.c.d.e g.b Manufactured products Boxed 1-200 16-3.200 *.i.j nerchandise. Packaged Soxed 15 240 a.b,i.j Gsr_ete Hanging 5 so i.j Metallic dusts Pul. 50-100 W00-1,600 Abrasive *,b,c.d,e g.b Sometimes difficult Niea, pulverized ulv. 20-30 320-480 Pree-f lowing a,b.c.d.e g,b,c Dusty Nolybdemua conc'te Pulv. 110 1.760 Abrasive *.b,d. b sticky Petroleum coke Lumpy 42 670 Mild abrasive a.b.c. g,b PMice Pull. 65 720 Mild abrasive .,b,c,d,e g.b,c Polisher Qa rts (ground) ulm. 110 1.760 Very abrasive a,b.c.d 5 Rabbet *crap Stringy s0 800 Sluggish a.b, g,b Difficult Balts coarse Granular 50 0OO Rygroecopic a,b,c,e g.b Corrosive if wet cake tul,. 75-95 1.200-1,520 Flous freely a.b.c,d., g.b Sands dry Cranular 90-110 1,440-1,760 AbraeLve ,e,f g,b dam Granular 90-110 1."0-1,760 Sticky a*,-f ,b Sawdust Granular 15-20 Z40-320 a,b,c,d,a g,b.c Savage Sludge Pulv. 60 960 Sticky if wvt a.b,,1f AbrasiVe SiliCa flour Pull. SO 1,280 Sluggish aeds g Abrastio Soap fl.5 Granular 10-20 160-320 Fragile a,c.e Sticky if hot Sods ebt: light Pulv. 25-35 *00-550 Flows freely o,b,c.d,e g.c Caustic hesvy Pulv. 55-65 880-1,.40 Flows freely a.b.c.d, g.c CCaustic Soybean flour Pulv. 30 480 Sticky a.b.c.e g.C Explosive dust Starch Puly. 30-40 *80-640 a.b.c.e g,c Ixploeive dual Sugtel raw Granular 55-65 880- 1,040 Sticky *.b ,c, , refined Granular 50-S5 300-860 *.b,c. O 9 Randle gently Sulfur Pulv. 55 880 Corrosive if vwt e.b,c,e g.b Explosion risk Tale Pulv. 50-60 800-960 Mild abrasive a,b,c,d,e gt,b Adheres to meotl Tobacco ste" Stringy 25 400 Sluqggih a.b,db,e Wheat Granular 48 770 Free-flowing a.c,d I g,c Keep clean Wood chips Granular 18-20 290-320 May arch a,c.dd e gc Corrosive if vet Zinc oxide kim. 20-35 320-360 ?ay pack e.b,c,d, 11 Arotd discoloration zinc eulfats Puki. 70 1,120 May pack b,c,d, egxplanetion of letter symbols: a-belt, b-flight, c-continuous flow, d-pnoeuatic, C-edrew, f-drag chain, g-btlt sa" bucket, i-ovsrhed straight power. J-overhead power and free. Source: World Bank. - 177 - maintain each type of conveyor can be a major factor in determining which conveyor is used.' Even though renewing a belt is a relatively simple operation, it is not necessarily an inexpensive one, so it is questionable whether belt conveyors are the least expensive type. Belt conveyors consist of a belt drive, tensioning unit, idlers - which periodically support the belt, a tail loading section, and the belt itself. Tensioning of the belts is frequently done by a hanging weight, but other means are available. One must clean the belt after the material is discharged, as a portion of the material carried will adhere to it. This cleaning is done with scraper cleaners mounted in the drive unit. Conveyors can be portable. Those used in mines, for example, are easily dismantled and reassembled at other sites. The drive unit and end loading unit are the most expensive parts of a conveyor. The unit cost of belt conveyors, therefore, falls as conveyors become longer. Unit maintenance costs are also reduced because in a long conveyor there is only one drive unit and one end loading unit. However, repairing a damaged belt takes longer and is more costly because the belt is longer. There are three types of belts: 1. Belts that have a tensioned carcass made of fabric covered with rubber. 2. Belts that have wire ropes embedded in the edges of the belt. 3, Belts that have wire ropes in the edges, but are covered with a built up rubber area (called flange belts). It is difficult to determine the magnitude of total cost differences between belt types without a detailed study. The capacity of a conveyor system depends on the width and speed of the belt, the density of the material and the shape of the pile of materials and the angle of repose of the material. Table VI.5.1.2 - 178 - Table VI.5.1.2 Maximum Belt Speed (ft/min) for Various Materials Belt Belt speed Belt Capacity and horsepower width normal max speed 50 pound/cu foot material (inches) (fpm) (fpm) capacity hp/10 ft hp/100 ft ________________ (tons/hr) lift centers 14 200 300 100 16 0.17 0.22 200 32 0.34 0.44 300 48 0.52 0.66 18 250 350 100 27 0.29 0.35 250 67 0.71 0.88 350 95 1.00 1.21 24 300 400 100 49 0.51 0.51 300 147 1.53 1.52 400 196 2.04 2.02 36 400 600 100 115 1.22 0.80 400 460 4.87 3.18 600 690 7.30 4.76 48 400 600 100 220 2.33 1.97 400 880 9.35 6.07 600 1320 14.00 9.10 60 450 600 100 360 3.82 2.49 450 1620 17.2 11.20 600 2160 22.9 14.95 Belt Capacity in Tons/Hr Belt Light Moderately Lump coal Heavy sharp width free flowing free flowing coarse stone heavy ores _____ materials materials crushed ore lump coke 12-14 400 250 16-18 500 300 250 20-24 600 400 350 250 30-36 750 500 400 300 42-60 850 550 450 350 Source: Chemical Engineering Handbook, Ed. R.H. Perry & C.H. Chilton, McGraw Hill, New York, 1973. - 179 - gives a rough estimate of the belt conveyor capacities. Differences in the shape of piles on the belt are not considered, and this must be checked in specific applications. It is common practice to design conveyors with 115 percent capacity of their rated or peak capacity, whichever is lighter. The peak capacity must be calculated considering periods that the conveyor is not running (e.g., when a ship loader is being moved to another hold). Conveyor systems are sometimes mounted on crawlers and are used to extend operational areas of crawler mounted stackers and reclaimers (see Figure VI.5.1.3). VI.5.2 Airslide Conveyors Airslide conveyors combine translational movement with the controlled falling of the material. The conveyor itself is a box divided into two sections separated by a membrane permeable to air, but not the material being carried. The material is put into the upper chamber and compressed air is put In the lower chamber. The air passes through the membrane and partially supports the material. The device is installed on an incline and the material slides down. The air is used to reduce the angle of incline which increases the distance the material is moved with each increment of height. Other than those in the air compressor, airslide conveyors have no moving parts. Table VI.5.2.1 lists comodities which have been handled with Table VI.5.2.1 Typical Materials Moved by Air Float Conveyors Alumina Plaster Bauxite Lime hydrate Bentonite clay Limestone Cement Magnesium oxide Flour Phosphate rock dust Gypsum Talc Kaolin clay Soda ash FIGURE VI.5.1.3 CRAWLER MOUNTED CONVEYOR 00 Source: American Hoist and Derrick Mechanical Excavators Division. - 181 - airslide conveyors. Figure VI.5.2.1 gives design and construction details of airslide conveyors and the diagram shows their use in a ship loading installation. VI.5.3 Bucket Elevators Material to be elevated can be easily processed with an inclined conveyor. However, such an approach requires a long conveyor and the space required to elevate material can be reduced by using a bucket elevator. Figure VI.5.3.1 is a typical bucket elevator. There are two kinds of bucket elevators: those that can be mounted on a rubber belt and those that can be attached to a chain. Continuous ship unloaders usually have bucket elevators and are frequently installed in grain terminals. Their capacity is the product of the bucket size times the filling efficiency times the number of bucket passes per hour. Table VI.3.2 in the section on ship unloaders provides this information in tabular form. VI.6 Ancillary Equipment Bulk ports require many special purpose devices to perform particular functions required by each commodity. Grain, for example, requires, provision, for medium-term storage, of fumigation enuinnent to kill insects. A common ancillary function required is weighing and sampling. There are many systems available to do this, but the simplest and most reliable is to mount a hopper on a scale and record its weight each time it is filled. Continuous weighing systems to measure the output of conveying systems are also available. When truck and rail cars are used, the quantity moved can be determined by weighing the - 182 - FIGURE VI.5.2 *1 AIR SLIDE CONVEYOR DESIGN INFORlMATION Material flow through enclosed conveyor D Co.weya Oim,,luarn (~~~~43 capaci y Six.fin) A a C 0 Iftlhr - - P 9 4-tl2 441t2 5.1/8 1,200 a 10-3/4 4-1/2 6.114 71s8 1.800 10 10-314 4-112 6I14 9-1/8 3,000 _ 12 12-3J4 4.1/2 8-114 t1-t/8 4,500 14 12-3/4 4-1/2 8-114 13118 6.500 c i t4 3/4 4- 0188 10.000 Ia 14^3/4 4-1/2 10.t/4 17. 1/8 14.500 qj A A 20 16-3/4 4.i1/2 12-1/4 1911/8 20.000 24 22 5-3/4 16-1/4 2ni1s 26000 B Air float conveyor sizes and capacities. 24"x85 ft. long boom Air Float Conveyor 12"x8O ft. long boom (750 tons/hr) type Air Float Conveyor Belt conveyor (100 tons/hr) 8"1x700 ft. long Open Air Float Conveyor 2) Telescoping 24x40 foot-long Loading Tespopg Enclosed Type Air Loaning Spouts Float Conveyors Ship loading installation (Harvey Alminum, St. Croix, Virgin Islands. - 183 - FIGURE VI.5.3.1 BUCKET ELEVATORS a b c d e f g 1 2 3 4 5 6 7 8 9 h i be,t4ev tpu ed acht dtails. (el Catnfml disarge. qpmd buAeb. (bh hPositive dmu . *Kvd bK*m. .c) CAtomao bhusket. (d) $upenpacit timmmno, b.rett. (e) Spaced I skets icive part of lad dast and part b% avoomp from bnttam. (I- Cojme' - shks ar Ailed tas thme pa ibrouO looding leg. with iced %.pout al%wr tail wh..d. ,g Contilmeu. nstb ith bottlo hoot, wt eanumt dor. (A) MaflkhAr.uw spwd bmaL&a for matvtuigl dm+darr. il Stewl buckstl Iw vont4imom4ucket fle%ators. tSfep%n.-A*aiu M&J C..) Source: Chemical En&ixieering HandbQok, E.D. R.fH. Perry & C.H. Chilton, McGraw Hill, New York, 1973. - 184 - truck or rail car before and after loading or discharging. The weight of material loaded on a ship can also be determined to within 150 to 400 tons by using the vessel's draft marks as a basis for calculation. Rail car loading and unloading is a common but important ancillary function in bulk handling. Rail car loading is easily accomplished by using conveyors to elevate the material and allowing it to fall into the car. This can be done using an intermediate hopper to allow weighing and to ensure uniform car loading. Figures VI.6.1 and VI.6.2 show two rail car loaders; Figure VI.6.3 shows a railcar unloader. Unloading of trains can be accomplished in three ways. Hopper cars can be turned over in a rotary car dumper depositing its contents in a bin below. Car dumpers are also available to unload box cars carrying grain. Bottom dump hopper cars can be moved over a reclaim pit and their doors opened to allow the cargo to fall onto the reclaim conveyor below. Finally side dump cars can be used. Side dump rail cars cost about twice as much as bottom dump cars and their use is justi- fied only for smaller volumes. Table VI.6.1 describes typical hopper and gondola car unloading equipment. Figure VI.6.4 shows the general regions of economic application of the different modes of rail car unloading. This information is situation dependent and is for guidance only. FIGURE VI.6.1. COAL LOADING STATION - KEY DIAGRAM = j F ~~~Tripper Travel HI Bunker 1~~~~ AI 36.32 m Source: World Bank. - 186 - FIGURE VI.6.2 RAIL CAR LOADER Spray Head Belt Plow Conveyor Flow Diverter Two Way Discharge Chute Water Line E n m , . .100 Ton Coal Car (Typ.) Elevation - Rail Car Loading Tower Source: World Bank. FIGURE VI .6.3 RAIL CAR UUNLOADER 3" x 3" Bar PLAN-R.R.CAR UNLOADING FACILITY Grizzly ~,*JJ JLSteel Pipe FRailinA Rodding Cone engaged Connect Feed Chute Steel Beam l._ Stringer Steel Liner Plate ReclaimAA Conveyor ELEVATION RR CAR UNLOADIf, FACIIBTY Source: World Bank. - 188 - Table VI.6.1 Comparison of Rail Car Unloading Equipment Design Capacity System Rail Car Cycle time Manpower (tons/hr) Comments Standard Random 3 min 4 2000 1 Rotary dumper Lengths&sizes train 3 dumper 1 Unit train Rotary coupler 2.5 min 3 2400 1 Rotary dumper equipped cars train 2 dumper 1 Unit train Rotary coupler 110 sec 2 3200 2 Rotary dumper equipped cars dumper 2 Car positioner Under track Random hopper 5 min 6 1200 3 Hopper with cars opening& car shaker closing doors 4 train 2 Under track Mechanical cars 2 min 3 3000 4 Hopper plug in air or handling electric air or el 1 (self clearing) train 2 Under track Automated cars 30 sec 1 up to 5 Hopper (self clearing) or less observer 6000 Trestle Automated cars 30 sec 1 unlimited 6 (self clearing) or less observer Comments: 1. Plant locomotive required as road locomotive and crew not available. Time includes switching to break into groups of 20 cars or less. 2. Road locomotives stay with train and crew rides through dumper. 3. Capacity assumes 100 ton cars. Plant locomotive required. Time includes switching to break train into groups of 20 cars or less. 4. If road locomotive can be used with radio control of starting and stopping crew can be reduced by one man. 5. Train in motion at speed dependent on takeaway capacity of under track hopper. Railroad locomotives and crew utilized. 6. The train is in motion at speeds of about 3-5 miles per hour. Railroad locomotives and crew used. Source: World Bank. - 189 - FIGURE VI.6 .4 RAIL CAR ECONOMIC ZONES 100 Bottom Dump or Rotar d 80 ~Unfavo able \ Rotary Dump ary E_ -Zone .H 60 X / Use Other $.4 _ Configuration u 40 Bottom Dump Unfavorable 20 o Se 0 50 100 200 300 400 TYPE, CAPACITY OR RANGE AND NOMINAL COST PE M IATERIAL NO. OF CARS CAR iROTARY DUMP Single car 75 miles $15,000 | A <5<> s 5000 tph and up $500,000 Fines to 50-100 cars run-of-mine SIDE DUMP About 40 up to 20 $27,000 cars maximum miles $0 ~ Fines to 15-40 cars run-of-mine BUITOM DUMP To 24,000 20-350 miles $18,000 tph 35-100 cars $300,000 ~~~ ~Size Jfree flow Source: Design Considerations for Storage and Reclaim s H. (oilJn. bulk materia1s Haniding7 Vol. 1, page 360, Univeristy of Pittsburgh, 1971. - 191 - YIU. TERMINAL DESIGN AND LAYOUT VII.1 General Design Considerations There are a number of steps to follow when designing and constructing a bulk port facility. The initial steps are general and emphasize generating and evaluating alternatives. As the project progresses, alternatives are eliminated and the remaining ones are examined both technically and financially until, at the contracting stage, only one remains. After this commitment to one design is made, the engineering emphasis shifts to detailed construction planning. One way to comprehend this process is to imagine it as a spiral with engineering work converging on the design as built by successive stages of refinement. This is illustrated in Figure VII.l.l. In the conceptual phase, the goal is to develop as many alternatives as possible to ensure that no feasible solutions are overlooked. These alternatives will be evaluated in successive phases. The conceptual stage requires a great deal of field experience and employs qualitative concepts. Ideas are evaluated and screened qualitatively without having to resort to quantitative analysis. In the preliminary design phase, rough requirements for land areas, machinery types and capacities, preliminary costs and revenue estimates are made. As the design becomes more detailed with more parameters quantified, preliminarv designs can be screened by using quantitative methods. However, because alternatives are still numerous, these methods are geared to a small computational investment per