~~~~~~~~~~~~~HI , )IPOhlOlS" lasD OJ R ISOIIS rhl? t S - . ',,_ -a1 :a g-#- -Th4- - 11 _ C=~~~~~~ ~ ~C "''d? A k-U- - =F V-W~-- ow~ w i ~~~~~~~~~~~~~~~~M W7z C= 8 C, KluTZ;-E __6 _ .. L t : _ ... . _ _ , ........ ._._ _ ...._ . .. . .. _ _ ._ _ ._ . ._ _ _ _ _ _ _ _ _ C a~M = C coo~~C cm MOST PROJECTIONS OF FERTILIZER USE around the world imply continuing rapid growth in fertilizer demand. To meet such increased levels of demand, substantial investments are being made in new pro- duction capacity. Many developing coun- tries have some or all of the iimportant raw materials for fertilizer production, and this circumstance, combined with the increased fertilizer demand and fear of shortages of supply, has led these coun- tries to consider producing fertilizer domestically. In several parts of the world plans are being made to gain self-suffi- ciency in this critical output. This second volume in the series, THE PLANNING OF INVESTMENT PROGRAMS, deals specifically with the formulation of such sectorwide investment programs in the fertilizer industry. The principal products and processes of relevance to fertilizer production are discussed, and the kind of planning problems that need to be addressed during the project iden- tification phase are described. Models analyze these planning problems, starting from a simple static model to evaluate the efliciency of an existing industry, to a continues of back flap The Planning of Investment Programs in the Fertilizer Industry A WORLD BANK RESEARCH PUBLICATION Volume Two of THE PLANNING OF INVESTMENT PROGRAMS Alexander Meeraus and Ardy J. Stoutjesdijk, Editors Ar'meane M. Choksi Alexander Meeraus Ardy J. Stoutjesdijk The Planning of Investment Programs in the Fertilizer Industry Published for the World Bank THE JOHNS HOPKINS UNIVERSITY PRESS Baltimore and London Copyright © 1980 by the International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, N.W., Washington, D.C. 20433 All rights reserved. Manufactured in the United States of America. The views and interpretations in this book are the authors' and should not be attributed to the World Bank, to its affiliated organizations, or to any indi- vidual acting in their behalf. The map on page 185 has been prepared exclusively for the convenience of readers of this book; the denominations used and the bound- aries shown do not imply, on the part of the World Bank and its affiliates, any judgment on the legal status of any territory or any endorsement or acceptance of such boundaries. Library of Congress Cataloging in Publication Data Choksi, Armeane M 1944- The planning of investment programs in the fertilizer industry. (The Planning of investment programs; v. 2) Includes index. 1. Fertilizer industry. 2. Fertilizer industry-Egypt. I. Meeraus, Alexander, 1943- joint author. II. Stoutjesdijk, Ardy J., 1938- joint author. III. Title. IV. Series: Planning of investment programs; v. 2. HD9483.A2C47 338.4'366862 78-8436 ISBN 0-8018-2138-X ISBN 0-8018-2153-3 pbk. Contents Editors' Note to the Series xi Preface xiii Part One. General Methodology I 1. Introduction 3 The Basic Approach 5 Previous Work 7 A Reader's Guide to the Volume 9 2. The Fertilizer Industry: Some Terminology 10 Fertilizer Nutrients and Grades 10 The Main Materials in Fertilizer Production 12 3. The Production of Nitrogenous Fertilizers 16 Ammonia 17 Nitric Acid 24 Urea 25 Ammonium Sulfate 31 Ammonium Nitrate 33 Ammonium Sulfate-Nitrate 36 v Vi CONTENTS Ammonium Chloride 37 Other Nitrogenous Fertilizers 38 4. The Production of Phosphatic Fertilizers 41 Phosphate Rock 41 Sulfuric Acid 43 Phosphoric Acid 47 Ground Phosphate Rock 50 Single or Normal Superphosphate 51 Triple Superphosphate 52 Dicalcium Phosphate 54 Thermal Phosphate Fertilizers: Basic Slag 55 5. The Production of Potassic Fertilizers 57 Production and Processing of Potash Ores 57 Potassium Chloride 58 Potassium Sulfate 61 6. The Production of Multinutrient Fertilizers 63 Granulation Processes 64 Multinutrient Fertilizer Materials 65 Granular Compound Fertilizers 71 Dry-Mixed Fertilizers or Bulk Blends 71 7. Problems in Fertilizer Sector Planning 72 Specifying the Planning Problem 74 Formulating an Investment Program 80 8. Planning Models for the Fertilizer Sector 88 The Transport Problem 89 Modeling an Existing Industry 92 The Static Capacity Planning Model 109 The Dynamic Capacity Planning Model 118 9. The Complete Investment Planning Model 122 The Specification of the Model 123 The Size of the Model 132 Alternative Specifications of the Constraints 134 Alternative Formulations of the Objective Function 143 Linking the Fertilizer Sector Model to an Agricultural Model 144 Links with Other Activities 145 Conclusion 149 CONTENTS Vii Part Two. The Egyptian Fertilizer Sector: A Case Study 151 10. The Fertilizer Sector in Egypt 153 The Demand for Fertilizers 154 The Domestic Production of Fertilizers 160 Raw Materials for Fertilizer Production 168 The Transport of Fertilizers 170 The Prices of Fertilizers 172 The Import of Fertilizers 178 Conclusion 180 11. An Analysis of the Egyptian Fertilizer Sector in 1975 181 The Structure of the 1975 Model 182 Results of the Basic 1975 Model: Solution 1 201 Introducing Interplant Shipments: Solution 2 204 Full Capacity Utilization: Solution 3 206 The Composition of Fertilizer Demand: Solution 4 208 Summary 209 12. A Medium-Term Planning Model of the Egyptian Fertilizer Sector 211 The Structure of the Dynamic Model 211 The Mathematical Formulation of the Dynamic Model 227 13. Results of the Dynamic Analysis 235 Basic Results: The Reference Solution 236 Alternative Scenarios at Suez: Scenarios I and II 249 Alternative Export Strategies: Scenarios III, IV, and V 254 A Complete Import Substitution Policy: Scenario VI 264 The Elimination of the Urea Consumption Restrictions: Scenario VII 266 Measures of the Domestic Resource Cost 269 A Sensitivity Analysis of the Results 279 Conclusions 285 Appendix. Summary Tables of the Results of Scenarios I-VII 288 Appendix. Computer-Readable Representation of the Model 303 Index 325 Viii CONTENTS Text Tables 1. Actual Use of Fertilizer Nutrients in Egypt, 1951-52 to 1974-75 155 2. Fertilizer Application Rate per Cropped Feddan in Egypt, 1955-56 to 1971-72 157 3. Estimated and Projected Use of Nitrogenous Fertilizer Nutrients for Various Crops, 1973-85 158 4. Estimated and Projected Use of Phosphatic Fertilizer Nutrients for Various Crops, 1973-85 159 5. Estimated and Projected Use of Potassic Fertilizer Nutrients for Vegetables and Fruits, 1973-85 160 6. Domestic Production of Nitrogenous Fertilizers by Type and Producing Unit, 1962-63 to 1974-75 162 7. Annual Output of Phosphatic Fertilizers and Percent Change in Output from Year to Year, 1950-71 167 8. The Transport of Fertilizers by Governorates Covered and Mode of Transport 171 9. Average Total Cost, Ex Factory Price, and Net Return for Selected Fertilizers, 1971-72 173 10. Prices and Margins of Distribution of the Locally Produced Fertilizers, 1971-72 174 11. Prices and Margins of Distribution for Imported Fertilizers, 1970-71 175 12. Cooperative Sales Prices of Nitrogenous Fertilizers as Determined by Ministerial Orders, 1960-72 176 13. Cooperative Sales Prices of Phosphatic Fertilizers as Determined by Ministerial Orders, 1960-72 177 14. Consumption of Fertilizers by Demand Region and Fertilizer Type, 1974-75 184 15. Design Capacity of Plants by Plant Location and Productive Unit 188 16. Input-Output Coefficients 189 17. Prices of Raw Materials from Domestic Sources by Plant Location, 1975 194 18. Prices of Miscellaneous Inputs, 1975 195 19. Road Distances from Plants and Alexandria (for Imports) to Marketing Centers for Final Products 196 20. Distances from Alexandria to Plant Locations for Imports of Raw Materials 197 21. Prices of Imports, 1975 197 22. Summary of Aggregate Results of the 1975 Model 204 23. Interplant Distances for Shipments of Intermediate Products 205 24. Summary of Results of the Static Model 210 25. The Investment Cost Functions for the Dynamic Model 222 26. Prices of Raw Materials Used in the Dynamic Model by Plant Location 224 CONTENTS iX 27. Road Distances from Plants to Marketing Centers 225 28. Rail Distances between Plants 226 29. Import and Export Distances from Alexandria or Suez to Plant Locations 226 30. Attainable Export Prices for the Dynamic Model 227 31. Summary of the Results of the Reference Solution 238 32. The Production of Final and Intermediate Products under the Reference Solution 240 33. Exports of Final Products under the Reference Solution 247 34. Transport Requirements by Mode under the Reference Solution 248 35. The Production of Final and Intermediate Products under Scenario I 251 36. The Production of Final and Intermediate Products under Scenario II 253 37. The Production of Final and Intermediate Products under Scenario III 256 38. Exports of Final Products under Scenario III 257 39. Plant Investment under Scenario IV 259 40. The Production of Final and Intermediate Products under Scenario IV 260 41. Exports of Final Products under Scenario IV 261 42. The Production of Final and Intermediate Products under Scenario V 262 43. Exports of Final Products under Scenario V 263 44. Plant Investment under Scenario VI 264 45. The Production of Final and Intermediate Products under Scenario VI 265 46. The Production of Final and Intermediate Products under Scenario VII 267 47. Exports of Final Products under Scenario VII 268 48. The Domestic and Foreign Exchange Cost Components of the Various Scenarios 271 49. Fertilizer Production and Consumption during the Planning Period, 1979-87 273 50. The Domestic Resource Cost for the Various Scenarios 274 51. The Ranking of the Scenarios According to Objective Function Value and Domestic Resource Cost 278 52. Comparison of Break-even Import Prices and Import Price Used in the Planning Model for Selected Products 283 53. Summary of the Results of Scenario I 288 54. Summary of the Results of Scenario II 290 55. Summary of the Results of Scenario III 292 56. Summary of the Results of Scenario IV 294 57. Summary of the Results of Scenario V 296 58. Summary of the Results of Scenario VI 298 59. Summary of the Results of Scenario VII 300 X CONTENTS Text Figures 1. The Chemical Elements Essential for Plant Growth 11 2. Alternative Processes for Ammonia Synthesis 20 3. Alternative Processes for the Manufacture of Urea 27 4. Alternative Processes for the Manufacture of Ammonium Nitrate 34 5. Alternative Processes for the Manufacture of Sulfuric Acid 45 6. Alternative Processes for the Manufacture of Phosphoric Acids 48 7. Alternative Processes for the Manufacture of Triple Superphosphate 53 8. Alternative Methods for Refining Potassium Chloride 60 9. Alternative Processes for the Production of Potassium Sulfate 61 10. ateh Investment Cost Function 81 11. Linearization of the Investment Cost Function 110 Map The Fertilizer Sector in Egypt 185 Editors' Note to the Series THIS IS THE SECOND VOLUME in a series dealing with the use of mathe- matical programming methods in investment analysis. The volume focuses on the use of such methods to analyze production and invest- ment problems in the chemical fertilizer industry. The exposition of the methodology follows closely that adopted in the first volume of the series, The Planning of Industrial Investment Programs. A Methodology, by David A. Kendrick and Ardy J. Stoutjesdijk. The series relies essentially on one among a number of possible approaches to investment planning; specifically, it employs mixed- integer programming to analyze investment problems in the presence of economies of scale. Alternative approaches, such as dynamic programming, are successfully used by other investigators to address selected aspects of the investment planning problem. For sectorwide investment analysis, however, we believe that mathematical pro- gramming offers the best prospects for operational use. ALEXANDER MEERAUS ARDY J. STOUTJESDIJK xi Preface MOST PROJECTIONS OF FERTILIZER USE around the world, and particu- larly in the developing countries, imply continuing rapid growth in fertilizer demand. To meet such increased levels of demand, sub- stantial investments are being made in new production capacity, in both developed and developing countries. Among the latter, many have some or all of the important raw materials for fertilizer produc- tion. That circumstance, combined with high growth rates of fertilizer demand and a fear of real or perceived shortages of supply world- wide, has led many developing countries to consider producing fer- tilizer domestically. In several parts of the world, investment plans for the fertilizer sector are being drawn up in the context of regional endeavors to gain self-sufficiency in this critical agricultural input. Against this background, it appeared appropriate to devote the first sector-specific volume in this series to the chemical fertilizer industry in the hope that it will assist planners in the developing countries in drawing up efficient investment programs in this sector. For those not familiar with the fertilizer industry, a few chapters that give the most important technical information regarding the sector have been included; without this information, it might be difficult to understand the rationale and structure of the planning models pro- posed. In many respects, however, these chapters are no more than xiii Xiv PREFACE an overview; more detailed technical background will often be desirable. The planning methods are set out in great detail in several chapters. The layout of these chapters is similar to the corresponding ones in volume 1, but all the models presented here are specific to the fertilizer industry. Moreover, attention has been paid to variants of the standard investment planning model to fit particular situations. Finally, a case study of the Egyptian fertilizer industry is presented in the second part of this volume. Several of the models discussed in the first part are used to address a number of typical problems en- countered by sector planners in Egypt. The study was carried out in close collaboration with Egyptian sector planners as well as with colleagues in the Industrial Projects Department of the World Bank. We would like to acknowledge the helpful suggestions and criticism of a number of individuals who have read part or all of the manuscript in various stages of preparation. These include, first of all, Frederick Moore and William Sheldrick, both (at the time) of the World Bank's Industrial Projects Department. Moore, a fervent supporter of the research program in general from its inception, made it possible for us to carry out the Egyptian case study. Sheldrick, head of the World Bank's Fertilizer Unit, provided us with numerous suggestions for improving our description of-the fertilizer sector. Although we do not assign any responsibility for the final product to him, we believe that, because of his help, the relevant chapters are a better and more up-to- date introduction to fertilizer technology than would otherwise have been the case. A number of other colleagues at the World Bank provided us with comments and suggestions; in particular, Donald Brown, Neithard Petry, and Harald Stier, all of the Industrial Projects Department, should be mentioned. Outside the World Bank, a number of individ- uals played an important role at various stages in the research. John Couston, of the Food and Agriculture Organization of the United Nations in Rome, commented on earlier drafts of this volume, as did Bernard Raistrick; we are grateful to them for their help. Numerous persons in Egypt, including staff of the (then) General Organization for Industrialization, the Organization for Chemical Industries, and the Ministry of Agriculture, were involved in various stages of the case study. Furthermore, we gratefully acknowledge the cooperation we received from the managers and staff of the fertilizer production facilities we visited in Egypt during the data collection phase of the study. PREFACE XV J. S. Rogers, of the University of Toronto, made many valuable sug- gestions for improving the manuscript. The final manuscript was edited by Robert L. Faherty. The charts were prepared by Pensri Kimpitak, proofs were read and corrected by Marie Hergt through the Word Guild, and the index was prepared by Nancy E. MacClintock. The map was compiled by Hans Stolle and was drawn by Larry A. Bowring under the supervision of the World Bank's Car- tography Division. ARMEANE M. CHOKSI ALEXANDER MEERAUS ARDY J. STOUTJESDIJK PART ONE General Methodology 1 Introduction THE PROVISION OF ADEQUATE FOOD SUPPLIES for a rapidly growing world population has become inconceivable without substan,ial increases in the supply of chemical fertilizers. Also, the increased use of fertilizers in the production of cash crops is often a contributing factor to increased income for the grower. Fertilizers must be made available to the farmer at the lowest possible delivered cost for two basic reasons. First, since fertilizer is a crucial input in the agricultural production process, its cost is a component of the production cost and thus of the supply price of agricultural produce. Second, the price of fertilizers is an important element in the individual farmer's decision about how much fertilizer to use. Not only does the purchase of a fertilizer constitute an important short-term investment, but its price is also a critical determinant of the expected return on its use. The delivered cost of fertilizers at the farm gate is the ultimate result of an intricate process of decisions at many different levels. Should a given country produce its own fertilizers or import them from abroad? Which fertilizers should be produced or imported? In the case of domestic production, which feedstocks should be used? At what scale and at which time and location should productive capacity be installed? If technological choice exists, how should the appropriate production process be selected? What is the most 3 4 INTRODUCTION efficient transport and distribution pattern for products in the fertilizer industry? These questions arise during the initial phase in the project planning process for the chemical fertilizer industry. This phase-that is, selecting the broad outlines of an investment, production, and trade pattern-should be distinguished from the phase in which a given project or set of projects are engineered and appraised in detail. At that time, many of the important decisions relating to the structure of the industry have already been made, either implicitly or explicitly, and although sensitivity analysis may frequently lead to modification of the project, such changes are small compared with those that should normally be allowed for during the project or program selection phase. One of the most difficult problems to handle during the project selection phase is that many of the questions which come up at that time are highly interdependent: decisions in one area affect decisions in other areas. To give one particularly obvious example, whether a given product should be produced domestically or imported from abroad depends, among other things, on the size of the market for the product in question since economies of scale, which are characteristic of the fertilizer industry, tend to lead to lower production costs per unit as the output level increases. In turn, since fertilizer demand can normally be expected to increase over time, the timing of new capacity construction will be of crucial importance in deciding on the appropri- ate scale. It is not difficult to see how other issues such as technology, location, and transportation would enter this decisionmaking process and influence the "make-buy" choice. Primarily because such interdependencies exist, the number of options at the project selection phase is frequently large, and literally thousands of project combinations can be technically feasible. The know-how and judgment of the sector specialist form an important guide to decisionmaking during this time. More often than not, rules of thumb assist the project planner in choosing from among the many alternatives normally available. Nevertheless, the validity of such rules of thumb can be doubted in any given situation, and sector specialists are increasingly dissatisfied with the absence of more systematic planning tools to assist in selecting the appropriate investment project or program. The primary objective of this volume is to describe a planning approach for the analysis of the most important aspects of the investment planning problem in the chemical fertilizer industry. THE BASIC APPROACH 5 The Basic Approach This volume describes the planning tools that are designed to analyze the implications of alternative investment, production, trade, and distribution patterns for the fertilizer industry over time and in a given geographical context. Moreover, it provides criteria on the basis of which the project planner can rank alternative programs and, ultimately, decide about the attractiveness of one program as com- pared with others. This planning approach forms the basis for subsequent analysis during the project engineering, appraisal, and implementation phases of the project cycle. The approach to be adopted is the following. The fertilizer industry is represented by a set of mathematical expressions that capture the essential technical and economic relations that characterize the industry. They concern primarily the relations among products at various levels of processing (in the form of input-output coefficients) and the investment and operating costs at various scales of production. The fact that fertilizer production costs are highly dependent on the scale of operations is explicitly modeled in the investment cost func- tion. A further set of expressions describes the cost of transporting the various products in the model between producing sites, import and export points, and marketing centers. Given a set of projections of the demand for fertilizer nutrients, over time, and by marketing center, the model is used to find the least-cost pattern of investment, produc- tion, and transportation to meet those demands. The possibility of importing from abroad versus producing domestically is normally taken into account. Furthermore, a range of area-specific restrictions can be represented in the form of constraints. Some of these may reflect industry-specific constraints, such as the availability of raw materials, whereas others may reflect government policies, such as a limitation on the investment resources available for expansion of the fertilizer industry. Also, choice among fertilizers may be constrained by more or less stringent fertilizer recommendations. Planning problems in the fertilizer industry do not, however, always cover the entire range of issues outlined above. The planner may be interested only in alternative transport and distribution patterns for an existing industry, or in an efficient plan to expand the capacity of a given firm. In such cases, a much less complex problem is posed and a simplified model would be applicable. In what follows, 6 INTRODUCTION this circumstance will be used and a series of planning models of slowly increasing complexity will be presented; the particular planning problem to which each variant of the model is applicable will be indicated. Moreover, guidelines will be provided about how each model can be modified further to handle specific problems. The sector planner may not always be interested in finding the least-cost production pattern for the industry as a whole, but simply in costing out the implications of a given project proposal or of competing proposals. As is demonstrated later, the model provides a highly efficient framework for analyzing such questions. One qualification must be noted. The models presented in this volume are simplified representations of reality, designed to guide decisionmaking, not to replace it. They are highly efficient tools to evaluate and quantify the implications of a certain understanding of the economic and technical relationships that typify the fertilizer industry and the environment in which the industry is supposed to function. Nevertheless, the decision process normally involves more elements than can or should be incorporated in a planning model. More or less important modifications of the initial investment program may therefore appear desirable in subsequent stages of the planning process. This is neither a weakness of the methodology nor something unique to this particular approach to project planning. All project planning methods proceed in phases, and it is inevitable that, as an investment project or program takes shape, greater attention is paid to detailed aspects. In the process, certain inconsistencies with earlier assumptions and judgments will frequently appear; in fact, one of the main advantages of the proposed approach is that, in such cases, a rapid and efficient reassessment of broad strategy options is possible. The approach to the investment planning problem advocated in this series has limitations. These are discussed extensively in volume 1, but it may be useful to recapitulate the main limitations briefly here. First and foremost, the approach requires a set of projections of demand for the final products that are relevant to the investment planning problem. As the supply price for final products is not known at the outset, the demand projections need to be based on price assumptions that may turn out to be incorrect. Depending on the price elasticity of demand, it may be necessary to revise the demand projections in an iterative manner. In theory, demand schedules that are responsive to price may be incorporated into the model, in combination with a more sophisticated formulation of the objective PREVIOUS WORK 7 function, but in practice this would lead to insuperable computational problems. Normally, another limitation should be made explicit: namely, that by definition the demand projection for final products excludes the possibility of substitution among products on the basis of supply price considerations. Interestingly, the fertilizer sector is one of the few cases in which this limitation does not necessarily apply; the simple reason is that final demand for fertilizers can be expressed in terms of nutrient content-as will be explained shortly-so that the model can be used to select the least-cost fertilizer types. Substitution among inputs and among transport alternatives is permitted, on the basis of explicitly stated supply prices; the limitation is that, normally, most inputs are assumed to be available at a given price either in unlimited quantities or up to a given maximum. Finally, the state of the art does not permit uncertainty to be incorporated. To a limited extent, sensitivity analysis can be carried out to determine the impact of projection errors on the least-cost investment program. The possible emergence of new products or new production technologies cannot be taken into account, but this is a limitation of investment planning in general rather than of any specific planning technique. Previous Work Until now, the major emphasis in investment analysis has un- doubtedly been on project appraisal, and a substantial body of literature is available on this subject.' Most of this literature is of a general methodological nature, and it does not explicitly address appraisal problems and methods in the fertilizer sector. In contrast, little systematic attention has been paid in the literature to formulating an investment program, and the fertilizer industry is no exception. It is true that the early experiments with the methodology 1. Several of the more important contributions in this area are: A. C. Harber- ger, Project Evaluation (Chicago: Markham Publishing Co., 1973), particularly chapter 2, pp. 23-67; 1. M. D. Little and J. A. Mirrlees, Project Appraisal and Planning for Developing Countries (New York: Basic Books, 1974); Lyn Squire and Herman van der Tak, Economic Analysis of Projects (Baltimore: Johns Hop- kins University Press, 1975); P. Dasgupta, A. Sen, and S. Marglin, Guidelines for Project Evaluation (New York: United Nations, 1972). 8 INTRODUCrION outlined in this volume included case studies relating to fertilizer manufacture, but these were specific, fairly restrictive, experimental formulations of the investment planning problem, and they are not sufficient to provide general guidelines to investment planning in the industry.2 Generalizing these earlier studies into a standard approach to the planning of investment projects and programs in the fertilizer industry appears to be a logical extension of this experimental work. The planning methodology outlined in this volume was developed by H. M. Markowitz and A. S. Manne;3 it was applied by Thomas Vietorisz and Manne to analyze the optimal location of capacity in the South American fertilizer industry.4 Manne and others took the methodology one step further, using it to study the optimal time- phasing, location, and scaling of nitrogenous fertilizer plants in India.5 Ardy Stoutjesdijk, Charles Frank, Jr., and Alexander Meeraus extended the methodology to analyze the optimal investment, produc- tion, and trade pattern for the fertilizer industry within the East African Community, substantially increasing the number of products and regions covered in the analysis.6 On the basis of this early work, fairly substantial advances have been made in the art of modeling the essential characteristics of the fertilizer industry. At the same time, rapid progress has been made in computer hardware and software, permitting extensive experimenta- tion with such models at continuously decreasing cost. Therefore, it appears to be the appropriate time to acquaint a wider audience of sector specialists with these modeling techniques so that their characteristics can be better understood and their applicability can be evaluated. 2. See Alan S. Manne, ed., Investments for Capacity Expansion: Size, Location and Time-Phasing (Cambridge, Mass.: The M.I.T. Press, 1967); and Thomas Vieto- risz and Alan S. Manne, "Chemical Processes, Plant Location, and Economies of Scale," in Studies in Process Analysis, Economy-wide Production Capabilities, eds. Manne and H. M. Markowitz (New York: John Wiley, 1963). 3. H. M. Markowitz and Alan S. Manne, "On the Solution of Discrete Pro- gramming Problems," Econometrica, vol. 25 (January 1957), pp. 84-110. 4. Vietorisz and Manne, "Chemical Processes, Plant Location, and Economies of Scale." 5. Manne, ed., Investments for Capacity Expansion. 6. Ardy Stoutjesdijk, Charles Frank, Jr., and Alexander Meeraus, "Planning in the Chemical Sector," in Industrial Investment Analysis under Increasing Returns, eds., Stoutjesdijk and Larry E. Westphal (forthcoming). A READER'S GUIDE TO THE VOLUME 9 A Reader's Guide to the Volume This volume is written in two parts: the first part presents the general methodology, and the second part reports on an application of that methodology. Chapter 2 provides a brief general introduction to the fertilizer industry, and offers definitions of commonly used terms in the industry. The next four chapters describe the products and processes associated with particular categories of fertilizers-the nitrogenous, phosphatic, potassic, and multinutrient fertilizers. Chapter 7 treats the major decisionmaking problems concerning fertilizers faced by a planner or project analyst. Chapter 8 presents a series of planning models designed to assist the analyst in solving the problems outlined in chapter 7. The models are presented in order of slowly increasing complexity; the presentation is self-contained, however, in the sense that no prior familiarity with methods and techniques is assumed. Those who are familiar with these methods can turn immediately to chapter 9, which presents a complete statement of the fertilizer model. Chapter 9 also includes a brief discussion of alternative specifications of the model, to suit situations other than the standard one assumed. In the second part of the volume, an extensive description is given of an application of the methodology to the Egyptian fertilizer sector. 2 The Fertilizer Industry: Some Terminology GENERAL BACKGROUND INFORMATION on the fertilizer industry is presented in this chapter. The information provides only a summary overview of the industry, but it is sufficient to understand the major sector-specific features of the planning models presented later in the volume.' Fertilizer Nutrients and Grades Broadly speaking, the function of a fertilizer is to furnish one or more of the chemical elements that plants need to grow. The chemical i. The following four chapters provide a more detailed description of products and processes in the industry. For more technical and comprehensive information, see: Vincent Sauchelli, ed., Chemistry and Technology of Fertilizers (New York and London: Holt, Reinhart, and Winston, 1960), a standard collection of reference papers on various aspects of fertilizer production; United Nations, Fertilizer Manual (New York: United Nations, 1967), a still largely up-to-date, compre- hensive source of information; and numerous publications of the Tennessee Valley Authority, Muscle Shoals, Alabama. 10 FERTILIZER NUTRIENTS AND GRADES 11 Figure 1. The Chemical Elements Essentialfor Plant Growth Natural ( Carbon nutrients Hydrogen Oxygen Primary Nitrogen Macronutrients Prinary Phosphorus nePotassium Secondary MCalcium nutrients i nSulfur Boron Chlorine Copper Micronutrients Iron Manganese Molybdenum Zinc fertilizer industry produces a multitude of fertilizer materials that contain these elements, also called plant nutrients, in a wide range of combinations and compositions. Fertilizer materials are usually classified on the basis of their plant nutrient content. The number of these nutrients has increased over the years and now totals sixteen. They may be separated into two categories: macronutrients and micronutrients, or trace elements, as shown in figure 1. The first three macronutrients (carbon, hydrogen, and oxygen) are supplied in sufficient quantities by air and water. The other macronutrients are subdivided into primary nutrients (nitrogen, phosphorus, and potassium) and secondary nutrients (calcium, magnesium, and sulfur). The seven micronutrients are required in much smaller amounts. A commercial fertilizer contains at least one primary nutrient (with or without secondary nutrients and micronutrients). In order for plants to grow, they need primary nutrients in sub- stantial quantities, although the specific quantities vary widely among crops, and primary nutrients are the main nutrients provided by 12 THE FERTILIZER INDUSTRY: SOME TERMINOLOGY chemical fertilizers.2 Fertilizer materials are therefore usually classified according to the primary nutrient or nutrients they contain; the presence of secondary nutrients and trace elements is often considered a bonus of somewhat varying importance. In this volume, fertilizer materials are classified by primary nutrient according to the following distinction: nitrogenous fertilizers, phosphatic fertilizers, potassic fertilizers, and multinutrient fertilizers. Two further terminological conventions should be noted in this context. First, fertilizer material containing only one primary nutrient is usually referred to as straight fertilizer. Second, a fertilizer product may be qualified as a low-analysis fertilizer if the ratio of nutrients to other components is relatively low, or as a high-analysis fertilizer if the ratio is relatively high. Commercial fertilizers are graded according to their content of available primary nutrients (and sometimes secondary nutrients), expressed as a percentage by weight, in the order presented in figure 1. Nitrogen is reported as the element N, whereas phosphorus and potassium are usually reported as the oxides P205 and K20, respec- tively. Thus, a 10-20-15 fertilizer contains 10 percent nitrogen, 20 percent phosphorus (expressed as P205), and 15 percent potassium (expressed as K20). When a fertilizer contains only two primary nutrients, the missing element is represented by a zero: for example, 10-20-0. The nutrient ratio is the proportion of the nutrients to each other. Thus, a 10-10-10 grade of fertilizer has a ratio of 1-1-1 and a 10-20-10 has a ratio 1-2-1. Thie Main Materials in Fertilizer Production The fertilizer industry covers a great number of raw materials, intermediate products, and fertilizer products. These products, as well as the main production processes currently employed, are described in some detail in the subsequent four chapters. In this section, a brief description is given of nitrogenous fertilizers, phosphatic fertilizers, potassic fertilizers, and multinutrient fertilizers. 2. The primary nutrients are also provided by animal manures. THE MAIN MATERIALS IN FERTILIZER PRODUCTION 13 Nitrogenous fertilizers Ammonia is the basis for almost all nitrogenous fertilizer produc- tion; in some technologically advanced situations, it is used directly as a fertilizer as well. The preferred feedstock to produce ammonia is natural gas; other raw materials include coal, naphtha, and fuel oil. Another important intermediate product in the nitrogenous fertilizer industry is nitric acid; at present, it is manufactured almost exclusively from ammonia. The most important nitrogenous fertilizer products are: * ammonia (82.5 percent N), which is produced primarily from natural gas, but also from naphtha, fuel oil, coal, refinery gas, liquefied petroleum gas, and the water-electrolysis process; * urea (45 to 46 percent N), which is produced from ammonia and carbon dioxide;3 * ammonium nitrate (33.5 percent N), which is produced from ammonia and nitric acid, often when carbon dioxide is un- available for urea production;4 . ammonium sulfate (20 to 21 percent N), which is produced either from ammonia and sulfuric acid, or as a by-product of the production of iron and steel in coking plants or of certain petrochemical processes such as the manufacture of caprolactam; . calcium ammonium nitrate (usually between 21 and 33.5 per- cent), which is produced by adding limestone to ammonium nitrate, thereby reducing its explosive as well as its hygroscopic character. Phosphatic fertilizers Phosphate rock is the basic raw material for the production of phosphatic fertilizers; in fact, finely ground phosphate rock itself is used directly as a fertilizer. In addition to rock-based phosphates, basic slag, which is a by-product of iron and steel manufacture, is also used as a raw material. 3. Urea is almost always produced in conjunction with synthetic ammonia because carbon dioxide can be obtained as a by-product of synthesis gas purifica- tion. 4. Ammonium nitrate is also used as an explosive. 14 THE FERTILIZER INDUSTRY: SOME TERMINOLOGY In most cases, phosphatic fertilizers are produced by reacting ground phosphate rock with either sulfuric or phosphoric acid. In turn, sulfuric acid is produced either from elemental sulfur, pyrites, or gypsum, or as a by-product of petroleum or metallurgical refining, while phosphoric acid is produced by combining phosphate rock and sulfuric acid. Depending on the type of acid used, and on the propor- tions of rock and acid, a variety of phosphatic fertilizers can be produced. The main ones, however, are single and triple super- phosphate. In short, the main phosphatic fertilizers are the following: * ground phosphate rock (variable P205 content, normally about 30 to 33 percent, but rarely of commercial interest unless more than 15 percent); * basic slag (usually between 10 and 20 percent P205) which is a by-product of iron and steel manufacture; * single superphosphate (16 to 21 percent P205), which is produced by combining ground phosphate rock and sulfuric acid; . triple superphosphate (43 to 48 percent P205), which is produced by combining ground phosphate rock and phosphoric acid. Potassic fertilizers Virtually all potassic fertilizers are produced from potash-bearing brines or from underground deposits of potash. The main potassic fertilizers are: * potassium chloride (60 to 62 percent K20); * potassium sulfate (50 to 54 percent K20). Multinutrient fertilizers Multinutrient fertilizers contain more than one primary nutrient. Their manufacture may or may not involve a major chemical reaction. Essentially, two routes are available to produce multinutrient ferti- lizers. The first is dry-mixing or bulk-blending, in which mutually compatible fertilizers, preferably in granular or prilled form, are mechanically mixed. In technologically advanced countries, bulk- blending is frequently integrated into the marketing and distribution system for fertilizers. The second possible route is granulation of several fertilizer intermediates; this process is more flexible in terms of input products, and it does not pose subsequent segregation problems. A variety of processes are in use, and they permit the production of a THE MAIN MATERIALS IN PERTILIZER PRODUCTION 15 wide range of fertilizer grades, including several of the highest concentration fertilizer products, such as diammonium phosphate (16 percent N, 46 to 48 percent P205) and monoammonium phosphate (11 percent N, 50 to 54 percent P205). These products enjoy a rapidly growing market around the world, both in direct use as a fertilizer and as material for further bulk-blending or granulation. Conclusion The chemical fertilizer industry comprises a large number of products and processes, and this number is continually increasing because of ongoing research efforts. Nevertheless, a rather small number of these possible products and processes predominate. Only the main products have been included in this chapter; products and processes that are used less frequently are included in the subsequent four chapters. 3 The Production of Nitrogenous Fertilizers THE ATMOSPHERE IS AN INEXHAUSTIBLE SOURCE OF NITROGEN. But, because atmospheric nitrogen is extremely inert chemically, it must be converted into an available form before it is suitable for plant use. Although some nitrogen is continually being fixed in the air and in the soil as a result of natural processes, this natural fixation is inadequate to supply the world's need for nitrogenous nutrients. Hence, increasing quantities of chemically fixed nitrogen are needed. Fixed nitrogen has three important natural sources: namely, by-product ammonium salts from the manufacture of coke, the Chilean deposits of sodium nitrate, and natural organic manures. Of the three well-known chemical processes for fixing nitrogen-the electric arc process, the cyanamide process, and synthetic ammonia production-only the last is important for the fertilizer industry. The electric arc process was formerly an important method for manu- facturing nitric acid, which is now produced chiefly by ammonia oxidation. The cyanamide process is currently important only in the manufacture of organic nitrogen compounds (nonfertilizer petro- chemical products), and hence it is not discussed here. Synthetic 16 AMMONIA 17 ammonia accounts for virtually all the nitrogen that is fixed chemically, and thus attention is focused on technological alternatives for producing it. Ammonia Ammonia has a crucial place in the manufacture of nitrogenous fertilizers. It is the principal form in which fixed nitrogen is available, and it is the basis for almost all nitrogenous fertilizers, including nitric acid, ammonium sulfate, ammonium nitrate, calcium nitrate, urea, ammoniating solutions, as well as fertilizer compounds and, indirectly, blends of fertilizer material containing nitrogen nutrients. Moreover, under certain conditions, ammonia-principally in the form of anhydrous ammonia-can be applied directly as a fertilizer.' Anhydrous ammonia contains 82.5 percent nitrogen by weight. Ammonia liquor (also known as aqua ammonia, ammoniacal liquor, or ammonia water) contains only 15 to 30 percent nitrogen, by weight in water; it is normally supplied to the fertilizer trade in a concentra- tion of 29.4 percent ammonia in water.2 Like the anhydrous form, the liquor has an alkaline reaction. The raw material used most often nowadays in the manufacture of synthetic ammonia is natural gas, which is present in many regions of the world. Natural gas can be classified as "associated" if it occurs with crude oil from which it is liberated, or as "nonassociated" if it comes directly from the well. Its composition varies, depending on the geographical location and the type of deposit, but essentially the gas is methane (60 to 96 percent) mixed with ethane (4 to 40 percent) and other higher hydrocarbons, gaseous impurities, and inert gases. It is preferable that the natural gas used as a raw material for synthetic ammonia contain a high concentration of methane because methane 1. The use of anhydrous ammonia by direct injection depends much more on the suitability of the soil and on the way the crop is grown than on most other factors; for example, the soil should be free of stones. This form of nitrogen ap- plication is most useful for wide-row crops. It is not good for grass or for the top dressing of cereals, no matter how technologically advanced the country is. 2. In Europe, the description "ammonia liquor" is usually reserved for the crude liquor obtained from coal carbonization. The aqueous solution made from anhydrous ammonia is usually described as "aqueous ammonia" or "aqua am- monia." 18 THE PRODUCTION OF NITROGENOUS FERTILIZERS has a lower ratio of carbon to hydrogen than higher hydrocarbons.3 In the synthesis process, all the carbon in natural gas is converted to carbon dioxide, which is removed in purification steps. Hence, smaller and less expensive purification units are required for plants that use natural gas with a high methane content. Until the early 1950s, fuel oils were commonly used in the produc- tion of ammonia, even in countries where natural gas was available, and even though fuel oil is more difficult to use than natural gas and the capital costs of plants using fuel oil are higher than those of plants using natural gas. Because of recent changes in relative prices, as well as in technology, however, there is currently renewed interest in using fuel oil. Naphtha is also an important alternative hydrocarbon raw material for ammonia synthesis. For economic and technical reasons, the sulfur content of the naphtha should be as low as possible. The use of cracked naphtha is not recommended because of the difficulty that might be encountered in removing the sulfur. Naphtha is produced during the refining of crude oil, and it can be processed further to yield gasoline. It can be transported by pipeline or bulk carriers such as trucks and ships. Until twenty to thirty years ago, coal was a major raw material for ammonia synthesis. Over the years, however, coal and lignite have given way to naphtha, natural gas, and fuel oil as the important feedstock. Consequently, old plants have been converted or shut down, and new ones have been designed for the use of these materials. Coal has not been economically competitive with liquid or gaseous petroleum materials; the quality of the competitive coal is low because of its high ash content and low calorific value. Further, the coal gasification processes are elaborate, and in general the plant investment, maintenance, and operating costs tend to be higher for them than for petroleum-based plants. The popularity of coal as a raw material for ammonia production may increase, however, because of the increasing costs of the petroleum materials.4 3. These considerations may not apply if the ammonia unit is part of a more comprehensive petrochemical complex. 4. Coal is only used as a raw material for ammonia synthesis today in those countries lacking oil or natural gas-South Africa is the main case. In the long term, however, producers are likely to be forced to use coal, and this is the reason for the keen interest in coal now being manifested in the United States as well as in other countries with substantial coal deposits, such as India, where two coal- based ammonia plants were scheduled to start production in 1978. AMMONIA 19 The manufacture of ammonia The synthesis of ammonia is of considerable historical importance in the chemical industry because it represents a significant application of thermodynamic principles to the solution of a difficult commercial process. The basis of ammonia production is the reacting of hydrogen with atmospheric nitrogen. The first ammonia plant came on stream in 1913 and used what became known as the Haber-Bosch process, after its inventors. Although variations on the basic process have been developed since then, the main differences lie in the method of preparing the synthesis gas, which is a 3-to- 1 mixture of hydrogen and nitrogen; the purification of synthesis gas; the design of the ammonia converter; and the method for recovering ammonia from the con- verter effluent gas. The four major steps in the manufacture of ammonia are synthesis gas preparation, carbon monoxide conversion, gas purification, and ammonia synthesis. Each of these steps can be accomplished by several processes, which will be discussed below. Figure 2 is a diagrammatic representation of the alternative processes available for the synthesis of ammonia. These processes are denoted by the rectangular boxes. Sy7nthesis gas preparation The synthesis ammonia process depends on the availability of large quantities of extremely pure synthesis gas.5 The major processes for preparing synthesis gas are steam reforming and partial oxidation; in addition, there are coal gasification, autothermal, and electrolytic processes.6 Several variations of each process are also available. Each process produces a gas that is rich in hydrogen and carbon monoxide; the carbon monoxide is later converted to carbon dioxide, generating hydrogen by reaction with steam in the shift converter. 5. Sulfur, phosphorus, carbon monoxide, and arsenic are irreversible catalyst poisons and can only be present in the order of parts per million (ppm). Water, oxygen, and carbon dioxide are reversible poisons that can be removed by heating the catalyst; even here, however, the catalyst suffers some permanent damage, so these gases must also be present in very low concentrations. 6. The electrolytic process is not very common; it produces hydrogen for syn- thesis gas by electrolysis of water. This process is used where electric power is (or was) inexpensive-for example, at Aswan in Egypt, and near Cuzco in southern Peru. 20 THE PRODUCTION OF NITROGENOUS FERTILIZERS Figure 2. Alternative Processes for Ammonia Synthesis Natural gas, coal, or naphtha SYNTHESIS GAS Coal Partial Steam PREPARATION gasifi- Autothermal oxidation reforming cation Impure synthesis gas CARBON Carbon MONOXIDE monoxide CONVERSION conversion CARBON Fluor Hot Monoethanol- DIOXIDE solvent Sulflnol potassium amine REMOVAL carbonate 0 0 FINAL Liquid Copper Metha- PURIFICATION nitrogen luor nation Purified synthesis gas Synthesis Synthesis AMMONIA SYNTHESIS converter converter #2 #1 Synthesized ammonia AMMONIA 21 STEAM REFORMING. Steam reforming of natural gas (methane) is usually carried out in two stages, using primary and secondary reformers with a nickel catalyst. Since the reforming catalyst would be poisoned by sulfur, the first step in the process is desulfurization of the natural gas. The purified gases are then mixed with steam and sent to the primary reformer. The exit gases from the primary reformer are mixed with air and are sent directly to the secondary reformer. Heat liberated by the partial oxidation of hydrogen and methane raises the temperature, which essentially completes the reforming of the natural gas. Until the early 1950s, the danger of tube failure restricted the process to a low-pressure operation. With the advent of new alloys and improved methods of fabrication, however, high-pressure operation is now common practice. This development has improved the efficiency of ammonia production by conserving the pressure of the incoming natural gas and eliminating the need to compress the process gas for the purification step, thus reducing the size of plant equipment and the volume of catalyst. Naphtha is an alternative feedstock for steam reforming, but there are essential differences between natural gas reforming and naphtha reforming: because of its high sulfur content, naphtha must undergo a preliminary acid treatment to remove most of the sulfur; naphtha contains more unsaturates and aromatics (that is, various other hydrocarbons), thereby increasing the possibility of carbon deposits; a vaporizer must be added in naphtha reforming; additional capacity to remove carbon dioxide is required in naphtha reforming because of the higher ratio of carbon to hydrogen in the feedstock; and naphtha reforming requires a catalyst that contains a promoter to inhibit carbon formation and that is more resistant to poisoning by sulfur. PARTIAL OXIDATION. Partial oxidation of hydrocarbons is another method of preparing synthesis gas, and several different forms of the method have been developed. One form of partial oxidation, which is very flexible, can be used for feedstocks ranging from natural gas to heavy fuel oils. The method does, however, require a source of oxygen, which is normally obtained from a liquid air plant. This method offers four advantages over the steam reforming process: no catalysts are required; the heat requirements are lower; impurities in the feedstock are tolerated; and the process can be adapted to a wide range of hydrocarbon feedstocks. The process also 22 THE PRODUCTION OF NrrROGENOUS FERTILIZERS has disadvantages: it requires a liquid air or air separation plant to produce oxygen, and the costs associated with such a plant are high; and substantial quantities of undesirable carbon are formed, and they must be removed from the combustion gases. AUTOTHERMAL. The autothermal process is a combination of the steam reforming and partial oxidation processes. The feedstocks for the process range from natural gas and naphtha to refinery gas and liquefied petroleum gas (LPG). Steam, air, oxygen, and the hydrocarbon feedstock constitute the feed material of the reactor. The process operates at high pressures without difficulty. The major advantage of the process is that it does not require catalyst tubes, which can present problems for high-pressure operations. Its disadvantage is that it requires an air separation plant to provide oxygen and nitrogen. COAL GASIFICATION. Many processes have been developed for coal gasification, but the two most important are the Lurgi process and the Koppers-Totzek process. Both use steam, oxygen, and coal or lignite as feed materials. In the Lurgi process, a mixture of steam and oxygen is passed into a bed of coal that is maintained at a high temperature. Small lumps of coal are fed in batches as the gas flows from the generator. This process for the formation of synthesis gas is used in ammonia plants in South Korea and Turkey. The Koppers-Totzek process uses pulverized coal as the feedstock. Oxygen and coal dust are passed into a gas generator into which steam is introduced. The gas flow is continuous. Gaseous or liquid hydrocarbons can be used instead of coal as feed material; this is a significant advantage. Among the countries using this process are France, Japan, and Spain. Carbon monoxide conversion The second step in the process of producing ammonia is the catalytic reaction between carbon monoxide, obtained in the first step, and steam to form hydrogen. Exit gases from the preparation unit (the secondary reformer in the case of the steam reforming process) contain appreciable quantities of carbon monoxide, which are converted to hydrogen by passing the gases through a converter containing a catalyst made of a mixture of iron and chromium oxide. For economic reasons, the conversion to hydrogen should be as high as possible. Variables affecting this conversion are: concentra- AMMONIA 23 tions of carbon dioxide, carbon monoxide, and steam in the gas entering the converter; temperature of the catalyst; pressure; catalytic activity; and gas velocity. Initial gas purification The third stage in the production of ammonia is to purify the process gas by removing carbon dioxide and carbon monoxide. In older ammonia plants, carbon dioxide was removed by scrubbing the process with water. This method, however, proved to be relatively inefficient and possessed several disadvantages; hence, all new processes employ solvents that absorb carbon dioxide more efficiently. Solvents commonly used are monoethanolamine, hot potassium carbonate, sulfinol, and fluorsolvent. Final gas purification Since carbon monoxide is an irreversible poison to the synthesis catalyst, its concentration must be reduced to a few parts per million in the synthesis gas. Carbon dioxide is still present in small amounts, and this must also be reduced before the gas is suitable for ammonia synthesis. Three processes are available to remove the traces of these oxides: methanation, copper liquor solution, and liquid nitrogen. METHANATION. Most new plants use catalytic methanation to remove carbon monoxide, carbon dioxide, and oxygen. In this process, the gas stream is heated and passed through a nickel-base catalyst. The carbon monoxide and carbon dioxide react with hydrogen to form methane and water. The equipment cost for this process is low, and the only operating cost is the initial charge of the catalyst. The major disadvantage is the hydrogen loss caused by the reaction with carbon monoxide and carbon dioxide and by the purging required to control the concentration of methane in the recirculating gas of the synthesis loop. COPPER LIQUOR SOLUTION. Scrubbing with copper liquor is one of the oldest processes for removing carbon monoxide, but it has been used very little in new construction. The operation of this process is more complex than the methanation process, and the corrosiveness of the solution results in higher maintenance costs. LIQUID NITROGEN. Scrubbing with nitrogen is economical only in 24 THE PRODUCTION OF NrTROGENOUS FERTILIZERS conjunction with partial oxidation plants, since the nitrogen is available from the liquid air plant. In this process, the gases are first dried and then washed directly with liquid nitrogen, which removes not only carbon monoxide and dioxide but also methane and argon. The resulting synthesis gas is so free of impurities that little purging is required. This is a low-temperature operation, however, and con- siderable heat exchange is needed to make the process economical. Ammonia synthesis Ammonia synthesis is a reaction between hydrogen and nitrogen at elevated pressure and temperature in the presence of a catalyst that is composed of iron oxides and contains promoters of aluminium, potassium, magnesium, and calcium oxides. Synthesis gas from the methanator is cooled, condensed water is removed, and the gas is then compressed to the final synthesis pressure. In the past, reciprocating compressors were used for the high-pressure compression; now, however, there is a trend toward centrifugal compressors and larger plants. The high-pressure synthesis gas is mixed with recycled gas from the ammonia converter; this mixture is passed through an oil trap to remove any entrained oil from the reciprocating compressors, and it is then sent to an ammonia refrigera- tion exchanger that condenses out ammonia from the recycled gas and removes nearly all traces of water from the synthesis gas. The purified synthesis gas passes to a separator and heat exchanger before entering the ammonia converter. The design of this converter is quite critical because fairly large quantities of heat are given off in the synthesis reaction. Two general types of synthesis converters are in use, and the major difference is in the method of temperature control. One type employs multiple beds of catalyst with provisions for cooling the gas between the beds by means of cooling coils or quenching with cold gas. The other type uses gas flow and heat exchangers to control the temperature. Nitric Acid Nitric acid is a strong acid and a powerful oxidizing agent that is normally manufactured as a product containing 55 to 60 percent acid. Approximately 75 percent of the manufactured nitric acid UREA 25 is used for fertilizer production, 15 percent for the manufacture of explosives (nitrates and nitro compounds), and 10 percent for numerous other purposes by the chemical industry. Three different processes can be used to manufacture nitric acid-ammonia oxidation, the electric arc process, and a process based on the reaction between sodium nitrate and sulfuric acid-but only the first is significant today. Basically, ammonia oxidation involves the oxidation of nitric oxide, which is itself produced by burning ammonia in air over a platinum catalyst, and the absorption of the oxides of nitrogen in water to form nitric acid. This process has three general forms: oxidation and absorption at atmospheric pressure; oxidation and absorption at elevated pressure; oxidation at atmospheric pressure but absorption at elevated pressure. There are numerous proprietory processes for nitric acid manufacture, which differ mainly in design details or selected operating conditions. Their major features are: * vaporization, superheating, and filtration of anhydrous ammonia; * preheating filtration and compression of process air; * catalytic oxidation of ammonia; * oxidation of nitric oxide to higher oxides; * absorption of the nitrogen oxides in water to form nitric acid; * acid bleaching; * tail-gas treatment; * energy recovery; * recovery of the catalyst. Urea Urea is the diamide of carbonic acid and in its pure state contains 46.66 percent nitrogen. It is sold in the form of crystals, granules (1.5 to 4 millimeters), or prills (I to 2.5 millimeters), with or without a mineral coating. Crystal urea and uncoated prills contain 46 percent nitrogen, and coated prills average 45 percent nitrogen. Synthetic urea is chemically identical with organically produced urea, and its high nitrogen content makes it the most concentrated solid nitrogen fertilizer material in the world fertilizer market.7 Accordingly, it is 7. For use as a straight nitrogen fertilizer, prills hold most of the world market. A typical specification for prilled urea is 95 percent concentration and 1 to 2.5 millimeters with no oversize. 26 THE PRODUCTION OF NITROGENOUS FERTILIZERS increasingly favored among manufacturers of high-analysis grades of NP and NPK fertilizers. Generally, urea must be used with caution in mixtures because it is hygroscopic and incompatible with triple superphosphate. Because of its hygroscopicity, urea should never be mixed with ammonium nitrate in solid fertilizers. One of the most popular uses of urea in the manufacture of mixed liquid fertilizers is its application in solution with ammonia as ammonia-urea liquor or as ammonia-ammonium nitrate-urea solution. Urea is sold under various trade names. Uramon is a processed form that contains 42 percent nitrogen. Ureor and Caliureor are trade names used in Europe to designate processed types of crystalline urea that are coated with a finely ground limestone to reduce the tendency to absorb moisture. The manufacture of urea Urea was the first organic compound to be synthesized from inorganic materials, and thus it is of considerable historical interest. At one time, substantial quantities of urea were made by the hy- drolysis of cyanamide produced from calcium carbide. At present, however, virtually all urea is based on the dehydration of ammonium carbamate. Most urea is manufactured in conjunction with synthetic ammonia since the necessary carbon dioxide is available from the synthesis gas purification system at essentially zero cost. Urea can be manufactured by various processes, which are classified according to the degree of recycling of the unconverted ammonia and carbon dioxide. The processes are referred to as the once-through process, the partial-recycle process, and the total-recycle process. These alternative processes are diagrammed in figure 3. The only difference among the processes is in the handling of the gases evolved during the decomposition of the ammonium carbamate. The choice of the process depends upon the location of the plant and whether or not the urea process can be integrated with other plant operations. Thus, the once-through and partial-recycle processes might be most economical if the urea off-gases could be used to produce ammonium sulfate or ammonium nitrate, and if carbon dioxide is available in ample supply at low cost. On the other hand, the total-recycle process would be used if there were no place to dispose of the off-gas ammonia. The most serious problems associated with any method of urea synthesis are the corrosion of equipment, the formation of biuret UREA 27 Figure 3. Alternative Processes for the Manufacture of Urea UREA SOLUTION Once through Ammonia and .Urea carbon dioxide Urea "Melt" solution reactor - ata ey:cle _ p _ Total recycle_ SOLID UREA A Atmospheric evaporation Urea Vacuum evaporation 1 Solid solution urea X Crystallization Granulation (which is toxic to some plants), and the presence of unconverted ammonia and carbon dioxide. THE ONCE-THROUGH PROCESS. In the once-through process, the solution or melt from the urea reactor is flashed to a lower pressure and is heated to drive off ammonia and decompose the unreacted carbamate. The resulting solution is 80 percent urea, which can be either utilized directly or concentrated to crystalline urea. Although this is the simplest process, it is the least flexible and cannot be operated unless some provision is made to utilize the large amount of off-gas ammonia. This off-gas can be absorbed in acid to produce ammonium nitrate, ammonium sulfate, or ammonium phosphates. Only about 32 percent of the ammonia is converted, and thus several 28 THE PRODUCTION OF NITROGENOUS FERTILIZERS tons of ammonium sulfate or nitrate are produced for each ton of urea. Even though the carbon dioxide in the off-gas is lost and hence must be available in sufficient supply, this is the least expensive process in both capital investment and operating costs. THE PARTIAL-RECYCLE PROCESS. In the partial-recycle process, part of the off-gas ammonia and carbon dioxide from the carbamate strippers is recycled to the urea reactor and part is used to produce a coproduct nitrogen material, as in the once-through process. Although the rate of coproduction is greatly reduced in this process, the operation of the urea plant must still coincide with that of the coproduct plant. THE TOTAL-RECYCLE PROCESS. In the total-recycle process, all the unconverted ammonia and carbon dioxide mixture is recycled to the urea reactor, where about 99 percent of the mixture is then converted. No by-products are produced, and hence no nitrogen coproducts are necessary. Of the three urea synthesis processes, this is the most flexible because it depends only upon its supporting ammonia plant for operation. It is also the most expensive process in investment and operating costs; hence, if the production of other materials requiring ammonia is planned, either of the other two processes would have lower investment and operating costs. Synchronizing the operation of the urea and the nitrogen coproducts plant presents difficulties, however, and this has often biased the decision in favor of the total-recycle process, despite the lack of by-products. Total-recycle systems are of two general types: the gas-separation system and the carbamate-solution-recycle system. Comparing these two types of processes is difficult, but there are indications that the requirements of the gas-separation unit for utilities such as water and electricity are significantly higher than those of some solution-recycle processes. Special processes Three special processes that have been developed deserve mention: the carbamate-recycle process, the thermo-urea process, and the "integrated-loop" process. THE CARBAMATE-RECYCLE PROCESS. In the carbamate-recycle process, most of the unreacted feed-gases (from the reactor effluent at synthesis UREA 29 pressure) are stripped by a fresh carbon dioxide feed in a special, indirectly heated stripper. The remaining unreacted gases are stripped and recycled by the aqueous carbonate solution method. Ninety percent of the ammonia can be converted in this way. It is claimed that the requirements for steam, electricity, and cooling water are reduced significantly in this process. A similar process has been developed with gaseous ammonia as the stripping agent. THE THERMO-UREA PROCESS. In the thermo-urea process, also known as the hot-gas recycle, sufficient heat is produced to eliminate the need for steam but the electrical power requirements offset some of this saving. Because of the high temperatures, centrifugal compressors rather than reciprocating compressors must be used, to avoid maintenance problems. But centrifugal compressors are not practical in small sizes, and thus 1,500 tons of urea a day would need to be produced for their best use. THE "INTEGRATED-LOOP" PROCESS. The integrated-loop process combines urea and ammonia production into a single unit. There are three combined steps: carbon dioxide is scrubbed from the converted gas in the ammonia train by carbamate-recycle solution (from the urea plant absorbers) and by fresh ammonia; the resulting scrubbing liquor is fed by its own pressure to the urea reactor; the hot gas from the carbon monoxide shift converter is cooled by heating the urea effluent in the decomposers to strip off the unreacted ammonia and the carbon dioxide. In general, more carbon dioxide is produced than is needed to convert all ammonia to urea; hence, an auxiliary system to remove carbon dioxide must be provided. The principal advantages of this process are the substantial savings in plant investment and operating costs that arise because: the need for a carbon dioxide compressor is eliminated; the size of the urea reactor is reduced; the requirements for carbon dioxide regeneration equipment and steam are eliminated; the heat of converted gas is utilized in the urea plant; the ammonia purge-gas recovery system is eliminated; and the ammonia pumping requirements are reduced. Some disadvantages are also associated with this process-namely, the ammonia plant is completely dependent upon the urea plant operation, and the production rate of the urea plant must be matched with that of the ammonia plant, unless a large standby carbon dioxide removal system is provided. 30 THE PRODUCTION OF NITROGENOUS FERTILIZERS Prilled urea The synthesis processes described above produce an aqueous solution containing between 70 and 80 percent urea, with the propor- tion depending on the extent of recycling. This solution must be concentrated if solid urea is to be produced, and it is in the concentra- tion step that biuret is formed, unless provisions are made to prevent a harmful combination of high temperature and long retention time. In practice, some compromise is made between the biuret content of the urea and the size and costs (both operation and investment) of the evaporation equipment. The method of evaporation and solidification should be chosen to give the maximum acceptable biuret content at the minimum cost. For most commercial crop applications, a biuret content of I percent is acceptable. Prilled urea is produced by four methods: atmospheric evaporation, vacuum evaporation, crystallization, and granulation. The differences among the first three lie in the method of producing the concentrated melt required for the prilling operation. ATMOSPHERIC EVAPORATION. The moisture from a continuously replenished film of stripped urea reactor effluent is evaporated. These evaporation units have either rotating discs to spray the feed against the jacketed walls or rotating blades that wipe the walls continuously. Both present maintenance problems associated with bearing lubrica- tion. In some units, a current of inert, dry gas is passed through to carry out the water vapor. The resulting "melt" is sprayed into the top of a prilling tower, and the droplets from the spray solidify upon cooling during their descent. Despite the short retention time (few seconds) in the evaporator, the biuret content is about 0.8 to 1.5 percent. VACUUM EVAPORATION. Water is evaporated at the melting point of urea in a conventional heat exchanger operating under a vacuum. Though the retention time is short because of the flash evaporation, the biuret content is generally 0.7 to 1.0 percent. Two or more stages of evaporation may be used. This process is widely used because it does not have the maintenance problems of the atmospheric evapora- tion unit. CRYSTALLIZATION. A circulating vacuum crystallizer is used to AMMONIUM SULFATE 31 produce a saturated solution of urea, from which urea crystallizes. A side stream of the suspension is withdrawn and fed into a centrifuge, where the crystals are separated. They are then dried in a rotary drier. The mother liquor is recycled through the off-gas absorbers to the urea reactor. Since the biuret does not crystallize with the urea, any that is formed is recycled with the molten liquor. Under the reactor conditions, the biuret is reconverted to urea. GRANULATION. The granulation process, which has been developed recently, uses a tilted-pan granulator.8 The stripped urea effluent solution is concentrated to about 95 percent in a vacuum evaporator. The solution is then sprayed onto a tumbling bed of urea granules. The sprayed solution agglomerates or coats the particles, the tumbling action rounds off the agglomerates, and the classifying action of the tilted pan produces urea granules that are substantially on-size-that is, 1.5 to 3.5 millimeters. These granules are then dried, cooled, and screened. Investments and operating costs of this process are lower than those of the prilling operation, and the particles are of larger size. The.biuret content is about the same, however. A granulation process has also been developed using rotary drums instead of the tilted-pan granulator. Ammonium Sulfate Pure ammonium sulfate contains 21.2 percent nitrogen. The commercial grade ordinarily used by fertilizer producers contains small amounts of moisture-free sulfuric acid and other impurities, and it is guaranteed to contain 20 to 21 percent nitrogen. It contains about 24 percent sulfur, and this is often seen as a valuable plant nutrient. In the formulation of granular mixed fertilizers and bulk blends, ammonium sulfate is highly esteemed as a nitrogen source because of its low hygroscopicity. It can be mixed in all proportions with practically all the solid raw materials of the trade. It is not, 8. The tilted-form granulator has been used in the fertilizer industry for at least twenty-five years, but it has only been applied to urea in recent years. Its main advantage is that it gives a product which is the same size as granular diammonium phosphate. Bulk blends can therefore be made without significant segregation problems. 