alternative and to the evaluation of any changes in major design parameters, such as the number of berths. Experience indicates that queueing theory fulfills many of the requirements for this stage of evaluation. - 192 - FIGURE VII.l DESIGN PROCESS FOR A BULK HANDLING FACILITY Service Design e Requirements Reduction in Number of Candidate Solutions Machinery l _ Selection l roject I's Evaluation Pier Operating Selection D Costs Area Preparation Determination Costs Dredging Machinery Costs Requirements ill Requirements Phase ____ Concept Designs 20 man-days xXe(many) Preliminary Design 300-man-days Contract Design 5,000-man-days (very few-l or 2) 0 Detail Design 60,000-man-days Source: Planning, Layout and Design of Bulk Terminals N.J. Ferguson, XXV International Navigation Conference, 10-16 May 1981, Edinburgh, Scotland, U.K. PIANC. - 193 - In preparing the contract design (i.e., the design used to prepare documentation for bidding), the goal is to develop a complete description of what is to be done and to provide information about the site and other factors that will allow cost estimates and construction schedules to be made by the bidding contractors and by the future owner of the port. More than one contract design can be prepared if the results of the preliminary design do not indicate a clear superiority of one concept. Actual bids are required to determine the proper course of action. Following the bids and the award of the contracts, construction plans are made which tell individual workers and sub-contractors what to do and to order materials and services. The conceptual design phase is the most important, as it determines the principal design parameters that will govern the completion of the entire project. It is also the time when the interactions between the project and the country's economy, commerce and infrastructure are most easily examined. Proper development of the conceptual design phase can reduce the cost of subsequent engineering steps. Often, through the detection of fatal flaws, many alternatives can be eliminated. The principal areas covered in the conceptual design phase are the development of logistics processes that will be used and the harbor geometry each one will subsequently require. In many cases, the required geometry of the harbor (especially dredging and fill requirements) has a major impact in ranking the desirability of a design concept. Figure VII.1.2 gives some quantitative features that are considered in conceptual equipment selection. Several of these features, such as foundation requirements and cost, are susceptible to analysis in more detailed design stages. Others, such as resistance to earthquakes, are unlikely Figure VII.1.2 Quantitative Factors for Terminal Design Design Model ervice Requirements Port Machinery Wharfs Area Civil Performance Capacity Required Requirements Engineer Measure Requirements Tests Ship Cost Machinery Wharf Site Prepara- Estimate Cost Estimate Cost Est. tion Cost Operating Cost Est. Capital Cost| Annual Cost Estimate |Project Evaluatio7n - 195 - to be considered further unless special reasons for concern emerge. The requirements placed on soils to support equipment and storage piles is easily delayed until later in the project. As a result, difficult engineering problems can arise when all candidate solutions turn out to have unacceptably high site preparation costs. Initially the evaluation of soils information can only be done in a gross fashion as good test information may not be available. At the conceptual stage, the major properties of the bulk material that must be considered are density, requirements to protect it from the elements and hazards that it may present. Many kinds of materials handling equipment, although rated in tons, actually move volumes and, varying the density of the commodity, also varies the capacity of the equipment. This is particularly true of equipment designed around buckets and grabs. Differences between the average and maximum shipment rate must be accommodated by buffer storage within the system (sometimes called surge storage). "Surges" can be of short duration, as when a grab unloader discharges to a conveyor, requiring a hopper to provide intermediate surge capacity. They can also be of long duration, as when ships arrive infrequently, requiring a pile the size of the ship's cargo as surge storage. It is important that the requirements for surge storage within the system be fully assessed. Otherwise the installed equipment will never be able to operate at its rated capacity for sufficient periods of time to achieve designed performance. Figure VII.1.2 indicates a general program for the organization of the work in each stage of the design process. The advantage of this type of approach is that the initial preliminary design phase forces attention - 196 - on areas where additional engineering and test information are required and on areas where existing information is sufficient and additional work is not required. The characteristics of this approach are as follows: 1. The design of the port is in levels, each subsequently more detailed. The level of detail of information in each area is balanced and to the same level. 2. Details are postponed to final stages. 3. Formalization of project status is required at the end of each stage. 4. The project is successively refined and formalized as the design progresses until easily translated into detailed design, construction plans and designs. The above approach assumes that the goal of the port project is clear at the start and can be used to provide objectives and evaluation criteria at each stage of the design. Experience demonstrates that, from time to time, the true goals of projects are not sufficiently well- defined to provide good guidance for engineers. For example, the goal of some design projects is to demonstate that- a project is not feasible or that actual alternatives are constrained politically. It is good to posit the goals of the project early in its development and to illustrate periodically the result that such goals have on the decisions made during the design. This permits non-technical personnel to see the guidance that the goals are providing. An important general point is that the above procedure is not the only one possible for the management of port design projects. Another common, potentially successful approach is called the "systems" or "black box" approach. Here the port design project is broken down into functional areas and each is assigned to a different party to design. Detailed design specifications are worked out to ensure that the pieces operate together smoothly. This approach is required if any area of - 197 - the port design is to be approached as a sub-contract. The problem with this approach is that the details of the individual areas may become firm before many of the details of the interface requirements. This can be especially true if one design group proceeds at a more rapid pace than the others. In this case, an entire area of the design can be completed with little input about the details of other areas. Hence, the efficient use of this method requires careful study of the inter- action between the sub-systems of the project in order to ensure that they are well-understood. This approach can be used successfully to manage large projects and is common in the defense and aero-space industries. A few other approaches to port design are possible. One might be called the linear approach. This consists of doing "first-things-first." The definition of the "first thing" may either be derived functionally from the cargo handling process being used or it may be the construction activity with the longest lead time. After the first item is designed, the remainder are approached in the order established by the selection of the first. This is a common way to approach the design of "simple" systems. One result of this approach is that only a single alternative for the port configuration is, in fact, developed and this as a series of pieces instead of as a complete package. Additionally, specific decisions are made about system configuration with little assurance that they are appropriate to the problem at hand. Often it is not one's intention to approach problems in this manner. This method, however, frequently emerges when large projects are undertaken on a crash schedule. Since the engineering for a large port will require at least two years, the temptation (or necessity) arises to design long lead time items (such as breakwaters) early in - 19R - the project so that their construction can proceed simultaneously with the design of the remaining areas of the port. While occasionally unavoidable, such a method should only be adopted with a defined purpose in mind and an understanding of the risks involved. Sometimes the risk can be reduced by delaying, slightly, the detailed design of the long lead time civil works until additional preliminarv engineering can he performed in other areas. A fourth method of managing a design effort is to formulate design goals, create a single preliminary design and then refine it. This reduces the number of seriously considered alternatives to a single alternative and works well only in straightforward situations. Engineering is also likely to take longer. This is primarily because the "refinements" may, in fact, be other alternatives which are being evaluated at a level of detail greater than that which is required in a demonstration of their feasibility and cost. VII.2 Mathematical Tools A variety of mathematical tools are available to assist in the design and evaluation of port facilities. Generally technical approaches can be described as predictive or optimizing. Predictive techniques estimate the results of selecting various design parameters, while optimizing provides suggestions for their selection. Predictive techniques can be used as part of an optimization procedure when coupled with parametric variation techniques. Different techniques are used to examine different facets of the problem. These can lead to different conclusions, in Dart, because thev are based on different assumptions or solution methods. Hence, even the best thought-out technique can only be a guide. Often the cost - 199 - of machinery is established by purely commercial considerations, such as order cancellations, and may affect decisions in a manner which is difficult to model with mathematics. Table VII.2.1 lists commonly used methods and their general areas Table VII.2.1 Various Terminal Design Methods and Their Applicable Area Predict Number of Type of Capacity Operational Inventory Area Layout and or Machines Machinery of Storage Storage Required Arrangement Method Optimize Machinery Capacity Capacity Queueing Theory P Yes No Yes Yes No No sO Flow Networks P Yes No Yes Yes No No No Simulation Both Yes Yes Yes Yes Yes NO No Rule of Thumb P Some No Yes Yes Yes Yes No Inventory Theory Opt No No Some No Yes No so Searcb Technique Opt Yes Yes Yes Yes Yes Yes No Pert Both Yes No Yes No No No NO Project Parameters Yes Yes Yes Yes Yes Yes Yes Craft No No No No NO NO Yes of applicability. Table VII.2.2 gives some of the advantages and dis- advantages of each method. A "disadvantage" of many alternatives is given as requiring a computer. Most personal computers which are available today are sufficiently powerful for use with these methodologies; the disadvantage is not that the hardware is required, but that good arrangements to provide the software and programming required are frequently not made. Instead, most good engineering time is spent writing programs rather than studying the real problems involved. Sometimes this is the brightest talent in the design effort; with little difficulty the computer can restrain rather than enhance the design project. If the programs required exist or are developed in a manner that indicates planning, then, of course, the computer approach can add tremendously to the ability of the design team to consider different alternatives and approaches. Rules of Thumb Rules of thumb are common in engineering. In the design of virtually - 200 - Table VII.2.2 Mathematical Models Available for Preliminary Design Rule of thumb Advantages Disadvantages *Quick *Requires very experienced *Solution likely to be feasible personnel *Easy to justify *System likely to be over de- *Coordination of different signed and expensive groups easier *Possibility (remote) that system under-designed and won't work Queueing Theory Advantages Disadvantages *Simple *Limited set of solutions *Easy Computation *Restrictive sometimes unre- *Easy parametric variation alistic assumptions *Easy to explain and visualize *Limited applicability to continuous systems *Complicated systems must be separated into components losing some of the syste- matic interaction Flow networks-GERT-Q/GERT-V/GERT Advantages Disadvantages *Complicated systems easy *Complicated systems require to mdel computer to solve *Easy parametric evaluation *Large data analysis require- *Hand solution possible (GERT) ments *Applicable to continuous systems Simulation Advantages Disadvantages *Easy to explain and understand *Requires Computer *Most features of system can be *Large fraction of effort de- modeled to detail required voted to making computer work less to studying the *Easy parametric evaluation problem *Random variable in answer (results frequently not repeatable) Inventory Theory Advantages Disadvantages *Stresses value of items *Parameter Estimation difficult. *Stresses operational proce- *Answer very sensitive to dures as effecting facility parameter estimates design *Simple *Useful at beginning stages *Forces thought about frequently *Personnel unwilling to make ignored issues value judgements required in particular about cost and acceptable probability of stockouts Layout algorithms-CRAFT Advantages Disadvantages *Considers many alternative *Considers few factors *Forces selection of objective *Simple relations between factors evaluation criteria - 201 - *anything they play a major role. A rule of thumb is simply a standard practice whereby previously successful experiences are codified. Since it is impossible to engineer everything fully, they save considerable time and discussion. Usually they provide conservative solutions which, in practice, prove satisfactory. They provide quick answers and allow personnel to concentrate on the matters which are of real importance to the design. Another advantage to rules of thumb and standard practice is that difficult value judgements are sometimes avoided by resorting to them. Table VII.2.3 offers a few rules of thumb that are of some use in port planning. Table VII.2.3 Generallv Accepted Rules of Thumb about Port Design *Average capacity of discontinuous machines is about 60% of their maximum performance *Design capacity of conveyors serving discontinuous machines should equal their maximum performance *The rated capacity of a conveyor should be 115% of the required maximum performance Storage Capacity 1. Ship loading and unloading 1.5 to 2.5 times maximum size of vessel expected 2. Train loading and unloading 1.5 to 2 times maximum size of train 3. Barge loading 2 to 3 operating shifts 4. Truck loading Variable-try one day's throughput 5. Overland conveyor 1.5 to 2 days supply at delivery end 6. Steel plant 1.5 to 2 month's supply 7. Coke plant I to 2 months supply 8. Power plant 2 to 3 months supply 9. Cement plant 1.5 to 2 months supply They should be used, however, under the presumption that they contain no factual input from the system which is being designed. Consider the common rule of thumb, for instance, that a power plant stack should be provided - 202 - initially with a two month supply of coal kept throughout its life for emergency purposes. First, this cannot be universally true of all power plants, as some are fitted to burn alternate fuels or are connected to power grids of such magnitude that temporary loss of the plant costs less than the holding cost of the coal. Second, the recommendation is made in ignorance of the price of coal and surely the correct choice must depend in some way on the value of coal. Third, it assumes that the problem causing the interruption in the supDlv of coal will be solved in two months and will not always be solved in a shorter period. In considering this situation, it must be noted that about ten percent of the cost of the power plant is the coal which is kept only for emergencies. This rule assumes, finally, that the accept- able risk of having a power plant, either one without fuel or one operating for extended periods at part load, is the same in situations throughout the world. Few people would be willing to address explicitly the acceptable level of risk of having a power plant without fuel; this, for fear that, no matter how unlikely it may seem, the situation could occur. The rule of thumb frees people from blame through the substitution of standard practice for judgement. When using rules of thumb a good practice is to evaluate system performance by a more sophisticated means; this is to ensure that reasons for deviation from standard practice do not exist. Queueing Theory Queueing theory is the mathematical study of waiting lines (or queues). Generally it is most useful when attempting to determine the congestion that results from various levels of system use as a function of the capacity of the equipment installed. With skillful - 203 - application it can also solve a variety of other problems. In port design it has almost universal application. The main advantage of queueing theory is that it offers solutions to complicated Droblems quickly and easily. Also it permits the expeditious evaluation of different design options. Users of the theory usually have little difficulty understanding the symbols in the formulas and, in conse- quence, which queueing systems are aDDlicahle to the svstem. In other words, it can be used by rote. Solutions to unusual queueing problems are difficult to achieve. Unless an appropriate queueing model can be found in the literature, it can be safely assumed that no solution is available using this simple approach. As a result, other approaches, such as simulation, must be tried. Appendix C gives the solutions to some common queueing systems which are sufficient for solving most queueing applications. The real drawback of queueing theory is that even mildly complex networks of queues cannot be solved without separating them into smaller problems and eliminating much of the systematic interaction between queues. In these situations, simulation is an alternative. Simulation Simulation consists of making a mathematical model of a physical system and then running it to determine the behavior of the system. Systems of any practical significance in port design can only be simulated with a computer. While indispensable for complicated systems, used on simple ones simulation generally arrives at the same conclusions ob- tainable from a queueing model. As simulation models become complicated, "debugging" them become a difficult problem and it is not easy to determine if the results of a simulation stem from the fundamental model - 204 - or from logic and coding errors in the program itself. Thus it is important to try verifying simulation models either with real data or from data derived from a queueing theory model. A severe drawback of simulation is its high cost especially if qualified personnel are limited. Inventory Theory Inventory theory stresses the general operation of the facility with special emphasis on the size of shipments and their frequency. The principal variables are the cost of holding inventory and of placing orders for it. The output of the method is the optimum order size and time between orders. While the ideas behind inventory theory have great importance to the design of bulk logistics systems, most of the technical work available stresses distribution svstems for manu- factured goods and the assumptions required for these systems reduce the value of the methodology for bulk systems. In particular, the cost of placing an order is critical to the optimum lot size. It is difficult to estimate the cost of ordering a shipload of coal. Appendix D provides a brief description of inventory theory. Flow Graph Techniques Flow graph techniques study materials flow as if the system were a network of rivers; They predict the flow through each one on the basis of its capacity and the demand for the system's use. The output is the total flow through the system at a given supply rate and the flow through each of the component branches. Sometimes internal inventories within the system can be estimated. Complicated flow graph models can often be solved bv hand, but with much more computation than aueueine theorv models - 205 - require. They can also model systems nearly as complex as simulation, but not in the detail a simulation model provides. CRAFT One interesting method available for port design is called CRAFT. The CRAFT method is used to generate and evaluate layouts of industrial facilities including ports. The computer program that does this is available in FORTRAN and works well. The user specifies the area required for each activity, the material flow and transportation costs between activities. Using this information, it produces a layout mini- mizing the transportation costs between activities. Conveyor centered systems conform closely to the assumptions used in the programs. Often it is the case when using CRAFT that the resulting layouts all have fatal flaws. They stimulate thought, however, and allow the manual generation of improved layouts. There are many optimization techniques available for port planners besides those already mentioned. These range from linear programming to search techniques. Appendix E describes a few of these. A necessary, but not sufficient condition for an optimum system is that the marginal returns from investment in any component be the same. This is not a sufficient condition since it can be true of the worst as well as of the best systems. Also it can be tbe-case that there are local as well as global optimums and this condition is true at every optimum. All optimization methodologies try, by various means, to locate conditions where this is true and, of course, to distinguish the best from the worst and to find global not local optimums. - 206 - Because equipment purchase decisions are "lumpy" and the equipment itself is available in discrete frame sizes, this condition cannot be exactly fulfilled for most real systems. However, the systems where the returns from additional investment in any component are most nearly equal are usually the best that can be designed. Tables VII.2.4 shows equipment selection factors. The mathematical idea has a physical illustration which is given in Figure VII.2.2. In system I, clearly overall system performance can be increased only by uniform investment in each component, while in the clearly non-optimal system II this is not true. VII.3 Process Selection Options The starting point of a port design is the clear delineation of the goals of the project and of the general processes available to accomplish them. This process can be organized as shown in Figure VII.3.1. It begins with the determination of goals and finishes with a description of the specific process options available to accomplish them. Work at this stage should be general with a sufficient number of options generated to ensure that no factors are overlooked and that objective standards rather than subjective means are used for evaluation. In specific logistics problems the number of alternatives will be small at this stage because existing technology will only make available a few options for overall systdm evaluation. The number of options will grow only when the details of overall alternatives are considered. Ports can be constructed with various goals or missions in mind. An increasingly common goal is industrial development which gives rise to the concept of the "industrial" port. This idea is equivalent to the - 207 - Table VII.2.4 Equipment Selection Factors I. Requirements for foundations and suitability of site for same. II. Allowable dimensions for equipment which is not part of the project III. Commitment to specific ship types * self unloaders * specially designed ships IV. Reliability V. Operating costs VI. Required operating skills VII. Cost and risk of shipment to site VIII. Required construction skills--especially qualified welders IX. Effect of the environment on operations * wind and rain * sea conditions X. Effect of the environment on survivability * wind * earthquakes * seas and flooding XI. Adverse environmental impact * dust * noise XII. Infrastructure requirements * power * water * housing XIII. Reversible operation XIV. Cargo loss and damage XV. Stevedoring damage to ships - 208 - FIGURE VII.2.2 PIPELINE ANALOGY OF TRANSPORTATION PLANNING rHIS INETET NTMN NV E5TMENT EQUAL EU' CAPABI TY INVESTMENT INSUFFICIFNT IVEWN IN VESTM*.ENT Source: The Total Transportation Concept, P.J. Maddex SoUrce* Bulk Materials IAndling, Vol. 2. University of Pittsburgh, 1973 FIGURE VII.3.1 STEPS IN INITIAL PRELIMINARY DESIGN Mission Type and of ~~~~Process an1 __| of | _ Requirements Flow Goals _.Material, Facilities - 209 - American idea of the industrial park - the idea being to stimulate the development of factories and industrial facilities through the provision of high quality logistics and usable industrial land. A second similar idea is to create a regional port to handle the logistics requirements of a large area. The EUROPORT complex in Rotterdam is an example of a port developed along the regional port concept. Planning a successful port along either line is difficult, as most, if not all, of the demand for the port's service will be generated by the port itself. Planning for a port is, in large part, market research to determine the types of industry that will be attracted, and the types of logistics services that should be provided to assist in the placement of industry. Until this market research is done, little can reasonably be said about the goals for the development of actual port facilities. On the other side of the spectrum is the bulk port to be integrated into a specific industrial development project, such as a steel mill or coal mine. Here the goal is to support the development of the industrial project and the goals of the port element should follow in a straight- forward manner from the plans for the industrial project. It is useful to structure the goals of the port development as performance specifica- tions expressed primarily in monetary and throughput capacity terms. These will establish at an early date the feasibility of the marine development. The goals should stress the market for the commodity in order to determine, among other things, the best estimate of shipment sizes resulting from commercial transactions rather than the optimum shipment size resulting from a cost minimization viewpoint. While effective project management and marketing may make the two the same, they need - 210 - "not be. Investment in ports to serve optimum ships that do not arrive is wasted. In the context of an export terminal, this will determine the maximum draft required to attain a given level of product sales. The analysis of the commercial sales implications for the product also establishes the maximum allowable cost for the export operation. The maximum and minimum production volumes of the industrial facility establish the throughput levels at which the performance of the ports must be evaluated. Table VII.3.1 gives what might be an initial per- formance specification for a steam coal export project. This can be expanded into somewhat greater detail by estimating the size distribution of the vessels which actually use the port as shown in Table VII.3.2. While it is of no use in determining the performance specification, the physical properties of the material should be assembled at this time since they are used in all subsequent steps in design. Table VII.3.3 gives the properties of coal. The general processes available for coal export are shown in the flow diagram Figure VII.3.2. Each can then be developed in greater detail as in Figures VII.3.3 and VII.3.4. One process table should be prepared for every alternative available. A preliminary design for the storage reclaim and shiploading sector of this system using rail transport from the coal mine to the port illustrates the procedure. In an actual study a preliminary design for competing approaches, including slurry and conveyor transport from the mine, would also be prepared. The questions having the largest effect on the terminals cost are the following: - 211 - Table VII.3.1 Initial Performance Specification for Coal Logistics System (1982 $/ton) 5,000,000 2,000,000 Throughput tons annually tons annually Delivered price coal at customer pier $35.00 $35.00 $35.00 Estimated cost at minehead $12.50 $17.00 Maximum logistics cost allowable $22.50 $18.00 Maximum ship size 150,000 customer pier DWT Estimated maximum marine freight $12.00 $12.00 Estimated minimum marine freight $5.00 $5.00 Possible range combined inland $10.50-17.50 $6.00-13.00 and port costs Source: World Bank staff. Table VII.3.2 Ship Performance Specification for Coal Logistics System 5,000,000 2,000,000 tons annually tons annually Average ship size 95,000 DWT Maximum ship size 150,000 DWT Minimum ship size 60,000 DWT Number ship calls annually (average ship size) 53 22 Mean intership arrival time 6.9 days 16.6 days Source: World Bank staff. - 212 - Table VII.3.3 - Important Physical Properties of Coal * Angle of repose 35-40 degrees * Material strength High * Density 48 pounds/cubic foot * Flow factor .63 m-tons/m3 * Material protection None requirements * Hazards imposed by Fire, dust material Source: World Bank staff. - 213 - FIGURE VII.3.2 FLOW DIAGRAM FOR LARGE SCALE MINERAL EXPORT Rail road cars Dumper Cl Conveyor Stacker (one) Storage reclaimers (two) Surge Piles C2 Conveyors C2 (two) Sample and Sample and Weigh Weigh C3 Ship loaders C3 (two) SHIP - 214 - FIGURE VII.3.3 COAL HANDLING USING RAILWAYS OR CQOEYOR TRANSPORT SYSTEM MINING SIZING CLEANING r - sq_t ~ UsT IJ_ TRANSPORT< W f~~~~~~~V STORAGE AND X t ..ts ,. RECLAIM | e SHIP LOADING SHIP UNLOADING STORAGE END USE PC .' C@E ,. IC"m(@ Source: "The Growing Viability of Continuous Systems in The Transport of Coal", E.R. Kennett, Coal Trade Transport and Handling, C,S, Publications, Surrey, England,1981. - 215 - FIGULL VII.3.4 SYSTEM FACILITIES COMPARISON INNI RAIL MIETO owattoo Facilities Envtron Facilities Evitron FactlittiS Aspects AspCets i1NN1C Winntnd Truck loading Truck loading Trucks Coarse slurry utrt. loam Traneport Trucks - haul rords Traffic Trucks - haul roads Traffitc lpelle Du p station Density Dump station Density Haul Noise Soles roeds Dust Daust Storage Open *ir aurge/esergency Visual Open air surge/eserger.cy Visual Surge t Surgalemraeay Dust Dnst emergency slurry sterWr Noise Noise storage Prepo. Crushiag screening Noise Crushing screentng Noise Critding slurry preparation. 