32 THE PRODUCTION OF NITROGENOUS FERTILIZERS however, compatible with alkaline materials such as lime, cyanamide, calcium nitrate, and basic slag. It can be mixed with urea, provided the mixture is used shortly after mixing. Ammonium sulfate is produced by six principal methods, which are described below. COMBINED REACTION-EVAPORATION METHODS. In the combined re- action-evaporation methods, anhydrous ammonia and sulfuric acid are reacted in a continuous saturator-evaporator under vacuum or at atmospheric pressure. The resulting crystals are recovered by means of a centrifuge or a filter. GAS-WORKS BY-PRODUCT METHODS. Before synthetic ammonia was available, most ammonia was obtained from solid-fuel carbonization. The bituminous coals used to produce gas and coke contain about 2 percent nitrogen, of which 15 to 20 percent can be recovered (at high temperature) as ammonia; that is, 5 to 6 pounds of ammonia per ton of coal. Hence, most by-product ammonia is associated with high-temperature carbonization units-for example, at coking plants for iron and steel production, about 35 to 40 pounds of ammonium sulfate can be produced for each ton of steel. There are three principal methods for ammonia recovery and the subsequent production of ammonium sulfate or other ammonium salts: the direct, indirect, and semidirect processes. The two main processes currently in use are the indirect and semidirect ones. THE AMMONIUM CARBONATE-GYPSUM PROCESS. Ammonium carbon- ate and calcium sulfate (anhydrite or gypsum) can be made to react in a series of wooden vessels or mild steel tanks fitted with steam coils and agitators.9 The gypsum can be derived from natural sources or by-product sources (for example, phosphoric acid production), and the ammonium carbonate is obtained by absorbing ammonia gas in water and carbonating the solution in carbonating towers. RECOVERY OF BY-PRODUCT LIQUOR. The waste liquor stream from the processes for the production of caprolactam, acrylonitrile, and some other products contain at least 35 percent ammonium sulfate in solution. This can be recovered as a nearly pure salt by crystallization 9. Almost all plants using this process have now been shut down. AMMONIUM NITRATE 33 of the waste liquor and subsequent centrifuging. Spent sulfuric acid from petroleum refineries, petrochemical plants, and soap factories can also be used, if the impurities present in the acid do not cause frothing or corrosion problems. An alternative is to ammoniate the contaminated acid and granulate the slurry on a moving bed or in a drum and to recycle the products in a drier-screening system to produce granules of the required size range. SPRAY TOWER AMMONIATION. Sulfuric acid can be sprayed into ammonia vapor inside a spray tower. The heat of reaction produces a dry, amorphous product, which is removed continuously from the base of the tower. This form of ammonium sulfate is particularly suitable for dry-mixed and granulated-mixed fertilizers. MISCELLANEOUS PROCESSES. One method utilizes an organic solvent that absorbs sulfurous gases. After the absorption, the liquor is blown with air to form a basic sulfate, and ammonia is added to produce ammonium sulfate. This salt separates from the organic base and is centrifuged, dried, and sent to storage. The organic absorbent (for example, xylidine or monomethylaniline) is recycled for further use. This process can be useful where sulfur is costly or where air pollution is a serious problem, since it does not require a source of sulfuric acid and it can operate on rich or lean sulfurous gases from roasters, boiler flues, and other sources. Other proposals for recovering sulfur from flue gases are based on scrubbing with ammonia to yield mixtures of ammonium bisulfite and ammonium sulfate. Ammonium Nitrate Next to urea, ammonium nitrate is the most concentrated solid nitrogen compound used in the fertilizer industry. It is available as granules, flakes or crystals, and prills, which, when mixed with kieselguhr or kaolin to improve the compound's caking behavior, contain about 33.5 percent nitrogen as a commercial product. One-half of this nitrogen content is in the nitrate form. Ammonium nitrate is hygroscopic and prone to fire or explosion unless some suitable precautions are taken. It is very soluble in water and aqueous ammonia and is the principal solid nitrogen material used in the preparation of ammoniating solutions. In its liquid form, it is available as an 80-90 percent concentrated solution. It is used in solution form 34 THE PRODUCTION OF NITROGENOUS FERTILIZERS Figure 4. Alternative Processes for the Manufacture of Ammonium Nitrate C eStengel PMelt processes .a_ fo dret ppictin o_ h Crsollatindisodfrmorbl-enng Ammonia and Ammonium nitric acid nitrate n epratining m T RI G. P ied ammoniunitratismadbyre Ca nitro circntreamk o Ts p ct Esolution a. Prilling is also a melt process. for direct application to the soil and in solid form for bulk-blending. Increasing quantities are used in conjunction with fuel oil for blasting purposes, and relatively sniall amounts are consumed by the brewing and chemical industries. Ammonium nitTate is produced from ammonia and nitric acid by several processes that vary only in their combinations of different neutralization, evaporation, drying, and finishing methods. The alternative manufacturing processes are diagrammed in figure 4. PRILLING. Prilled ammonium nitrate is made by reacting nitric acid with ammonia in a circulating stream of ammonium nitrate solution. This produces an 83 percent solution of ammonium nitrate, which salts out at 71 degrees centigrade. The solution is then pumped to an evaporator. Concentrated ammonium nitrate solution containing less than 5 percent water is pumped from the evaporator to the top of the AMMONILUM NrrRATE 35 prilling tower and sprayed down inside the tower. The droplets of nitrate cool and harden during the fall to spherical pellets (or prills).'0 These are carried off from the base of the tower by a conveyor and, if necessary, are further dried to below 0.5 percent moisture. The pellets are coated with a dry powder to keep them free-flowing. MELT PROCESSES. Ammonium nitrate can also be produced in molten form. In one process, ammonia vapor and 58 percent nitric acid are preheated separately in specially designed heat exchangers and are fed continuously and simultaneously to a reactor. Ammonium nitrate and a trace of ammonia flow from the reactor into the separator, and the molten ammonium nitrate from the separator then enters a wire box that distributes it on an endless, stainless steel, watercooled belt. The chilled nitrate flake is fractured, coarse ground, screened, coated, and bagged as a 33.5 percent nitrogen, semigranular product. In another process, gaseous ammonia and concentrated nitric acid are pumped into molten ammonium nitrate. The anhydrous molten salt is obtained after removal of the water vapor formed. The salt is then used to produce granules, prills, or other forms according to need. CRYSTALLIZATION. In crystallization, ammonia is reacted with 60 percent nitric acid in a circulating loop, and the ammonium nitrate formed is dissolved in the recirculating molten liquor from the crystallizer."1 To avoid the possibility of hazard, a vacuum crystallizer is used at low temperatures. Evaporation is divided between the concentrator (evaporator) and the crystallizer. Concentrated liquor from the evaporator is fed into the crystallizer, where the solution is cooled under vacuum to effect crystallization. The crystals are centrifuged, passed through a rotary drier, coated, and bagged. The finished product contains about 33 percent nitrogen. GRAINING. In graining, the nitrate solution is evaporated in batches to about 98 percent (almost a molten salt) and is then discharged into kettles equipped with heavy plows that knead the material as it cools 10. One of the major problems of prilling ammonium nitrate is that many of the spherical pellets are not solid. They contain internal cavities and the prill is correspondingly weak. 11. This process is not significant for fertilizers, but it is mentioned for the sake of completeness. 36 THE PRODUCTION OF NITROGENOUS FERTILIZERS and solidifies.12 Steam or cooling water is used to control the cooling rate. Additional moisture is evaporated during this treatment, and the mass first "fudges" and then breaks apart into small, rounded pellets or grains which are subsequently cooled, screened, mixed with clay, and bagged. GRANULATION. Ammonium nitrate is also produced in granular form using rotary drums, pug mills, or combinations of these. The granules are then dried by the latent heat of crystallization. Air-swept rotary drums are used for final drying, cooling, and hardening. This is followed by coating and bagging in moisture-proof bags. NITROCHALK. Adding powdered limestone or calcium carbonate to ammonium nitrate improves its storage properties and minimizes the risks of fire or explosion. The addition is made to the concentrated solution before prilling or granulating. The product, which is called calcium ammonium nitrate (CAN), "calnitro," or lime ammonium nitrate, contains between 15 and 30 percent nitrogen; the most common grade is 26 percent nitrogen. AMMONIUM NITRATE SOLUTIONS. There are several ammonium nitrate solution processes. In one process, part of the ammonia from the once-through urea process is used to make nitric acid, and part is reacted with tail gas from the nitric acid unit to yield ammonium nitrate. This product is used alone or in combination with urea in the form of a fertilizer solution containing 32 percent nitrogen (UAL 32). Ammonium Sulfate-Nitrate Ammonium sulfate-nitrate, which is sold under the trade names Leunasalpeter and Montansalpeter, is a double salt with a nitrogen content of 26 percent. It is not hygroscopic, and its storage properties are superior to those of ammonium nitrate or mixtures of solid ammonium sulfate and ammonium nitrate, since no free ammonium nitrate is present. The importance of ammonium sulfate-nitrate has, 12. This process is also not significant for fertilizers. AMMONIUM CHLORIDE 37 however, diminished in most countries because of the large-scale manufacture of urea as well as of binary and tertiary high-analysis fertilizers. Ammonium sulfate-nitrate can be produced by ammoniating mixtures of sulfuric and nitric acids or by combining ammonium nitrate and ammonium sulfate in special ways. In one process, it is made by ammoniating the requisite mixture of sulfuric and nitric acids, followed by evaporating and adding ferrous sulfate (to reduce caking), and then cooling, chilling, and flaking the products. After further conditioning by spraying with diluted ammonia solution, the salt is granulated, dried, cooled, and bagged. The product is a double salt containing 62 percent ammonium sulfate and 38 percent ammo- nium nitrate; its total nitrogen content is 26 percent. In a simpler process, ammonium nitrate solution is evaporated under vacuum to a 95 percent concentration, cooled, and reacted with solid ammonium sulfate in a pug-mill granulator. The product is then dried, cooled, and bagged. Ammonium Chloride Ammonium chloride is the ammonium salt of hydrochloric acid and contains an average of 23.4 percent nitrogen-the range is from 18.6 to 26.1 percent. It is not used as a fertilizer in the United States, but in Japan and other Far Eastern countries it is applied to paddy on an appreciable scale and is usually produced when there is an abundant supply of hydrochloric acid. Its major disadvantages as a fertilizer material are: the resulting high acidities and chloride content in most soils, unless they are well irrigated and limed; the poor storage properties, unless the material is granulated and packed in moisture- proof bags; and a tendency to corrode handling and application equipment, unless certain components are suitably modified and protected with acid-resistant materials. If these precautions are observed, ammonium chloride fertilizer could become a useful outlet for surplus supplies of chlorine. Further, it can be safely applied to rice in the presence of certain fungi that would reduce ammonium sulfate to toxic sulfides. Ammonium chloride is also used in the manufacture of dry-cell batteries and as a flux for soldering and brazing. 38 THE PRODUCTION OF NrTROGENOUS FERTILIZERS Other Nitrogenous Fertilizers Besides the nitrogenous fertilizers that have been treated above, others that are less important in the fertilizer industry deserve some mention because of their special nature. Calcium nitrate Calcium nitrate is the calcium salt of nitric acid and contains 17 percent nitrogen and 34.2 percent calcium, calculated as oxide. Because of its extreme hygroscopicity, even in moderately humid climates it is not used as a fertilizer, nor is it used in conjunction with other fertilizer intermediates. Also, precautions are usually taken to avoid impregnating organic material with calcium nitrate because of its tendency to explode in the presence of heat. It is used in explosives, fireworks, and inorganic chemical operations. Calcium nitrate is made in two ways. In one method, calcium carbonate is reacted directly with nitric acid; in the other, calcium nitrate is produced as an important by-product in some nitro- phosphate processes. Sodium nitrate Sodium nitrate has a nitrogen content of 15.4 to 16.5 percent. Although it has long been applied to the soil as a surface dressing for some vegetable crops, cotton, and tobacco, its use as a "straight" nitrogen fertilizer has declined. Like other nitrates, it tends to leach the soil. Sodium nitrate can be produced from natural deposits or as a synthetic product. Nitrogen solutions A nitrogen solution is an aqueous solution of ammonia, ammonium nitrate, or urea-separately or in combination-used either for the manufacture of mixed fertilizers or for direct application. These nitrogen solutions include aqua ammonia but not anhydrous ammonia or liquid mixed fertilizers containing potash or phosphate. They have certain advantages over solid fertilizers: they can be applied OTHER NITROGENOUS FERTILIZERS 39 accurately to the soil without problems of caking or dusting; they can be incorporated in irrigation water more readily than solids; and they are easy to handle since there is no need to lift or move fertilizer bags. They can be manufactured in an independent process or in an accessory operation integrated with the manufacture of synthetic ammonia, ammonium nitrate, or urea. Their disadvantages are: they are expensive to store, they need expensive application equipment, and they often contain less nitrogen than solids. At the moment, it is unlikely that these solutions will be used significantly in developing countries. The most important nitrogen solution is aqua ammonia. Other relatively less important nitrogen solutions are: solutions containing ammonia and ammonium nitrate; solutions containing ammonia and urea; solutions containing ammonia, ammonium nitrate, and urea; and solutions containing ammonium nitrate and urea but not ammonia. Aqua ammonia is used primarily as a direct-application fertilizer. It is a good source of nitrogen under many conditions and is some- times superior to other sources. Its nitrogen content is between 15 and 30 percent by weight. It is generally applied subsurface to avoid loss by evaporation. A simple operation that can be carried out independently of an ammonia synthesis plant is to add anhydrous ammonia to water. The operation requires facilities for proportioning the flows of water and ammonia, cooling the freshly formed aqua ammonia, and measuring its concentration. The rough proportioning of the flows of water and anhydrous ammonia can be regulated by the heat of solution using a temperature recorder-controller that adjusts the water flow based on temperature. Other methods of proportioning the flows of water and ammonia are based on the concentration of a recirculating stream of aqua ammonia, as indicated by a hydrometer. The aqua ammonia can be cooled using heat exchangers or refrigeration. Local conditions such as the availability and temperature of the cooling water may affect the choice of method. Ureaform Ureaform is a generic name for a type of nitrogen fertilizer that has the valuable property of controlled nitrogen release to plants. These urea-formaldehyde fertilizer materials are reaction products of urea and formaldehyde and contain at least 35 percent nitrogen, largely in 40 THE PRODUCTION OF NITROGENOUS FERTILIZERS insoluble but slowly available form. Most ureaform products contain a small amount of unreacted urea which, along with the simpler methylene ureas, make up a soluble and relatively quickly available nitrogen portion. The more complex methylene urea molecules constitute the less soluble and predominant fraction of ureaform, and they provide the slow-release supply of nitrogen for plants. Ureaform of high quality is, therefore, a mixture of methylene ureas, ranging from methylene diurea to forms containing about six urea molecules. In general, this type of fertilizer tends to be expensive and is most often used for horticultural purposes.13 13. Ureaform is much too expensive to be used in agriculture. A more promising approach to controlled-release nitrogen fertilizers may be use of sulfur-coated urea or use of very large urea granules. 4 The Production of Phosphatic Fertilizers PHOSPHORUS IS ESSENTIAL TO THE GROWTH OF PLANTS, and fertilizers containing this element are applied to improve yields and hasten plant maturity. Phosphate is fixed quite quickly by most soils, and the resulting phosphorus complexes are only slightly soluble in the soil solution. Phosphorus is absorbed from the soil solution by plants as the orthophosphate ion rather than as the element. Deposits of phosphate rock exist in nature because of the low solubility of the phosphorus compounds. A primary objective of the fertilizer industry is to convert the rock phosphates into more available compounds- that is, into forms of the element that can be more readily absorbed by plants. The processes involved in this conversion are the subject of this section. Phosphate Rock Phosphate rock is a term applied to naturally occurring phosphate minerals containing 10 percent P205 or more. Chemically, the phos- 41 42 THE PRODUCTION OF PHOSPHATIC FERTILIZERS phate minerals present are variants of apatite, with fluoropatite and hydroxyapatite predominating. Certain iron ore deposits also have a significant phosphorus content, and a fertilizer (basic slag) is ob- tained as a by-product of the steel industry. Phosphate rocks can be converted into a form that is available to plants by various methods, but in general these methods can be classified as thermal or acidula- tion. The latter is far more important because it produces a fertilizer that is cheaper and agronomically more effective. Most phosphate rock undergoes three general stages of treatment -mining, washing, and beneficiation-in order to upgrade the P205 content of the phosphate rock. Although the methods and tech- nologies vary according to the type of deposit, general descriptions of the three stages can be given. MINING. The mining operations can be surface mining or under- ground mining.' In surface mining, the tasks of removing overbur- den (deposits of unconsolidated quartz sand), digging the ore, and delivering it to a sump are normally carried out by dragline opera- tions, bulldozers, or other digging methods. In underground mining, several methods can be used, depending on the type of ground and the wall conditions. Where both walls are firm, room-and-pillar or stull stopes are employed. If the bed is thin, however, underground mining may be uneconomical, and the mines may either close down or resort to strip mining. Strip operations have become the major method of mining in the western United States. WASHING AND BENEFICIATION. After the matrix has been broken up in the pit and sluiced to the suction of the sand pump, it is pumped to the head of the washer plant.2 In the United States, it used to be customary to move a washer plant to the site of the mine. The changes in plant construction, however, and the upgrading of equipment may now require that the feed be transported to the plant. Under such circumstances, careful selection of the construction material of both pumps and pipelines is an important economic decision. Clearly, 1. Most mining operations are surface mining because underground mining is usually much more costly. The proportion of underground mining is falling, and this trend is likely to continue. 2. Washing and beneficiation can be done together or separately, depending on the P205 content of the rock. SULFURIC ACID 43 this need not be the case in all countries. After the slurry has reached the feedbox on top of the washer plant, it is generally split into three fractions: the oversized particles go to the log washers and hammer mills, the fines go to the flotation plant, and the middlings are scrubbed and screened. When the middling fraction is screened, clay, sand, and fines are separated from the pebbles, which are retained on the screens. The discharge through the screens contains all the clay slimes in the matrix, in addition to the silica and the phosphate particles. These slimes tend to absorb reagents and make flotation costs prohibitive. The removal of the slimes before further concentration is thus very important. Before the flotation process was discovered, the pebble fraction was the only phosphate value recovered. Since there was no known method to separate the small phosphate particles from sand and clay, they were discarded with the slimes. In the method of fatty acid flotation, however, the fine phosphate particles are separated from silica sand by a froth-flotation process. After the pebble rock is recovered, the remaining plant feed is separated into two fractions: agglomeration flotation of the material larger than 29 mesh, and cell flotation of the material smaller than 29 mesh.3 The first fraction is conditioned with flotation reagents and is treated by mechanical separation to separate the phosphate from the silica sand. The tailings from this separation join the tailings from the cell-flotation fraction and go to the general mill-tailing pond. The concentrate is washed and then dried for shipment. Sulfuric Acid To be converted to an available form, phosphate rock must be acidulated; hence, sulfuric or phosphoric acid is needed. The ferti- lizer industry consumes more than 45 percent of the world produc- tion of elemental sulfur and sulfur equivalents from pyrites and other sources, primarily as sulfuric acid for the manufacture of phosphatic fertilizers. The cost of producing sulfuric acid is lower 3. The mesh number refers to the openings per linear inch. 44 THE PRODUCTION OF PHOSPHATIC FERTILIZERS than that of any other mineral acid. Sulfuric acid is normally manu- factured from elemental sulfur, pyrites, anhydrite, and the sulfur- containing gases produced by roasting sulfide ores such as zinc and copper. Crude sulfur is produced mainly by the Frasch process, which involves drilling a well and running three concentric pipes down to the sulfur deposits. Water at high temperature and pressure is pumped down the center pipe to melt the sulfur, which collects in a pool at the base of the well. Air at high pressure is forced down the central pipe and the mixture of air and molten sulfur is brought to the surface in the annular space of the two outside pipes. The molten sulfur is distributed to vats, where it solidifies and can later be broken Up.4 Crude sulfur mined in this way is 97 percent pure. Al- though this sulfur can be shipped in open railroad cars and barges, for economic reasons it is often shipped in liquid form in insulated tank cars, barges, and ocean vessels. The sulfide minerals-such as iron, zinc, and copper sulfide-are normally obtained as concentrates by flotation from the crude ore. The chief problem with using sulfide ore as the basis for sulfuric acid is that the mineral industry is often located at a considerable distance from potential markets for the acid. Sulfuric acid is so inexpensive that it is usually impractical to ship the chemical over large distances because the shipping costs would be more than the value of the acid. Thus, large quantities of sulfur are wasted in the refining of the ore.5 The main processes for manufacturing sulfuric acid are from sulfur by means of either the contact process or the chamber process; they are diagrammed in figure 5. The differences between them lie in the manner in which sulfur dioxide is converted to sulfur trioxide and then reacted with water to produce sulfuric acid. Sulfuric acid can also be made as a coproduct in the manufacture of cement. Anhydrite is used in place of limestone in a cement kiln, coke is 4. The increasing tendency is to convert the molten sulfur into forms that do not need to be broken up and that can be handled easily. Two important forms are flakes and pellets. 5. Of the three sulfide minerals mentioned, zinc and copper are smelted to re- cover the metals, and the sulfuric acid is a by-product. In the case of iron sulfide (pyrites), the mineral is normally shipped to the location where the sulfuric acid is required. SULFURIC ACID 45 Figure 5. Alternative Processes for the Manufacture of Sulfuric Acid Sulfur pyrites Chamber Sulfuric acid: process 77 percent Sulfur Contact Sulfuric acid: Pyrites, metallurgical Su9498 percent gases, or acid sludge Contact process added to reduce the sulfate, and the sulfur dioxide in the kiln gases is converted to the trioxide by the contact process.6 THE CONTACT PROCESS. The success of the contact process depends on the catalytic oxidation of sulfur dioxide to sulfur trioxide. The common catalysts are iron oxide and vanadium pentoxide, although new plants use the latter almost exclusively. The sulfur trioxide is then passed to a tower, where it is absorbed in recycling concentrated sulfuric acid to which an equivalent amount of water is added. 6. When sulfuric acid is made via the cement route, anhydrite is normally used as the source of calcium sulfate; using gypsum results in a more costly production operation because of its water content. But phosphogypsum can be regarded as a by-product, and using it is often less expensive than mining anhydrite. The pro- duction of acid by the cement/sulfuric route is tending to diminish at present be- cause the capital costs of the plant are much higher than those for making sulfuric acid from sulfur and the quality of the cement produced tends to be lower than that of conventional cement. Moreover, the quantity of cement made annually by a reasonably sized plant is large; therefore, some of the product has to be shipped long distances and the cost of transport becomes progressively higher. 46 THE PRODUCTION OF PHOSPHATIC FERTILIZERS The basic design of the plants has become more or less stand- ardized, and descriptions of modern plants are easily available. All plants perform the following operations: produce sulfur dioxide from a suitable raw material; cool, purify, and dry the sulfur dioxide gas; preheat the gases to the kindling temperature for conversion to sulfur trioxide; catalytically oxidize the gases to sulfur trioxide; cool the converted gases; and absorb the sulfur trioxide in strong sulfuric acid. Contact sulfuric acid plants are of two types, depending on the raw material used. The first type, referred to as the "hot-gas-purifi- cation" plants, includes those plants operating on Frasch or recov- ered sulfur. These are the least expensive and easiest to operate if there is an abundance of sulfur deposits. In countries where pyrites are the major source of sulfur, however, the "metallurgical" type of plant is built. These plants can also use metallurgical gases and acid sludge as the source of sulfur. In metallurgical plants, sulfur dioxide is contaminated with water vapor, dust, sulfuric acid mist, and other impurities which must be removed by cooling the gases, condensing out moisture, filtering, and drying before sending the purified gas to converters. Although the initial cost and the operating cost are higher for such plants, the cost of the sulfuric acid product may be lower if sulfur dioxide is available as a by-product or waste product from some other operation or if the pyrites are available at low cost. THE CHAMBER PROCESS. The chamber process, which is the oldest method for manufacturing sulfuric acid, utilizes brimstone or pyrites as the raw material. This process gets its name from the fact that the reactions which produce sulfur trioxide, and then sulfuric acid, take place in lead chambers, in which oxides of nitrogen are the catalysts. The gases from the final chamber pass to a Gay-Lussac tower, countercurrent to cold sulfuric acid, which recovers the oxides of nitrogen before discharging the gases to the stack. The chamber process is becoming obsolete in the United States, partly because of the increase in demand for strong acid and oleum which cannot be supplied by this method, and partly because contact plants with high throughput capacity can be constructed at a lower cost, require less floor space, and produce a pure, concentrated acid. Thus, contact acid has become less expensive than chamber acid, even for use in the manufacture of fertilizer, which does not usually require the stronger acid. PHOSPHORIC ACID 47 Phosphoric Acid Phosphoric acid may be categorized either as wet-process acid or as furnace or thermal acid. The former is less expensive but quite impure; it is used for the manufacture of fertilizer and, after up- grading, for phosphate chemicals. Furnace acid is more expensive, but it is very pure and can be produced from certain low-grade phosphate rocks. The use of thermal acid has become much more limited as new processes for cleaning wet-process acid have been developed.7 A third category is superphosphoric acid, which is a concentrated form of the acid that can be made by the two alterna- tive processes. These processes are diagrammed in figure 6. WET-PROCESS PHOSPHORIC ACID. Phosphoric acid can be made by completely acidulating phosphate rock, using a strong acid such as hydrochloric or sulfuric acid.8 For economic and technical reasons, sulfuric acid is the one most commonly used. Since the purity of the sulfuric acid is unimportant, contact acid, chamber acid, and even spent acid from petroleum refining can be used, as long as no appreciable amount of organic sludge, chloride, or metallic im- purities is present.9 A fairly good grade rock is required, however, or else water-insoluble phosphates may be produced for fertilizers.'0 The acidulation process has used nitric and hydrochloric acid, but these acids are more expensive and there is the problem of sepa- 7. The article of commerce normally sold as wet-process acid contains 54 per- cent P205-that is, 75 percent H3P04. A common grade of furnace acid is 1.75 density-that is, 90 percent H3P04. Within reason, any concentration can be pro- duced by either process, because there are no constraints limiting the strength, as occurs, for example, with nitric or hydrochloric acid. But no dilute acids find their way into the market-mainly because it would be too costly to transport them. 8. Phosphoric acid is never made by acidulating phosphate rock with nitric acid-the separation problems are too great. 9. The purity of the acid is becoming important in some locations, mainly be- cause of the gypsum disposal problem-for example, acid from pyrites sometimes contains cadmium and mercury. 10. The question of whether difficult impurities are present in rock is often much more important than the question of grade. For example, a low-grade rock can be tolerated if the diluent is silica, but not if it is a component of iron or aluminum. 48 THE PRODUCTION OF PHOSPHATIC FERTILIZERS Figure 6. Alternative Processes for the Manufacture of Phosphoric Acids FURNACE PROCESS WET PROCESS Phosphate Phosphate rock rock and acid Furnace wet process process Elemental Phosphoric acid: phosphorus 30 percent P2,O Concentrator Two-step One-step process process Phosphoric acid: 54 percent P205 Vacuum or atmospheric Super- Phosphoric or evaporator phosphoric superphosphoric acid acid Superphosphoric acid PHOSPHORIC ACID 49 rating calcium salts from the acid product. In the case of hydro- chloric acid, however, the calcium chloride has been separated from the acid by means of a solvent that is subsequently separated from the acid and recycled to the process. Although this use of solvent provides phosphoric acid of higher purity than that produced by the standard wet process, this process is not significant for fertilizer manufacture. CONCENTRATION OF WET-PROCESS PHOSPHORIC ACID. Most wet- process acid is used to manufacture triple superphosphate, for which a strength of 54 percent P205 is normally required, and to manu- facture monoammonium phosphate (MAP), diammonium phosphate (DAP), and multinutrient fertilizers (NPK). The usual wet-process acid has a strength of only 30 to 32 percent P205 when it leaves the gypsum filter. Hence, the acid must be concentrated to the strength of up to 54 percent P205, depending on the use of the material. This evaporation used to be difficult and expensive because of the cor- rosive nature of the acid and the presence of other impurities, but it is no longer so. FURNACE PHOSPHORIC ACID. The furnace process grew rapidly in the past, particularly during the period from about 1945 to 1955. The growth has moderated over recent years, however, and virtually all new fertilizer phosphoric acid has come from the wet process. A world shortage of sulfur or the development of low-cost nuclear or hydroelectric power could result in greater use of the furnace method for fertilizer production. But large wet-process phosphoric acid plants today are so economic that the power would have to be very inexpensive indeed to justify using the furnace method for fertilizer production. SUPERPHOSPHORIC ACID. Superphosphoric acid can be prepared either from elemental phosphorus (the furnace method) or by con- centrating the wet-process acid (the wet method). In the former process, the acid is prepared merely by limiting the amount of water added to the phosphorus pentoxide. The resulting acid contains 76 percent P205, and the production cost per unit of P205 is essentially the same as that of the furnace acid already described. The wet method produces a similar acid of about 71 percent P205, equivalent to 98 percent phosphoric acid. The wet-process phos- 50 THE PRODUCTION OF PHOSPHATIC FERTILIZERS phoric acid (54 percent P205 content) is concentrated in vacuum and atmospheric evaporators, and an acid of improved quality is thus obtained."1 Superphosphoric acid has six advantages over phosphoric acid manufactured by either the furnace method or the wet method: the acid contains 40 percent more P205 than ordinary acid; large savings in freight and other transport costs are possible; it is less corrosive; the trend to higher analysis fertilizers has created a demand for stronger acids ;12 the presence of the polyphosphoric acids is an advantage for the production of these high-analysis fertilizers ;" and the acid can be used directly for liquid fertilizer applications. Ground Phosphate Rock Phosphate rock used for direct application must be ground to the proper particle size (200 to 300 mesh). The grinding can be done at the mining site or near the location of use.'4 The rock is stored until it is transferred to the millfeeder which controls the flow of the rock to the mill-either a roller mill or a ball mill. After the rock is ground to the desired size, it is carried out of the mill by an air stream to a centrifugal separator and then it is discharged to a storage silo. Some types of phosphate rock, however, are not suit- able for direct application. Rocks of igneous origin, for example, are virtually useless. The usable ones are those with a high specific surface containing 3 to 5 percent carbonate substituted in the rock structure. i 1. When making superphosphoric acid from either the wet process or elemental phosphorus, the acid produced can have any strength in a range. It is easiest to make the highest concentrations from elemental phosphorus because, in the case of wet acid, impurities increase viscosity and prevent the highest concentrations being achieved. 12. High-strength acids are necessary to minimize the cost of transportation, evaporation, and the like. They do not lead to fertilizers of significantly higher analysis. 13. Polyphosphoric acids are advantageous mainly because of this sequestering property, which makes it easier to obtain liquid fertilizers and suspension fertilizers of higher strength than can be achieved with orthophosphoric acid. 14. If the grinding is done at the mining site, the problem of dust in transit is aggravated. SINGLE OR NORMAL SUPERPHOSPHATE 51 Single or Normal Superphosphate The term "superphosphate" applies to the mixture of finely ground phosphate rock and sulfuric acid, phosphoric acid, or the mixture of the two. The product obtained using sulfuric acid is called ordi- nary or single superphosphate (ssP) and contains between 16 and 21 percent P205. The phosphoric acid product is referred to as double, treble, multiple, concentrated, or triple superphosphate and contains between 43 and 48 percent P205. Both single and triple superphos- phates are used as straight fertilizers; large proportions, however, are further processed into multinutrient fertilizers. Single superphosphate is a product obtained by mixing finely ground phosphate rock with sulfuric acid. It has a composition of 30 percent monocalcium phosphate, 10 percent dicalcium phos- phate, and 45 percent calcium sulfate, with the remainder being iron and aluminum phosphates plus some moisture."5 The manufacture of superphosphate involves four mechanical or physical operations: . Preparation or grinding of the phosphate rock for acid treat- ment. When the rock is finely divided, quick and complete reaction with sulfuric acid can occur. . Mixing of the finely ground phosphate with sulfuric acid. A thorough mixing of the rock and the acid is necessary to obtain a good reaction and the maximum yield of available P205. Unless this operation is efficiently performed, a larger quantity of sulfuric acid may be used than is actually required. As a result, the product will have a high content of free acid, will be in poor mechanical condition, and will be difficult to handle. . Curing and drying of the acidulated material. This is performed by the closed-chamber, or den, method which promotes a good reaction between the rock dust and the acid, resulting in the production of a dry, porous type of superphosphate.16 . Excavation, milling, and bagging of the final product. 15. Superphosphate manufacturers that produce to a water-soluble specification could not tolerate 10 percent dicalcium phosphate in the product. 16. Some types of dens can be used for both single superphosphate (ssp) and triple superphosphate (Tsp) and even, with some modification, for monoammo- nium phosphate (MAP). 52 THE PRODUCTION OF PHOSPHATIC FERTILIZERS These various mechanical processes can be conducted as a num- ber of separate steps, or they can be carried out continuously in complete, self-contained units. Also, several different makes and designs of mixers are available. Regardless of the mixing process employed, the manufacture of this product is a relatively simple operation. Triple Superphosphate Triple superphosphate (TSP) iS obtained by acidulating finely ground phosphate rock with phosphoric acid. It consists almost entirely of monocalcium phosphate and thus differs from single superphosphate, which contains a large proportion of calcium sulfate. Although TSP may be made with any phosphoric acid, it is most commonly made from wet-process acid. The P203 content of the product may range from 43 to 48 percent, depending on the purity of the acid and the rock and on the efficiency of the manu- facturing process.17 Triple superphosphate is manufactured by three principal processes, which are diagrammed in figure 7. "RUN-OF-PILE" TRIPLE SUPERPHOSPHATE. The "run-of-pile" (ROP) process mixes the phosphoric acid (50 to 54 percent P205) and phos- phate rock in a cone mixer.18 The slurry produced is discharged either into a den or onto a conveyer belt, where it is hardened.'9 It is then disintegrated mechanically and stored on the curing pile. Although this product is normally used for ammoniated fertilizers, it can be granulated after the curing period is over. Normally, the capital investment and labor and maintenance costs for ROP triple superphosphate are lower than those for the granular variety phos- phate, but an extensive gas scrubbing system is frequently required for ROP, and then the cost differential might be narrowed. 17. The efficiency of the process is not a significant factor in determining the P205 content of TSP. The efficiency is important, however, to ensure good conver- sion to water-soluble or citrate-soluble P,05. 