1U Cleaning C. Injection Storage Surge storage Visual Surge storage Visual AuLomatic in la ne coal clertng Dust Dust Cleaning pleat - voter WNste Cleaning plant - water Waste Pupa - water - mineral matter - mineral matter - Bimetal _tteU Conveyors Storage Open air storage - stackers Visual Dust Surge or emergency only Surge storage Capacity - depends train size Small surge train frequency taeel.. - reclaner - gullet feeder WRtAND Rail loading bin Dust In line main feed Pump station !3SPOtT ialloon loop - service siding Noise - workshops Right of my Land tight of way Land light of way none cleaning use cleaning use Pipeline wig fencing Noise fencing Noise route retured to bridges (heavy) Spil ge bridges (light) Spil 1e original use overpass Visual overpass Visual underpass underpass Access occasional track Access rad - full length Access to pump 1uses Material handling I Material handling Slurry handling Rolling stock (!ocos wagon) Counvyors - fire service Fipeline Ballast track rwintenance Traasfer points-power rupply Pump houses - powar spply Controls signalling eaintenance spares 6 equip Controls S co_munictiqs Counicati-as Controls Dump station Communications Conveyors - sanpling Sampling Sampling ?QT STORAGZ 6 Storage area - open air Storage area - open sir Stcrag ponds uirer itr ILTERIALS SLacking mactl.ni Visual Stacking machine Visual Ate storage NDLING Reclaim machires Dust Reclaim merLines Dust LSustr mrate Conveyors - sampling Noise Conveyors - sampling Noise Pump reclaim dust Tire Fire Pipeline EIPLOADINC Harbour - protection Seashr' harbour - protection Seashr None - dredging use - dredging use Wharf - fendering Recrea. Wharf - fendering Recre-. - services-power supply - ervizes-power supply Conveyors - fire Dust Conveyors - fire_t Dust Undersea pipeline to bsy dustI protectior. Noise duet- protection Noise Mechanical loader Fire Mechanical loser Ftre _IPPN 60,000 - 160,000 D.W.T. 60,000 - 160,000 D.W.T. 250,000 - 300,000 D.W.T. UIP UILOADING drbo.r - protectionr Se*shr Harbour - protection Seashr' None - dredging use - dredgiog use Whaif - fenderio g Racres. Wharf - fendering Recrea. Offshore buoy - erv1ces-p..er supply - servtces-powr supply Mechanical u-loacers - bac,h Dust l>chanicl laders - batch Dust Undersea piplinre to buoy Conveycas :loise Conveyor. Noise Fire Fire STOtAGC Storage area - open air Visual Stnrrgc reas - open air Visual Pond storage - inter stereo Mach. stacker Dust Mech. stack.r Dust Nech. reclaimsr Meh. reclalncr Danveyore - fire Noise Corvey.rv - fire.1 Not3e dustJ protection Ftre a uscl protection Fire _ _ _ _ __ __ _I,_- ---_ ___ _ _ - ___ _____ STRIdUTION looding bin Tralt1c load'rg bin Traffic .__ -- -- , deniLLy r _ density1 Pump raciste. road conveyor ro ad Conveyor trucks tail R.O.W. Notse tr,.ke ratl l O.W. Noise FLpeline dump i_Duet d -Dust uItU Opetn air st.rage - fire; Notse Open air trro6c - fire 11, _ ,.t]protect Visual dtlprotect Vioual Pond storage Crfisdng Dust Crinding Dust de Water Use Fire U.e Fire Use Source; "The GrojinR ViabilitY of Continuous Svsteas in The Trtnsport if Coalj , 'ennett, Coa3 Trade ransport Knd Handl.ing, C.S'u. Publicatiolts, Wurrey, FfingLanfs, lw L - 216 - 1. Number of ship loaders 2. The use of conveyor c-2 intended to allow direct leeding of coal without stacking and reclaiming 3. The use of the surge storage pile 4. The number of stackers and reclaimers installed Table VII.3.4 identifies the types of decisions necessary during preliminary design. Table VII.3.4 - Major Decision Variables for Coal Port Design Install Type Number Size Variable or not Number of Berths X X Ship loader X X X Conveyor (C-2) X Reclaimers X X X X Stackers X X X X Surge Storage (S-1 X X Storage (S-2) _ _ X VII.4 Terminal Design Optimization VII.4.1 Number of Berths It is impossible to precisely schedule ship arrivals at a port, because a vessel's average speed is heavily dependent on the weather. The variance in over the ground speed accounted for by the weather is easily ±1.5 knots which indicates that the arrival time of a bulk carrier on a 5,000 mile voyage will not be known in advance to an accuracy greater than three days. In addition, for heavily laden vessels requirements to wait for tides create additional random delays expected to be hours long in duration. - 217 - However, the average occurrences of winds, waves, and currents on the various portions of a route are known by season and can be included in the calculations of the voyage time. Deviations from the average always occur and, therefore, off-schedule arrival is a normal phenomenon. When taking the arrival of all vessels at a terminal facility into account, it is obvious that this phenomenon of early or late arrivals will create a pattern that rather closely fits that of a random distribu- tion. The selection of the number of berths required for a particular alternative will be based on Mettam's ship queueing theory. This theory establishes the relation between berth occupancy and the waiting time ratio. This relation is graphically shown in Figure VII.4.1.1. The broken lines present the case of uniform service time, whereas the full lines present the case of varying service time (the variation is exponential). An important consideration is the manner in which ships are procedurally entered in the queue. As the days remaining in a voyage decrease, the ship's arrival time becomes increasingly more certain. During the last four days of the voyage some flexibility exists to advance or retard the physical arrival time by changing the vessel's speed. In the last stages insufficient time exists to allow a sub- stantial advancement of the vessel's arrival time, and only delaying arrival through slow steaming is possible. Ports differ in the way they deal with this. One philosophy is that entrance to the berth should be on a first come, first served queue with entry to the port being the same as entry into the queue. Other ports allow entry into the queue to be made while the vessel is - 21S - FIGURE VII.4.1.1 QUEUEING ANALYSIS OF WAITING TIMES FOR SHIPS IN PORT 3.0 uniform service time 2.5 X j -----exponential service time QH~ InraigI D. E--4,, 4-)~~~~~~~~~~~~i bGo Source: IForeasting Dly oSisi ot .D etm 2.0a) 0 Cd ~ ~ ~ ~ ~ ~ r4 0) 1.5> Source: Inreasting De /y to Sipsi ot .D et The k and ar thory, London, England, .April 1967. - 219 - at sea allowing it to wait in the queue while underway, enabling it to-wait in the queue at no financial cost. Systems using this approach grant "preferential" days to vessels on the basis of their planned arrival schedules with the ship losing its place if late. Svstems such as this allow higher berth utilization for a given wait than the first arrival discipline. Utilizing the results of the aueueina model together with rough estimates of the marginal costs of additional berths, a graph similar to Figure VII.4.1.2 can be developed. In the case of the given example, a single berth is adequate for the high production figure of 5,000,000 tons and the low of 2,000,000. VII.4.2 Capacity of Shiploader The required capacity of the shiploader can be calculated using queueing theory as shown in an example in Table VII.4.2.1. The service time of the system is modeled with an Erlang distribution which allows the shape of the distributions of service times to be varied (see Figure VII.4.2.1). The value of K ranging from 1 (if the service time is exponentially distributed) to 0 (if the service time is constant). In this case, the difference in service times caused by the differing ship sizes is critical. The maximum ship sizes is critical. The maximum ship size is 150,000 dwt, the smallest 60,000 and the mean 95,000. To fit the Erlang distribution, the standard deviation of the ship sizes arriving at the loader must be estimated. This can be achieved either by setting policies for ship procurement or by taking the information available and estimating the standard deviation using the beta dis- tribution. The constant, K, for Erlang distribution is then calculated - 220 - FIGURE VII.4.1.2 QUEUEING THEORY IN PRELIMINARY PLANNING $M/Yr 180- Total Annual Cost 160 - including ship time in port-i berth system 140 - 120 - 100 / 100~ ~ ~ ~ _o'ota~l Cost including ship time 80 _ / in port - 2 berth _0 - system Capital + Operating 60 - Costs (2 berth) ~-~~zaptal+ OeraingCosts 40 - (l berth) Capital Cost - 2 berth system 20 - Capital Cost - 1 berth system Expected operating range 0 , , I I I 0 10 20 Millions of tons Note: In practice stockpile capacity, rail capacity, etc. would limit the ultimate throughput. - 10,000 tph reclaim and loading system - maximum ship 150,000 DWT - 221 - Table VII.4.2.1 Example of Estimation of Number of Berths and Optimum Size of Ship Loader A. Estimate Arrival Distribution Largest Ship = 150,000 DIT Smallest Ship= 60,000 DWT Mean Ship = 95,000 DWT To get the standard deviation, simply use the Beta distribution -2 _ [150,000 - 60,000]2 6 a2 = 15,0002 calculate K for Erlang Distribution K = 95,0002 = 40.11 '--2 1F5, 000' X 5,000,000 = 52.6 ships/year = .14 ships/day 95,000 Loader Capacity Unloading Time i (ships/day) (tons/hr) (hours) 2000 47.5 + 12 - 59.5 .4 4000 23.8 + 12 = 35.8 .67 6000 15.8 + 12 = 27.8 .86 10000 9.5 + 12 = 21.5 1.12 12000 7.9 + 12 = 19.9 1.21 Mean waiting = R +1 x days 2K u(' -A) Loader Capacity Ship waiting cost ($20,000/ship-day) 2000 .68 days x 20,000 x 52 - 707,000 4000 .20 days x 20,000 x 52 - 208,000 6000 .115 days x 20,000 x 52 - 119,000 10000 .06 days x 20,000 x 52 = 62,000 12000 .055 days x 20,000 x 52 - 57,000 Source; World Bank staff. - 222 - FIGURE VII.4.2.1 TOTAL COSTS VS. SHIP LOADER CAPACITY 0 Op timum T \ Total Costs \ 8 =;=er.-'--':~ e~t1h Costs Ship Wait ing i - C~~~~~~osts 1 2 3 4 5 6 7 8 9 1() 11 12 SHIP LOADER C-APACITY (000 's t/hr) Source: World Bank staff . - 223 - &nd the formulas given in Appendix C are used to calculate the waiting time and total system cost as shown in Figure VII.4.2.1. Minimum total cost is achieved with shiploading capacity in the region of 5500 tons/ hour on the average. Assuming the vessel has eight holds of 15,000 tons, 2.73 hours are allowed per hold and half an hour is required to move the loader to the next hold. This leaves 2.2 hours for the actual loading operation. Thus, the rated capacity of the loader must be 6,818 or 7,000 tons per hour, in order to achieve an average rate of 5500 tons/hr. The size of the vessel which is being loaded allows either one slewing loader on a finger pier, one quadrant loader (if the ship is moved during loading) or two quadrant loaders (if the ship is not to be moved). For the same loading time, it is found from the results of a comparative analysis of the three alternatives, two quadrant loaders of 4,000 tons per hour capacity each is the best choice. VII.4.3 Storage System Selection The coal waiting for shipments can either be stored at the mine and loaded directly aboard the ship after rail car unloading via the bypass conveyor, or it can be stored in a stacking reclaim yard. A third alternative of combining the two, that is, storing only a portion in the reclsim yard, also exists. The option of silo storage would be worth considering were land at a premium and annual volumes small. Storage in parked rail cars is not required, because blending require- ments are not substantial. Since this system will be at least twice;as expensive as the other alternatives, it will not be considered. The basic options available are shown in Figure VII.4.3.1. - 224 - FIGURE VII.4.3.1 STORAGE OPTIONS t OPTION I - OPTION II OPTION III - 225 - Table VII.4.3.1 presents the equipment and equipment utilization as a function of the number and size of the reclaiming capacity installed. The direct loading option with no in-port storage is for the system with no reclaiming capacity installed. Only costs and information not common between options are presented. Table VII.4.3.1 Equipment Capacities as Function of Reclaimer Capacity Reclaimer D of Yard Yard Max Average # of Dumper Surge Average Production Reciaimers Production Surge Dump Car Duiup D umpers Storage Size (tono/hr) (tons/hr) (tone) (tons) (tons) (tons) 2,000 1 2,000 0 5,000 3,500 2 x 2,400 1100 2 4,000 0 3,000 1,500 1 x 2,400 0 3 6,000 1,500 0 0 1 x 1 7001/ 0 4 8,000 0 0 0 1x 1,70t/ 0 3,000 1 3,000 0 4,000 2,500 1 x 4,000 1100 2 6,000 1,500 0 0 1 x 1,7001/ 0 3 9,000 0 0 0 1x 1,700t/ 0 None 0 -- - 7.000 5,500 2 x 3,200 3300 Source: World Bank staff. Table VII.4.3.2 Storage System Costs (1000's USD) _Capacity_ Stackers Reclaimers Yard Cars Dump Pit Sw Pit Total (tonis/hr) 2,000 1 800 2,750 0 10,227 1,200 2/ 330 3/ 15,300 2 1400 5,50C 0 7.404 600 0 14,900 3 2000 8,250 550 4,500 400 0 15,700 4 2000 11,000 0 4,500 400 0 17,900 3,000 1 1000 3,500 0 8,590 1,50G 330 14,920 2 2,000 7,000 550 4,500 400 0 0 14,450 3 2,000 10,5Q0 0 4,5G0 400 0 0 17,400 None 0 - - 13,500 3,000 660 17,160 1/ Dump train in 4 hours 2 trains - 1 shift 2/ Estimated from dump pit costs at Puertod E dana 3/ Estimated from Surge storage pit cost Puerto de Rain& Source: World Bank staff. From the calculations of the shiploader capacity, the maximum requirement will be for 7,000 tons/hr of material and average 5,500 tons/hr. For systems which have a delivery capacity in excess of 7,000 tons per hour, no surge capacity is required. For systems with delivery capacities between 5,500 and 7,000 tons/hr surge storage is required at some point in the system. When the option utilizes direct loading from rail cars this surge storage is most cheaply incorporated into - 226 - the railcar unloader. For other systems this surge storage is incorporated at the output of the reclaimers. As the contents of the surge storage goes to zero just prior to the completion of the loading of the hold, its size is determined by the time required to load each hold and the difference between the capacity of the system and the peak of 7,000 tons per hour. This means that the maximum hold size of vessels calling is an important parameter which should be verified. The costs of the options are given in Table VII.4.3.2. of the previous page. Note that the direct loading and the purely port storage option are very close in cost. In a real situation, additional factors would have to be considered. Among these are the following: 1. The probable higher cost of stock pile site preparation when the