18. Run-of-pile triple superphosphate is a mixture of all sizes of phosphate rock, from small lumps to dust. 19. Most TSP can be made in the same den that is used for ssp. TRIPLE SUPERPHOSPHATE 53 Figure 7. Alternative Processes for the Manufacture of Triple Superphosphate Process #1 Run-of-pile TSP Phosphoric acid and phosphate rock >- Process #2 Granular Single-step Process #3 granular TSP GRANULAR TRIPLE SUPERPHOSPHATE. The granular process produces a granulated product in a multistep process. It is well-suited for direct application as a phosphate fertilizer or for inclusion in mixed bulk blends made by dry-blending solid raw materials. The process employs 45 to 48 percent phosphoric acid and phosphate rock, which are reacted in a series of reaction tanks. Quick drying is facilitated by blending the slurry into a bed or by recycling crushed, oversized particles and product. The partially dried granules are then fed to a rotary drier. A multilayered, dense product with excellent handling qualities is thus built up. SINGLE-STEP GRANULAR TRIPLE SUPERPHOSPHATE. The single-step granular method combines the features of quick drying and granula- tion in a single step. Acidulation and granulation are accomplished simultaneously in a rotary-drum mixer. Wet-process phosphoric acid and steam are fed through distributors under a rolling bed of material in the drum. The product has a short curing time, can be readily ammoniated, and has good storage properties. The process has the advantage of low equipment and processing costs, and it can be adapted to use existing continuous ammoniators. ENRICHED SUPERPHOSPHATE. The manufacture of enriched super- phosphate is similar to that of TSP, but a mixture of sulfuric and phosphoric acid is used for acidulation. The acid mixture is chosen 54 THE PRODUCTION OF PHOSPHATIC FERTILIZERS so that the product will have an adequate amount of sulfur which is not as diluted as in the case of single superphosphate. Dicalcium Phosphate Dicalcium phosphate has valuable properties as a fertilizer,20 as an ingredient in animal feedstocks, and as a source of calcium and phosphorus in human foods and restoratives. It is normally mar- keted as the dihydrate, containing 41.26 percent P205 when pure. In this form, it is citrate-soluble, but if dehydrated, it can be citrate- insoluble. When dicalcium phosphate with a high standard of purity is required, it is usually made from thermal acid and limestone or lime. Less pure grades can be made from bones or from phosphate rock. Normally, bones are not used to make this fertilizer, but they are frequently used to produce feedstock supplements. PREPARATION FROM BONES. When dicalcium phosphate is prepared from bones, the grease in the bones is usually extracted by a solvent, and they are then leached with a weak solution of hydrochloric acid to separate the tricalcium phosphate component from the organic material. The tricalcium phosphate passes into solution as mono- calcium phosphate and calcium chloride. In another process for manufacturing dicalcium phosphate from bones, a sulfur dioxide solution is used to extract the calcium and phosphorus as monocalcium phosphate and calcium sulfate. Subse- quent treatment with steam, followed by hydrochloric acid, leaves a precipitate of dicalcium phosphate in a solution of calcium chloride. The dicalcium salt is filtered, washed, dried, and packed. PREPARATION FROM MINERAL PHOSPHATES. Dicalcium phosphate is also prepared on a large scale from mineral phosphates by the action of hydrochloric acid on phosphate rock. One process utilizes hydro- chloric acid available from the production of potassium sulfate. The phosphate rock is mixed with the acid in agitators. With many African rocks, grinding is unnecessary, because relatively coarse rock dissolves quite easily and the reaction can be controlled to get com- 20. To date, however, it has not found much use as an agricultural fertilizer. THERMAL PHOSPHATE FERTILIZERS: BASIC SLAG 55 plete dissolution of the P205 in the rock. The clear liquor overflows from the agitators to a first stage of dicalcium phosphate precipita- tion, where lime slurry is added. The solution overflowing is treated with more lime slurry, which precipitates the remaining P205, pri- marily as tricalcium phosphate. The precipitate is dewatered and returned to the dicalcium phosphate precipitation step, where the tricalcium is converted to dicalcium phosphate. If total recovery of the P205 is attempted in a single liming stage, the final product will contain tricalcium phosphate that is low in both grade and citrate solubility. The dicalcium phosphate dihydrate is filtered and care- fully dried to avoid losing water of hydration, which in turn would cause loss of citrate solubility. Small amounts of some type of binder can be added during the drying stage to reduce the tendency of the material to dust. Since this material is insoluble in water, coarse granules are undesirable for use as a fertilizer. Thermal Phosphate Fertilizers: Basic Slag One of the oldest thermal phosphate fertilizers is basic slag, a by-product of iron and steel manufacture. Basic slags differ con- siderably in quality, depending on the process and raw materials used. A converter slag might contain 16 to 20 percent P205, whereas an open-hearth slag might contain 10 to 12 percent P205. The liberal use of fluorspar may reduce the P205 content to 2 to 3 percent, unless tapping is undertaken before the spar is added. The increased use of scrap metal can also reduce the yield and value of slag. Al- though the reactions between impurities in molten iron and basic furnace linings are quite complex, iron phosphide is removed from the metal, with the assistance of oxygen, lime, and the basic lining of the converter or furnace. Since many steelmakers believe that the production of a superior slag is of secondary importance, developments in steel technology have not resulted in improvement of the quality of basic slag. In particular, the use of fluorspar as a flux is of great benefit to the steel industry, but it gives an inferior basic slag, since fluorapatites are formed, and these compounds are insoluble when applied to the soil. Attempts to raise the P205 content of slags, by adding ground phosphate rock either to the furnace charge or to the slag, have not been widely made except in Europe. Such attempts have not been 56 THE PRODUCTION OF PHOSPHATIC FERTILIZERS made in the United States because of the fear of handicapping steel production. The Duplex process, however, is one method of com- promising between high steel production and good quality slag yields. In this operation, iron that is high in phosphorus is first treated in an acid-line converter and blower. The charge is then treated in a basic open-hearth furnace to yield a slag high in basic phosphates, which is cooled or chilled, cracked, and ground to a fineness of 95 percent or better. Other thermal phosphate fertilizers include Rhenania phosphate, Rochling phosphate, calcium mag- nesium phosphate, and calcium metaphosphate. Since these are not important products of the fertilizer industry, they are not discussed in detail. 5 Production of Potassic Fertilizers POTASH, OR POTASSIUM, which is one of the major nutrients, is widely dispersed in nature, occurring in highly soluble salts such as potas- sium chloride, in insoluble minerals such as potassium-bearing silicates, and in marine and land plants. The principal minerals cur- rently used as fertilizers are potassium chloride (muriate of potash), potassium sulfate, potassium nitrate, and a mixture of potassium and magnesium sulfate. The potassium content of the minerals is stated in terms of potassium oxide (K20). Production and Processing of Potash Ores Virtually all commercial potash is produced from potash-bearing brines or from underground deposits of soluble minerals. The brines are available from surface deposits, and their compositions are widely known. The underground deposits, which are often difficult to exploit, are vastly more important because of their high content 57 58 THE PRODUCTION OF POTASSIC FERTILIZERS of potash and generally more favorable locations. Subterranean deposits of potash occurring as sylvinite (a mixture of potassium chloride and sodium chloride) or carnallite (a mixture of potassium chloride and magnesium chloride) are usually exploited as under- ground mines. There are also a few operations that mine solutions. Mineral langbeinite, a double salt of potassium magnesium sulfate, occurs in underground beds mixed with halite and sylvite. Langbeinite and the chloride salts are both soluble in water, but the chlorides dissolve at a much faster rate, and this is the basis of the separation. After the minerals are crushed, they are washed at a rate sufficient to remove most of the chloride salts, while the langbeinite remains largely unchanged. The langbeinite is then separated from the wash waters by centrifuging, and it is kiln-dried and conveyed to finished product storage. Potassium Chloride Pure potassium chloride has an equivalent K20 content of 63.1 percent; sylvinite usually contains 15 to 30 percent K20. The muri- ate of potash used for fertilizer manufacture is not pure potassium chloride, and its K20 content is between 60 and 62 percent (95 to 98 percent KCI). Fertilizer and agricultural grades of muriate of potash are sold in various particle sizes and concentrations, usually containing 60 to 62 percent K20. Crude salts containing 20 to 25 percent K20 are also available, and they find a market for direct application to agricultural land. Two principal processes are used to produce potassium chloride: solar evaporation and mining methods. SOLAR EVAPORATION OF BRINES. Solar evaporation is used mainly in Israel and the United States. Basically, in this process brine is evaporated in solar pans. Sodium sulfate and chloride are the first salts to crystallize. At this stage, the solution contains magnesium and potassium with some sodium chloride. The concentrated brine is then moved into harvesting pans where carnallite-a double salt of magnesium and potassium chloride-crystallizes. After the carnallite has crystallized, it is sent to a plant to be separated by flotation. MINING METHODS. Two types of mining methods are employed: POTASSIUM CHLORIDE 59 shaft mining and solution mining.' There are several forms of shaft mining. The room-and-pillar mining method is generally used to recover sylvinite and carnallite; it is also used for langbeinite. Coal mining methods are used for underground recovery; continuous mining machines are used to extract the ore, and belt conveyers are employed for underground haulage. Solution mining is utilized at depths considered less practical to mine by conventional shaft mining. Essentially, water is pumped through drilled holes to the potash deposits. The water dissolves the potash and sodium chloride, and the solution is forced to the surface. The difficulty with this method is to obtain a relatively pure solution of potash. A solution to this problem that has been used in Canada is to inject a blanket of natural gas over the potash deposit to pre- vent the dissolution of the sodium chloride. Refining methods for potassium chloride Two methods are used to refine the potassium chloride: one is solution and recrystallization; the other is flotation. These methods are diagrammed in figure 8. SOLUTION AND RECRYSTALLIZATION. Solution and recrystallization is an old, refined process that, with some modifications, can produce relatively pure potassium chloride from a wide variety of ores. Among its disadvantages are that it demands high-temperature processing of corrosive brines, expensive equipment, and large fuel and cooling requirements. It was favored in Europe, but it is being superseded by flotation separation methods. The process is based on the different solubilities of sodium and potassium chloride in hot and cold saturated brines. Sodium chlo- ride is an unusual salt because its solubility increases only slightly with temperature; with most salts, the solubility increases much more as the temperature increases, and this is the case with potas- sium chloride. The procedure is to take a brine that is saturated with sodium and potassium chloride at a high temperature and to cool it so that potassium chloride crystallizes and can be separated as a solid. The brine mother liquor is reheated and used to dissolve more 1. Shaft mining usually has high capital costs and low operating costs; the re- verse is true for solution mining. 60 THE PRODUCTION OF POTASSIC FERTILIZERS Figure 8. Alternative Methods for Refining Potassium Chloride Impure recrystallization Refined potassium | potassium chloride Flotation chloride recovery crushed sylvinite. The last brine is removed along with small quanti- ties of insoluble clays and similar impurities. The solid is usually discarded or sold for ice control on roads and similar applications. The hot solution is cooled once more, and the cycle of operations continues. Crude carnallite ore containing sylvite is leached in a vat with a hot saturated brine solution. Potassium chloride and the carnallite dissolve, leaving common salt and insoluble matter in the residue. The hot liquor is cooled in vacuum crystallizers that yield a low- analysis potassium chloride product. The mother liquor is evapo- rated and cooled to form carnallite. Potassium values are recovered from the crystallized carnallite by a hot leach followed by vacuum crystallization. The potash deposited in this step is of high purity. FLOTATION RECOVERY. A widely practiced method is to separate potash from its ore by froth flotation.2 In this separation process, a selectively reagentized solid is suspended in an aerated aqueous liquid. The potassium chloride adheres to the bubbles of the air and is removed from the suspension as a froth. The froth is mechanically removed, and the unreagentized sodium chloride is removed as a pulp. The floated froth is concentrated in a thickener, centrifuged or filtered, washed with water to displace the brine, and then dried. The product generally averages 97 percent purity, and the recovery efficiency of this process is greater than 96 percent. 2. In the United States, the term "mineral flotation" is widely used rather than "froth flotation." POTASSIUM SULFATE 61 Potassium Sulfate Potassium sulfate, also known as arcanite, is a white crystalline compound. The dissolved solids in water form a neutral solution. The potash content is 54.06 percent K20, and the compound con- tains 55.13 percent sulfate. Potassium sulfate is the least hygroscopic of the common fertilizer materials. It is often associated with other salts in potash deposits. When it is present with calcium and mag- nesium sulfate, the stable salt is called polyhalite, which has a 15.5 percent K20 content. When calcium is not present, the double salt of potassium and magnesium sulfate is called langbeinite, which has a K20 content of 22.6 percent. With the exception of -specialized crops, potash fertilizers are equally effective. Tobacco is a common exception; it is sensitive to chlorides and is often fertilized by potas- sium sulfate.3 The latter is also considered to be a superior fertilizer for potatoes and citrus fruits in certain soils. Three methods can be used to produce potassium sulfate, and they are diagrammed in figure 9. 3. In fact, any chloride-free fertilizer can be used for chloride-sensitive crops, and the nitrate is widely used in the United States and Denmark. Figure 9. Alternative Processes for the Production of Potassium Sulfate Sulfuric acid and potassium chloride Mannheim Do furnace process Sulfur and air Hargreaves Potassium process F_ sulfate Potassium chloride and langbeinite lon exchange process 62 THE PRODUCTION OF POTASSIC FERTILIZERS THE MANNHEIM FURNACE PROCESS. Sulfuric acid is made to react with potassium chloride in the Mannheim furnace process. First, potassium bisulfate is formed and then potassium sulfate. Hydro- chloric acid is evolved as the by-product. THE HARGREAVES PROCESS. The Hargreaves process depends upon the burning of sulfur in air to produce sulfur dioxide. Water vapor is added and the composite gas is then passed countercurrent through a series of chambers. Each chamber contains approximately 80 tons of potassium chloride, which are converted by the sulfur dioxide to approximately 90 tons of potassium sulfate. PRODUCTION BY ION EXCHANGE. A third process produces potassium sulfate from naturally occurring salts by an exchange of ions. It requires the slurrying of potassium chloride and ground langbeinite and water. The potassium sulfate liquor is saturated with potassium chloride and leonite; after allowing time for the solution to ap- proach equilibrium, potassium sulfate is separated from the slurry, dried, and stored. 6 The Production of Multinutrient Fertilizers MULTINUTRIENT FERTILIZERS-that is, those containing more than one of the primary nutrients-comprise a wide range of chemical compositions and nutrient ratios, and they can be produced by many different processes. In their final form, however, they can usually be categorized either as compound fertilizers or as dry-mixed or bulk-blended fertilizers. Compound fertilizers are prepared by granulating together solid or liquid fertilizer ingredients or intermediates.' In some cases, the intermediates are prepared at their sources immediately before granulation. The composition of each granule is the same, and a wide range of ingredients can be used and fertilizers produced. Dry-mixed or bulk-blended fertilizers are obtained by mechanical mixing of fertilizer materials already in the granulated or prilled This chapter has benefited from substantial redrafting by William T. Sheldrick, Head, Fertilizer Unit, World Bank. His help is gratefully acknowledged. 1. Compound fertilizers are also known as complex or mixed fertilizers. 63 64 THE PRODUCTION OF MULTINUTRIENT FERTILIZERS form. Individual granules do not necessarily have the same composi- tion, and the technique requires careful matching of sizes to avoid segregation and ensure an acceptably uniform product. Granulation Processes In order to produce fertilizers that have good storage and distribu- tion properties, it is now customary to produce compound fertilizers or blend components in a granular form. In this context, the term "granulation" is used to describe processes that produce large ferti- lizer particles within the range of I to 4 millimeters: for example, prilling, agglomerating, solidification from the melt, compacting, and flaking. SLURRY GRANULATION. In the slurry granulation process, most of the feed enters as a slurry or liquid, and granulation is achieved by operating a large recycle and building up successive layers of solid on each recycled granule. This process is most commonly used for compounds based on ammonium phosphate and for nitrophos- phates. Granulation is achieved in a pug mill or paddle mixer oper- ating with high recycle ratios. AGGLOMERATION GRANULATION. In the agglomeration granulation process, the feeds are normally in powder form, although liquid feeds may be used. Granulation is achieved by adding steam or water to cause the particles to adhere. The granulation recycle ratios are small-normally of the order 1:1-and granulation is carried out in a rotary drum or pan granulator. MELT GRANULATION. Molten mixtures of salts are granulated by controlled solidification in a rotating drum or pan granulator. This process is usually applied to ammonium-phosphate-nitrate mixtures. PRILLING. Melts of ammonium nitrate phosphate or other pure salt melts can be solidified by spraying the melt down a void or dust-filled tower. Prilling can also be achieved by spraying the melt into oil followed by centrifugal separation. COMPACTING. Compacting is used on a small scale to prepare multinutrient fertilizers, although its major use is to upgrade potash MULTINUTRIENT FERTILIZER MATERIALS 65 to a size in which it can be bulk-blended. Powder material is forced between rolls under pressure to produce a sheet of compacted ma- terial. This material is disintegrated and screened to the desired size range. Multinutrient Fertilizer Materials Although a large number of compound fertilizers can be produced, their formulation generally tends to be based on two or more chem- ically compatible materials. For example, potassium chloride and, to a lesser extent, potassium sulfate are used as a source of potash; superphosphate or ammonium phosphates, as a source of phos- phate. Urea, ammonium nitrate, and ammonium sulfate are the principal sources of nitrogen. Ammonium phosphate Ammonium phosphate-which can exist as monoammonium phosphate (MAP), diammonium phosphate (DAP), or a combination of the two-is one of the most concentrated fertilizers. Because of this, it has become an important fertilizer intermediate in recent years, and it is used both as a bulk-blending component and as an ingredient for compound fertilizers. Monoammonium phosphate contains 11 percent nitrogen and 48 to 54 percent P205, depending on the purity of the ingredients. It is made by granulating wet-process phosphoric acid. The degree of ammoniation can be controlled in the process to yield MAP or a mixture of MAP and DAP. Diammonium phosphate usually contains 16 to 18 percent nitrogen and 46 to 48 percent P205. RAW MATERIALS. Ammonium phosphates are now normally pro- duced from wet-process phosphoric acid. There is a growing trend for both phosphoric acid and ammonium phosphates to be produced in large plants near the phosphate rock sources. Ammonia is usually provided to the site as anhydrous liquid ammonia, although either liquid or gaseous ammonia can be used for neutralization. In a few cases, phosphoric acid is used to neutralize ammonia present in the gases from coke ovens. PROCESSES FOR THE MANUFACTURE OF AMMONIUM PHOSPHATES. Most 66 THE PRODUCTION OF MULTINUTRIENT FERTILIZERS ammonium phosphate is made from wet-process phosphoric acid using slurry granulation to produce a granular product or by one of the newer processes developed in the last few years to produce am- monium phosphate powders. Some ammonium phosphates, mainly from furnace-grade acid, have been produced using crystallization techniques, but this process is now rarely used. The slurry granulation process was developed by Dorr Oliver more than forty years ago and, although improvements and changes have been made, the basic process remains the same. The slurry granula- tion plant can also be used to make granular triple superphosphate or ammonium phosphate nitrates. In the ammonium phosphate process, wet-process phosphoric acid of between 38 and 50 percent P205 is fed to a reaction system and neutralized with ammonia. Ammonia leaving the reactors is scrubbed with incoming phos- phoric acid. In some variations of the process, the second-stage ammoniation from MAP to DAP can be carried out in the granulator. Normally, a high recycle ratio is employed in this process, and after granulation the product is dried and screened. In recent years, several processes have been developed to produce ammonium phosphate powder as a fertilizer intermediate. These processes usually have low investment cost and are inexpensive to operate; their main object is to provide a simple phosphate inter- mediate for transport as an alternative to phosphoric acid. The MAP powder is used as an ingredient in compound fertilizers. The first process of this type was the Scottish Agricultural In- dustries' (SAI) "Phos Al" process; another is Fison's "Miniphos" process. Other not-so-well-known processes have been developed by Swift, Norsk Hydro, and Gardinier. About forty plants of this type have been built in the last ten years. One major variant of this process is to ammoniate phosphoric acid under pressure in either a pipe or a tank reactor, followed by a simple spray-drying process. The product is normally of a coarse powder or minigranular form with good properties for agglomeration granulation. SOME ECONOMIC CONSIDERATIONS FOR AMMONIUM PHOSPHATE. The main advantage of ammonium phosphate is that it can be produced simply as a high-concentration fertilizer which has good storage properties and can be shipped in bulk either as a fertilizer inter- mediate or as a finished product. Increasingly, the trend is to build ammonium phosphate plants near the phosphate rock mines so that MULTINUTRIENT FERTILIZER MATERIALS 67 low-grade rocks can be utilized. Most such developments are occur- ring in North Africa and the United States. Ammonium phosphate is a valuable ingredient for mixed ferti- lizers because of its compatibility-superphosphates, for example, do not have this property to the same extent because they contain calcium. Three useful products of ammonium phosphate mixed with other salts are given below. Ammonium phosphate sulfate is formed by ammoniating mixtures of phosphoric acid and sulfuric acid to form products such as 13- 39-0 and 16-20-0. Potash can be added as required, and this process is popular where sulfuric acid is inexpensively available as a by- product from smelting operations. Ammonium phosphate nitrate is produced by neutralizing mix- tures of phosphoric acid and nitric acid-or, alternatively, mixtures of ammonium nitrate and phosphoric acid-with ammonia. Under certain pH conditions, ammonium phosphate nitrate forms a eutec- tic mixture that can be melt granulated, and a number of processes have been developed to achieve this. Typical grades that can be achieved are: 30-10-0, 27-14-0, and 25-25-0. Adding potash yields nitrogen-phosphorus-potassium fertilizers such as 17-17-17. Ammonium phosphate nitrate fertilizers are popular in Europe and the United States, where nitrate-based compounds are preferred to urea-based compounds for agronomic reasons. Finally, concentrated fertilizers can be produced using urea and ammonium phosphate, and several plants have been built recently to make these compounds. In view of the rapid growth of urea and ammonium phosphates as fertilizer intermediates, the manufacture of ammonium phosphate urea (APU) fertilizers is expected to in- crease. These materials are particularly suitable for paddy and, since they are the most concentrated compounds possible from con- ventional materials, they offer savings in transport and storage. Unlike ammonia-nitrate-based compounds, they offer no hazard from fire or decomposition. Ammoniated superphosphate Ammoniated single and triple superphosphate have been pro- duced for many years in the United States and elsewhere. Ammo- niation improves the storage properties of fertilizers containing superphosphates, but it also reduces their water solubility. Nor- 68 THE PRODUCTION OF MULTINUTRIENT FERTILIZERS mally, about 3 or 4 units of nitrogen are added to 20 units of P2O5 in this process. Nitrophosphates Nitrophosphates are produced by acidulating phosphate rock with nitric acid. The main advantage of this process is that it elimi- nates the use of sulfur. Nitrophosphates were developed mostly in Europe during periods of sulfur shortages, but the availability of inexpensive sulfur and sulfuric acid within Europe and certain in- herent disadvantages have tended to reduce the general impact that these processes once had. The major disadvantage of nitrophosphate processes is that cal- cium nitrate, which is a product of the acidulation stage, is extremely hygroscopic and must be treated in some way to produce a fertilizer than can be stored. Although many processes have been developed to do this, they fall into four main types: * Separation method. In variations of the Odda process, excess calcium nitrate is crystallized from the acidulation mass by cooling and is separated by centrifuging. Various degrees of concentration and water solubility of product can be achieved, depending on the quantity of calcium nitrate removed. The calcium nitrate so removed can be sold separately as a fertilizer, or it can be reacted with carbon dioxide to produce ammonium nitrate, a more concentrated fertilizer. * Sulfonitric and phosphonitric processes. These are mixed-acid processes in which the ratio of calcium to phosphate is adjusted by attacking the phosphate rock with a mixture of either nitric and sulfuric acids or nitric and phosphoric acids. * Nitrocarbonic process. Free calcium nitrate is reacted with carbon dioxide after ammonia is added to convert the mixture into calcium carbonate. Products of this process are of low analysis and are water insoluble. * Sulfate recycle process. In the sulfate precipitation process, either an outside source of ammonium sulfate is used to pre- cipitate the calcium as calcium sulfate (by-product ammonium sulfate from caprolactam production is sometimes used for this purpose) or, alternatively, the ammonium sulfate can be regen- erated from the calcium sulfate, in which case it is referred to as the sulfate recycle process. The excess calcium is eliminated as calcium carbonate. MULTINUTRIENT FERTILIZER MATERIALS 69 These processes can be used in various combinations. After the acidulation and separation stages, the ammonia and potash are added to produce fertilizers of the required analysis. Nitrophosphate processes have two further disadvantages. First, they are not flexible. It is not easy to obtain a wide range of products in these processes and N: P205 ratios of less than about 1.5:1 can only be achieved by using supplementary phosphoric acid or by removing the calcium nitrate as a concomitant product. Second, most nitrophosphate processes produce fertilizers with phosphates that are water insoluble or only partly water soluble; these are regarded as less desirable than completely water soluble phosphates. Several new nitrophosphate processes can produce high degrees of water solubility either by additional separation of calcium nitrate or by use of the sulfate recycle process, but these processes are complex and have to be operated on a large scale to be economic. SOME ECONOMIC CONSIDERATIONS FOR NITROPHOSPHATES. The con- troversy over the relative merits of nitrophosphates and ammonium phosphates has gone on for many years. The claims for nitrophos- phate processes were probably justified when sulfur and phosphoric acid were expensive raw materials and when fertilizer concentrations were low. General international trends in the fertilizer industry, however, have made the arguments academic. First, the need for high concentrations has, to a large extent, elimi- nated the low-grade nitrophosphate process as well as the arguments about the water insolubility of its products. Concentrated fertilizers now have most of their phosphate in the form of ammonium phos- phate. Second, the availability of relatively inexpensive phosphoric acid and ammonium phosphates and the differential pricing of phosphate rock and the intermediates makes it less favorable now than it was formerly for producers to import rock and process it with either sulfuric or nitric acid. Third, in many areas urea has become the principal nitrogenous fertilizer for both technical and agronomic reasons, and it seems likely that this trend will continue. The worldwide availability of urea and ammonium phosphates-both of which are produced in large plants at the most economic locations and conditions-provides fertilizer intermediates that can be compounded to give a wide range of fertilizer analyses. Recent evaluations by the Tennessee Valley Authority (TVA) indicate that in most countries urea and diam- 70 THE PRODUCTION OF MULTINUTRIENT FERTILIZERS monium phosphate would be a less expensive form of fertilizer than nitrophosphates.2 Moreover, under certain conditions, nitrophosphates undergo de- composition reaction and this necessitates more careful storage and handling than would be required for compounds based on ammo- nium phosphate urea. As a result of these considerations, it seems unlikely that the use of nitrophosphate processes will grow significantly. An exception might be in Eastern Europe and the Soviet Union, where state con- trol and operation of fertilizer plants is more favorable to large, single-product plants than elsewhere. Potassium nitrate Potassium nitrate (KNO3) can be used as a component of solid or solution fertilizers when nitrate nitrogen and potassium without the presence of chloride ions are required. The application of potassium nitrate has been limited, however, because of its relatively high cost in terms of plant-food units and its high-temperature solubility coefficient in water. The main nonfertilizer uses include explosives, fireworks, and chemical manufacture. It is available in various grades; the technical quality is 99 percent pure, and the fertilizer quality is 95 percent pure. The latter contains 13 percent N and 44 percent K20. Alternative names for the salt are niter and saltpeter. The chemical means used to produce potassium nitrate are similar to those used to manufacture sodium nitrate, described in chapter 3. One recently developed process is based on the reaction of potash with nitric acid to yield potassium nitrate and chlorine. In another process, solid potash is added to a hot aqueous solution of sodium nitrate; the sodium chloride formed crystallizes in the hot mother liquor and is separated by filtration or centrifuging. The potassium nitrate liquor is concentrated and cooled to yield the crystalline product.3 A method that reacts solid potassium chloride and 60 to 65 per- cent nitric acid, without the use of heat, has been developed in Israel. It is based on the use of an organic solvent that separates the solid 2. T. P. Hignett, Technical and Economic Comparison of Nitric and Sulphuric Acid Routes to Phosphate Fertilizers (International Fertilizer Development Center). 3. The process is not generally used for agricultural fertilizers. DRY-MIXED FERTILIZERS OR BULK BLENDS 71 and liquid phases of the reaction, yielding solid potassium nitrate. The hydrochloric acid formed can be used for other purposes such as phosphoric acid manufacture or dicalcium phosphate production. Granular Compound Fertilizers The process of producing multinutrient fertilizers was developed as a convenient method to prepare solid materials from a solution or slurry and also because granulation allows fertilizer materials to be handled and stored more easily than mixed powders. Granulation of different nutrients into granules of the same composition also ensures that the growing plant receives the recommended nutrient ratio and that the effects of fertilizer segregation are minimized. The various processes that can be used to produce compound fertilizers are outlined above. Dry-Mixed Fertilizers or Bulk Blends Bulk-blending provides the best results if two or more granular materials having similar particle size are blended. The resulting blend, then, has little tendency toward segregation. The fertilizers used most often for bulk-blending are ammonium nitrate, ammonium sulfate, urea, triple superphosphate, diammonium phosphate, and potassium chloride. If mutually compatible fertilizers (in terms of particle size and moisture content) are used, bulk-blending is a simple and inexpensive operation; it can produce a wide range of mixtures. In the United States, bulk-blending is largely integrated into the fertilizer distribution sector, and it is used to supply fertilizer mixes that correspond exactly to the user's nutrient needs. Since the mixing plant itself is usually inexpensive, does not ex- hibit significant economies of scale, and does not transform the intermediate (fertilizer) products into a new product, bulk-blending is not usually specified as a separate activity in the planning models described in this volume. The location of blending plants can appro- priately be dealt with once the primary producing plants have been analyzed. 7 Problems in Fertilizer Sector Planning THE GREAT VARIETY of raw materials, intermediate and final products, and productive processes that characterize the chemical fertilizer industry may pose a difficult planning problem for those selecting the appropriate products and processes for a sectoral investment program. This problem is aggravated by the wide range of possibilities for product substitution on the demand side; that is, specific nutrients can usually be provided by many different fertilizer products. Other factors also complicate investment planning in the sector. Since most fertilizer products can be traded internationally, the choice between producing domestically and importing from abroad needs to be investigated.' Further, since most products associated with fertilizer manufacture are produced under economies of scale (that is, production cost per unit tends to fall as output level in- 1. Some products cannot be traded internationally today. The two main prod- ucts are single superphosphate, which is too dilute, and fertilizers that are at a grade below international standards but are perfectly good for home use: for ex- ample, low-grade triple superphosphate made from a low-grade rock without much beneficiation. 