site is located near the water rather than at the mine. 2. The cost of the direct loading option is not burdened with the cost of additional coal storage facilities at the mine. These will cost at least $5 million (U.S.) and should indicate that the best choice is Option III. As the design with two surge piles at the discharge is the cheapest, it will be selected for further development. The surge pile must have a size sufficient to hold the production of the reclaimers for the period required to move the ship loader from one hold to the next. This was estimated to be one-half hour. The piles must then hold 1,5OOM tons each, the production of a single reclaimer for half an hour. A simple conical pile with tunnel reclaim is chosen. Because of the large throughput utilization of materials, the dead-storage area in the pile is not possible, and the pile must be made larger. The pile - 227 - itself is described in Table VII.4.3.3 below. Table VI.4.3.3 Stockpile Size and Volume Size of surge storage pile capacity of pile = 1,500 tons density of coal 45 lbs/cubic/foot angle repose 400 Volume Live Storage = 1,500 x 2,280 76,000 cubic feet 45 ¶rr2H -irr3 tann a 28r3 Volume Cone = wr H ___ana2 3 3 Dead Storage Section Area - 1/2 * - * r * tan a * 2 = _4_ 2 2 4 Length = irr Live Storage Volume = r 3w(0.28 - 0.185) 76,000 = 0.095wr3 3 r 253,000 r = 63 feet h = 63 x 0.84 = 53 ft Source; World Bank staff. The last storage area to design is the attached reclaim stock yard. The inventory carried in the stock pile follows a pattern shown in Figure VII.4.3.2. The minimum level of inventory in the yard is such that, if a ship should arrive perchance at more frequent intervals than the mean, then there would be sufficient ability to load it. The maximum level is set by a desire not to be forced to shut down the production of coal at the mine should the ship arrive late. - 228 - FIGURE VII.4.3.2 TYPICAL VARIATION IN DRY BULK CARGO TERMINAL INVENTORY LEVEL EXPORT STOCKPILE I N …_ __ _____ Maximum V L E E -_ -- Mean N V T E u RR L R O L _linimum R Time L - Shiploading period R = Replenishment period Source: Port Development, UNCTAD Secretariat, U.N. 1978. - 229 - To do this, acceptable levels of risk must be established for each event. The early arrival of a ship at port has a higher acceptable level of risk because of the following: 1. The railroad supplying coal was designed for operation on a single shift and has a larger capacity if operated for longer periods. 2. The loss, if sufficient coal is unavailable, is demurrage for the longer period required to load the vessel (about $10,000/day). On the other hand, the loss from a ship which is late is the value of the coal which could have been produced; this amounts to about $80,000 per day. In addition, relations with customers are sometimes jeopardized if production schedules cannot be met. Finally, as this maximum inventory level is only the capacity, not the actual inventory level, there is no holding cost associated with it, only the additional capital investment required to install the larger storage capacity. To estimate the minimum level of coal inventory, one must calculate the probability that the time between vessels will be less than a given value. Storage must then be equal to the amount that would be supplied during the interval between the early and expected arrival of the ship, since this is the position of the replenishment cycle not available for uninterrupted loading of the ship. The minimum time between ships is, of course, zero, as it is always possible that one will be loaded immediately after the completion of the loading of the first. The maximum time between vessels is probably close to the case where a ship was lost at sea or delayed due to mechanical problems. Thus, the maximum time between ships is probably twice the mean. With this information, it is possible to estimate the standard deviation of the inter-arrival times using the beta distribution. - 230 - For this distribution: Standard deviation = (Maximum time - Minimum time)/6 This can then be used with the cumulative normal distribution to determine the probability that the time between ships is under a given value. This is used next to estimate the stock required, as shown in Table VII.4.3.4. Table VII.4.3.4 Safety Stock Calculation Minimum Time Between Arrivals = 0 Mean Time Between Arrivals = 6.9 days Maximum Time Between Arrivals = 13.8 days Standard deviation = (13.8 - 0)/6 = 2.3 days It is decided that a probability of 0.99 that shiploading is not delayed for lack of stock is acceptable. This is equivalent to a time between stockouts of 2 years. Using the normal distribution area = .01 k = 2.33 when the shaded area is .01. The time period is then 2.33 x 2.3 = 5.36 days The production during this period is 72,800 tons which is the minimum safety stock. _j, _i.__safety stock = 72,800 Source! Wbrld Bank staff. The matter of late ship arrival requires a more stringent view. It is important to protect the system from becoming inoperable because of excess stocks so that such will be the case only once in twenty years. - 231 - This occurs when the probability of the inter-arrival time being greater than the mean is .9991. When this is the case K=3.1, the number of days is 3.1 x 2.3 or 7.13 days. The required stock to meet this criterion is 7.13 times the railroad's mean delivery rate or 98,400 tons. The mean level of material in the stockyard is 95,000 tons or the average ship size. In a real problem more scrutiny should be given to the assumptions behind this analysis. In particular, the probabilities of a stock- out or being full should be treated as parameters and the costs of varying levels of protection estimated. On the basis of this analysis, the maximum capacity of the stock- yard should be 95,000 tons to load one ship and 98,400 tons for protection against late ship arrival or 193,400 tons, say 200,000 tons. 72,800 tons will be "permanently" stored in the facility to protect against early ship arrivals and the inventory level will fluctuate by 95,000 tons as a result of ship loading operations. General options for the layout of the yard are given in Figure VII. 4.3.3. These arrangements derive directly from increasing the number of stackers serving the two reclaimers from the minimum of one to the maximum of 3. The capacity of the storage yard is given in the following formula: capacity = L x A x H x K x N where, L = Length of piles * ~~A = Width of piles H = Height of piles K = factor depending on the angle of repose to account for the shape of piles - 232 - FIGURE VTI.4.3.3 OPTIONS FOR STOCKYARD LAYOUT O Stacker © = Reclaimer = R ~~~~~~~ o ~~ ~ ~~~~~~~~ _____, Source: World Bank staff. TABLE VII.4.3.5 GEOMETRIES OF PILES Cross Total 3 Piles 4 Piles Section Capacity for Length of Length Length Width Height Area Coal Storage Pile Required 1 Pile I Pile (E) (m) (m3) (m. tons/m) (m) (m) (mn) 35 13.2 254 160 1,650 550 412 40 15.0 332 209 956 318 239 45 16.0 415 261 766 255 191 Source: World Bank staff. - 233 - N = Number of piles. This formula neglects the semi-conical shaped portion on the end of the stacks. This is a small error at the initial design stages. Table VII.4.3.5 of the previous page, gives the geometry of the pile system required by stackers and reclaimers of various widths. The basic geometries of the piles are given in Table VII.4.3.6. This information was then used to estimate the costs of each alternative. This comparative cost is given in Table VII.4.3.7. Note that the land development cost is an important parameter in bulk ports as much or all of the stockyard may be built on filled land and the alternatives differ greatly in land requirements. If the land cost is $50 a square meter, the lowest cost solution is for one stacker and two reclaimers. An effective method for pile capacity computation is described here. Volume of material that can be stockpiled in elongated form in an area of 115 feet wide and 415 feet long is calculated as follows: a) Assuming that the material has 350 an le of repose, the ends volume of the conical pile is 5050 yd b) The volume of prism in yd3 is V= (length)(width)(height) (415 - 115)(115)(40) 2 x 27 ft3/yd3 2 x 27 = 22,550 yd3 c) Total Volume = 5050 + 22,550 = 27,600 The method of computing radial stockpile volume can be readily compared with that of elongated stockpiles. A computation method for the amount of bulk material that can be withdrawn from a conical stockpile, is similarly easily developed. - 234 - TABLE VII.4.3.6 GEOMETRY OF PILE SYSTEM < < rA 0 -.4~~~~ E-.O0 PILE WIDTH f 1 e0 W W M 3 V. A B C D E F G H hI h2 RR SR 30 10.0 7.0 1.5 8.0 6.0 1.0 11.5 10.5 1.0 30 20 35 10.0 7.0 1.5 8.0 6.0 1.0 13.2 12.2 1.0 35 22.5 40 11.0 8.0 1.5 10.0 7.0 1.5 15.0 14.0 1.0 40 25 45 11.0 8.0 1.5 10.0 7.0 1.5 16.0 15.0 1.0 45 27.5 50 14.0 10.0 2,0 10.0 7.0 1.5 16.0 15.0 1.0 50 35 55 14.0 10.0 2,0 11.0 8.0 1.5 16.0 15.0 1.0 55 40 RR Reclaimer Boom Radius SR Stacker Boom Radius C A .~ ~~ Pil iidth (Reclaimer PStacker Gauge C-auge Source: World Bank staff. - 235 - Table VII.4.3.7 Comparative Costs for Stockpile Alternatives COST OF LAND DEVELOPMENT $25 PER SQUARE METER PILE WIDTH 35 AREA OF STOCKYARD 70,000 NUMBER PILES LENGTH TRACK # STACKERS # RECLAIMERS COST 2 2,475 1 2 15,987.5 3 2,200 2 2 18,850 4 2,062.5 3 2 21,781.25 PILE WIDTH 40 AREA OF STOCKYARD 45,000 NUMBER PILES LENGTH TRACK # STACKERS # RECLAIMERS COST 2 1,434 1 2 16,042 3 1,274.667 2 2 19,362.33 4 1,195 3 2 22,722.5 PILE WIDTH 45 AREA OF STOCKYARD 40,000 NUMBER PILES LENGTH TRACK # STACKERS # RECLAIMERS COST 2 1,149 1 2 18,774.5 3 1,021.333 2 2 23,110.67 4 957.5 3 2 27,478.75 COST OF LAND DEVELOPMENT $50 PER SQUARE METER PILE WIDTH 35 AREA OF STOCKYARD 70,000 NUMBER PILES LENGTH TRACK # STACKERS # RECLAIMERS COST 2 2,475 1 2 17,737.5 3 2,200 2 2 20,600 4 2,062 3 2 23,531.25 PILE WIDTH 40 AREA OF STOCKYARD 45,000 NUMBER PILES LENGTH TRACK # STACKERS # RECLAIMERS COST 2 1,434 1 2 17,167 3 1,274.667 2 2 20,487.33 4 1,195 3 2 23,847.5 PILE WIDTH 45 AREA OF STOCKYARD 40,000 NUMBER PILES LENGTH TRACK # STACKERS # RECLAIMERS COST 2 1,149 1 2 19,774.5 3 1,021.333 2 2 24,110.67 4 957.5 3 2 28,478.75 (Table continues on the following page) - 236 - Table VII.4.3.7 Comparative Costs for Stockpile Alternatives (cont'd) COST OF LAND DEVELOPMENT $75 PER SQUARE METER PILE WIDTH 35 AREA OF STOCKYARD 70,000 NUMBER PILES LENGTH TRACK # STACKERS # RECLAIMERS COST 2 2,475 1 2 19,487.5 3 2,200 2 2 22,350 4 2,062.5 3 2 25,281.25 PILE WIDTH 40 AREA OF STOCKYARD 45,000 NUMBER PILES LENGTH TRACK # STACKERS # RECLAIMERS COST 2 1,434 1 2 18,292 3 1,274.667 2 2 21,612.34 4 1,195 3 2 24,972.5 PILE WIDTH 45 AREA OF STOCKYARD 40,000 NUMBER PILES LENGTH TRACK # STACKERS # RECLAIMETS COST 2 1,149 1 2 20,774.5 3 1,021.333 2 2 25,110.67 4 957.5 3 2 29,478.75 Source: World Bank staff. 237 - The stacker's average production is 1,725 tons per hour when 13.8 tons of coal are unloaded from trains in an 8 hour shift. Thus, the rated capacity of the stacker should be 1.15 x 1,725 - 1,983 or about 2,000 tons per hour. The rail car dumper must be able to dump cars at a rate of 1,800 tons per hour. The conveyor C-1 must have a capacity of 2,000 tons per hour. The conveyors C-2 and C-3 must be sized to accept the maximum digging output of 3,000 ton per hour average per reclaimer. This maximum digging output is about 3,000 tons per hour. The conveyor loaded by the reclaim tunnel must be of sufficient capacity to supply the shiploader. This is 4,000 tons per hour. The belts in the system are selected on the basis of Table VII.4.3.8. The rest of the equipment installed is listed in Table VII.4.3.9. A cost estimate is then prepared; a model is shown in Table VII.4.3. 10. Given the gross nature of the design and failure to consider dredging and other costs in any detail, the contingency portion of this cost estimate is large. As the design progresses to the eenter of the design spiral, this contingency is replaced with estimates for specifica- tions. Prior to the refinement of this design there are several areas where improvements may be possible. First, it would be desirable to reduce the speed of operation of Belts C-2 and C-3, and to eliminate the surge piles at their discharge. This could be done by installing 3-3,000 ton per hour reclaimers instead of 2. The evaluation of this option would be done in a manner similar to the example used in this text. Figures VII.4.3.4 and 5 show examples of bulk loading and unloading systems. - 238 - TABLE VII.4.3.8. Guide to Belt Selection Belt Capacity* Speed Width (tons/hr) (ft/min) (inches) C-1 2,000 500 60" C-2EC-3 6,000 900 72" C-3EC-5 4,000 700 72" *Assuming 40 lbs/ft2 belt loading Source: World Bank. TABLE VII.4.3.9 Other Terminal Equipment Installed Berths One Dredged to 55 feet to accept 150,000 ton ship Ship Loaders Two - quadrant type Required cream capacity 3,500 tons/hour Pated capac-ty 3,0nn _ 1.15 = b,000 tons/hour Outreach of hooii AN neters Reclaimers Two - pile width 40 meters Required rated capacity 3,000 tons/hour Theoretical capacity - 6,000 cubic meters/hour Stackers One - pile width 40 meters Rated capacity 2,000 tons/hour Rail Car Dumper One Rated Capacity 1,800 tons/hour Source; World Bank. TABLE VII.4.3.10 Initial Coal Port Estimate (USD) Berth 5,000,000 Shiploader 12,000,000 Stacking/reclaim yard 17,167,000 Rail dumper 3,200,000 Misc. and contingency 26,160,000 (70% of above) Initial Estimate Port Cost 63,520,000 Source: World Bank. - 239 - FIGURE VII.4.3.4 IRON PELLET IMPORTS *-To Steel Mill 2 Reclaimers 2 Stackers 2 Ship IJnloaders Design Capacity 3,200,000 tons/year Equipment 2 1000 ton/hour grab unloaders 2 3000 ton/hour belt conveyors Subtotal (ship unloading) 2 3000 ton/hour stackers 2 500 ton/hour reclaimers Stacking belt conveyor Reclaiming belt conveyor Subtotal (storage and reclaim) Dredging and quay construction Source: World Bank Project Files - 2q0 _ FIGURE VII.4.3,5 COAL EXPORT TERMINAL -J~~Lt C-z '.'3 L~~ 'X':;" 'i'UWLUL I Layout of Pacific Coal Corp's Portland, Oregon, coal terminal showing: (1) total storage capacity of 1.7 mt; (2) 3.8 million gallon retention pond; (3) stacker/reclaimers (the second unit will be added in the event of a Phase 2 expansion); (4) rail loop serving Burlington Northern and Union Pacific unit trains; (5) wagon dumper shed; (6) 6600 travelling shiploader. Source: Bulk Systems International, Sept. 1982. - 241 - VIII. TERMINAL COST ESTIMATION VIII.1 Introduction To estimate the cost of bulk terminals, one should proceed by refining initial comparative prices to decision prices, then to contract prices, and finally, to actual costs. Comparative prices are used to judge the relative desirability of alternatives and to evaluate the feasibility of a project at the initial stages. In later stages, one needs more accurate information to arrange for contracts, equipment purchases, and financing. Costs, other than the initial comparative costs, must be obtained by quotations and bids. Real world prices vary continuosly with business conditions, currency exchange rates, and work loads at particular factories. Most bulk port equipment must be custom-designed for specific situations, and that which is not, is available with many optional features. Factors, such as the power source (on board or external electrical supply) and a sophisticated control system easily can account for 30% of the variations in equipment cost. The initial comparative costs have to be only accurate enough to guide designers. While a high degree of accuracy is in itself desirable, too great an emphasis on the accuracy of costs in the initial stages can force design decisions to be made that are better postponed to later stages of the design. Port development costs can be divided into site preparation and development costs, and equipment costs, although further division within these groups is possible. Figure VIII.1.1 shows a break do-rm of the costs for the Port Kembla coal loader. The figure provides the initial cost estimate for the port, and the results of an analysis of bids for the orolect by construction companies. For this new Dort, about severnty Percent of - 242 - the cost is for equipment and equipment installation (excluding foundation construction which is accounted for separately). The exact proportion of construction costs in each group varies among projects. Some port development projects re-use existing berths, channels, and other portions of existing ports, and others require complete turnkey development. Table VIII.1.1-VIII.1.3 analyze cost estimates from the engineering feasibility studies of three ports. The tables give the percentage of groups of major project costs and shows the wide divergence in the break down between equipment and sitework. Non-equipment related costs vary from a low of 12 percent to a high of 52 percent. FIGURE VIII.1.1 PORT KEMBLA COAL LOADER ESTIMATES AND TENDER RESPONSES (Figures in Millions of Australian Dollars) MATERIAL SITEVORKS FOUNDATIONS KARINE HADLING SRIPLOADER STACERS RECLAIMS 25 25 20 20 15 1 10 10 5 5 0 ~~~~~~~~~~~~~~~~~~~~0 g Tender sums received 0 Estimate Source: Planning, Layout and Design of Bulk.Terminals, N.J. Ferguson, XXV International Navigation Conference. 