72 PROBLEMS IN FERTILIZER SECTOR PLANNING 73 creases), the question of the optimal scale and timing of new capacity must always be addressed during the planning phase. Because the demand for fertilizer material is by nature geographi- cally dispersed, the question of the optimal location of producing facilities, blending plants, granulation facilities, warehouses, and distribution points can have important implications for the supply price of the final product at the farm gate. Similarly, since alternative means of transportation (rail, highway, water) are often possible, the selection of the least-cost means and pattern of transportation should be based on careful analysis. Thus, decisions may have to be made on major strategy options during the early project planning phase. Which products should be produced? Which are the appropriate feedstocks? Should specific raw materials and products be imported from abroad or produced domestically? Which is the most efficient investment, production, and trade pattern over time for the products to be produced do- mestically? The analysis of these broad issues requires assumptions and judg- ments on many aspects of the planning problem. During later phases in the project planning cycle, several of these assumptions and judg- ments may be found not to stand the test of more detailed scrutiny. This is an inevitable characteristic of planning in phases, and it occurs frequently in investment project planning. One of the main advan- tages of the methodology described in this volume for planning investments in the fertilizer sector is that the initial broad strategy can be reformulated conveniently so that an intensive and meaning- ful interaction between different phases in the planning process is possible. This chapter provides a systematic discussion of the major decision- making problems faced in fertilizer sector planning. Broadly speak- ing, these can be placed in two categories. The first category is de- cisions related to specifying the planning problem. Which feedstocks and end products should be included in the investigation? How many and which potential production sites should be specified? How many and which marketing centers should be selected to represent adequately the geographical dispersion of demand for fertilizer material? Which transport alternatives should be considered? The second category of decisions relates to formulating an efficient investment program once the scope of the planning problem has been specified. The planning models described in this volume are specifically designed to address the second set of decisionmaking 74 PROBLEMS IN FERTILIZER SECTOR PLANNING problems. But, because this second set of problems is an efficient screening device, they permit the basic planning problem to be specified more comprehensively than would be feasible with more conventional methods of project selection. Hence, many more prod- ucts, production processes, transport alternatives, and the like, can be specified and assessed on their relative merits. Specifying the Planning Problem This section treats various factors involved in specifying the plan- ning problem: feedstocks, fertilizer products, production processes, production sites, by-products, and means of transport. In addition, the questions of regional breakdown and number of time periods in the model are discussed. Feedstocks Many products in the fertilizer industry can be produced from several feedstocks. The choice may be among basically different raw materials or products (for example, coal, natural gas, or naphtha for ammonia manufacture) or among different grades of a given product (for example, grades of phosphate rock with different phos- phorus and impurity contents).2 In most cases, common sense sug- gests which feedstocks from among the likely candidates should be specified. Clearly, if abundant natural gas is available domestically, whereas coal is not, the latter need not be specified as a possible feedstock. Similarly, if high-grade phosphate rock is available locally, one does not need to consider imported low-grade rock. The situa- tion is not always so clear-cut, however. For instance, even though pyrites are locally available for the production of sulfuric acid, it may be desirable to specify imported elemental sulfur as an alterna- tive feedstock. Fertilizer products Since the fertilizer industry is characterized by a wide variety of end products, one of the most important problems at the project planning stage is selecting the appropriate product mix to be con- 2. Wet phosphoric acid is preeminent in the phosphate fertilizer industry. The grade of rock-that is, its P20 content-is usually much less important than the nature of the impurities it contains. SPECIFYING THE PLANNING PROBLEM 75 sidered. A related consideration is selecting a plant that is flexible enough to produce the proper mix. The simplest case is the one in which stringent fertilizer recom- mendations have been drawn up for agricultural crops grown in the planning area. In that case, the fertilizer material that will be in demand is known in advance, and the product mix to be produced domestically depends fully on whether or not domestic production meets a given efficiency criterion. Although the choice of end products to be produced by the ferti- lizer industry is frequently made on this basis, the validity of the approach can be questioned. Fertilizer recommendations are nor- mally based on agronomic and economic considerations, and the price at which fertilizer is delivered to the farmer plays a consider- able role in determining the most profitable type of fertilizer to be used in any one region for specific crops. Clearly, however, the attainable delivered price of fertilizer products is not known at the outset of the planning exercise. If fertilizer requirements are ex- pressed in terms of major nutrients, then a priori a wide choice of products should be considered. In principle, therefore, as wide a range of end products as possible should be considered in the project selection phase. The number of end products is very large, however, and normally it will be neither possible nor desirable to include all products in the planning problem. No simple rules can be proposed to make a final choice, and a combination of elements must enter the decision. These include: the range of fertilizers currently used, the range of fertilizers for which area-specific fertilizer recommendations have been drawn up, agronomic advice, and expert opinion of the sector specialist. The selection of the final products to be considered in the planning model-not to be confused with the selection of the prod- ucts recommended for importation and production-is a typical example of an area in which judgment plays an important role. The use of a planning model is often mistakenly considered to be a proce- dure that does not allow for expert judgment. Nothing could be further from the truth, because both the specification of the planning problem and the interpretation of the results are primarily based on such judgments. Production processes Fertilizer manufacture has a much greater range of technological alternatives than is commonly assumed. In principle, if alternative 76 PROBLEMS IN FERTILIZER SECTOR PLANNING manufacturing processes for the products selected for investigation exist, these should be explicitly taken into account. This can greatly expand the scope of the planning problem: if ten products can each be manufactured with five alternative processes, the total choice set amounts to fifty processes. In practice, it will hardly ever be neces- sary to complicate the planning problem to such an extent. First, many alternative processes for producing a given product may be competitive in the sense that they have similar implications for investment and operating cost and are simply different proprietary processes that represent minor modifications of a basic manufactur- ing process. In other cases, alternative processes are clearly asso- ciated with plant scales, so that for a given capacity range one proc- ess is dominant, whereas for another capacity range an alternative process is dominant. A typical example is the production of am- monia: motor-driven reciprocating compressors are used for capaci- ties when the capacity is less than 600 short tons, whereas turbine- driven centrifugal compressors are used when the capacity is greater. Sometimes, the choice of production process may be limited because of the nature, availability, and quality of the feedstock, or the quality requirements of the end product. For example, in the production of sulfuric acid for phosphatic fertilizers, processes using pyrites normally need not be specified if elemental sulfur is locally available. In the case of end products, the example of multinutrient fertilizers that can be produced by bulk-blending plants as well as by compounding plants can be cited. If transport distances are great, the compounding process may often be preferable, since today's bulk blends exhibit segregation problems. The availability of processes may also restrict the number of processes that need to be considered in specifying the planning model. Some processes previously used in developed countries may no longer be operated. Even though such processes, under certain circumstances, may still be of interest to countries with different resources, difficulties in procuring equipment, replacement parts, and managerial expertise may make it undesirable to include them among the possible technological alternatives for a particular coun- try. Finally, certain processes may be more technically complicated than others, and hence require more highly trained manpower. Shortages of such manpower may often justify the a priori elimina- tion of these processes. In fact, a host of other local conditions may render certain production processes more attractive than others, SPECIFYING THE PLANNING PROBLEM 77 although it is difficult to generalize about this. As in the case of product choice, the judgment of the sector specialist is of great im- portance in deciding which production processes should be con- sidered explicitly in the sector planning problem. Production sites Since all products in the fertilizer industry can be transported, there is, in principle at least, a wide choice of potential plant sites. Economies of scale in the production of most products will tend to restrict the number of sites where production actually takes place. On the other hand, because the demand for fertilizer products is dispersed geographically, there will be a countervailing tendency to disperse the location of plants in order to lower transport costs. The major problem in site selection is to decide whether to locate a productive activity close to the raw material supply, close to the demand center, or close to transport facilities. Although the planning models discussed later are designed partly to answer this question, the answer depends to a large extent on which potential sites have been specified. In view of the dispersion of demand for fertilizer material, it will usually be desirable to specify the model as a region- ally disaggregated planning model, with a number of demand regions and corresponding marketing centers. In principle, each marketing center can be specified as a potential site. Moreover, the major ports should often be regarded as potential sites. If the planning area is large, this approach can result in a large number of potential sites. Since this greatly complicates the planning problem, it is desirable to attempt to restrict the number of potential sites as much as pos- sible. Once more, judgment plays an important role. Remote regions, with relatively modest fertilizer requirements and no relevant raw materials, will normally not have to be considered as sites for pro- ductive facilities. Two ports located in close proximity can be con- sidered as one potential site. Other, similar decisions that limit the number of potential sites to a manageable few can also be made. By-products Certain by-products of the fertilizer industry find uses in other industries. Similarly, certain products required in the manufacture of fertilizers are required in the production of other commodities as well; for example, sulfuric acid is extensively used outside the fertilizer industry. In such cases, the market for such products out- 78 PROBLEMS IN FERTILIZER SECTOR PLANNING side the fertilizer industry should be specified. First, the sale of by- products directly benefits the operation of the fertilizer industry. Second, explicitly incorporating the sales volume outside the ferti- lizer industry allows larger production capacities and thus lower production cost per unit. Both activities can be specified as "by- product sales." Their inclusion provides an approximation for the possible links of the fertilizer sector with other industrial activities and is a potentially important aspect of the specification of the plan- ning problem. Means of transport The correct specification of the transport alternatives facing the fertilizer industry is of considerable importance for several reasons: * To a varying degree, all raw materials and products in the in- dustry are mobile. * Transport cost is usually a substantial portion of delivered cost. * Transport cost is the main inhibiting factor for large-scale, geographically concentrated production units.3 * Even in some undeveloped regions, a number of transport alternatives are usually available.4 When the transport alternatives are being specified, it may often be useful to carry out a number of back-of-the-envelope calculations, and to decide on the basis of these whether any one form of trans- port is clearly more efficient than any other. With this approach, the least-cost means of transport between any two points in the planning area can usually be specified. If, however, the cost impact of economies of scale in transportation can be specified correctly, ignorance of the volume of trade between supply points and market- ing centers (or plants) would require that alternatives be explicitly specified. Marketing centers Since demand for fertilizer material is by nature geographically dispersed, and since transport cost is an important component of 3. In many locations, however, the cost of providing infrastructure for the con- struction of several plants is even more inhibiting. 4. In the most underdeveloped regions, a dirt road is often the only possibility. SPECIFYING THE PLANNING PROBLEM 79 delivered cost, a number of demand regions within the planning area and a number of marketing centers within each region should be specified as the terms of reference for determining transport cost. It is difficult to generalize about the appropriate number of demand regions.5 Often, the specification of demand regions will be based upon the availability of data, because a projection of fertilizer de- mand is required for each demand region. Administrative regions or districts can frequently form a reasonable basis for regional disag- gregation. Time periods The dynamic properties of alternative investment programs in the fertilizer industry can be investigated, using the planning models discussed in this volume. In certain cases, it may not be necessary to analyze these properties, and a static model will suffice; for ex- ample, in investigating the efficiency of the domestic transport pat- tern for an existing fertilizer industry. In most cases, however, the planner is interested in drawing up an investment and production program for a particular period of time, and questions arise about which planning horizon should be chosen and for what time inter- vals results should be reported. The first aspect to be taken into account is that a certain amount of time will elapse between the moment an investment decision is made and the moment the new capacity is installed and in operation. Although local circumstances may be responsible for variations in the length of this so-called gestation lag, a reasonable approximation of the average lag in the fertilizer sector is three years; that is, if a decision to invest in a new fertilizer plant is taken in early 1980, the plant can be expected to be in operation in early 1983. If a static model is not deemed appropriate by the planner, how many time periods should be specified? The answer depends on the expected rate of growth of demand for end products. If demand is projected to grow very slowly-say, at 2-3 percent a year-a two- period model may be sufficient to draw up a reasonably efficient investment program for the medium term. On the other hand, if demand grows fast-and in developing countries annual rates of 5. In terms of the computations involved in the planning models, it does not make much difference whether ten or fifteen demand regions are decided upon. 80 PROBLEMS IN FERTILIZER SECTOR PLANNING growth of over 10 percent are not unusual-this phenomenon needs to be incorporated into the planning model, so that the initial in- vestment decisions are made against the background of a rapidly growing market. As will be explained later, the presence of econo- mies of scale can, under such conditions, justify the installation of temporary overcapacity. In that event, four or five time periods of three years each may be a more appropriate model specification. Formulating an Investment Program Once the scope of the planning problem in the fertilizer sector has been specified, the next planning task is to formulate an efficient and internally consistent strategy for investment, production, foreign trade, and distribution. Several problems have to be addressed during this phase; they often involve issues that are highly inter- dependent and thus must be evaluated simultaneously. Among these issues are site selection, the scaling of capacity, the timing of ca- pacity expansion, and the transport pattern for raw materials, inter- mediate products, and fertilizer material; also, there are problems associated with selecting the appropriate feedstock, production process, and type of fertilizer material. Site selection If more than one potential site has been specified in the planning problem, a decision must be made about the site or sites where the capacity will actually be installed. Several factors affect this decision. First, investment cost, as well as operating cost, may differ among potential sites. Second, each potential site has different implications regarding the transport cost for both raw materials and final prod- ucts; in the case of a nonintegrated plant, there may be different implications for intermediate products as well. Third, the trade-off between transport cost and economies of scale will have to be deter- mined when decisions are made about the appropriate plant site. Clearly, evaluating the impact of these factors on site selection can involve cumbersome calculations. This is true particularly if fertilizer demand is heavily dispersed geographically and if it will grow rapidly over time. In that case, the selection of the appropriate plant site cannot be viewed as a static planning problem; rather, the whole lifetime of the plant must be taken into consideration. This FORMULATING AN INVESTMENT PROGRAM 81 raises the possibility of trade-offs over time, since a specific site may be the most efficient, even though during the initial years more attractive alternatives seem to exist. The question of site selection is interdependent chiefly with the question of whether economies of scale are present in fertilizer pro- duction. This latter question raises difficult planning problems con- cerning the selection of the appropriate scale and timing of new capacity. Capacity scaling Most activities within the fertilizer sector exhibit economies of scale in some form or other. This implies that the scale at which each of these activities is carried out directly influences the structure of production costs. Hence, selecting the appropriate scale is vitally important in designing an efficient investment program. Typically, the relation between the investment cost and the in- stalled capacity can be represented by a curve, as shown in figure 10. When capacity is doubled from X1 to X2, the investment cost in- Figure 10. The Investment Cost Function 0 ~ ~ ~ ~ 0~~ Xi X2 Capacity 82 PROBLEMS IN FERTILIZER SECTOR PLANNING creases, but it does not double. Algebraically, this relationship can be represented by an exponential function of the general form: I = aC', where I is total investment, C is plant capacity, a is a constant, and b is the constant elasticity. The parameter b will be less than unity when there are economies of scale, and larger than unity when there are diseconomies of scale. In the chemical sector, the so-called six- tenth rule is often used to approximate capital requirements; in such cases, 0.6 is the value assigned to the parameter b. The planning problems associated with the capacity-scaling deci- sion can be described, focusing first on the static case. To simplify the example, it is assumed that the decision to be made is about the optimal scale of a fertilizer plant that faces a given, regionally disag- gregated market. The possibility of market growth over time is ex- plicitly ignored for the moment. The main planning questions are the following: * How many plants should be constructed? * Where should they be located? * Which plant scales should be selected? For this static capacity-scaling decision, the trade-off between trans- port cost and economies of scale is the crucial determinant. In the presence of economies of scale, a few large plants are preferred to many small productive units because of production cost advantages. With the former structure, however, the distribution costs would be higher if the market is geographically dispersed. It is conceivable that distribution costs for certain remote markets would be so high that the cost advantages of a large plant would be eliminated, and that such regions could be better supplied from either imports from abroad or other domestic plants. But caution should be exercised in reaching this judgment, as the following simple numerical example may make clear. Assume a country that is broken down into two regions, A and B. In region A, the market for a specific fertilizer is 50 units; in region B, the market for the same product is 100 units. In terms of the previously stated questions, the planning problems are the fol- lowing: Should domestic production be recommended? If so, should two plants be constructed-one in region A and the other in region B? Or should there be just one plant-either in region A or in region B? Furthermore, should the total capacity to be constructed equal the total market demand? Or should one or both regions be supplied FORMULATING AN INVESTMENT PROGRAM 83 through imports from abroad? To answer these questions, informa- tion is needed on production cost per unit at various capacity levels, the import price, and transport cost. Let the production costs vary with scale so that, at a production level of 50 units, the per unit production cost is 110; at 100 units, it is 100; and at 150 units, it is 90. The transport cost between A and B is 10 per unit. The delivered price of the imported product in A is 95 per unit, and in B it is 105. A simple set of calculations shows that the least-cost solution to the problem is to establish a plant in B that produces 150 units and serves the entire domestic market. Although total costs of meeting domestic demand under this program are lower than under any other alternative, the delivered cost of the product in region A is higher than could be achieved through imports if full-cost pricing were practiced.6 In this case, region A loses all the cost advantages of larger scale production because of the transport cost between A and B. In contrast, region B reaps the full advantage of its plant having access to the market of region A. To guarantee continued access, it may therefore be to region B's advantage to sell its product at less than full price to region A. This exceedingly simple example can be made more complex by specifying the problem more realistically. There may be several com- peting fertilizer products or modes of transport; there may be some question about the availability of infrastructure, the location of raw materials, or the manufacture of intermediates. With these additional specifications, detailed hand calculations become increasingly more difficult, and some efficient screening device becomes necessary to evaluate the alternatives. Such a device becomes even more urgent if the element of time is brought into the picture. The timing of capacity expansion If demand for a given fertilizer product is projected to grow over time, and if the domestic production of the fertilizer is contemplated, the selection of the appropriate scale of production is closely related to the timing at which capacity expansion should take place. Since 6. Imports in A have a delivered cost of 95, whereas production in B involves a delivered cost in A of 100 (production cost of 90 at a scale of 150 units, plus 10 for transport cost). 84 PROBLEMS IN FERTILIZER SECTOR PLANNING a larger plant is normally preferred to a smaller plant in the presence of economies of scale, all other factors being the same, it may pay to postpone construction of a plant until the size of the market has reached a certain level.7 If the market continues to grow, the con- struction of the plant could conceivably be postponed until some technological maximum becomes feasible and diseconomies of scale begin to appear. It is possible, however, that long before this point the size of the domestic market would have permitted domestic production to take place at a scale that results in production costs below import prices. In that case, the planning problem is to decide on the most efficient timing of capacity construction. This decision has to be made on the basis of a comparison between, on the one hand, the present value of the total cost saving that can be made by an earlier and smaller plant in relation to import cost and, on the other hand, the saving of a later and larger plant. A further possibility is that it may be profitable to install a plant that will initially have some excess capacity. As the market grows, capacity utilization will gradually increase so that, eventually, lower production costs per unit can be achieved than would be the case with a smaller plant. Clearly, the rate of growth of demand, the extent of the scale economies, and the appropriate rate of interest at which future costs and benefits ought to be discounted play the determining role in this decisionmaking process. The transport problem Transport cost constitutes the major mitigating factor to cen- tralizing economic activity that exhibits economies of scale. In this context, the transport problem can be defined as one of finding the transport pattern that minimizes the cost of delivering the required quantities of raw materials and intermediate products to the various plant sites, and of fertilizer products to the marketing centers. This is achieved by selecting not only the shortest route but also the least-cost mode of transport among plants, between plants and marketing centers, and between importing points, marketing centers, and plants. Solving the transport problem requires that distances for each mode of transport and rates for each product be correctly 7. In addition, in many developing countries, the technical strength to com- mission and run several plants starting at the same time is seldom available, thus necessitating phased construction. FORMULATING AN INVESTMENT PROGRAM 85 specified. The planning models presented in subsequent chapters are well geared to handle the cumbersome calculations required to select the most efficient transport pattern. The choice offeedstock Many products in the fertilizer industry can be produced on the basis of several feedstocks. The choice may be among basically different raw materials or products or among different grades of a given product type. In most cases, the question of feedstock selection will be resolved by comparing the cost of alternative feedstocks. Several problems can arise. First, the comparative cost analysis requires the appro- priate pricing of alternative feedstocks. If the feedstock is imported from abroad, the c.i.f. (cost including insurance and freight) import price forms the basis for evaluation. If the raw material is available domestically, the evaluation problem is more difficult because alter- native uses-for example, use as an export-should be taken into account.8 If the raw material is a depletable resource, as will often be the case in fertilizer manufacture, the opportunity cost from the point of view of the entrepreneur may diverge from that of society as a whole. Other, more indirect problems may arise in the context of feed- stock selection. For example, a specific feedstock may produce coproducts or by-products that can be used elsewhere in the econ- omy, whereas alternative feedstocks may not exhibit this feature. This element should obviously be entered into the calculations, because the benefits derived from the sale of by-products can be attributed to the investment program. Second, the use of a particu- lar feedstock may affect a number of economic policy objectives, such as job creation, regional development, or improvement of the environment. These considerations might play a role, for example, in selecting coal as a feedstock for nitrogenous fertilizer production. So far, the problem of feedstock selection has been discussed as if it could be discussed independently of the other aspects of the 8. Imported phosphate rock is almost invariably beneficiated to a high degree, and its cost is correspondingly high. The advantage of local rock, in this case, would be that it can be used without undue beneficiation and that its cost is cor- respondingly lower. In such cases, the imported and domestic phosphate rock can be treated as two (substitutable) products. 86 PROBLEMS IN FERTILIZER SECTOR PLANNING planning problem. It is not difficult to conceive of circumstances in which this may be an inappropriate approach. To begin with, the comparative cost analysis of alternative feedstocks cannot be carried out unless the location of processing facilities is considered simul- taneously. Since the latter will also be influenced by demand factors, the geographical dispersion of demand must be considered as well. In the same vein, if the cost of imported raw materials or intermedi- ate products is a function of the quantity required, the scale of op- erations of the processing facility must be considered. In turn, differ- ent feedstocks may entail different technologies in the processing stage, with different investment requirements and operating costs. Finally, the optimal timing of capacity construction in processing facilities may be affected by the gestation periods required to open mines or to provide adequate transport facilities or infrastructure. Process selection If a product that is a candidate for inclusion in the investment program can be produced by more than one process, another dimen- sion is added to the planning problem. When the number of alterna- tive processes is large, this modification of the planning problem can complicate the identification of the optimal program considerably. The major difficulty of selecting the appropriate production proc- ess is that the scale of production often determines which production process is the most attractive. Since the scale of production depends greatly on cost of production and location, among other factors, the choice of process has to be made simultaneously with decisions about other aspects of the investment program. If many products are to be considered at the same time, the selection of production processes may render the planning problem even more complex than it al- ready is. Even using the investment planning models presented in the following chapter, the introduction of alternative production processes greatly increases the difficulty of the planning problem. Thus, the number of alternative processes to be specified in the planning problem should be as limited as possible. The choice offertilizer types Normally, the planning problem should specify a variety of ferti- lizer products that can be used to meet given nutrient requirements. This specification may or may not be combined with fertilizer recom- FORMULATING AN INVESTMENT PROGRAM 87 mendations of various degrees of stringency. In practice, at the problem specification stage, fertilizer types are often selected on the basis of technical or agronomic considerations, but that cannot fully account for the interdependencies between supply price, on the one hand, and the various aspects of the planning problem, on the other. High-analysis fertilizers such as urea are often said to be preferred to lower-analysis fertilizers such as ammonium sulfate because the transport cost per ton of nutrient is lower. Clearly, however, trans- port cost is not the only element that should be taken into account, and what counts is the total delivered price and agronomic effective- ness of the various products. Circumstances may well exist under which a country would be better off producing its own low-analysis fertilizer material rather than importing high-analysis substitutes. The investigation of this aspect of the program becomes rapidly impossible-particularly if the planning problem involves many products, sites, marketing regions, and so forth-unless an efficient screening device is available. As will be demonstrated, the planning models to be described can be used effectively to address this difficult and, at the same time, crucial problem of fertilizer product choice. Special problems The foregoing discussion of decisionmaking problems focused on the major issues related to designing an investment and production strategy. Clearly, in specific situations, other problems can come up that have not been covered so far. For example, a planner may simply want to evaluate the efficiency of an existing industry or, even more specifically, the transport pattern that has evolved over time for fertilizer material. Similarly, a planner may be asked to recommend whether a particular plant should be scrapped, main- tained, or modernized. An endless number of area-specific problems can be conceived, and little is served in attempting to enumerate all possibilities. Rather, the next two chapters, which focus specifically on models that can be used to address the planning problems de- scribed above, contain guidelines on how to modify the model struc- ture to address specific problems not covered in detail. 8 Planning Models for the Fertilizer Sector A SERIES OF LINEAR AND MIXED-INTEGER PLANNING MODELS is de- scribed in this chapter; those models may be useful in addressing the decisionmaking problems that need to be resolved when formu- lating an investment and production strategy for the fertilizer sector. The presentation is self-contained, in the sense that prior familiarity with mathematical planning models is not required. The models are presented in order of gradually increasing complexity, from a simple transport model to a full-fledged, dynamic project selection model. Those who are already familiar with such models may wish to skip this chapter and proceed immediately to chapter 9, which starts off with a full statement of the complete model and a list of its variables and parameters.' 1. The present chapter follows closely the layout of chapter 3 of volume 1 in this series-David Kendrick and Ardy Stoutjesdijk, The Planning of Industrial Investment Programs: A Methodology, The Planning of Investment Programs, ed. Alexander Meeraus and Stoutjesdijk, vol. 1 (Baltimore: Johns Hopkins University Press, 1978) -except when the structure of the fertilizer sector calls for a sector-specific formu- lation (for example, to account for product substitution on the demand side). 88 THE TRANSPORT PROBLEM 89 In this chapter, a model that can be used to analyze transport problems relating to the fertilizer sector is presented first. This model can be expanded to include existing production facilities, so that the efficiency of an existing fertilizer sector can be assessed. Next, a series of capacity-expansion models are presented; as new dimensions to the planning problem are gradually added, these models increase in complexity and realism. Toward the conclusion of the chapter, enough background information will have been provided so that one can understand a planning model of the fertilizer sector that is dynamic; covers several potential production sites; explicitly includes the choice of feedstocks, processes, and fertilizer products; and permits the determination of the investment, production, foreign trade, and transport pattern that leads to the least-cost strategy to meet specified fertilizer requirements. The Transport Problem Because of the simplicity of the transport model, most textbooks on linear programming begin by describing it. For the fertilizer sector, the transport model has significance beyond its instructive value: transport costs of fertilizer products constitute a significant proportion of total delivered cost of the product to the farmer, and hence they are of considerable importance in determining the least- cost supply pattern. The transport problem can be described as follows: given a num- ber of plants that can serve a number of markets, which transport pattern will minimize transport costs? Two conditions have to be borne in mind during the solution of this problem. First, no plant can ship more products than it produces (hence, the possibility of inventories is ignored). Second, no market should receive less than it requires (hence, it is assumed that the plants' capacity to produce is sufficient to meet total market requirements). Let: i = an individual plant in the set of plants I; j = an individual marketing sector in the set of market- ing centers J; xii = the amount of fertilizer shipped from plant i to mar- keting center j ki = the capacity of plant i. 90 PLANNING MODELS FOR THE FERTILIZER SECTOR Then, the first building block of the transport model is the following constraint: (8.1) Xii < ki. e j6J Sum offertilizer 1 1 shipments from < Capacity of plant r each plant plant i to all _ marketing centers j In (8.1): . the symbol e indicates membership in a set; * the statement E indicates that one should sum over all mar- jts keting centers in the set J; * the symbol < indicates less than or equal to; . the symbols i e I on the right indicate that there must be a con- straint of type (8.1) for each plant i. Consider a country with two single superphosphate plants and three marketing centers. Then, the set I can be written as I = 11, 2} and the set J can be written as J = { 1, 2, 3 }, so that (8.1) is written as: Xll + X12 + X13 < k1, x21 + x22 + x23 < k2. In other words, the shipments from plant 1 to marketing centers 1, 2, and 3 cannot exceed the capacity of plant 1, and the shipments of plant 2 to marketing centers 1, 2, and 3 cannot exceed the capacity of plant 2. The second constraint of the transport model stipulates that no marketing center should receive less fertilizer than its specified re- quirements. Let: dj = the fertilizer requirements of marketing center j. Then, the fertilizer requirement constraint can be written as: (8.2) Exj > di. jEJ Sum of shipments Fertilizer For all from all plants i > requirement of marketing to each marketing _ marketing centers center j _ _ centerj j _ THE TRANSPORT PROBLEM 91 In (8.2), the symbol > means equal to or larger than. It can be ques- tioned why an inequality sign is specified here rather than an equality sign. Why, indeed, would shipments to any marketing center exceed market requirements? In most cases, shipments to a region would exactly match requirements. Sometimes, however, market require- ments can be met by composite products that contain fixed propor- tions of several elements. Diammonium phosphate (DAP), which contains both nitrogen and phosphorus, provides an example for the fertilizer industry. Unless market requirements to be met by DAP are specified in the same proportion as nutrient content, it is clearly impossible to meet both nitrogen and phosphorus requirements exactly. If the market requirements constraint were specified as an equality, it would not be possible to use DAP to the fullest extent to meet demand; if no other product were available, it might even render the solution of the model infeasible. For the example given above-that is, a country with two single superphosphate plants and three marketing centers-(8.2) can be written as: xi1 + X212 di, X12 + x22 > d2, x13 + x23 > d3. Finally, shipments of fertilizers, xij, cannot assume negative values. This is formally stated in the model by nonnegativity con- straints: (8.3) Xij > 0. i jfj The objective or criterion of this problem is to minimize transport cost while meeting the constraints (8.1), (8.2), and (8.3). Let: gij = the constant unit transport cost for shipping fertilizer from plant i to marketing center j. Then, the transport cost of shipment xij from plant i to marketing center j is equal to: lijixi, and the objective of minimizing total transport cost can be stated as minimizing i, where: 92 PLANNING MODELS FOR THE FERTILIZER SECTOR (8.4) =E E Aijxij. Total Sum of the transport traporta = costs for all shipments cost from all plants i to all L Cost _ L marketing centers] j For the two-plant, three-market example, (8.4) would be written as: min t = (,axIll + A12xI2 + pu13x13) + (p12x21 + g22x22 + JU2x23). In summary, the transport model is to select xij to minimize total transport costs: (8.5) min = pijxij, id j.J subject to the capacity constraints (8.1), the market requirements (8.2), and the nonnegativity constraints (8.3). Modeling an Existing Industry The transport model can be gradually expanded to a model of an existing industry by introducing in turn production cost, multiprod- uct plants, production processes, intermediate products, interplant shipments, exports, and imports. Finally, a description is given of how product substitution and fertilizer recommendations can be incorporated into the structure of the model. The possibility of capacity expansion is explicitly ignored for the time being. Production cost The simplest way to modify the transport model into one that incorporates production cost at plant i is to assume for the time being that production cost per unit of product is constant. Let: Oi = the average production cost for a given fertilizer at plant i. Then, the total production cost for the product at plant i is: oi E xij. j3j MODELING AN EXISTING INDUSTRY 93 The criterion function of the model must now be modified to in- clude the total production cost at all plants i, as well as the transport cost: (8.6) min = E E jxij + E (o E x iel jfj isI j J _ Sum of the trans- _ Total port cost for Total production cost = all shipments from + cost for all plants i to all plants i Li _ _marketing centers j_ =E E Xij(gij + °i)@ icI jeJ Constraints (8.1), (8.2), and (8.3) remain unchanged. The simple model above permits one to analyze the trade-off between production and transport costs. If plant 1 has lower pro- duction cost than plant 2, it may be able to absorb the transport cost and still deliver the product to a relatively distant market at a lower total cost than plant 2. In that case, because of the cost mini- mization criterion, shipments from plant I to the distant market will be preferred to shipments from plant 2. A multiproduct model Normally, more than one fertilizer product is produced in the industry, and thus the model should be generalized to a multiproduct form. This is done simply by adding a subscript c to represent the fertilizer types that can be produced. Let: C = the set of fertilizers produced in the industry; x,ij = the amount of fertilizer c shipped from plant i to marketing center j. Then, the model is to select x,jj so as to minimize: (8.7) min = = (Z E icijxcij + E O,i E xci) ceC if IgJ ifl jeJ _ i_ Shipment cost Total _ from all plants i +Production cost cosJ to all marketing at all plants i JL _ _ centers j subject to the following constraints. 94 PLANNING MODELS FOR THE FERTILIZER SECTOR CAPACITY CONSTRAINTS (8.8) E xcii < kei c eC MARKET REQUIREMENTS (8.9) E xcij 2 d1j c e C iei jeJ NONNEGATIVITY CONSTRAINTS (8.10) x.ij 2 0 c C jJE Productive units Until now, plant capacity has been linked to a single fertilizer product. This treatment is unsatisfactory, however, whenever plant capacity can be used to produce more than one product-for example, a bulk-blending plant that can produce a large number of fertilizer mixes of widely varying nutrient content. This possibility can be in- corporated into the structure of the model by introducing the notion of productive units that have their capacity expressed in terms of hours of availability or volume of throughput. It now becomes necessary to redefine I as a set of plant locations, with i referring to a particular plant site rather than the plant itself. To specify the model in this form, let: z,i = the production of fertilizer c at plant site i; M = the set of productive units, where m is an individual machine or productive unit in the set; bm.,i = the number of units (for example, hours) of capacity used on machine or productive unit m per unit of output of fertilizer c at plant site i. The model can now be stated as: select x.ij and zci so that the sum of production and shipment costs are minimized. That is: (8.11) min e =E (y E yijx-ij + E 6ciz9i, ceC id ,{J ieI =E E ( E ccijxcij + *czi)i s oC iai jeJ subject to a set of constraints. MODELING AN EXISTING INDUSTRY 95 MATERIAL BALANCE CONSTRAINTS (8.12) Zci 2 S Xcie Ce C j6J icel 1 _ Shipments of_ 1 Production of fertilizers c For all fertilizer c at > from plant fertilizers and I plant site i site i to all alpansie pnsemarketing la[ L _ L centers j _ CAPACITY CONSTRAINTS (8.13) E bmcizci < kmi m eM eEC ieI Production a pacity For all levels for all < nanalbly pro productive fertilizers at s< I Ln y prodc u nits and a] plant site i -at plant site i _ plant sites Constraint (8.13) replaces the previous constraint (8.8) and reflects the change in the definition of capacity from a commodity charac- teristic (kai) to a machine characteristic (kji). The coefficient bjci in (8.13) will be positive when a commodity requires a particular pro- ductive unit for its production and zero when the productive unit is not required. Finally, two constraints correspond to the previous constraints (8.9) and (8.10). The only difference is that there is an additional nonnegativity constraint for zci. MARKET REQUIREMENTS (8.14) 2 x.ij2d 4 ceC i6s jeJ NONNEGATIVITY CONSTRAINTS (8.15) Xij, Zci > ° jEJ ce C A process model The mathematical formulation of the process model is completed by introducing variable inputs required for the production of ferti- lizers. To do this, the notion of a productive process must be gener- 96 PLANNING MODELS FOR THE FERTILIZER SECTOR alized, from the output of commodity z,i to the activity level of a process zPi, where zpi = the activity level of process p at plant site i. This modification of the model permits production methods that create more than one product to be described. For example, carbon dioxide, which is produced as a by-product in the synthesis of am- monia, is required for the production of urea. In addition, the process notation enables one to introduce into the model alternative processes for producing the same good: for example, the use of fuel oil, natural gas, naphtha, or coal for the production of ammonia. To specify processes, let: a,,i = the input (-) or output (+) of commodity c by process p at plant site i per unit level of activity of process p. For example, consider the particular process p for producing the chemical fertilizer single superphosphate. The process has phosphate rock, sulfuric acid, and labor as its main inputs, and electric power and unspecified supplies as its miscellaneous inputs. In such a case, one could have: a.pj = 1.0 ton c = single superphosphate api = -0.6 ton c = phosphate rock aepi = -0.36 ton c = sulfuric acid acpi = -0.4 manhour c = labor api = - $0.50 c = miscellaneous material inputs That is, the process combines 0.6 ton of phosphate rock, 0.36 ton of sulfuric acid, 0.4 manhour, and $0.50 worth of miscellaneous material inputs to produce I ton of single superphosphate. In order to specify a process model, commodities used or produced in the fertilizer industry are divided into five groups: fertilizers that are shipped from plants to markets; intermediate commodities that are either produced within the plant or shipped from other plants; raw materials that are purchased from outside sources; material inputs for which no separate specification is considered necessary, referred to as "miscellaneous material inputs"; and inputs of dif- ferent types of labor. The latter three categories are considered to be directly variable with output. Accordingly, the set of inputs and outputs C is subdivided into the following subsets: CF = fertilizer types; CI = intermediate commodities; MODELING AN EXISTING INDUSTRY 97 CR = raw materials, miscellaneous material inputs, and labor types. Since cost is no longer attached to a final commodity, but rather to the various input categories, it is necessary to define: pci = the price of input c at plant site i; u,i = the amount of purchases of input c at plant site i. Thus: E p,ju,i = the cost of inputs to plant site i. CeCR With this notation, the process model can specified as: select zpi, xefj, and uca so as to minimize the sum of total shipment cost and the total cost of raw materials, labor, and miscellaneous ma- terial inputs. That is, (8.16) min = E E , c3ixcs + E pciuc, CeOF iel jeJ CECR subject to the following constraints. MATERIAL BALANCE CONSTRAINTS ON FERTILIZERS (8.17) E acpizpi 2 E xcij C E CF P' J'> ~~~~~~ieI _ _ ~~~~Shipments of__ Production of fertilizer c For all fertilizer c by > from plant fertilizers all processes p _ site i to all and all plant at plant site i marketing sites _ _ centersj The typical process p that provides fertilizers can be assigned a coefficient a,,i = 1.0 in the fertilizers, or final commodities, con- straint (8.17) because the unit of capacity can be defined arbitrarily in terms of one of the inputs or outputs. MATERIAL BALANCE CONSTRAINTS ON INTERMEDIATE COMMODITIES (8.18) E acizis 2 ° c e CI pe,p iel: Production of For all inter- intermediate product c > 0 mediates and by all processes p at all plant sites plant site i a l i 98 PLANNING MODELS FOR THE FERTILIZER SECTOR The production of fertilizers and some intermediate products (for example, phosphoric acid) requires the use of intermediate products. At least one process p in constraint (8.18) therefore has a negative coefficient a,j, while at least one process for producing an intermediate good must have a positive coefficient in the intermediate product constraint for the latter to hold. This assumes, for the time being, that there are no interplant shipments of intermediates. MATERIAL BALANCE CONSTRAINTS ON RAW MATERIALS AND LABOR (8.19) E apiz,i + uci 2 0 c e CR peP ieI Use of raw Purchases of [Forallraw1 material c by raw materials F>0 materials and all processes p c at plant all plant sites at plant site i _ site i l The coefficient a,pi in constraint (8.19) is normally negative, be- cause the production of all intermediate and final products requires raw materials and labor. Purchases of raw materials and hiring of labor u,j therefore have to be positive for the constraint to hold. CAPACITY CONSTRAINTS (8.20) E b.,iz,i < k-i m e M Total number _ 1 _ capacity used Capacity of For all pro- in .prctived productive ductive units In productive < unit m at and all plant unit m for all plant site i sites processes p .Lat planit site i J L _ L MARKET REQUIREMENTS (8.21) E xcij 2 djc c CF "i jEJ NONNEGATIVITY CONSTRAINTS (8.22) X¢ij, zpi, ui > 0 ieI j eJ c e C p e P The description of the process model is now complete, although MODELING AN EXISTING INDUSTRY 99 several features of the fertilizer sector are still missing from the model structure: namely, interplant shipments of intermediate products, the possibility of foreign trade in the commodities of the sector that can be traded, and, in particular, the phenomena of fertilizer prod- uct substitution and fertilizer use recommendations. These elements are introduced into the model in the next few sections. The remainder of the chapter then focuses on the main subject of this volume- that is, the structure of planning models for expanding capacity in the fertilizer sector. Interplant shipments In principle, almost all intermediate products of the fertilizer industry can be shipped among plants, and this possibility should be provided for in the mathematical structure of the model. Two simple changes need to be made to achieve this purpose. First, the balance equation for intermediate products needs to be changed. Second, a transport cost element needs to be added to the objective function. Let: xcii, = shipments of commodity c from plant site i to plant site i'; uci, = unit cost for transporting commodity c from plant site i to plant site i'. Then, constraint (8.18) of the process model must be modified to read as follows: (8.23) E acpizpi + E (xXwi - xcw) > 0. c e C Shipments of intermediate product c Production of from all other For all inter- intermediate plant sites mediate product c by + minus ship- > 0 products and all processes p ments of all plant at plant site i intermediate sites product c to all other L J L plant sites . The expression ij34i means that only shipments of intermediate products among different plants, not shipments within the same plant, are included. 100 PLANNING MODELS FOR THE FERTILIZER SECTOR The change in the objective function of the model requires the addition of transport cost for interplant shipments, so that equation (8.16) now takes the following form: (8.24) mine = EE Eucijxij + E E E ucii xei, ceCF itl jeJ CCI ier iltr _ _T nTransport cost for Transport cost for Total _ all fertilizers from + all interplant cost all plant sites to all shipments of L_ _ marketing centers L intermediates + E E Pc,Uc- ceCR ie F Total cost of raw materials, labor, + and miscellaneous material inputs _ at all plant sites The other elements in the model are not affected by the introduction of interplant shipments of intermediate products. Foreign trade The model must now be modified to permit the import of raw materials and the import and export of intermediate products and fertilizer products. In this context, imports and exports refer to ship- ments from and to areas outside the planning area. To specify imports and exports, the following new variables and parameters need to be defined: v,i = imports of commodity c (raw materials, interme- diates, and labor) to plant site i; v,j= imports of commodity c (fertilizers) to marketing center j; e,ii = exports of commodity c from plant site i to export market area 1; L = set of export markets; A,i = unit transport cost for shipping imported commodity c from the appropriate port of entry to plant site i; puj = unit transport cost for shipping imported commodity c from the appropriate port of entry to marketing center j; juij = unit transport cost for shipping commodity c from MODELING AN EXISTING INDUSTRY 101 plant site i to the appropriate port of exit when the commodity is bound for export market area 1; Pei= import price of commodity c at the appropriate port of entry when the commodity is bound for plant site i; pj = import price of commodity c at the appropriate port of entry when the commodity is bound for marketing center j; pci, = export price of commodity c at the appropriate port of exit when the commodity is bound from plant site i to export market area 1. Exports are labeled by destination I in order to permit the model to include discriminatory pricing and, as explained below, to enable the specification of area-specific limitations to the volume of ex- ports. The criterion, or objective, function can be simplified by being divided into the following components: = domestic raw materials, miscellaneous material in- puts, and labor costs; transport cost; = import cost; = export revenues. Then, the criterion function can be defined as the algebraic (plus and minus) sum of these terms, of which the first three are positive (costs) and the last one is negative (revenues). The model can now be restated as: select z,i (process levels), xj (shipments of fertilizers), x,ip (shipments of intermediate goods), eeil (exports), v,j (imports of raw materials, intermediates, and labor), v,j (imports of final goods), and uei (purchases of domestic inputs), so as to minimize: (8.25) min + + where INPUT COST (8-26 ) +;=EEpdic,, CeCR i61 Input Total cost of domestic raw 1 cost = materials, labor, and s = [miscellaneous material inputs] 102 PLANNING MODELS FOR THE FERTILIZER SECTOR TRANSPORT COST (8.27) _ = E (Z E Acijx,i + E AjVe +E ) Cfcp if I jfj ijf if I IfL + E (E E Acii'xcii ci + E E: cilecil cfCI ifeT i,' if ier I eL + ( civi a rwR ig d _Trans-1 Total cost of transporr of fertilizers, intermediates, labor,_ Trans- and raw materials, inicludinig domestic transport cost of port = imports and exports; domestically purchased raw materials cost are assumed to be priced inclusive of domestic transport L JL~~cost IMPORT COST (8.28) = E vj + E EPciv6i CeWe j,J Ce(CIUCR) ifr _ ] _ -Total cost of imports of fertilizers and intermediate goods- im- as well as raw materials and labor, where the notation port = CIU CR means the union of the sets Cl and CR; that is, cost the summation is over all commodities that are either Li L_intermediate products, raw materials, or labor EXPORT REVENUE (8.29) p= E Z ZPce.il, ce(CFUCI) ifd If L Exort 1 Total revenue resulting froml Export = exports of fertilizers and revenueJ L intermediate goods subject to MATERIAL BALANCE CONSTRAINTS ON FERTILIZERS (8.30) E acizi > E xij + E e.il c fECF pfp iJ IfL ieI Production Domestic Exports of 1 of fertilizer > shipments of +feriie c For all Lc by all >1fertilizer c from plant ~ fertilizers atpln sier om pall r xotn+ln ie processes p rsioe ipto a + site i to all and all site i marketing e re ing plant sites sie J L_ centers 1 MODELING AN EXISTING INDUSTRY 103 MATERIAL BALANCE CONSTRAINTS ON INTERMEDIATE PRODUCTS (8.31) X a,,iz,i + X, xc,:. + V Zx 2 E: x,ii + E e ci E C CI Pep i'r$ .i'iI teL iE I Production Shipments of of inter- intermediate Imports of mediate product c to intermediate product c at + plant site i + product c plant site i from all shipped to by all other plant site i Lprocesses pi L plants _ Shipments of i ir intermediate Exports of For all product c intermediate intermediate 2 from plant + product c to products site i to all all export and all other plant regions plant sites L sites J L L MATERIAL BALANCE CONSTRAINTS ON RAW MATERIALS, LABOR, AND MISCELLANEOUS MATERIAL INPUTS (8.32) , aC,izj + u,j + v.j > ° c fECR PEP ieI Use of Domestic Usebor, raw hiring of Imports labor, raw labor and of labor, materials, purchases raw ma- For all and mis- of raw terials, raw ma- caerlanou materials + and mis- > teil material + and mis- + cellaneous and all inputs C cellaneous material plant at plant material inputs c sites all pro- inputs c for plant cesses p at plant site i cesss L siteiJ LL CAPACITY CONSTRAINTS (8.33) , biz,i < ki m E M peP i E I _ at _ 1 For all productive Capacity < Capacity units and all Irequired_ available ]_ plant sites I 104 PLANNING MODELS FOR THE FERTILIZER SECTOR MARKET REQUIREMENTS (8.34) i xczj + vej 2 dej c E CF iel jeJ fShipments of io Require- For all fertilizer c Imports of ments Of fertilizers, from all + fertilizerc > marketing and all plant sites i for market- _ centerfor marketing to markentng ing center] fertilizer c centers EXPORT BOUNDS (8.35) eci 0. p e P i', icel ce C je J Ic L I e L By-product sales Several products of the fertilizer industry can be sold outside the industry as by-products, and this possibility of domestic sales may affect investment decisions. The notations used in the model to denote domestic sales and prices are: w.j = domestic sales of commodity c as a by-product; PC'. = domestic price of commodity c at plant i. MODELING AN EXISTING INDUSTRY 105 The revenue from the domestic sale of by-products, k, should be added to the criterion function (with a negative sign, since it repre- sents a revenue) where: (8.37) Ox E E Pi. ceCI idl Moreover, the material balance constraint for intermediate products (8.31) changes by adding w.j to the right-hand side of the con- straint: (8.38) E a.pjzj + E xci,j + v. > E xcii, + E ecjj + wcj. C e rI Finally, a nonnegativity constraint has to be added for w,j: (8.39) wci > 0. c E CI ieI Product substitution and fertilizer recommendations Until now, the development of the planning model for the ferti- lizer sector has followed closely the general industrial planning model. With the introduction of product substitution and fertilizer recommendations, the model structure assumes a form specific to the fertilizer industry. If fertilizer market requirements are stated in terms of nutrients, a large number of fertilizer types can, in principle, meet market demand. On the other hand, fertilizer recommendations, drawn up by agronomists and soil chemists, tend to limit the scope for choice. In order for these phenomena to be incorporated in the model, the following notations must be introduced: CQ = the set of all nutrients-usually this set will be re- stricted to the three primary nutrients (nitrogen, phosphorus, and potassium), although secondary and micronutrients can also be included; d j = requirements of nutrient c' at marketing centerj; a,,, = the content of nutrient c' in fertilizer c; dc.j = recommended use of fertilizer c at marketing center j. 106 PLANNING MODELS FOR THE FERTILIZER SECTOR The market requirements constraint is now modified as follows: (8.40) E E ac¢xcij + E (x,,,v,j > d]Ej. je J id ceCF c6CF C e CQ Quantity of Quantity of - 1 nutrient c' nutrient c' Requirement For all shipped to supplied to of nutrient c' marketing marketing + marketing >1 at marketing centers, and centerj center j center j all nutrients from all from Lplant sitesi L imports J L i Moreover, a recommended use constraint is added to the model: (8.41 )E xcij + vcj > dj. c e CF id j e J Quantity of Recom- For all fertilizer c mended use fertilizers supplied to > offertilizer and all marketing c at market- marketing centerj j L ing center j L regions _ Recapitulation The model of an existing industry is now complete. At this point, it might be helpful to give a list of all the symbols used and a state- ment of the model as it has been developed. First, the symbols are categorized as the sets and indexes, the parameters, and the variables: Symbol Definition THE SETS AND INDEXES ceC Commodities used or produced in the fertilizer industry CF Fertilizer types CI Intermediate products CQ Nutrients CR Raw materials, miscellaneous material inputs, and labor types i, i'fI Plant sites jeJ Marketing centers leL Export regions meM Productive units peP Processes MODELING AN EXISTING INDUSTRY 107 Symbol Definition THE VARIABLES z Process level (production level) x Domestic shipments e Exports v Imports u Domestic purchases w Domestic sales of by-products 0 Cost groups 0^b Domestic raw materials, miscellaneous material inputs, and labor cost 0), Transport cost 0. Import cost oe Export revenue 4AX Revenue from domestic sale of by-products THE PARAMETERS a Process inputs (-) or outputs (±) b Number of units of capacity required k Initial capacity d Market requirements pI Domestic price Ov Import price PI Export price Unit transport cost a Nutrient content d Recommended fertilizer use Export bound Second, the mathematical structure of the model of an existing industry is described in terms of the constraints and sets as follows. MATERIAL BALANCE CONSTRAINTS ON FERTILIZERS (8.42) X,a.pjzpj 2 E x.j + E ecil c e CF pe6p jeJ leL LeI MATERIAL BALANCE CONSTRAINTS ON INTERMEDIATE PRODUCTS (8.43) , aCpjzpj + E xcj,j + v.j 2 E xcjj, + E e.ij + w.j P,p i"el il~r ItL iIOdi i'sdi c Ie C1 i e I 108 PLANNING MODELS FOR THE FERTILIZER SECTOR MATERIAL BALANCE CONSTRAINTS ON RAW MATERIALS, MISCELLANEOUS MATERIAL INPUTS, AND LABOR TYPES (8.44) , a,izpi + uei + Vci 2 0 c e CR p6P ieI CAPACITY CONSTRAINTS (8.45) bmpizpi < ki mEM p6P ieI MARKET REQUIREMENTS (8.46) E E acgcxc,j + E a,,'v,j > dc,j je J iel ceCF C6CF C' CQ RECOMMENDED USE CONSTRAINTS (8.47) E x.ij + v,j 2 4i c e CF i6l jeJ EXPORT BOUNDS (8.48) e.il < ec, C e (CI U CF) ieI leL NONNEGATIVITY CONSTRAINTS (8.49) Zpi, xcij, xcij' ecil, vcj, v,j, uaj, wIj > 0 p e P i', ieI jeJ ce C le L OBJECTIVE FUNCTION (8.50) min + A + where DOMESTIC INPUT COST (8.51) = OECR i'I THE STATIC CAPACITY PLANNING MODEL 109 TRANSPORT COST (8.52) X = E ( > ,Lcijxcij + E ucjv>2 + E0 ± >je c6CF ipcii'Xcii' + E tciVci + E E liciecil) c6CI i61 s'61 i61 iel leL + E (E2scivci) c 6CR i6 IMPORT COST (8.53) ½ = E E pjvcj + E E Pvci Ceff CjiJ c6(ClUCR) i6e EXPORT REVENUE (8.54) p6 = E E >p1jecj ce(CFUCI) ie6 teL REVENUE OF BY-PRODUCT SALES (8.55) 4x E E P24WC ceCI i61 The Static Capacity Planning Model Planning models that incorporate the possibility of capacity ex- pansion can also be developed. This section treats the capacity plan- ning problem in a situation in which the element of time is not mod- eled. Such a situation can occur when the planner is asked to draw up an efficient investment plan for capacity expansion in the fertilizer sector for a given year. For example, the planner may be asked in the year 1979 to plan the sector for the year 1985, without having to give details about what happens to the sector during the inter- vening years, nor having to account for prospective developments beyond 1985. Before the appropriate planning model is discussed, however, the difficulties associated with the presence of economies of scale in fertilizer production should be noted. The treatment of economies of scale Economies of scale are exhibited in most activities within the fertilizer sector, especially in the construction cost for new plants. 110 PLANNING MODELS FOR THE FERTILIZER SECTOR Figure 11. Linearization of the Investment Cost Function f(h) Fixed (/ chargei/ (w)~~~~~~~~~~~~~~~~fh Size (hz) In chapter 7, an example of an investment cost function that incor- porates economies of scale was given. Unfortunately, the nonline- arity of that function renders it not directly amenable to treatment in a linear programming model, and a linear approximation needs to be formulated. Figure 11 shows how this linear approximation can be carried out. Algebraically, the linear function can be de- scribed as: (8.56) fK= ± + vh, where fK= approximate capital cost for a productive unit of size h, =fixed-charge portion of investment cost, v= slope of linear portion of investment cost function, h = size of productive unit to be installed. The fixed charge, ct,, ensures that economies of scale in construction cost will be realized, in the sense that, the larger the capacity con- THE STATIC CAPACITY PLANNING MODEL 111 structed, the smaller the incidence of the fixed charge per unit of capacity. The linear approximation to the investment cost function is not, however, without its problems. At small plant scales, the fit may be very bad; in fact, no observations may exist below some minimum capacity level. Also, beyond a given maximum scale, it is necessary to begin to duplicate facilities and, consequently, to incur the fixed charges again. Therefore, the linear approximation must be specified up to the technical maximum capacity. Between these two extremes, the quality of the linear approximation can vary by productive unit. In some cases, it may be necessary to proceed by iteration or piece- wise linear approximation;2 in many cases, however, it can be as- sumed that the linear approximation as shown in figure 11 is close enough to the actual investment cost to yield an estimate of invest- ment cost that is within the margin of error that must normally be considered as admissible in estimating capital cost. When capacity construction is considered in the model, the cost to be incurred for such capacity expansion needs to be included in the objective function. Two complications arise at this point. The first is related to the fixed charge in the capital cost function. The second refers to the fact that, once capacity is constructed, it is avail- able for a certain length of time. If the capital cost associated with capacity expansion were to be entered into the objective function in the formulation given in (8.56), the fixed charge X would always be incurred, regardless of whether or not capacity is constructed, because no mechanism is present in the equation that would delete it from total cost. This is in contrast to the variable charge, which is capacity-dependent, so that vh = 0 whenever h is 0. This problem is solved by introducing a new vari- able, y, into the function in the following manner: (8.57) fK = cy + vh. Here, y is a variable that can assume only two values, either 1 or 0, where I is selected if capacity expansion takes place (so that the fixed charge is incurred), and 0 is selected if no capacity is con- structed (so that the fixed charge is deleted). Hence, (8.58) y = 0 or 1. 2. See Kendrick and Stoutjesdijk, The Planning of Industrial Investment Programs: A Methodology, pp. 43-45. 112 PLANNING MODELS FOR THE FERTILIZER SECTOR The next question is how to select the appropriate value of y. This requires one additional constraint: (8.59) h < hy, where h = upper bound on the size of capacity that can be con- structed; it may be some arbitrary large number or, if available, a technically defined maximum plant size. The effect of (8.59) is obvious. Capacity can only be constructed- that is, h can only be a positive number-if y = 1. If y is placed at the only other value allowed, y = 0, the constraint can only hold if h = 0 as well. The second complication has to do with the fact that capacity, once installed, is available for productive purposes in subsequent years. Even in a static model, this circumstance has to be taken into account, because it affects the way investment cost is charged in the model. Clearly, if the entire investment outlay was charged to the planning year under consideration, it would impose a heavy burden on that year, making it highly unlikely that, in terms of total cost, domestic production would be preferable to imports from abroad. It would appear desirable, therefore, to allocate only a proportion of total investment cost to that year. Based on a device known as the capital recovery factor (CRF), investment cost can be fairly allo- cated over the years during which the new capacity can be expected to be productive. The simplest way to explain the significance of the capital re- covery factor is to imagine that the newly constructed production capacity is available at a constant annual rental charge. The rental charge can be computed by augmenting the original investment cost by the interest charges attributable to the investment capital incorporated in the production capacity, and by dividing the total by the number of years that constitute the expected productive life- time of the plant. The formula for calculating the CRF is the same as that for computing an annuity with present value = 1:3 (8.60) ( + ± p) 01 p 1 - (1 + W)-r 3. A complete derivation of the cRF is given in ibid., appendix to chapter 3. THE STATIC CAPACfIY PLANNING MODEL 113 where am = capital recovery factor for productive unit m, p = discount rate per time interval, tm = useful life of productive unit m. The periodical investment charge on a capacity increase is then obtained as follows: (8.61 ) . = o-( ±y. + vmhm). The structure of the model With the description of the 0-1 variables and the capital recovery factor, the most important elements that need to be incorporated into the planning model of an existing industry to transform that model into a capacity planning model have been discussed. Most of the constraints of the previous model remain unaltered. The changes occur in the capacity constraints, the nonnegativity constraints, and the objective function; in addition, the new model contains capacity expansion constraints, and 0-1 or integer constraints. Until now, the capacity constraint limited the capacity that could be utilized to the amount of the installed capacity. In a planning model that allows capacity expansion, two factors need to be taken into account: first, an increment in capacity is possible; second, during the plan period, a certain amount of capacity may be retired. The new capacity constraint must incorporate both factors. Thus: (8.62) > bmpizi < kmi + hmi - Smi. m e M Capacity use Initial of productive capacity of unit m < productive for all unit m at processes at _plant si.te i _ plant site i [increment toL Capacity For all capacity of retirement of productive + productive _ productive units and unit m ai pnt it all plant sites _plant site i _ _plant site i_alpansie 114 PLANNING MODELS FOR THE FERTILIZER SECTOR Both kmi and smi must be exogenously specified; alternatively, kmi can be given in net terms. As described above, the capacity expansion constraints perform a dual function in the model. First, they force the 0-1 variables y to assume the appropriate value (O or 1). Second, if there are technical limitations on the magnitude of the capacity expansion, h provides a meaningful bound on plant size. Thus, the capacity expansion con- straints have the following form: (8.63) hmi < h,iymi, meM Increment Upper _ F to capacity bound on For all ofproductive < capacity of 0-1 invest- productive unit m at _ productive ment variable units and lntsite unit m at allplant sites plant ste Z _plant site i_ __ _ where (8.64) y 0i = or I. Two additional nonnegativity constraints are required in the model: (8.65) hmi > °eE I meM (8.66) ymi > 0. ieI mME To accommodate investment costs, the objective function of the model needs to be modified substantially. As this modification has been discussed extensively above, only the additional element of the objective function is stated here: (8.67) OK= E ¢ om(&rinYmi + ± mihmi). F _F _ _Total variable Total fixed portion of the Capital Capital charge for capital cost cost _ recovery all productive + for all produc- charges factor units at all tive units _ 2 _ _ plant sites at all plant L J L J L sites THE STATIC CAPACITY PLANNING MODEL 115 Recapitulation The complete static capacity expansion planning model can now be stated. First, the symbols are categorized as the sets and indexes, the parameters, and the variables: Symbol Definition THE SETS AND INDEXES CeC Commodities used or produced in the fertilizer industry CF Fertilizer types CI Intermediate products CQ Nutrients CR Raw materials, miscellaneous material inputs, and labor types i, i'el Plant sites idi Marketing centers leL Export regions meM Productive units peP Processes THE VARIABLES z Process level (production level) x Domestic shipments e Exports v Imports u Domestic purchases w Domestic sales of by-products y 0-1 investment decisions h Capacity expansion 0 Cost groups Domestic raw materials, miscellaneous material inputs, and labor cost q9x Transport cost Capital cost 4,,- Import cost 0. Export revenue ox Revenue from domestic sale of by-products THE PARAMETERS a Process inputs (-) or outputs (+) b Number of units of capacity required k Initial capacity d Market requirements 116 PLANNING MODELS FOR THE FERTILIZER SECTOR Symbol Definition THE PARAMETERS (continued) pd Domestic price pI Import price p6 Export price p Unit transport cost a Nutrient content s Retirements of capacity p Discount rate am Capital recovery factor for productive unit m wo Fixed-charge portion of investment cost v Linear portion of investment cosi Recommended fertilizer use Export bound h Maximum capacity expansion Second, the mathematical structure of the static capacity expansion planning model can be described in terms of the constraints and sets as follows. MATERIAL BALANCE CONSTRAINTS ON FERTILIZERS (8.68) E a,pizpi > E xij + E eril c f CF pep jj IteL ieI MATERIAL BALANCE CONSTRAINTS ON INTERMEDIATE PRODUCTS (8.69) > a0pjzpj + E xGi½i + Vi > E Xii' C E CI pep i'er ii'ol i e I + E e0jj + wcj tEL MATERIAL BALANCE CONSTRAINTS ON RAW MATERIALS, MISCELLANEOUS MATERIAL INPUTS, AND LABOR TYPES (8.70) E acpjz± + u,, + Vej > 0 c c CR pep E CAPACITY CONSTRAINTS (8.71) E bmpizpi < ki + h -mi-Smi meM pEP iel THE STATIC CAPACITY PLANNING MODEL 117 INVESTMENT CONSTRAINTS (8.72) hmi < hmiymi ieI m e M (8.73) yri = 0 or I i e I mME MARKET REQUIREMENTS (8.74) E E rcctxc3j + r a,,'v,j 2 dc'j jeJ ie ceCF ceCF c e CQ RECOMMENDED USE CONSTRAINTS (8.75 ) X cij + vEj 2 dj c e CF tel jEi EXPORT BOUNDS (8.76) e.i1 < ec, c e (CI u CF) iel leL NONNEGATIVITY CONSTRAINTS meM p eP (8.77) zpi, xcij, xcii , eciE, Vci, vc;, uci we hin mi ymi2 0 i', i e I j c- J ce C I e L OBJECTIVE FUNCTION (8.78) mine = O+ + + o + (P - - Ox, where CAPITAL COST CHARGES (8.79) a.L =in(co,miYmt + vmihmi) itE meAl DOMESTIC INPUT COST (8.80) , =pciUc, CeCR iel 118 PLANNING MODELS FOR THE FERTILIZER SECTOR TRANSPORT COST (8.81) x = E ( E cijxij + Ei Acve, + E E cileil ceCF ie1 j6J j6J iel leL + E (E E Acii'xcii' + E' civci + E E ucileciz C~c6 i6I i'61 id i6d 1L ± E (2 E ciVci), IMPORT COST (8.82) 0 = E Zpcvjvcj + E EPcvivci c6CF j6J -6(ClUCR) if EXPORT REVENUE (8.83) = ilecil c6(CFUCI) iJ l6L REVENUE FROM BY-PRODUCT SALES (8.84 ) , E E PCiw.i cCI ie6 The Dynamic Capacity Planning Model Although a static capacity expansion planning model may be the appropriate model in specific circumstances, the element of time will often need to be modeled explicitly in fertilizer sector planning. The presence of economies of scale in fertilizer production, on the one hand, and the likelihood that the demand for fertilizer products is growing over time, on the other hand, result in interdependence that should normally be taken into account in drawing up an efficient investment program for the sector. Introducing the time element into the model To introduce time into the model, the subscript t is added to the variables and parameters of the previous model, where: t E T = the set of time periods specified in the model. Each time period t can consist of one or more years. Since the construction period for new investments in the fertilizer sector is THE DYNAMIC CAPACITY PLANNING MODEL 119 often approximately three years, it is convenient to select time pe- riods of three years each, so that investment decisions in period I result in operational capacity in period 2, and so on. The fact that each variable and parameter now has an additional subscript t means that discount rates must be introduced to render costs and benefit streams comparable over time. This would be relatively simple if the model were specified in annual time periods. Consider, for example, a model that covers a nine-year interval, has annual time periods, and has a cost 4t in each year. The discounted cost for the first year would be: (1 + p)-1 0,; for the second year, (I + p)2 )It; and for the entire period: 9 (8.85) = E (I + ptt4 t=l1 where p = the annual discount rate, t = time index. As was stated above, however, the time periods in the model will usually include more than one year, and that makes the discounting procedure somewhat more difficult. To outline the procedure, as- sume that the model is specified, not with nine annual planning pe- riods, but with three time periods of three years' duration each. The cost charges for each time period would now be: Kic= , T = 1, 2, 3, K2t = , r 4, 5, 6, :K3 = 0j,, r 7, 8, 9, where K, = the average annual cost for each year during time period t. Then, (8.85) can be written: 3 6 9 (8-86) =Ki (I + p)- + K2 E I+ p)+ K3 E (I + p) s, where the first term on the right-hand side of (8.86) is the dis- counted sum of the investment charges during the first three-year period, the second term is the discounted sum of the charges for the second three-year period, and so forth. In its general form, (8.86) can now be rewritten as: 120 PLANNING MODELS FOR THE FERTILIZER SECTOR 3 (8.87) S= Z Kt, t=l1 where 0 at= E (I±+p)-' ,*= +( t- 1) = E (1 + p)-(-l where 0 = the number of years per time period. The dynamic model Most of the constraints of the static planning model remain the same in the dynamic specification of the model, except that the subscript t is added to the variables and parameters. For example, the activity level zi in the dynamic formulation is written as Zit, exports ecit as e,iu, and so forth. In view of the simplicity of this change, the complete dynamic capacity planning model is not re- stated here; it is described in the next chapter. Since the capacity constraints and the objective function undergo a modification that may require some further explanation, however, that change is de- scribed here. The capacity constraints limit the use of any productive unit in the model to the available capacity. In a dynamic model, it must be taken into account that capacity can be constructed during the planning period, so that at a given point in time capacity use is limited to capacity at the beginning of the planning period, aug- mented by new capacity constructed during the planning period. The capacity constraints of the dynamic model have the following form: tE (8.88) , bpizit < kmi + (hi - Smi). meM t e T THE DYNAMIC CAPACrrY PLANNING MODEL 121 Total num- Capacity ber of units Capacity berof capacits Initiadded during For all ofcapacity Initial previous plant sites, used of pro- capacity periods in all productive ductive unit f< oproduc + the planning units, all m for all - tive unit m period minus processes, processes p at plant retirements and all time at plant site i during pre- periods site i in vious years _time period t_ L I L _ L Constraint (8.88) ensures that capacity, once constructed, is car- ried over into the next time periods of the model and is available for productive purposes. Similarly, it ensures that capacity, once retired, is no longer available in later periods in the model. The objective function of the dynamic investment planning model, in summary notation, differs from that of the static model because the discount factor is added. That is: (8.89) YE A (O., + X + + - l - [_ FDiscounteda Capi- Recur- Trans- Sales Total _ sum over tal + rent + port +Import Export of by- costj all time cost cost cost cost revenue prod- L _ _ ~~periods __uct_ where (8.90) += v E ,h(c mr + v ih). reT iel meM 7:5t [Totall1 [ Capital l Fixed +Variable capital = recovery i charge charge I L cost J [ factor JL Constraint (8.90) ensures that the investment charges related to capacity constructed during the previous time periods in the model are charged in any time period t. With these changes and the earlier modifications to the static model, the dynamic investment planning model for the fertilizer sector is fully specified. In the next chapter, a complete description of this model is presented. 