10-16 May 1981, Edinburgh, Scotland, U.K. Permanent International Association of Navigation Congresses. - 243 - Table VIII.l.1 Cost Breakdown--Port 1 Equipment % Total 2-1000 ton/hour grab unloaders 22 2-3000 ton/hour belt conveyors 13 2-3000 ton/hour stackers 8 2-5000 ton/hour reclaimers 9 stack in between conveyor 17 reclaiming belt conveyor 19 Sitework dredging and quay construction 12 100 Table VIII.1.2 Cost Breakdown--Port 2 Equipment % Total Shiploader 20 Stackers 9 Reclaimers 16 Conveyors 27 Sitework Sitework 11 Foundations 8 Marine 9 Subtotal sitework 100 Table VIII.1.3 Cost Breakdown--Port 3 Equipment % Total Coal unloader 16 Electricity 5 Conveyors & transfer towers 27 Sitework Pier construction 16 Conveyor foundations 9 Miscellaneous Land 4 Contingency 14 Coal pile 9 100 Source: For Tables III.1 1-3 World Bank project files. - 244 - VIII.2 Civil Engineering Costs Civil engineering costs include dredging, pier and foundation construction, land fill, road building and other construction activities. In the initial design phases of a project, one should estimate unit construc- tion costs at the site(s) in question with each alternative port using the same costs. In estimating these costs, it is important to consider the cost of construction materials, such as rock and cement at the site, mobilization, and construction costs per unit of work done. A good way to estimate dredging costs is to use dredging cost figures from similar, already-existing ports, but the cost of moving the dredging equipment to the port and differences in the cost of removing the spoils (the material removed by the dredge) must be considered on a case by case basis. The cost of removing the spoils usually cannot be estimated from previous dredging work, because the distances the spoils must be transported before they are dumped vary. Bottom soils to be dredged can be grouped into four categories: Type 1 - Loose soils, such as lagoon deposits and sand Type 2 - Somewhat cemented soils, which can be dredged with a suction dredge, cutter dredge or bucket ladder dredge Type 3 - Hard soils, that require a powerful cutter dredge Type 4 - Very hard ground, which requires loosening, usually with explosives, before it can be dredged. Most port projects involve some of each type of dredging. A recent port project, financed by the World Bank established the following unit costs (in 1983) for dredging: - 245 - Type 1 - Loose soils $1.67 per cubic meter Type 2 - Somewhat cement soils $5.00 per cubic meter Type 3 - Hard soil $11.97 per cubic meter Type 4 - Very hard ground not estimated These costs illustrate only the order of magnitude of the dredging costs. Unit costs, for example, vary depending on the volume dredged (they fall as more is dredged). Thus, if the amounts vary significantly among port alternatives, it could result in different project costs. Site exploratory work is required before dependable estimates can be made. Land fill is, in a sense, the opposite of dredging. After the fill is deposited, another layer of fill, called a surcharge, is usually required to be put over the original fill to compact it. It may be several years before the surcharge can be removed and the land used. Frequently, the dredged spoils are not suitable for land fill, and filling material must be brought to the site. As a result, cost estimates for land fill are even more site-sensitive than dredging estimates. When earth fill is obtained dry, transported by truck and put into place with a bulldozer, a figure of $1.67 per cubic meter was used to estimate the cost in the project above. The cost to compact the fill, remove the surcharge, pave, and pay for utilities was assumed to be $25.00 per square meter. The construction costs of breakwaters can be estimated from the costs of delivering and installing rock on the site. Breakwaters are usually constructed of an internal rubble fill with an exterior facing to protect the interior. Figure VIII.2.1 shows the cost estimate for such a breakwater. - 246 - FIGURE VIII.2.1 BREAKWATER COST ESTIMATION - Quarry run 0 to 500 kg for breakwater laid from land 7 $/m3 - Quarry run 0 to 500 kg for breakwater laid by sea 12 $/m3 - Quarry run 50 to 1000 kg for breakwater laid by sea 13 $/m3 - Rock material 20 to 200 kg or 50 to 1000 kg laid by sea 30 $/m3 - Rock material 1 to 3 ton 25 $/m3 For example, for a breakwater -10,00 m deep rising to + 2.00 above sealevel: 6.00 -0.00 Quarry run core cost per meter: $13 x (6,00 + 39) x 11 = 247.50 m3 x 13 -$ 3,217 2 Facing of natural blocks cost per meter: Total volume (6 + 42) x 12 -247.5 = '288.00 - 247.5 = 40.5 n3 2 Cost per meter - 40.5 x $25 = $1,012 giving an average price per m3 of :$14.7/m3 approx. - 247 - FIGURE VIII.2.2 QUAY COSTS PER LINEAR METER USD (1983 USD) 14000 . 13000 12000 2nd 11000 ~~~~Category 71000 10000 lst 80001 9000 2000 1 I I I L I I 10000 12000 14000 16000 18000 20000 22000 240000 Depth (m) Source: World Bank staff. - 248 - The cost of both breakwaters and quays are highly dependent on the type of support provided by the bottom. In some cases, the bottom is too hard, which requires special provisions to anchor the breakwater or quay to the hard bottom. Pavement construction costs for container yards vary with type of construction. Table VIII.2.1 shows costs per square meter for asphalt and gravel construction in various countries. For a recent project financed by the World Bank, cost data has been obtained for bulk terminal construction. Table VIII.2.2 shows civil engineering costs for this project. Table VIII.2.1 Unit Construction Costs for Pavement (1982 USD) Country USD/m of 10 cm USD/m2 of 40 cm asphalt, with 20 cm gravel bed base and 15 cm sub-base Cyprus 11 3.5 Greece 9 3.0 Malaysia 11.5 4 U. K. 20 8.2 Indonesia 17.3 7.7 Oman 17 8.4 Netherlands 18.3 9.8 Saudi Arabia 17 8.2 U.A.E. 16.5 9 Source: World Bank staff. - 249 - Table VIII.2.2 Bulk Terminal Civil Works Costs (USD 1983) Depth Alongside Width USD/m Bulk Berth (Piled) for lOm 12m 28,719 heavy shiploaders/unloaders including rails, etc. 12m 14m 32,724 14m 16m 40,920 Approach Causeway (Piled) for 12m 6m 10,370 conveyors, roadway, etc. 12m 8m 11,200 1Om 6m 8,700 1im 8m 9,400 8m 6m 7,667 8m 8m 8,280 Source: World Bank project files. Table VIII.2.3 Bulk Terminal Storage Construction Costs (USD 1983) Size Unit Cost Storage shed (top conveyor fged) with load capacity 2-4 ton/m 100,000 m2 $150/m2 Paving for open storage - $ 10.8/m2 Silo (High rise) 100,000 m3 $180/m3 Source: World Bank project files. - 250 - VIII.3 Equipment Costs A number of manufacturers were surveyed to develop price ranges on bulk handling equipment. Figures VIII.3.1 to VIII.3.5 graph the responses. Prices do not include shipping costs. Prices, not displayed as a range, are for equipment built to an "economical" standard. We note that these price ranges are based on budgetary estimates. For the World Bank project mentioned in the last section, the equipment costs are summarized in Table VIII.3.1. A compilation of some additional bulk handling equipment data is provided in Table VIII.3.2. This is a sampling of data being gathered as part of an ongoing effort to create a database of port project costs. Table VIII.3.1 Bulk Terminal Equipment Costs (1983 USD) Capacity Cost Ship unloaders 25 t. grab 500 tph $4.92 million 50 t. grab 800 tph $7.15 million Continuous 800 tph $4.39 million Conveyors including support 300 tph $780/m Frames, Drives, Etc. 500 tph $960/m 800 tph $1200/m 1000 tph $1420/m Scraping/reclaimers $1 million each Bagging machines $100,000 each Bag stacker/reclaimer $700,000 each Bag loaders/unloaders $108,000 each Source: World Bank project files. - 251 - FIGURE VIII.3.1 SHIP UNLOADER COSTS (1983) 5.0 -., 2.5 0 375 750 1125 1500 Rated capacity (tons/hour) Note: Costs are in millions of US dollars per unloader converted from quotation in yen at rate of 239 per $ Source: Mitsubishi Heavy Industries FIGURE VIII.3.2 COST OF RAIL Mv1OUNTED PNEUMATIC GRAIN uNLOADERS (1983) 5.0 2.5 0 100 200 300 400 Rated capacity (tons/hour) Note: Costs are in millions of US dollars per unit converted from a quotation in yen at rate of 239 per $ Source: Mitsubishi Heavy Industries - 252 - FIGURE VIII.3.3 PRICE OF CRAWLER. MOUNTED BUCKET WHEEL EX3YAVATORS (1983) 5.0 2.5 0 0 1500 3000 4500 6000 Rated capacity (cubic meters/hour) Note: Quotation in millions of US dollars per unit Source: American Hoist and Derrick-Mechanical Excavators Division FIGURE VIII.3.4 COST OF CRAWLER lYOUNTED CONVEYORS(1983) 5.0 0~~~~~~~~0 2.5 _ , __ -- 100 150 200 250 300 Length nf conveyor in feet Note: Cost in millions of dollars per unit Source:American Chain and Hoist-Mechanical Excavators Division - 253 - FIGURE VIII.3.5 BELT CONVEYOR PRICES (1983) 10 5, 01_ I I I I 0 25 50 75 100 Width conveyor in inches Note: Cost in thousands of US dollars per foot including gallery and installation-excluding land costs Source: Louis Berger Inc. - 254 - Table VIII.3.2 Selected Bulk Equipment Project Cost Data Description Capacity Unit Cost(USD) Year Ship Unloaders 600 tph each 1.6 million 1979 540 tph each 2.0 million 1981 1000 tph each 2.24 million 1979 1800 tph each 3.47 million 1979 Ship Loaders 500 tph each 0.54 million 1981 II 1000 tph each 0.90 million 1981 Conveyors 300 tph metre 816 1979 600 tph 1,122 1979 600 tph 2,000 1981 1000 tph 1,530 1979 1200 tph 1,735 1979 1500 tph 2,040 1979 1200 tph 2,200 1981 1500 tph 2,500 1981 1800 tph 2,347 1979 2000 tph 2,551 1979 Stackers 600 tph each 0.61 million 1979 1000 tph each 0.73 million 1979 1800 tph each 0.97 million 1979 Reclaimers 600 tph each 1.22 million 1979 1000 tph each 1.53 million 1979 1800 tph each 2.14 million 1979 Stacker/Reclaimer (Rec. Cap) 1000 tph each 1.8 million 1981 Tripper Cars - each 0.13 million 1979 Wagon Loading Station - each 0.35 million 1979 Truck Loading Station - each 0.30 million 1979 Major Surge Bin - each 0.31 million 1979 Minor Surge Bin - each 0.09 million 1979 Transfer Cars - each 0.27 million 1979 Rail Weigh Bridge - each 0.055 million 1982 Road Weigh Bridge - each 0.055 million 1981 Electronic Scales - each 0.011 million 1981 Source: World Bank Project Files - 255 - VIII.4 Bid Invitation-and Evaluation This section provides some general guidelines for procurement of port equipment. Some comm n conditions and evaluation factors are highlighted and can serve as a check-list in organizing procurement activities. The following factors must be covered by the purchasing department in evaluating bids: 1. Comparison of net price 2. Comparison of payment terms 3. Comparison of transportation charges 4. Existence of escalation terms and the probable impact of final cost 5. Charges for optional items including: - spares requirements - service and erection engineers fees - prices of additional work options 6. Availability of spares with respect to: - extent to which locally available - factory stocking policy of manufacturer - need to order for factory production and lead time 7. Comparison of rights to manufacturing drawings and ability to purchase locally 8. Equipment delivery and erection schedules 9. Manufacturer's shop loading and capacity available for this project 10. Extent of warranties and engineering responsibility - scope of coverage - manufacturers liability in event of failures or performance inadequacies 11. Financial responsibility of bidders. - 256 - At times evaluation procedures may eliminate a manufacturer who is normally competitive, reputable and qualified. The purchasing department should review the specifications for lack of clarity or ambiguity when such a manufacturer fails to make the competitive range. Invitations for bids on material handling equipment should comprise three sections broken down as follows: 1. General Project Description: a. Outline of the material handling requirements; b. Location with applicable environmental condition and site limitation; c. Arrangement drawings; d. Statement defining responsibilities of supplier for engineering, schedule, and financial aspects. 2. Equipment Specifications: a. performance or insistence on specific types or features; b. minimum acceptable design conditions; c. component performance; d. requirement to provide drawings and technical descriptions; e. equipment evaluation criteria. 3. Commercial Specifications: a. procedure and timing for proposed submission; b. requirements to be satisfied before contract award and procedure for award; c. estimate of project implementation schedule; d. special requirements. - 257 - The purchase contract must cover the following provisions. 1. Engineering Provisions a. requirements for successful handling of material and interface with other systems; b. limits of engineering responsibility of supplier; c. list of drawings to be provided; d. requirements for engineering coordination; e. safety features to be provided; f. installation and testing responsibility - obtaining permits and certification; - site preparation, excavation and laying of foundations; - provision for storage and safekeeping of delivered equipment; - temporary power and wiring provisions; - procedures for examination, testing and acceptance; - requirements for clean-up and operational start-up; g. responsibility to train operators and provide training materials. 