9 The Complete Investment Planning Model A COMPREHENSIVE INVESTMENT PLANNING MODEL for the fertilizer sector is presented in this chapter. The model is dynamic, and it permits the explicit specification of economies of scale in the pro- duction of fertilizers and intermediate products. It can incorporate a variety of intermediate and final products, regionally specified ferti- lizer requirements, alternative production processes, and alternative means of transport. Finally, it permits import and export policies for the sector to be analyzed. The most exhaustive use of the model is to determine the least- cost investment, production, and trade strategy for the fertilizer sector, over a given planning period and with specified market re- quirements for fertilizer nutrients. Other uses are also possible. For example, the model can be used to determine the cost implications for the sector of a given investment project or program as compared with alternative project configurations. With the least-cost objective, the values that minimize the total discounted net cost of meeting market requirements can be selected 122 THE SPECIFICATION OF THE MODEL 123 for the following variables:' . the production levels of fertilizer products and the production process used; . the production levels of internediate products and the produc- tion process used; * investments in the sector; . the shipment of final products from plants to markets and of intermediate products among plants; . domestic purchases of raw materials, labor, and miscellaneous inputs; . exports of final and intermediate products within specified bounds; * imports of intermediate products and raw materials; * imports of final products to each marketing center in each time period. The complete model is described in the next section. This is fol- lowed by an examination of the size of the model. Then, alternative specifications of the model to suit particular situations and conditions are given. The chapter concludes with a discussion of linking the fertilizer planning model to an agricultural planning model and to other activities. The Specification of the Model The complete dynamic investment planning model for the ferti- lizer sector is described in this section. First, the symbols used in the model are presented. They are categorized as the sets and in- dexes,2 the parameters (or exogenous variables),3 and the (endoge- nous) variables: 1. Other objectives are, of course, possible; for example, maximizing the rate of return on investments or maximizing the sum of consumers' and producers' surplus. 2. The elements of these sets are explicitly defined in the case study that appears in the second part of this volume. 3. These parameters are specific to both plant location and time. They are plant- location specific to take into account any locational inefficiencies that are unlikely to be remedied during the planning period; they are time specific to take into account any expected improvements in the technology of a particular process at time t (the expectations are those held at the beginning of the planning period). The case study does not take these two considerations into account; hence, the parameters are denoted as a,p and bm., 124 THE COMPLETE INVESTMENT PLANNING MODEL Symbol Definition THE SETS AND INDEXES ceC Commodities used or produced in the fertilizer industry CF Fertilizer types Ci Intermediate products or commodities CQ Nutrients CR Raw materials, miscellaneous inputs, and labor types i, il'I Plant locations or sites jeJ Marketing centers IEL Export regions meM Productive units4 PEP Production processes' r, t, t'eT Time periods THE VARIABLES z Process level (production level) x Domestic shipments e Exports v Imports u Domestic purchases w Domestic sales of by-products y 0-1 investment decision variable h Capacity expansion Total discounted net cost 4. The concept of a productive unit is flexible and depends on the level of dis- aggregation desired. At the aggregate level, a productive unit may be a complete plant or a set of unit operations (for example, crystallization, evaporation, and drying-that is, a subset of the plant). At the micro level, a productive unit may refer to separate pieces of equipment or machinery-that is, a crystallizer or an evaporator or a drier. In such cases, productive units are numbered sequentially; no distinction is made among the processes to which each unit refers. Hence, an evapo- rator in one process will be denoted by m = 1; the same evaporator in another process will be denoted by m = 2. A constraint on disaggregation may be the availability of cost data for each individual unit. 5. A production process refers to the production of final or intermediate commod- ities. Several processes may be available for the production of a single commodity (for example, ammonium nitrate), with each differing in either the inputs used or the unit operations involved. In some cases, the same plant and equipment may be used to produce two final products (for example, monoammonium phosphate and diammonium phosphate); in such cases, the production of each product is referred to as a process. THE SPECIFICATION OF THE MODEL 125 Symbol Definition THE VARIABLES (continued) Cost groups X. Capital cost X0 Domestic recurrent cost (raw materials, miscellaneous inputs, and labor cost) Transport cost Import cost Export revenue Ox Revenue from domestic sale of by-products THE PARAMETERS a Process inputs (-) or outputs (+) b Number of units of capacity used k Initial capacity s Retirements of capacity d Market requirements pd Domestic price P. Export price p, Import price 3 Number of years per time period p Discount rate per year Discount factor per time period Capital recovery factor P (1 + P)~ (I + ( ) 1 where cm = life of productive unit m ,., Fixed-charge portion of investment cost v Linear portion of investment cost A Unit transport cost a Nutrient content d Recommended fertilizer use Export bound h Maximum capacity expansion per time period' The constraints of the model can now be specified, using the nota- tion listed above. A full derivation of the constraints, and a detailed description of their meaning, is given in chapter 8. Here, the con- straints are described in conventional summary notation. 6. If the productive unit m is defined as a complete plant producing a commodity c using a process p, then the upper bound on capacity expansion of the plant would depend upon the particular process used. In this case, hj, is defined as the maximum of the set of such upper bounds. 126 THE COMPLETE INVESTMENT PLANNING MODEL MATERIAL BALANCE CONSTRAINTS ON FINAL PRODUCTS (9.1 ) , acpizpit > Z X,ijt + E e.ij, c e CF p,p jeJ IeL ieI t e T [ Totalproduction of Quantity offertilizer fertilizer c by all c shipped from plant processes p at plant > site i to all market site i during time centers j during time period t _L period t Quantity of fertilizer ] F c exportedfrom plant For all fertilizers, + site i to all export all plant sites, and regions during time all time periods _ period t ___ MATERIAL BALANCE CONSTRAINTS ON INTERMEDIATE PRODUCTS (9.2) Lacpizpit + 22Xci'it [ Total production of Quantity of inter- 1 intermediate product c mediate product c by all processes p + shipped from plant at plant site i during site i' to plant site i time period t L during time periodt t +- Veit > EXcii'e i'tr Quantity of inter- iF Quantity of inter-] mediate product c mediate product c + imported to plant > shippedfrom plant site i during time site i to plant site i' period t _ during time period t - + ± ecjjt + w.it c e CI .,L ieI t eT THE SPECIFICATION OF THE MODEL 127 Quantity of _ intermediate Domestic sales For all in- product c ex- D omsc termediate ported from of by-product s products, all + plan sitei + cat plant site plant sites, ± planzt site ii during time patsts to all export period t and all time regions during periods L timte period t J L L MATERIAL BALANCE CONSTRAINTS ON RAW MATERIALS, MISCELLANEOUS INPUTS, AND LABOR TYPES (9.3) a,piZPit + Ucit + Vcit 2 0 c e CR Pep iel te T Raw materials, labor, Raw materials, labor, or miscellaneous or miscellaneous inputs used in all + inputs domestically processes p at plant purchased at plant i during time site i during time period t _ period t -Raw materials, labor, _ ora matel, labr,o For all raw material [ or miscellaneous and labor, all plant + Liplanptsimpteduringperiod to importedinputs1>0 [ sites, and all time plant site i during peid _ time period t _peid CAPACITY CONSTRAINTS (9.4) ) bm,izpit < kmi + E (hmr - Smir) meM eT~ ~~~e t f- T Capacity of productive Initial capacity of unit m for all processes < productive unit m at p at plant site i _ during time period t plant site i Total capacity Total capacity _ expansion of retiremenit of For all pro- productive productive ductive units, + unit m at unit m at all plant sites, plant site i plant site i and all time L periodtperitup to time L up to time periods _ period t period t_ 128 THE COMPLETE INVESTMENT PLANNING MODEL MARKET REQUIREMENTS c' e CQ (9.5) oecc'Xcijt + E acc,vcjt > dc jt j E J ceCP ied ceCF t e T [ Quantity of nutrient uny t t c' shipped to mar- Qaimpornutr-ent kein cetrjfo. c' imported to mar- keting centeri from + keting center]j dur- all plant sites during ketingtimeperiodtr- time periodt Ing tme peod t Requirement of nu- For all nutrients, trient c' at market- all marketing re- > ing centerj during gions, and all time L time period t _ L periods RECOMMENDED USE CONSTRAINTS (9.6) E Xcijt + Vcj' > de, c e CF jEl t e T [ Quantity offertilizer FQuantity offertilizerl c shipped to market- c imported to mar- ing center j from + keting centerj dur- all plant sites i dur- ing time period t ing time period t _ Recommended use For all fertilizers, >1 offertilizer c at all marketing re- > marketing center] j gions, and all time L during time period t _ periods EXPORT CONSTRAINTS (9.7) E'ecitt < PcIt Ce (CFu CI) i< I~~~~~~eL Quantity of final or in- Maximum For all fer- termediate quantity of tilizers and product c ex- exports of intermediate ported from < product c to products, all all plant sites export region I export areas, i to export re- during time and all time gion I during period t periods L time period t J L THE SPECIFICATION OF THE MODEL 129 CAPACITY EXPANSION CONSTRAINT: MAXIMUM CAPACITY EXPANSION (9.8) hmit < hmitymit m e M i4cI t e T of pro v uUpper bound on Increment to capacity capacity of p- of prdcieui. < ductive unit m at at plant site i dur- plant site i during ing time period t time period t F 1 investment F For all productive variable units, all plant sites, I I I and all time periods _ CAPACITY EXPANSION CONSTRAINT: ZERO-ONE (9.9) ymit = O or 1 meM ie I e T NONNEGATIVITY CONSTRAINTS (9.10) Zit > 0, iel teT peP XjJt > 0, ceCF ieI teT jeJ X¢wt > 0, ceCI ieI teT t I i% Fdi v,it > 0, cc (CI uCR) ieI teT Vyt > 0, ceCF teT jeJ ecilt Ž 0, ce (CFu CI) ieI t eT leL U¢it > 0, ceCR ieI teT hmit > 0, ieI leT meM Wit > 0, ce CI ieI teT 130 THE COMPLETE INVESTMENT PLANNING MODEL OBJECTIVE FUNCTION (9.11) min = E (O. + 4, - Ot + Okt - Oi t- t-Xt) ttT Total capital Total domestic Discounted net Total capital recurrent cost cost time period t during time period t tTotals iF Total Total Total trnpr Tmorta expor hp Ofu J cost during + import cost revenue by-product time period during time during time sales during time period period t period t sales where DISCOUNT FACTOR (9.12) at = E (1 + p)-6(t-l_, CAPITAL COST CHARGES (9.13) i = E E E am (rniryri + Vmirhj,i) Ze T il meMl ,ct = _ _ _ ~~~~~~Totalfixed Total of the vari- Total Capital charge for all able portion of the captalcos recovery productive units capital cost for all capital cost = factor for installed at all + productive untis period t productive plant sites installed at all unit m up to time plant sites up to L L __ period t time period t where (9.14) p(I + p)t_I THE SPECIFICATION OF THE MODEL 131 DOMESTIC INPUT COST (9.15) c't EUi CfCR id F Cost of domestic raw Total domestic recurrent materials, labor, and cost during time period t pumiscellaneous inputs purchased at all plant (9.16) Cot _sites during time period t] TRANSPORT COST (9.16) it=E cijtxciit ctCF ifI jfJ _ _ - Cost of shipping all Total transport cost final products from all during time period t plant sites to all domestic markets _ during time period t + L ucjtv,jt + L E Mciltecile jfj ie. t L / Cost of transporting all final productsin all imported final exported from all products from the + xplttes to the + border to all domestic + plant sites to the morder a doms tic border for all export markets during time regions during time LIperiodt period t + E cii} txciit t + E pitvcit "CfC i{ 'tl ifd = _ ~~~~~~Cost of transporting Cost of all inerplant alimotdner shipments of mdate iprortduintsr + all intermediate + frmedithe borodertosl Prout duridtig tieplant sites during period t ~~time period t + A A.piltecilt) + E citvcit ifl f; ceCR idl Cost of transporting Coto_rasotn all intermediateall imported raw + poutexotd + materials and labor to from all plant sites all plant sites during to all export regionstieprot _during time period t_tieprot 132 THE COMPLETE INVESTMENT PLANNING MODEL IMPORT COST (9.17) >2 E cj'vej + E2 >2i-tVc,t ceCF jeJ eCtCUCR id Cost of im- Cost of im- porting aol porting all intermediate Total import final products products, raw cost during sold in all + materials, and time period t market areas labor used at during time all plant sites period t during time L J L J L period t EXPORT REVENUE (9.18) fpEt = E E E> i eilt ceCFUCI iI IL _ 1F Revenue from the export of all final and inter- Total export revenue mediate products from during time period t all plant sites to all export regions during time period t REVENUE FROM BY-PRODUCT SALES (9.19) I= 9 Epcwit cCI iel 1 Revenue from domestic Total revenue from by-product sales of by-product sales during = commodity c from all time period t plant sites during time L _ period t The Size of the Model The size of the model cannot be indicated by the mathematical specification of the model; rather, the size depends on the specifica- tion of the sets. The sets to be specified are: commodities, C;' plant 7. The commodities include final and intermediate products, raw materials, and labor types. THE SIZE OF THE MODEL 133 sites, I; domestic marketing regions, J; export market areas, L; nutrients, CQ; productive units, M; production processes, P; and time periods, T. Clearly, the potential size of the fertilizer planning modql is very large. Normally, more than one fertilizer type has to be considered, and more than one plant site may be a plausible candidate for pro- duction facilities. To capture the trade-off between economies of scale, on the one hand, and transport cost, on the other, it is desir- able to break down the entire market for any plant site into a num- ber of domestic marketing regions and export market areas. More- over, although the various nutrients are not directly competitive, the existence of multinutrient fertilizers makes it necessary to con- sider different nutrients simultaneously. Similarly, it may often be necessary to specify various productive units and processes. Finally, to investigate the desirability of constructing temporary overca- pacity, so that the advantages of economies of scale can be exploited more fully later, it is necessary to specify several time periods in the model. Although disaggregation may be desirable, large-scale planning models have a distinct disadvantage: they become extremely expen- sive to solve, even on modern, high-speed computers. This is par- ticularly the case for models that incorporate economies of scale in the form of 0-1 variables. In the absence of such 0-1 variables, the planning models in this volume reduce to linear programming models. Large models of this type-3,000 to 5,000 constraints-are now routinely solved on large-scale computers, although even they pose serious problems if their structure is complex. Mixed-integer linear programming models of that size can only be solved at reason- able cost if the number of 0-1 variables is limited to a handful, say, 20 to 30. This number is easily reached: one only needs to specify 2 time periods, 5 products, and 2 locations before the number of 0-1 variables reaches 20. Why does the number of 0-1 variables play such an important role? A more complete explanation of this aspect of the planning problem is given in volume I of this series;' here, only an intuitive 8. David Kendrick and Ardy Stoutjesdijk, The Planning of Industrial Investment Programs: A Methodology, The Planning of Investment Programs, ed. Alexander Meeraus and Stoutjesdijk, vol. 1 (Baltimore: Johns Hopkins University Press, 1978), chapter 8. 134 THE COMPLETE INVESTMENT PLANNING MODEL explanation is given. Consider, for example, a project planning problem in which the plant site for the production of a given product is to be selected. There are two potential sites, A and B, and four possibilities: plants in both A and B; a plant in A but not in B; a plant in B but not in A; and no plant at all (only imports). Thus, although there are only two 0-1 variables, there are four possible combinations-that is, 22. This example can be generalized to the statement that the number of project combinations in a 0-1 invest- ment problem is equal to 2', where n is the number of 0-1 variables. If n = 20, the number of project combinations is more than 1 mil- lion! And each project combination constitutes a separate linear pro- gramming problem. Fortunately, procedures have been developed so that it is not necessary to solve each individual linear program- ming problem; in fact, only a fraction of them is ever solved. These procedures, however, are still sufficiently cumbersome to encourage the user to limit the size of the planning model as much as possible, by carefully specifying the sets and by deleting any decision problems that can be solved or approximated at the outset. Alternative Specifications of the Constraints The constraints of the model are specified with sufficient generality to encompass most of the situations that may be considered in plan- ning the fertilizer sector. Some situations may, however, be unique to the specific problem under consideration, and this section ad- dresses such cases. In particular, the constraints discussed here are export constraints, investment constraints, domestic labor con- straints, foreign exchange constraints, import constraints, and re- gional preference constraints. This section is not intended to be ex- haustive, and other constraints may be relevant to a particular problem. Export constraints The export constraint (9.7) places an upper bound on the export of a given product to a given export market for each time period, for all producing sites taken together. This may not always be the appropriate way to specify the export constraint. Alternative speci- fications of the export constraint are given below. ALTERNATIVE SPECIFICATIONS OF THE CONSTRAINTS 135 TOTAL BY EXPORT MARKET (9.20) E E ealt < ejt IEL cfCFUCI iJ t e T Quantity of all final and intermediate Maximum quantity of commodities exported < exports for region from all plant sites to < during time period t export region I during time period I This constraint limits total exports from all producing locations taken together to each export region during any given time period, without specifying the constraint for each product separately. As a result, the use of constraint (9.20) involves a smaller total number of constraints than (9.7). TOTAL BY COMMODITY (9.21) eciul< t c eCFUCI ieI leL teT Quantity of final or intermediate product c Maximum quantity of exported from all < exports of final or plant sites to all < intermediate product export regions during c during time period t time period t It may be desirable to focus on specific commodities rather than on export markets. Constraint (9.21 ) restricts the total amount of com- modity c that can be exported from all plant sites to all export re- gions during any given time period. It differs from constraint (9.20), which placed an upper bound on the exports to any particular export region. As before, substituting constraint (9.21) for (9.7) decreases the total number of constraints in the model. TOTAL BY COMMODITY, LOCATION, AND EXPORT REGION (9.22) ecilt < ccilCe CF U CI iEI I e L t - T [ Quantity of final or - -Maximum quantity of 1 intermediate product commodity c from c exported from plant < plant site i to export site i to export region region I during time I during time period t - period t 136 THE COMPLETE INVESTMENT PLANNING MODEL This strong constraint restricts the export of a commodity from each producing location to each export region in each time period. The use of constraint (9.22) instead of (9.7) increases the total number of constraints in the model. AGGREGATE EXPORT CONSTRAINTS Constraint (9.22) specifies a restriction on the amount of exports for each product, from each plant site, to each export market, and during each time period. It is conceivable that, even though the planner considers this specification the appropriate one, he is in- terested in knowing which export market, which product, which plant site, or which time period is the most attractive, if a choice has to be made. The simplest way to make such a determination is to formulate an additional constraint which specifies that the arith- metic sum of the individual bounds is larger than a given number. For example, assume that the individual bounds of constraint (9.22) are placed at 10,000 tons, and that there are 2 products, 2 plant sites, 2 export markets, and 2 time periods. This means that there are 16 individual constraints within the constraint set (9.22); in principle, exports could thus total 160,000 tons. If an additional constraint were placed on this total so that exports could amount only to, say, 100,000 tons, the model would be forced to choose product, plant site, export market, and time period for which ex- ports make the greatest contribution to the objective of the model. The combination of constraint (9.22) with (9.20) would focus on the choice among products and plant sites. A similar purpose would be achieved by restricting the value of Jet: (9.23) elt < I EL ceCFUCI id t e T _ _ Maximum quantity of exports of allnfial Maximum quantity of and intermediate all exports to region I < products from all during time period t plant sites to export region I during time L iL period t Similarly, if the planner wants to focus on the most attractive export market or plant site for exports, he can add a constraint of ALTERNATIVE SPECIFICATIONS OF THE CONSTRAINTS 137 the following form to (9.21): (9.24) e < E ecilt. c eCF UCI iel 6eL t e T _ _ Maximum quantity of aximum quantity of exporis ofmfinal or exports offinal or < intermediate product c intermediate product c from all plant sites to during time period t all export regions _durinmeerodt _ lduring time period t_ Investment constraints The model specified above does not take into account constraints on investment funds. But there may be situations under which the shortage of capital poses a real constraint, and the optimal program must take this into consideration. IN EACH TIME PERIOD (9.25) 0.t < tX J tT Maximum amount of Total capital cost dur- < capital available for ing time period t _ debt servicing during time period t where +ba is defined as in (9.13). This constraint restricts the maxi- mum amount of funds available for debt servicing in each time period, &,, which is given exogenously. OVER THE PLANNING HORIZON (9.26) E at6ct < $ teT r Maximum amount o Present value of total < total capital available capital cost over the relevant planning horizon where at is defined as in (9.12). Here, constraint (9.25) is modified to take into account the entire planning period, and the constraint on capital expenditures is specified as an upper limit that can be incurred over the planning horizon (q3,). This figure (given exoge- nously) is usually specified at the beginning of the planning period 138 THE COMPLETE INVESTMENT PLANNING MODEL in current units of currency; hence, the left-hand side of the con- straint incorporates the discount factor. AGGREGATE CAPITAL CONSTRAINTS If the planner wishes to investigate in which period it is most attractive to begin to incur periodic capital charges, he can combine constraint (9.25) with a constraint on the value of 0: (9.27) < E X.K- teT Maximum amount of T he sum of the maxi- total capital available < mum amount of capi- over the relevant < tal available in each planning horizon L time period This will force the model to choose between various time periods. Domestic labor constraints The labor requirements of the chemical fertilizer industry are insignificant, as compared with the capital requirements. There may, however, be constraints on highly skilled labor such as managers or plant operators; in particular, these constraints may be location- and time-specific. In other words, if a plant is set up away from urban centers, it may be difficult to obtain the necessary number of managers and plant operators to run the plant at that location during that time period. Such constraints can be introduced into the model. The elasticity of supply of labor is assumed to be infinite during each time period for each location up to the constraint. The salary or wage rate is thus assumed to be location- and time-specific as before-that is, p' . If the set of labor or skill types is defined by CO, and if subsets are specified: CO = COI, C02, C03, . . . , it is possible to specify skill-specific bounds. For example: COI = subset of managers, C02 = subset of plant operators, C03 = subset of skilled laborers, and so on. This breakdown of labor by type of skill is intended merely to be illustrative; clearly, there is no limit to the extent of disaggregation, except perhaps data availability. ALTERNATIVE SPECIFICATIONS OF THE CONSTRAINTS 139 The general constraint can be formulated as: (9.28) Uat < ffcit. i E I C e CO te T Maximum number of 1 L Total number of workers of skill class c workers of skill class c < availle at plant at plant site i during < site i during time time period t period ti Generally, these constraints are applicable to the more highly skilled class of workers rather than the unskilled; that is, these constraints may only be relevant for managers and plant operators rather than unskilled labor. Further, equation (9.28) assumes no economies of scale in labor inputs. Foreign exchange constraints The model specified above does not consider the possibility of foreign exchange shortage that may be faced by a country and that may thus impose a restriction on the fertilizer sector. It is assumed that this restriction is exogenously imposed by the central planning authority and is hence taken as given by the planner; its value is denoted by the symbol G. The foreign exchange component of the total cost manifests itself as some proportion ('y,) of the capital costs (0K,) in time period t and all of the import costs less export revenues in that time period. The proportion y, of capital costs is made time-specific to permit the ratio to alter over time. An approximate value of this ratio is 2:3 for less developed countries, but more accurate values of this frac- tion can be calculated for each time period. The constraints are given by time periods and over the planning horizon. IN EACH TIME PERIOD (9.29) -yt+Kt + OXt- Oa < G3 t e T Foreign i Maximum exchange Export amount of component revenue foreign of capital + cost dur- during < excihange cost during itg time time available pei period t p dridng time L d_ _ L perod t period t 140 THE COMPLETE INVESTMENT PLANNING MODEL OVER THE PLANNING HORIZON (9.30) E t('yttt + - 4a) < G teT [ Present value of the M a _ foeg.ecag Maximum amount oj' foreign exchange o foreign exchange component of the 1< available over the capital cost and net aalbeoe h aimport cost a entire planning period NONBINDING (9.31) G< teT Maximum amount of Sum of the maximum foreign exchange < amount of foreign available over the exchange available in entire planning period] each time period Constraint (9.31) along with constraints (9.29) and (9.30) ensure that there are nonbinding constraints in some of the time periods. Import constraints Import constraints can be added to the system of equations if the central planning authority wishes to restrict the amount of imports and thereby follows a policy of partial or complete import substitu- tion in the fertilizer sector. Constraints can also be imposed on the imports of intermediate products and raw materials, particularly if the central authority wishes to emphasize self-sufficiency in the production of intermediate products and complete exploitation of the locally available raw materials. Hence, the following constraints may be needed to test the effects of import licensing. FINAL PRODUCTS: TOTAL BY COMMODITY AND MARKET LOCATION (9.32) V.jt < pi.t C E CF jEJ t e T _ Maximum allowable Quantity of imports i.mports of final of final product c at < product c to market- marketing center j _ ing center j during during time period ttieprot _ _ _ tlme penod tt~ ALTERNATIVE SPECIFICATIONS OF THE CONSTRAINTS 141 FINAL PRODUCTS: TOTAL BY MARKETING CENTER (9.33) E Vcjt < Vjt jEJ ceCF teT Quantity of allfinal [Maximum quantitynof products imported to Imports of allfinal marketintg center j < products to marketing during time period t centerj during time duringtimeperiod] _period t FINAL PRODUCTS: TOTAL BY COMMODITY (9.34) E VCjt < it C e CF j 1, iEIN m0M t,T where IN, the set of chosen plant sites, is a subset of I. According to ALTERNATIVE FORMULATIONS OF THE OBJECTIVE FUNCTION 143 this constraint, at least one plant must be built in the chosen location IN during any time period. In this case, the temporal dimension is not important to the central authority, so long as at least one plant is erected in the desired location. As an example, consider the case of 2 plants and 3 time periods- that is M = 2, T = 3; then, constraint (9.40) can be written as: 2 3 (9.41) E ymit 1, i E IN m=l t=l that is, Ylil + YlU2 + ysi3 + Y2il + Y2i2 + Y2i3 2 1. i e IN Assume further that there is only one desired location-that is, IN = 1. If more than one location is desired, there will be an equation like (9.41 ) for each location. Since the y variables can take on only 0 or I values, equation (9.41 ) permits all the y's to take the value 1, but it insists that at least one y be equal to 1. AT MOST ONE PLANT Analogous to constraint (9.40) is one that can reflect the central authority's preference to consider at most one potential location. This can be expressed by: (9.42) E EYmit < 1. ielN M,M teT Such a situation may arise if the central authority believes that, in the interest of regional balance or income distribution, location k should be avoided as a potential plant site if possible. In this case, the central authority exhibits some flexibility, in the sense that the location k should be avoided only if it is possible to do so; if it is not, then only one plant may be built there. If the authority has definite negative feelings about such a location, then the constraint can be rewritten as: (9.43) Z.ivt = o. ieIN mnM teT Alternative Formulations of the Objective Function In theory, it is possible to modify the objective function of the model stated in equations (9.11) through (9.19) from a cost-mini- 144 THE COMPLETE INVESTMENT PLANNING MODEL mizing objective to one that either maximizes profits or maximizes the sum of producers' and consumers' surplus. These latter two objectives require that the model be specified in terms of demand that is responsive to price rather than fixed demands. Although both methods of specification have been employed in linear programming models, there is as yet no computational experience with the use of investment planning models involving 0-I variables.9 Linking the Fertilizer Sector Model to an Agricultural Model The principal limitation of the sector planning methodology pre- sented in this volume is its partial equilibrium nature; that is, the environment in which the fertilizer sector is supposed to function is exogenously specified, even though significant interdependencies with other sectors-primarily, the agricultural sector-are likely to exist. Clearly, it would be of considerable importance to relate the fertilizer sector model formally to a model of the agricultural sector. From the perspective of the fertilizer sector model, such a linkage would imply that the demand for fertilizer material or nutrients is no longer given; in turn, the agricultural model would not need to be specified with given supply prices for fertilizer types. In theory, the link between the two models is not difficult to con- struct. The objective function of the fertilizer model is simply de- leted, and the market requirements constraint is replaced by a model of the agricultural sector that generates the market demand for fertilizer types on the basis of the relative contribution each fertilizer type makes to the objective function of the agricultural model. This requires specifying agricultural production technologies that ex- plicitly incorporate fertilizer use, and that show the physical yield from the use of each fertilizer on each crop in each region; such response schedules of physical yield are required for each combina- tion of agricultural inputs that is deemed appropriate. In practice, linking the two models may run into formidable 9. For a description of the mathematical structure of these alternative formula- tions, as well as a discussion of their computational problems, see Kendrick and Stoutjesdijk, The Planning of Industrial Investmenit Programs: A Methodology, chapter 7. LINKS WITH OTHER ACTIVITIES 145 difficulties. First, the data requirements are great; in particular, it may be impossible to specify the set of agricultural production tech- nologies accurately. Second, even if it were possible to obtain all the information required, it is far from certain that this system of models could be solved, even on modern, high-speed computers. It is diffi- cult enough to solve the fertilizer model alone; to link this model to a detailed agricultural sector model, in the hope of obtaining a simultaneous solution to the entire system, is probably outside the realm of current technical possibilities. For the time being, therefore, it would appear that less formal links between the two sector models have to be attempted; the most promising of these may be to adopt an iterative procedure. First, the fertilizer sector model is solved on the basis of varying sets of market requirements by region in terms of nutrients. This gives a supply price schedule for fertilizer nutrients at a range of output levels. Next, the agricultural sector model is solved on the basis of a selection of such supply prices for nutrients by region, resulting in a demand schedule for fertilizers. Provided the marginal product of fertilizer use is decreasing, or the production cost of fertilizers is falling at a declining rate, this procedure must eventually permit a market equilibrium to be identified. The procedure may be extremely time-consuming, however, and therefore costly. Links with Other Activities Apart from the agricultural sector, the fertilizer sector may have technological relations with other sectors in the economy. For ex- ample, the fertilizer sector requires intermediate products such as sulfuric acid that are extensively used in other industrial activities. Thus, the investment program should be designed so that it covers the possibility of either sales or purchases of intermediate products to or from other sectors. Such intersectoral transactions can apply to fertilizer products as well as intermediate products; for example, ammonium nitrate is used as an explosive in mining operations. Sometimes the fertilizer industry produces by-products that are of value in other sectors; for example, phosphate rock can have a high iron content that needs to be lowered before the rock can be used in fertilizer manufacture. The resulting by-product-magnetite-can be sold to a local iron and steel industry. In turn, other sectors may produce by-products that are important to the fertilizer sector, either 146 THE COMPLETE INVESTMENT PLANNING MODEL as intermediate or as final products. An example of the former is carbon dioxide, which is important in the manufacture of urea; an example of the latter is basic slag, a by-product of the iron and steel industry and a phosphatic fertilizer in its own right (Thomas slag). The formal modeling of these intersectoral links is not difficult as long as demand (in the case of sales from the fertilizer sector) and supply (in the case of purchases by the fertilizer sector) can be exogenously specified. Moreover, except in the case of sales of final and intermediate products from the fertilizer sector, prices have to be stated explicitly. If these conditions are not fulfilled, the same linkage problems are encountered as in the case of the agricultural sector. To introduce these links with other sectors into the mathematical structure of the fertilizer model, the following notation is introduced: w = domestic sales of by-products, and of final products for nonagricultural use; u = domestic purchases of by-products, or of fertilizers pro- duced outside the fertilizer sector itself. 