2. Equipment Provisions a. detailed specifications of material handling equipment and accessories; b. references to control drawings; c. listing of buyer furnished equipment and their specifica- tions; d. specifications of painting, protective coatings, and corrosion control equipment; e. provisions for acceptance of substitute equipment and materials. * 258 - 3. Commercial & Legal Provisions a. Identification of supplier and buyer, and individual points of contact for technical and contractual matters; b. summary of work description and location of work identifying all applicable specifications; c. delivery and erection schedules; d. applicable penalties for delays and failure to complete; e. provision if supplier is delayed by buyer; f. price and payment schedule including withholding and final payment; g. assessment of transportation charges; h. insurance requirements; i. responsibility for person and property; J. procedures for negotiating work deletions, additions or change orders; k. provision for buyer approval of sub-contracts; 1. inspection of work in progress by buyer and correction of deficiencies; m. responsibility for work in progress and associated equipment; n. procedure for buyer's acceptance of work and transfer of ownership; o. liability for taxes including payroll, sales, excise and other taxes; p. provision for payment of Workmen's Compensation, social security, unemployment contribution, and employee insurance; q. responsibility for permits, licenses, certificates, and payments for same; r. definition of warranty terms, including provisions for correcting inadequacies; responsibility for use of patents and copyrights; t. buyer's use of supplier drawings and rights to data; - 259 - Organizing for a project requires careful delineation of functions between the various groups involved. The breakdown of activities among the groups are as follows: 1. Buyer's Engineering Department Functions: a. complete definition of material handling problem (could include preliminary arrangement drawings); b. develop detailed specifications; c. evaluate bids for technical parameters; d. coordinate engineering activities; e. supervise installation and fine tuning of equipment; f. implement training programs for operating and maintenance personnel. 2. Buyer's Purchasing Department Functions: a. handle all commercial aspects; b. send inquiries and receive proposals; c. make commercial evaluations of bids integrating technical evaluations made by engineering; d. select supplier and award contract; e. monitor progress and take expediting actions. 3. Equipment Manufacturer Functions: a. prepare and submit bid, coordinating with buyer's technical group to ensure responsiveness; b. refine preliminary engineering - review and correct as necessary c. design equipment and accessories for fabrication; d. coordinate engineering with other engineering contractors; e. purchase necessary equipment; f. schedule fabrication and shipment; g. fabricate, install and turn over equipment; - 260 - h. train operators and provide operating manuals; i. stock spares inventory; 4. Consulting Engineer Functions: a. provide expert personnel on temporary basis; b. engineering services; c. procurement services; d. installation and training services. Quality and performance of material handling equipment will depend on a number of factors. Even very tightly written specifications can be variously interpreted by vendors. Consequently, bids on identical specifications can vary a good deal. Evaluation of bids on first cost alone is usually not desirable; equally important are the life cycle costs which will be incurred. The life cycle costs will depend on inherent design features as well as the application in which the equipment is used. - 261 - APPENDIX A. REGIONAL DEVELOPMENT ANALYSIS As mentioned in Chapter V.6, it is in the national interest to balance regional development by increasing the income of depressed regions. Development of a new port in a region will generate a new source of income for the region and can thus contribute to the objective of balancing regional development. Therefore, it should be emphasized that, when deciding on a port site, planners should consider a broader aspect of balancing regional development. This can be achieved by undertaking such regional development analysis as the one described below while trying to minimize the necessary total investment costs. Parameters Used in the Analysis Rn = Regions (1 to n). aij = the marginal average propensity of region i to spend its income in region j. n (Note that I aij = 1). j=1 A = Exchange Matrix = all a12 . aln a21 a22 . . . a2n ani an2 . * . ann Yi= Total income of Region i n = I aijYj = Sum of income of region i spent in i and j=1 incomes of other regions spent in i. Y = Equilibrium income vector of regions which satisfies the equation A'Y = Y. - 262 - Methodology The following simple example based on a Markov chain analysis with three regions will illustrate the method of computing the equilibrium incomes of regions relative to others. Assume that we have the exchange matrix of 0.5 0.2 0.3 A =0 0.6 0.4 0.2 0 0.8 The above system of homogeneous linear equations can be solved by writing it into A'Y = Y formulation as below: 0.5Y1 + 0.2Y3 = yl 0.2Y1 + 0.6Y2 = Y2 0.3Y1 + 0.4Y2 + 0.8Y3 = Y3 and we get the equilibrium income vector of Y = (Y1, 0.5Y1 , 2.5Y1) in terms of Y1. Interpretation The resulting equilibrium income vector shows that, with given pro- pensities to spend and using the income of region 1 as the basis, the equilibrium income of region 2 is half of the income of region 1 and the income of region 3 is 2.5 times the income of region 1. From the equilibrium income vector, one can see that region 2 is the most depressed region in terms of the total income and, to increase the absolute amount of its income, the incomes of region 1 and 3 should be increased by an even greater amount to maintain the equilibrium conditions. The equilibrium growth can take place in this model as long as the growth rate is the same for each region, but this will increase the relative income difference among regions further. - 263 - Dynamic Regional Growth In order to increase the income of a depressed region, it is necessary to inject one period or continuous income into the region and the development of a port in a region can provide a continuous source of increased income for the region. The following analysis, based on a Markov chain solution method, will indicate the effect of this new investment on the total equilibrium income of the region and on that of others. By introducing index t to represent different time periods the following results: Vij(t) = element of the row i in the exchange matrix in time period t which satisfies the condition of n I Vi (t) = 1 and, n Vij(t + 1) = E vij(t)aij j =1 Then, the equilibrium income vector becomes Y(1) - Y(O)A Y(2) - Y(1)A - Y(O)A2 Y(3) - Y(2)A = Y(O)A3 which are generalized into Y(t) - Y(t-l)A = Y(O)At. As an example, consider a simple two regions case with the following exchange matrix A { 1 -p and by performing Markov's chain analysis, the equilibrium income vector becomes. y.{ Y1, ( q Y1} - 264 - If a capital investment, such as the development of a port, is put into a region so that its income increases by an amount equal to x in period 1 and all following periods, the region's relative income will be increased. Following are the resulting equilibrium income vectors for each time period. Y(O) = [1y, (1-p) Yi q Y(1) = [(Yi + x), (1 - p) Y1] q Y(2) =(UY1 + x (1 + p)), ((1 - p) (Y1 + x))] q Summary The method of analysis presented so far assumes that the propensities to spend remain constant throughout the periods. From the static analysis of the equilibrium income for regions, one can determine the most relatively depressed region. From the dynamic growth analysis, the effect of continuous additional income on the equilibrium income distribution among regions in each time period resulting from capital investment to a depressed area, can also be determined. Methodology summarized from: Smith, Paul E., "Markov Chains, Exchange Matrices, and Regional Development," in Regional Analysis and Development (Chapter 9), edited by John Blunden and others, The Open University Press, Harper & Row, Publishers, New York, 1973. - 265 - APPENDIX-3. TECHNIQUES FOR PORT DEVELOPMENT ANALYSIS In this appendix a number of computer-based optimization techniques are briefly described. Then, simple linear programming models are developed which will support the decision-making process of choosing a port site from available alternatives (see Chapter V) and the desired inland transportation modes from the site to the demand centers. B.1 Description of Available Optimization Techniques Linear programming and dynamic programming methods, which are two of the most widely used computer-based optimization techniques, are described below along with the simulation method, which does not produce optimum solutions, but compares alternative solutions. In applying the above mathematical methods to real world problems, such as port logistics problems, the users should bear in mind that these methods are to be used only as guidelines to support human judgement in their decision-making process. Linear Programming (LP) Linear Programming is a powerful optimization method developed which can be applied to a wide variety of problems. By optimally allocating constrained resources among activities, it maximizes profit or minimizes the cost. Although the objective function is restricted to being linear, non-linear functions can be linearized through piecewise linear approxima- tion. Due to the development of an efficient simplex algorithm and the availability of high-speed digital computers, LP problems can be solved - 266 - quickly and inexpensively when compared to other methods. Another major advantage of LP is the ability to perform sensitivity analysis. It gives ranges on the objective function coefficients and in the righthand side values of constraints for which the optimum solution remains unchanged. It also calculates shadow price (or opportunity cost) of a constraint which is defined as the amount of change in the optimal objective function value per one unit increase in the righthand side value of that constraint, given all others remain constant. This is valuable information since, by pricing out an activity with shadow prices, one can tell whether the activity is worth undertaking. Dynamic Programming (DP) Dynamic Programming is an optimization technique that simplifies a complex problem into a series of multi-stage, more easily analyzable problems. The stages are treated as independent of each other. For each stage optimal value for different states is calculated using a recursive relationship that is applicable to every stage. Sub- sequently when the optimal solution is found at the final stage, one can track back to the beginning for the states of each stage that consti- tute the optimal solution. Simulation When a given problem for which mathematical programming models cannot be formulated, such as LP or DP, the simulation approach is resorted to. The difference is that simulation models do not generate alternatives or produce an optimum answer to the problem under study. Rather, it merely evaluates alternatives developed by the decision-maker. - 267 - Simulation computer programs are usually a series of logical arith- metic operations which are arranged and executed in a sequence closely resembling the actual operation of the system. Thus, it usually requires extra time and effort to develop simulation models and it may be more difficult to track back to the source of the error. The formulation of simulation models is very flexible and it can be made as realistic as needed by incorporating conditions which can be expressed in logical arithmetic expressions. One disadvantage of simulation is that decision-makers are required to provide flexible alternatives to be evaluated by the model. The decision-makers should also have a good understanding of the optimal solution for the problem under study in order to judge the accuracy of the model. X.2 Analysis for the Optimum Choice of the Inland Transportation Model In the general methodology for bulk terminal logistics described in Chapter V.2, the information on the total inland transportation costs from each prospective port site to demand centers are required in order to choose a minimum cost port site. A method of determining the minimum cost inland transportation mode from each port site using a Linear Programming (LP) model is illustrated by the following simple example. More detailed LP description is pre- sented in Appendix B.3. Example Assume that there exist the following prospective port sites of A, B and C and demand centers of 1,2 and 3 in a region and that they are located as in Figure B-1. Assume also that the available alternate inland transportation modes - 268 - are road, rail and barges with transportation cost per ton as in Figure B-2. Figure B-1 Alternate Port Sites and Demand Centers Figure B-2 Inland Transportation Costs/ton for Commodity 1 Demand Centers . 1 2 3 road ) A Cjkl rail )=k barges) road Ports B rail barges road C rail barges - 269 - bbjective: The objective is to determine the minimum cost trans- portation mode from each prospective port site to demand centers. Methodology: The optimal way to solve the above transportation problem is to set up a simple LP model that minimizes total in- land transportation cost subject to constraints on the capacities of each transportation mode for each link, and the demand sufficiency condition for each demand center. LP Formulation: By utilizing cost and capacity information, LP model is formulated for each prospective port site of A, B and C. Minimize I k E CjklXjkl Subject to (1) Xjkl = dji for all J and 1 (2) 1 X. kl < CAP1 for all 1 jk jkl all Xjkl > 0 Where subscripts j refers to demand centers, k refers to transportation modes, and 1 refers to commodities, and Cjkl = transportation cost from a site to demand center j through inland transportation mode k for commodity 1. Xjkl = amount of bulk flow from and to demand center j through mode k for commodity 1. djl = amount of bulk handling (inflow and outflow) required at j for commodity 1. CAP1 - capacity of a port for handling commodity 1, - 270 - The constraint (1) states that bulk handling requirement at demand center has to be satisfied for all commodities and the constraint (2) states that the total amount of bulk flow frost/to a port for commodity 1 cannot exceed the port's handling capacity for that commodity. If xjkl = 0, it means that the mode k is not used in transporting commodity I from a terminal site to ji This LP model determines the minimum total inland transportation cost from each port site and the resulting choice of inland transportation mode for each transportation link. This model assumes that the transportation costs are linear and different for each commodity. In addition to the total inland transportation cost, initial investment costs for port and inland transportation facilities have to be considered in making the final choice of the port site and the inland transportation mode from it. B.3 Detailed LP Formulation for Bulk Terminal Logistics In Appendix B.2 a simple LP model was developed to determine the minimum cost inland transportation mode. This is a simple LP formulation because it only considers demand and capacity constraints. In this Appendix the usefulness and power of the LP method is illustrated by presenting a more detailed formulation for solving bulk port logistics problems using the same example as in Appendix B.2. Define the subscripts and variables as follows: Subscripts i - prospective port sites j - demand centers or regions - 271 - k = modes of inland transportation. (Depending on the quality of a mode, they can be subdivided further - e.g. gravel road, paved road, highway) 1 = commodities t = time period Variables COSTijklt and VOLijklt = Inland transportation cost per unit volume and amount of volume transported respectively, from port i to demand center i tbroup'k transDortation mode k for commodity 1 in time period t. PEXCAPi,l,t and PNEWCAPi,l,t - existing and new additional handling capacity, respectively, in port i for commodity 1 during time period t. DEMANDI1,t = required bulk handling (from/to) in a demand center j for commodity 1 during time period t. BUDGETt = available financial resource for port expansion during time period t. INVCOSTi,l,t - investment cost per unit capacity increase in port i for commodity 1 during time period t. ICAPi,j,k,t - inland transportation capacity of mode k from port i to demand center j during time period t. - 272 - The LP formulation is as follows: Objective Function Minimize I X (COST x VOL ) i j k 1 t ijklt ijklt This objective function minimizes the total inland transportation costs. Constraints (1) Demand Constraints i k ijklt DEMAND 15t for all J,l,t i k ijt lt These constraints assure that the total amount of bulk volume transported from all ports (i) to demand center j through all transportation modes (k) for commodity 1 during time period t should be equal to DEMANDj,l,t-. - 273 - (2) Port Capacity Sufficiency Conditions VOLE iklt I PNEWCAP 1't + PEXCAPi 1 for all i,l,t These constraints prevent the total amount of bulk volume transported from exceeding the available (existing plus new) capacity for handling commodity 1 in port i during time period t. Here, we assumed that new capacity become available instantly. (3) Updating the Port Capacity PEXCAPi,l,t + PNEWCAPi,l,t = PEXCAPi,l,t + 1 for all i,l,t Existing port capacity during time period t+1 for handling commodity 1 in port i is updated by adding new capacity, which is to be determined by the LP model, to the existing capacity in time period t. - 274 - (4) Budget Constraint for Port Capacity Expansion I (PNEWCAPi 1 t x INVCOSTii,) ' BUDGETt for all t. These constraints prevent the total investment for expanding bulk material handling capacity of all commodities (1) in all ports (i) from exceeding the available financial resource during time period t. (5) Inland Transportation Capacity Sufficiency Conditions VOLJklt < ICAPijj,k,t for all i j k t These constraints prevent the total amount of bulk volume transported from port i to demand center j through mode k during time period t from exceeding the available transportation capacity of mode k from port i to demand center j during time period t. The inland transportation capacities can be updated, and the required investment costs for inland transportation capacity expansion are added to the left hand side of constraint (4). - 275 - Output The output of this LP formulation is somewhat different from the output of the general methodology for bulk terminal logistics presented in Chapter V.2. The general methodology selects one port site among several alternatives and the minimum cost inland transportation mode from that port site. The LP model formulated in Appendix B-3 does not just select one port site, but may select a number of port sites, which may already exist or which should be newly developed. This is done in order to achieve the minimum total port investment and inland distribution costs. This model can also determine the timing and the amount of port and inland transportation capacity expansions subject to budget con- straints. More importantly, by incorporating the time period into the model it is possible to evaluate the proposal for its entire project life and the availability of efficient LP packages in most computers is a great advantage for other similar applications. Shortfall One shortfall of this LP formulation is in constraint (2). This constraint assumes that new capacity can be added whenever the required bulk handling demand exceeds the available handling capacity in the ports. But, as pointed out in Step 4 of the methodology for bulk terminal logistics in Chapter V.2, this may not be the optimal port capacity expansion strategy. Due to inflation, fixed cost involved in starting port expansion project and economies of scale, it may be better to expand the capacity so that it can satisfy the bulk handling demand for several years into the future rather than expanding the - 276 - capacity every one or two years. A dynamic programming approach can solve this expansion problem most efficiently. It essentially considers all the available expansion strategies and selects the one which results in the minimum total investment cost over the project life. The optimum expansion strategy obtained from a dynamic programming model specifies when and by what amount the port capacity should be expanded subject to demand require- ment and budget constraints. - 277 - APPENDIX C. QUEUEING THEORY This appendix is intended to provide a summary of the more common queueing theory models. The cases selected are sufficient renderings of ballpark estimates of the utilization and waiting times experienced for most queues commonly encountered in facility design. By using this information a rough evaluation of a system is possible with a few minutes' work. It is not intended as an exposition of the mathematics behind the queueing process, but merely as a presenta- tion of summary formulae actually used. Readers wishing to learn more about the subject should consult a book such as, Elements of Queueing Theory, by T.L. Saaty. Four classes of queues are considered: 1. Single channel-exponential interarrival time and exponential service rate. Here there is a single waiting line for a single service facility. The time between arrivals in the queue and the time required to service them are both assumed to be exponentially distributed. The exponential distribution assumes that any service time is possible, both very long and very short (although very long times are unlikely as the probability of a service which requires a great time falls rapidly as the time increases). For the exponential distribution the mean is equal to the variance. 2. Single channel-exponential arrival arbitrary service. This queue is similar to the simple queue above except that any service time distribution can be used. This includes a fixed service time. The parameter varied to model the arbitrary service time affects standard deviation. When this is zero a constant service time results, when one, an exponentially distributed time, etc. 3. Limited queue length. This is a single channel queue similar to the first, where waiting line can only accommodate a maximum number of units. When the queue is full, additional arrivals are lost and never pass through the system. An example of this would be the departure of patrons at a cinema when a very long line is present. In many situations the results from a queue of limited length approach that of unlimited for a small maximum queue length. - 278 - 4. Limited number of customers for a single queue This system can be used to model systems for machinery repair. Here a few machines break down randomly and are repaired by a single repairman. Other interesting applications are possible. 5. Multiple servers serving a single queue Here one line feeds a number of services as is common at highway toll booths and other situations. 6. Single server-Erlang arrivals Poisson service This is similar to case 2 except the distribution of inter- arrival times is the parameter rather than the service time. The Erlang distribution has two parameters, the mean and "K " when K is one, the Erlang distribution is identical to the exponential. As "K" increases the variability in the inter- arrival time is reduced. For large values of "K," the time between arrivals is nearly constant. Table C.1 gives the pro- cedure to calculate "K". The meaning of symbols used in the table is given below: Ls = Mean number of units in the system including those being served Lq = Mean length of the queue. This is the mean number in the system Ls minus the mean number being served P(O) = Probability there are no units in the system (counting those being served) N = Number of units in the system (counting those being served) P(N) = Probability there are N units in the system (counting those being served) = mean interarrival rate p = mean service time a = standard deviation, in case 2 that of the arbitrary service time M = maximum length of queue, case 3 or maximum number of customers, case 4 S = number of servers in a multi-channel queue K = coefficient of Erlang distribution mean2 K = variance TABLE C.1 - QUEUEING FORMULAE mean NuMber Haan Number Probability Mean Wetting Mean Time Probabilitty on. Probability I In System L l a Queue Lq All Busy Time In System is Systgm P(O) tn System *(1) LS Single Channel Ixponential 12 1( Arrivals toisson Service - - a V V - I (v - ~ ~ ~ ~ ~ ~ pA ( -. Single Channel Ltponential 2 (1%2 L Arrival Arbitrary Service e A, l - 1 L 1 -o Limited Q e of Length N a 2 + A2 L - L.S - >i5 X Ls - (I-Fl0)) q LP I I 'D _.~~~~~~~~~~~~~~~~~_ 2 A(x-pH) I(1-Fm) lNe Limited Number of Customers x (!.)(i-TO)) _ _ a" (1-1(o)) I-F(O) KL5 I _I (4 N Custom.e a . L(M-L5) j S) . fMI(A) (M-N)I multiple Server*. A (a) (h.) 1 L L4 + 10_ _ _ _ _ - 1 ( 0 ) 009 Po:oa3n Service S S- SrIgle Channel £rlang 2 A A l A .+1 A I Arriv is Poisson ServCe K+i I 2S A 2 - i~~~een2 a~Z V61)* g .(.) 2K 0-0 NOA) a~pA - 280 - Table C.1 gives the formulae describing the principle parameters of these queues. It is sometimes required to perform queueing calculations when good estimates of the mean and standard deviations are not available. Estimates of these can be obtained using the Beta distribution. This distribution provides an estimate of the mean and standard deviation if the fastest time, "most" likely time, and slowest time can be estimated. The formulae to do this are given in Figure C.2. While this way of estimating mean and standard deviations is not completely justified in theory, it can give useful results in most situations. The Beta Probability Density Function originated from making PERT computations. The review of Beta distribution will facilitate our calcula- tions of means and standard deviations for such parameter as ship inter- arrival time. A Beta-PDF looks as in Figure C.2 and by estimating: a = Optimistic time b = Pessimistic time m = Most likely time The mean and the standard deviation of the activity can be calculated by using the formulas shown in this figure. - 281 - FIGURE C..2 BETA DISTRIBUTIONS r.ean = to = (a+4mi+b)/6 standard -(b-a)/6 deviation probability density Actual Performance a m t b time = t* 0 Optimistic Most Pessimistic time likely time time - 283 - APPENDIX D. INVENTORY MANAGEMENT It is important to keep enough inventory of raw materials to ensure smooth and continuous production activities. However, keeping excessive amounts of inventory is costly and it should be avoided to keep much needed capital from sitting idle in the form of raw materials inventory. Part of the success of many Japanese manufacturing industries is the result of efficient production planning, which required very small amounts of inventory. The inventory management also affects the facilities that will be required, such as storage area and transportation requirements. In this appendix some inventory management techniques, based on the detailed examination of demand patterns and other related costs, are reviewed. The port planner should consider these techniques in their designing effort rather than make arbitrary inventory management decisions. In general there are four functional classifications of inventory. They are as follows: Cycle stock Due to economies of scale or technology requirements, orders are placed in batches and the resulting inventory is called cycle stock. Safety stock To account for the variability in demand and delivery schedule of raw materials, a certain amount of inventory is kept in each time period as safety stock. Anticipation stock As with peaks in sales, when situations, such as strikes or wars, are expected, anticipation inventory is built up. Pipeline stock Work in process and in transit inventories The simplest inventory management model is the Economic Order Quantity (EOQ) model, which is also known as Wilson's lot size formula. The assumptions are the following: - 284 - 1. Demand is continous at a constant rate. 2. Process continues infinitely. 3. Replenishment is instantaneous. 4. All costs are time invariant. 5. No shortage or quantity discounts allowed. Based on the above assumptions, the total ordering and inventory holding cost is: TC = A *D + r * c * Q = (Annual Ordering Cost) + (Annual Inventory Holding Cost) where A = ordering cost D = amount of annual demand Q order quantity r = inventory carrying charge per year per unit c = unit price of the item _= average inventory 2 By differentiating the above total cost equation with respect to Q and setting it equal to zero, we get Q*, the optimal order quantity, to be Q*= /2A-D Essentially, this EOQ formula is obtained by r.c trading off the annual ordering cost with the annual Inventory holding cost. As an example, for A = $10/order D = 10,000 units/year r-= 10%/year/unit c = $100/unit then, Q* = 2 x 10 x 10,000 = 141 0.1 x 100 The EOQ formula can be modified in order to apply it to cases where demand varies, supply rate is finite and backlogging and quantity discount is allowed. (See "Decision Systems for Inventory Management and Production Planning" by R. Peterson and E. Silver). Table D.1 is a summary of those cases. This EOQ scheme is represented as follows. Inventory,i time - 285 - TABLE D-1 SUMMARY OF INVENTORY CONTROL CASES 1. EOQ Syste,n d* = 9 2A-D Inventory A t ime 2. Finite_ Supply 2A= -D Inventory Rate Q* Nr(-D) P = supply rate/year time 3. Backlogging Allowed ___ Allowe-D rdc+b Inventory =r -c -, r c*b tm b = annual backorder cost/unit *= optimal amount of backorder 4. EOQ with a. Find T2 d(T) - T = Length of tire Varying r-c period that the Oeta,and b. Find T.*, where T >d(T )current replenishment (Silver - Meal 1 1 _ 1 should last 2Ashudls r-c d(T) demand rate in neriod i c. Solve T - 2A r-c d(Ti) 5. Quantity a. Compute Discounts for all QiThj.12EA Cost lJnits rc b. Qi' bi.1 if Qi <+7 IN~~~~~~~~~~~~~~~I 12000 . _ I