0,t = total cost of purchases from outside the fertilizer industry during time period t. With this notation, several constraints and the objective function of the model can be modified as follows. MATERIAL BALANCE CONSTRAINTS ON FERTILIZERS (9.44) E acpizpit + ucit > E xeijt + E ecilt + W.it ce CF pep jeJ eL ijtl t e T Total production Domestic pur- Quntity of offertilizer chases of ferti- fertilizer c product c from lizer c from shipped to all all processes at + outsde the > market areas I .plant site i fertilizer indus- from Plant during time Ztry via plant site i during during time site i during time perioing L period t time period t time period t 10. Fertilizer produced as a by-product by another industrial sector is not neces- sarily produced at one of the plant sites specified for fertilizer production. Instead of introducing new sites, it can be assumed that such fertilizer is shipped to the consuming region via the nearest fertilizer production site. LINKS WITH OTHER ACTIVITIES 147 F Quantity of ex-Quantity of Quat 1t of fertilizer c sold fertilizer c ex- outside the For all ferti- ported from agricultural lizers, all plant plant site i to + sector from sites, and all all export plant site i time periods regions during during time time period t _ period t J L MATERIAL BALANCE CONSTRAINTS ON INTERMEDIATE PRODUCTS (9.45) E acDizpjt + E Xwit pep ~~i'ed i'5# i Total production of [ Quantity of inter- intermediate product c mediate product c from all processes at + shippedfrom plant plant site i during site i' to plant site i time period t _ during time period t _ + mcat + Ueit > E Xcwt i'd i'SS Domestic pur- chases of inter- Quantity of Quantity Of mediate product intermediate intermediate product c + product c im- + the frrtltizer > shipped from ported to plant idth ryti plant site i to inidustry, atplnsie' site i durintg plant site i, pdant site i time period t during time perigod t _ periodtpeidt _ + e, ilt + Wcjt C e CI 1lel ieI te 1T Quantity of Quantity of intermediate intermediate product c For all inter- product c ex- domestically mediate prod- + ported from + sold outside ucts, all plant plant site i the fertilizer sites, and all to all export sector from time periods regions during plant site i time period t during time L L period t L 148 THE COMPLETE INVESTMENT PLANNING MODEL NONNEGATIVITY CONSTRAINTS Add to (9. 10): (9.46) wciŽ > 0 ce(CFuCI) iEI teT Ucit Ž0 ce(CFU CI) ieI teT OBJECTIVE FUNCTION (44) min = tt + oo + Oxt + Ort + C1t - O-a xt), teT _ 1 Total Total Total Discouwted capital recurrent transport net cost = cost during + cost during + cost during time time time L L _ period t Lperiod t L periodt t Total cost of pur- Total chases revenue morta from out- Tot ex- from sales + cost dur- + side the enue dur- outside the ing time fertilizer ing fertilizer period t industry lng t sector dur- period t duin period t' igtm during ing timne time period t L J _Lperiodt I L J _ where '>rat = E EpditUCit, ce(CFUCI) ie Total cost of purchases from Cost of purchases from outside outside the _ the fertilizer sector offertilizers fertilizer industry and intermediate products during during time time period t period t and tkxt e l PciWcii* ce(CFUCI) iJ _ _ ~~~~JRevenue from sales offJinal and- Total revenue intermediate products of the from by-product fertilizer industry for use in sales during nonagricultural sectors during time period t _ time period t CONCLUSION 149 Conclusion This presentation of the fertilizer sector model and the model's possible variants has not been exhaustive; rather, it was attempted to demonstrate that the basic model can be modified to suit many different situations. PART TWO The Egyptian Fertilizer Sector: A Case Study 10 The Fertilizer Sector in Egypt THE PRESENT STRUCTURE AND DEVELOPMENT POTENTIAL of the Egyptian fertilizer sector is evaluated in this case study, using several plan- ning models. The primary purpose of the study is to demonstrate that these planning models can be useful analytical tools, and to show that the results described have considerable operational rele- vance. An attempt has been made, in collaboration with Egyptian sector specialists, to obtain a reasonably accurate statistical repre- sentation of the fertilizer sector at the present time. Moreover, the sector's development potential is considered in the light of forecasts of demand and prices consistent with those made by Egyptian plan- ners and fertilizer sector specialists.' The written account of the case study does not explicitly address the difficulties associated with solving the planning models. Mixed- integer programming models are usually difficult to solve; they re- quire sophisticated analysts, as well as computer software and hard- 1. Acknowledgements in connection with the case study are incorporated in the preface to the volume. 153 154 THE FERTILIZER SECTOR IN EGYPT ware. The User's Guide in this series is designed to address this aspect of the proposed planning methodology. The case study is organized in the following way. First, a concise description of the fertilizer sector and its role in Egypt is given. Next, the present structure of the sector-including the existing production pattern of the various fertilizer plants in the country, as well as the transporting and importing of raw materials, intermediate products, and fertilizer materials-is analyzed. This analysis is carried out against the background of regionally specified fertilizer use. The third chapter in the case study is perhaps the most important. Given that fertilizer use will continue to increase from present levels, and that existing production capacity is not sufficient to meet even cur- rent demand, Egyptian planners are faced with the question: What is the best policy for Egypt to adopt in order to meet future fertilizer demand requirements? Would it be preferable to import fertilizers, to produce them domestically, or both? If some fertilizers are to be imported, which ones should be imported? Which locations should be preferred for fertilizer production? At what scale should fertilizer production take place? Which feedstocks should be used? What is the least-cost transport pattern for both imported and domestically produced fertilizers? Should intermediate products be shipped be- tween plants? A dynamic, mixed-integer planning model of the Egyptian fertilizer industry is used to address these questions, from the point of view of the sectoral planner.2 In the final chapter of the study, a number of policy experiments are described, to illustrate the impact of various government policies in the areas of fertilizer pro- duction, importation, distribution, and pricing on the structure of the fertilizer sector in Egypt. The Demand for Fertilizers The construction of the Aswan High Dam effectively marked the end of Egypt's traditional form of irrigation-that is, the seasonal flooding of the Nile. The immediate effect was a significant increase in the demand for chemical fertilizer. Since the turn of the century, 2. This qualification is quite important, since the type of models discussed in this volume are not primarily designed to analyze investments from the point of view of the individual enterprise. THE DEMAND FOR FERTILIZERS 155 Table 1. Actual Use of Fertilizer Nutrients in Egypt, 1951-52 to 1974-75 (thousands of metric tons of nutrients) Total Nitrogenous Phosphatic Potassic fertilizer Year fertilizers fertilizers fertilizers nutrients used 1951-52 110 27 0.0 137.0 1955-56 114 21 0.0 135.0 1956-57 102 24 0.0 126.0 1957-58 132 27 0.6 159.6 1958-59 137 27 0.8 164.8 1959-60 172 33 1.9 206.9 1960-61 181 34 2.0 217.0 1961-62 186 38 0.9 224.9 1962-63 204 39 1.0 244.0 1963-64 227 43 1.0 271.0 1964-65 253 45 0.6 298.6 1965-66 280 53 0.4 333.4 1966-67 264 44 0.6 308.6 1967-68 259 38 1.5 298.5 1968-69 275 51 1.4 327.4 1969-70 330 57 1.4 388.4 1970-71 299 56 1.8 356.8 1971-72 327 66 1.5 394.5 1972-73 372 65 3.2 440.2 1973-74 n.a. n.a. n.a. n.a. 1974-75 403 74 n.a. n.a. n.a. = data not available. Source: Compiled from unpublished data provided by the General Organization of Agricultural and Cooperative Credit. Egypt has been importing fertilizer from Chile in the form of natural sodium nitrate. By 1960, the consumption of fertilizer per acre of cultivated land surpassed all developing countries except Taiwan and South Korea. In 1972-73, by far the greatest demand was for nitrogenous fertilizers (84.5 percent of total nutrient demand), followed by phosphatic fertilizers (14.8 percent of total demand); potassic fertilizers represented only 0.7 percent of total demand. Table 1 shows how fertilizer demand has changed over time. Nitrogenous fertilizers are applied to almost all crops in Egypt; about 90 percent of these fertilizers are used for rice, wheat, maize, cotton, millet, and vegetables. Phosphatic fertilizers have been used 156 THE FERTILIZER SECTOR IN EGYPT in Egyptian agriculture since the early 1930s. They are used mainly for three crops: berseem (Egyptian clover), which accounts for about 33 percent of total consumption; cotton, 32 percent; and rice, 24 percent. The remainder is used for other crops such as sugarcane, beans, sesame, onions, flux, and ground nuts. Potassic fertilizers are used only for fruits, potatoes, and vegetables. Total consumption of potassic fertilizers is very small compared with the other two types, and it tends to fluctuate from year to year; since 1972-73, consump- tion has averaged around 3,000 metric tons a year. Fertilizer application rates The fertilizer application rate is the quantity of fertilizers applied to a unit of land being cultivated with a certain crop. In Egypt, three different rates are identified: actual, assigned or recommended, and economic. The actual rate represents the quantity of fertilizer actually applied by the farmer. The major variables determining this rate are the individual farmer's intuitive judgment, his financial position, and the availability of the appropriate fertilizer. Table 2 shows how this actual rate has changed over time. The general trend is toward in- creasing rates of application; from 1966 to 1968 and from 1970 to 1971, however, the absolute level of fertilizers used per feddan declined. 3 The assigned or recommended rate is drawn up by the Ministry of Agriculture for different crops in the different regions of the country. These rates are based on experimental studies undertaken by the ministry's soil chemists, and they take into consideration the var- ious climatic, regional, and soil conditions for different crops. They were put into effect in 1960-61 and are subject to annual revisions, which are determined in great part by the total supply of fertilizers. The present trend is toward increasing the recommended rates.4 Nitrogenous fertilizers have recommended rates that differ by geo- graphical region for several crops, including wheat, maize, millet, cotton, onions, and sugarcane. For other crops, the rates differ according to the responsiveness of different regions to nitrogenous fertilizer. Phosphatic fertilizers have recommended rates that differ according to the crop rather than the region, whereas potassic ferti- lizers do not have any recommended rates. 3. A feddan is equal to 1.038 acres. 4. All fertilizer distribution is based on these rates. THE DEMAND FOR FERTILIZERS 157 Table 2. Fertilizer Application Rate per Cropped Feddan in Egypt, 1955-56 to 1971-72 Total fertilizer Total Fertilizer nutrients used, cropped area nutrients used per (thousands of (thousands of cropped feddan Year metric tons) feddans) (kilograms) 1955-56 135.0 9,962 13.6 1956-57 126.0 10,100 12.5 1957-58 159.6 10,089 15.8 1958-59 164.8 10,289 16.0 1959-60 206.9 10,370 20.0 1960-61 217.0 9,974 21.8 1961-62 224.9 10,366 21.7 1962-63 244.0 10,358 23.6 1963-64 271.0 10,377 26.1 1964-65 298.6 10,261 29.1 1965-66 333.4 10,487 31.8 1966-67 308.6 10,462 29.5 1967-68 298.5 10,742 27.8 1968-69 327.4 10,732 30.5 1969-70 388.4 10,733 36.2 1970-71 356.8 10,700 33.3 1971-72 394.5 10,799 36.5 a. Nitrogen, phosphorus, and potassium. The Institute of National Planning has been working on the con- cept of an economic rate of fertilizer use, which takes into account the economic as well as the technological aspects of fertilizer ap- plication. In general, this rate is higher than the recommended one. The projected demand for fertilizers The objective of the national agricultural policy is to increase agricultural production by 1982 to 150 percent of the 1972 produc- tion, primarily through improvements in agricultural technology, including further increases in fertilizer use. Correspondingly, a new set of planned fertilizer application rates has been announced. These rates, which are based upon several considerations, recommend different levels of application for various crops. It is intended that these targets will be attained by 1980 and will remain reasonably 158 THE FERTILIZER SECTOR IN EGYPT stable thereafter. The Ministry of Agriculture has projected the de- mand for fertilizers on the basis of these planned fertilizing rates and the planned or projected cropping pattern. The cropping pattern in Egypt is determined by several factors, of which the most important are available land and water. The pat- tern that has emerged is the triennial rotation system-that is, a three-year rotation of crops. The crops are grown in a sequence according to the seasons. There are three basic seasons: the winter season, when crops are cultivated during October and November and are harvested during April and May; the summer season, when crops are cultivated from March to June and are harvested from August to October; and the Nili season, when crops are cultivated from the second half of June to the end of July and are harvested from October to December. The importance of the Nili season has de- clined, however, because of the completion of the Aswan High Dam, which has provided more water for the summer season. Table 3. Estimated and Projected Use of Nitrogenous Fertilizer Nutrients for Various Crops, 1973-85 (thousands of metric tons) Crop 1973 1974 1975 1976 1977 1982 1985 Barley 3.5 3.2 3.5 4.0 4.0 6.5 6.9 Cotton 69.3 69.3 74.3 74.3 79.2 82.5 80.0 Flax 2.6 2.5 2.7 2.8 3.0 2.9 2.5 Fruits 22.0 23.0 24.8 26.0 27.5 32.3 34.9 Legumes 3.9 4.0 4.4 4.5 4.8 5.5 5.7 Maize 86.8 85.3 94.3 91.7 98.6 116.3 120.0 Millet 23.1 21.9 22.6 22.2 23.6 33.5 34.4 Oilseeds 3.0 3.2 3.7 4.1 4.6 5.5 5.7 Onions 5.0 5.1 6.0 6.1 6.2 7.0 8.0 Rice 40.5 40.5 43.1 43.2 45.7 48.3 50.9 Sugarcane 22.9 24.1 25.7 26.7 29.5 30.4 31.9 Vegetables 37.9 38.9 39.8 41.2 41.9 53.1 54.9 Wheat 54.7 54.8 59.1 59.1 64.0 75.0 76.2 Others 5.5 4.8 4.9 4.8 5.0 7.3 1.5 Total 380.7 380.6 408.9 410.2 437.6 506.1 513.5 Note: Years refer to cropping years. Source: Ministry of Agriculture, quoted in R. Hassan and others, An Economic Study of Problems and Prospects of Fertilizers' Utilization, Production, and Marketing in the Arab Republic of Egypt (Cairo: Institute of National Planning, 1973). THE DEMAND FOR FERTILIZERS 159 Table 4. Estimated and Projected Use of Phosphatic Fertilizer Nutrients for Various Crops, 1973-85 (thousands of metric tons) Crop 1973 1974 1975 1976 1977 1982 1985 Cotton 24.8 24.8 24.8 37.1 37.1 37.1 36.0 Flax 0.7 0.7 0.7 1.1 1.1 1.1 0.9 Fodder 48.4 48.4 48.2 71.6 71.0 72.5 72.4 Fruits 4.5 4.7 9.8 10.2 10.6 12.3 13.3 Legumes 7.2 7.5 7.7 7.9 8.0 12.4 12.8 Oilseeds 2.7 2.8 4.5 5.0 5.2 7.8 8.1 Onions 1.0 1.0 1.1 1.2 1.2 1.3 1.5 Rice 18.4 18.4 18.5 27.7 27.8 28.6 30.2 Vegetables 12.2 12.6 12.8 26.6 27.1 34.3 35.4 Others 1.7 1.5 1.4 2.7 2.6 3.6 0.8 Total 121.6 122.4 129.5 191.1 191.7 211.0 211.4 Note: Years refer to cropping years. Source: Ministry of Agriculture, quoted in Hassan and others, An Economic Study of Problems and Prospects. The cropping pattern is also related to the area upon which the demand projections are based. Because of the increase of nonagri- cultural uses of land, the total agricultural land has declined by 60,000 feddans over the past few years. Legislative measures have been instituted to prevent further loss. Moreover, land reclamation programs have been instituted to increase the area of new land to be put under cultivation. The planned increase is from about 480,000 feddans in 1972 to about 668,000 by 1977. Based on the above considerations, the ministry has projected the demand for fertilizers for the period 1976-85. These projections are presented in tables 3, 4, and 5. Over the period 1973-85, nitrogenous fertilizers are projected to retain their dominant position and, in terms of nutrients, their use is expected to increase at an average annual rate of 2.5 percent. At the same time, the use of phosphatic fertilizers, in terms of nutrients, is projected to increase by 4.7 percent a year. The implied rate of growth in potassic fertilizer use is highest-that is, 8.1 percent a year. In that case, however, the base-year estimate for total use is so low that, in spite of the high growth rate, projected 1985 usage is still relatively small. Four crops-namely, wheat, maize, rice, and 160 THE FERTILIZER SECTOR IN EGYPr Table 5. Estimated and Projected Use of Potassic Fertilizer Nutrients for Vegetables and Fruits, 1973-85 (thousands of metric tons) Crop 1973 1974 1975 1976 1977 1980 1985 Fruits 1.1 1.1 1.2 1.2 1.7 1.9 2.1 Vegetables 2.0 2.0 2.1 2.1 4.3 5. 5 5.7 Total 3.1 3.1 3.3 3.3 6.0 7.4 7.8 Note: Years refer to cropping years. Source: Ministry of Agriculture, quoted in Hassan and others, An Economic Study of Problems and Prospects. cotton-account for about two-thirds of the total nitrogen require- ment. Sugarcane, vegetables, and fruits account for another one-fifth. Thus, the projections do not entail any significant change in the consumption pattern of fertilizers. The increase in the consumption of potassic fertilizers results primarily from the increase in the area for vegetables and fruits, which are the only sources of demand for such fertilizers. In addition, the completion of the Aswan High Dam has given rise to an increase in the use of such fertilizers; the reason is that it has deprived the soils of the richness of the Nile "silt" or "'alluva," which was naturally rich in potash. Thus, chemical ferti- lizers must now substitute for the natural ones. The Domestic Production of Fertilizers Only nitrogenous and phosphatic fertilizers are produced do- mestically in Egypt; there is no production of potassic fertilizers, and the relatively small potash demand is met entirely by imports. In 1975, there were three nitrogenous fertilizer plants and four phos- phatic fertilizer plants in operation. The products and the technology used varied according to plant location. The nitrogen fertilizer industry Nitrogen fertilizer production began in Egypt in 1951 when the El Nasr Company for Fertilizers and Chemicals produced its initial output of 83,000 tons of calcium nitrate (15.5 percent nitrogen). Until 1961, this continued to be the only nitrogenous fertilizer pro- THE DOMESTIC PRODUCTION OF FERTILIZERS 161 duced in Egypt. Although this product is well suited to the high acidity of Egyptian soils, its low nitrogen content means that the transport cost per ton of nutrient is high. It also requires large amounts of inert matter, and it is highly subject to leaching. During the 1950s, imports of calcium ammonium nitrates in- creased rapidly, and in 1954 it was decided to produce these ferti- lizers domestically. Calcium ammonium nitrate has a higher nitrogen content (ranging from 20 to 33.5 percent nitrogen) than calcium nitrate, and it is resistant to leaching. In 1963, the company's plant at Suez began producing ammonium sulfate. These three types of fertilizers are the only nitrogenous fertilizers produced in Egypt, as is shown in table 6. They are produced domestically at four plants, as described below. THE EL NASR COMPANY FOR FERTILIZERS AND CHEMICALS AT SUEZ. The El Nasr Company plant at Suez began producing calcium nitrate (15.5 percent nitrogen) in 1951 and reached a peak level of 275,000 tons in 1963-64. In 1963, this company began producing ammonium sulfate (20.6 percent nitrogen) at an initial experimental output of 365 tons and reached a level of 98,300 tons in 1966-67. The 1967 war with Israel led to a reduction in the production of calcium ni- trate and ammonium sulfate to 140,000 tons and 46,600 tons, re- spectively, in 1967-68. In the following year, production was further reduced to 111,000 tons of calcium nitrate and 32,000 tons of am- monium sulfate. Both plants were completely shut down on March 9, 1969. The ammonia production at the complex utilizes the partial oxida- tion process, using the refinery gas supplied by the Suez refineries; the gas reforming unit was added in 1958. In addition, the plant has six units producing nitric acid (55 percent concentration) from ammonia oxidation. Five of these units require 8,250 tons a year of ammonia and the sixth, installed in 1958, requires 13,860 tons a year. Calcium nitrate is produced using the direct process;5 two lines are used, each producing 125,000 tons a year. Construction was begun on the unit in 1946, and the plant was ready in 1950; it is completely manual. The damage to this plant caused by the 1967 war was scheduled to be repaired by mid-1976, after which production was expected to resume. 5. That is, the direct reaction of limestone with nitric acid. Table 6. Domestic Production of Nitrogenous Fertilizers by Type and Producing Unit, 1962-63 to 1974-75 (tons) Calcium ammonium nitrate Calcium 26 percent; nitrate: Ammonium sulfate: 20.6 percent KIMA Co., 15.5 percent; 20.5 percent, Aswan, and 31 percent; 33.5 percent; El Nasr Co., El Nasr Co., El Nasr Co., 1UMA Co., El Nasr Co., KIMA CO., El Nasr Co., Year Suez Suez Helwan Total Aswan Helwan Aswan Helwan Total 1962-63 267,000 365 0 365 320,819 0 0 0 320,819 1963-64 275,026 64,500 0 64,500 417,010 0 0 0 417,010 1964-65 270,655 79,994 3,837 83,831 96,619 266,123 0 0 362,742 1965-66 266,789 83,887 3,593 87,480 0 379,509 0 0 379,509 1966-67 262,813 98,352 3,914 102,266 0 394,143 0 0 394,143 1967-68 140,276 46,611 4,434 51,054 0 437,937 0 0 437,937 1968-69 111,251 32,124 4,624 36,748 0 0 370,538 0 370,538 1969-70 0 0 4,500 4,500 0 0 377,107 0 377,107 1970-71 0 0 4,200 4,200 0 0 379,564 0 379,564 1971-72 0 0 6,800 6,800 0 400,000 400,255 38,000a 838,255 1972-74b n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 1974-75 0 0 8,000 8,000 0 0 347,000 90,000 437,000 gource: Compiled from the records of General Organization for Chemical Industries. a. Production of El Nasr Company for Coke and Chemicals at Helwan. b. Data for 1972-74 are not available. THE DOMESTIC PRODUCTION OF FERTILIZERS 163 Ammonium sulfate is no longer being produced at Suez, and the original sulfuric acid plant, which was built in 1962, has been sold to the Abu Zaabal Fertilizer Company and moved to Abu Zaabal. In 1976, the ammonium sulfate plant was still damaged as a result of the 1973 war with Israel, but repairs, which should take about two to three months, were expected to be made in the near future. No ammonium sulfate will be produced, however, until a new source of sulfuric acid is available, either from a new plant or in the form of shipments from another plant location. In 1964, the El Nasr Company at Suez began to expand the ferti- lizer complex by adding new units for the production of calcium ammonium nitrate, ammonia, and nitric acid. Construction was completed in 1967 but, because of the war in that year, production never began. These units were subsequently disbanded and trans- ferred to a new location at Talkha. The refinery gas required for ammonia production is obtained from the Suez refineries, which process crude oil from the Sinai and other oil fields. In 1967, the crude from the Sinai was preferred because it yielded the most desirable composition of refinery gas. This composition, however, depends upon the depth from which the crude is extracted. Since 1967, the Sinai oil fields have been ex- ploited by the Israelis, and thus oil will have to be extracted from a greater depth than before, this will affect the composition of the refinery gas used for ammonia. This factor, among others, has led some to consider using natural gas, rather than refinery gas, as the input into the existing plant. This natural gas will have to be piped in either from the Red Sea wells (pipe length, 300 kilometers) or from Helwan which receives it from the Western Desert. THE EGYPTIAN CHEMICAL INDUSTRIES (KIMA) AT ASWAN. The loca- tion of the Suez plant and the increasing demand for fertilizers in the late 1950s posed certain problems. The Suez refineries supplied all their gases to the existing calcium nitrate plant and could not accommodate a greater capacity. Appreciable reserves of natural gas had not been discovered. It was therefore decided to carry out an old plan for electrifying the Aswan Dam and to use an electrolytic process to obtain the required hydrogen.6 The construction of the plant, 6. In 1937, at the initiative of the government, the English Electric Company submitted a proposal to electrify the dam for 6.9 million Egyptian pounds. The project, which only needed the approval of the Parliament, was shelved. 164 THE FERTILIZER SECTOR IN EGYPT KIMA, was well coordinated with the electrification of the dam, and it started producing calcium ammonium nitrate (20.5 percent nitrogen) in 1960 at an initial rate of 75,000 tons a year. This was increased to 417,000 tons a year in 1964. In response to the increasing demand for high-analysis fertilizers, the 20.5 percent variety was abandoned in favor of the 26 percent form, and in 1968 the plant started pro- ducing a fertilizer with an even higher concentration-namely, cal- cium ammonium nitrate of 31 percent nitrogen. The ammonia required to manufacture this high-analysis fertilizer is produced from electrically generated hydrogen. The water electroly- sis plant is claimed to be the largest in the world. In recent years, this plant has been undergoing renovation; some of the electrolysis units were replaced by new ones in 1974, whereas others have been renewed by replacing some parts. The long-term plans at KIMA are to replace all units with new ones. The newer units are more efficient than the old ones (each new unit replaces two old ones and has a lifetime that is 50 percent longer than the old ones).7 There are four units for ammonia production with a total throughput of 150,000 tons a year. Part of this ammonia is utilized in the manufacture of nitric acid (265,000 tons a year), and the remainder is used in the produc- tion of prilled calcium ammonium nitrate (363,000 tons a year). Given the dry climate of the region, it appears possible that the cal- cium ammonium nitrate can be upgraded from 31 percent nitrogen to 33.5 percent nitrogen. The large quantities of water required to produce hydrogen are obtained from the Nile, and the water is demineralized on the site by resins and ion exchange. The electricity is obtained from the Aswan Dam, the nitrogen (for ammonia production) is obtained from the atmosphere using an air-liquefication plant, and the limestone (used in calcium ammonium nitrate) is obtained from quarries owned by KIMA.8 THE EL NASR COMPANY FOR COKE AT HELWAN. The El Nasr Company for Coke was established in 1960 to produce coke and small amounts 7. There are very low economies of scale in this process of ammonia production. 8. In addition to fertilizers, ferrosilicone alloys are also produced at KIMA. This location was chosen because of the availability of cheap electricity (1 million kilowatt hours). The required iron scraps are obtained from Cairo factories and the coke from Helwan. Some of this product is sent back to the steel complex at Helwan and the remainder is exported. THE DOMESTIC PRODUCTION OF FERTILIZERS 165 of other chemicals. Initially, the coke plant was intended to serve the forging and power requirements of the Helwan iron and steel complex. In 1964, however, the company began using part of the by-product gas to produce ammonium sulfate (20.1 percent nitrogen). Then, in 1971, the company started producing calcium ammonium nitrate, initially at a concentration of 26 percent nitrogen, and later at 33.5 percent. At present, ammonia is produced by the partial oxidation of coke- oven gas at the rate of 42,000 tons a year. This is less than the design capacity of 56,000 tons a year because of bottlenecks in the coke-gas plant and in the cracking department (that is, in the production of hydrogen for ammonia synthesis). Thus, only two out of the three ammonia streams are in operation. There are plans to pipe natural gas to this location from the Abu Gharahib fields in order either to convert the present process of ammonia production to one using natural gas or to permit capacity expansion of ammonia.9 The cal- cium ammonium nitrate (33.5 percent nitrogen) is produced at a rate of 90,000 tons a year, which is less than the design capacity because of the bottlenecks mentioned above. The amount of ammonium sul- fate produced is only 8,000 tons a year, and it is obtained as a by- product of the scrubbing of coke gas with sulfuric acid. The major raw material that drives this complex is the coal used in the coke plant, which is independently operated. The coal is im- ported primarily from the Soviet Union, but Poland, Australia, and Czechoslovakia serve as other sources. The fertilizer complex then buys the coke gas from the coke plant. THE EL NASR COMPANY FOR FERTILIZERS AND CHEMICALS AT TALKHA. The El Nasr Company for Fertilizers and Chemicals at Talkha began producing calcium ammonium nitrate in July 1975; it thus represents the most recent nitrogen fertilizer complex in Egypt. The units involved were initially erected at Suez as an extension to the existing complex there. Because of the war in 1967, however, produc- tion of calcium ammonium nitrate was interrupted, and the entire set of units (that is, the ammonia, nitric acid, and calcium am- monium nitrate plants) was dismantled and moved to Talkha. Con- struction was completed in early 1975, and the production of cal- cium ammonium nitrate commenced soon after. 9. The economics of this conversion to natural gas are discussed in chapter 13. It is shown that this conversion process would be an uneconomical proposition. 166 THE FERTILIZER SECTOR IN EGYPT The ammonia is produced by the partial oxidation of natural gas. At present, this is the only location in Egypt that uses natural gas in ammonia manufacture. The plant requires a liquid-air plant from which nitrogen and oxygen are obtained, and its design capacity of 400 tons a day has already been attained. Part of the ammonia is oxidized to produce nitric acid (55 percent concentration). There are three such units, each producing nitric acid at a rate of 350 tons a day. A fourth unit will be attached to the ammonia-urea complex being built at Talkha (known as Talkha II). The remainder of the ammonia, along with the nitric acid, is utilized to make calcium ammonium nitrate at the rate of 600 to 800 tons a day. This is less than the rated design capacity of 1,000 tons a day because of caking problems encountered in the storage depart- ment and bottlenecks in the transportation sector. The nitrogen con- tent of the calcium ammonium nitrate was initially 26 percent; by the end of 1975, however, this was raised to 31 percent, and it is hoped to raise this even further to 33.5 percent. The phosphate fertilizer industry The phosphate fertilizer industry in Egypt grew out of both the strong domestic demand resulting from the cultivation of beans and clover and the abundant domestic supply of phosphate rock. Some of this rock was initially applied directly to the soil in areas with sufficient soil acidity. Later, as more deposits were discovered, most of the rock was exported to Japan, and by 1936 total exports of phosphate rock had reached 500,000 tons. In that year, Egypt began exploiting her natural advantage in producing superphosphates. The output of the industry has shown an upward trend at an average annual growth rate of about 8.8 percent; the annual changes in output are shown in table 7. In the late 1950s, the output was sufficient to meet domestic demand and permit some exports. From 1961 to 1966, however, imports were needed to meet domestic demand; after that period, imports of superphosphate ceased, and Egypt reemerged as an exporter, mainly to the Arab countries. THE EL NASR PHOSPHATE COMPANY AT KAFR EL ZAYAAT. The first superphosphate factory in Egypt was established in 1937 at Kafr El Zayaat with an annual capacity of 25,000 tons. This capacity was expanded in 1949, 1956, and 1964, until it reached a level of 200,000 tons of single superphosphate a year. THE DOMESTIC PRODUCTION OF FERTILIZERS 167 Table 7. Annual Output of Phosphatic Fertilizers and Percent Change in Output from Year to Year, 1950-71a Output of Percent superphosphate: change from Year 15 percent P206 year to year 1950 69,283 - 1951 88,486 2.7 1952 106,136 19.9 1953 66,723 -37.1 1954 107,900 61.7 1955 137,050 25.1 1956 156,516 15.9 1957 177,474 11.9 1958 178,630 2.0 1959 167,410 -5.7 1960 189,620 12.6 1961 179,779 -5.2 1962 171,000 -4.9 1963 158,000 -7.6 1964 216,205 36.8 1965 258,609 19.6 1966 271,016 4.8 1967 265,000 -2.2 1968 309,000 16.6 1969 344,000 11.3 1970 447,000 29.9 1971 522,000 16.8 - - not applicable. Source: For the period 1950-60: Ministry of Agriculture, Monthly Bulletin of Agri- cultural Economics and Statistics, various issues. For the period 1961-70: Public Agency for General Mobilization and Statistics, Annual Book for General Statistics, various issues. For 1971: General Organization of Chemical Industries, unpublished data. a. Excluding ground rock phosphate. All the superphosphate produced at Kafr El Zayaat has a nutrient content of 15.5 percent P205; it is in powdered form and is produced using the conventional den method described in chapter 4. Initial production of the required sulfuric acid was accomplished using imported pyrites, but the subsequent expansion of the sulfuric acid plant utilized elemental sulfur. The total annual sulfuric acid pro- duction now stands at 100,000 tons. THE EL NASR PHOSPHATE COMPANY AT ABU ZAABAL. The second 168 THE FERTILIZER SECTOR IN EGYPT superphosphate plant in Egypt was erected at Abu Zaabal in 1948. Like the factory at Kafr El Zayaat, it produces powdered single superphosphate with a nutrient content of 15.5 percent P205 using the den method. Its capacity is 200,000 tons a year. Four sulfuric acid plants, yielding a total annual output of 150,000 tons, are lo- cated at this site. Three of these plants produce acid using pyrites; the most recent one, however, uses elemental sulfur. Future expansion under consideration at this site includes a new sulfuric acid unit producing 100,000 tons a year, a phosphoric acid plant with a capacity of 60,000 tons a year, and a superphosphate plant with a capacity of 200,000 tons a year. The phosphoric acid is to be used in the production of enriched superphosphate and, later, triple superphosphate. Some thought is being given to the production of monoammonium phosphate, diammonium phosphate, and ammonium sulfate using ammonia from Talkha.'0 THE EL NASR PHOSPHATE COMPANY AT ASSIOUT. The most recent superphosphate plant operates at Assiout. It started production in 1969 and produces granular single superphosphate (15.5 percent P205) using the den method. The design capacity of this plant is 200,000 tons a year, and a capacity expansion of an equivalent amount is planned for the future. Unlike the other two factories, it produces sulfuric acid from imported elemental sulfur. The amount of sulfuric acid that can be produced at this location is 80,000 tons a year and, to accommodate the future expansion of single super- phosphate, a capacity expansion of 80,000 tons a year of sulfuric acid is planned." Raw Materials for Fertilizer Production The major raw materials required in fertilizer production in Egypt are pyrites, elemental sulfur, phosphate rock, and natural gas. Of these, pyrites and sulfur, which are used in the production of the intermediate product sulfuric acid, are imported. No deposits of sulfur or usable pyrites are known to exist in Egypt. Pyrites are ob- 10. These plans are evaluated in chapter 13. It is shown there that the most effi- cient project would be the erection of a new phosphoric acid plant and the produc- tion of powdered triple superphosphate. 11. These plans are also evaluated in chapter 13. Capacity expansion at this location turns out to be uneconomical because of the low phosphorus demand in the surrounding governorates. RAW MATERIALS FOR FERTILIZER PRODUCTION 169 tained from Cyprus, Yugoslavia, and the Soviet Union; sulfur is imported from the United States, Poland, and Iraq. Substantial reserves of phosphate rock and natural gas do, however, exist in Egypt. Phosphate rock reserves Considerable deposits of phosphate rock are located in Egypt, but rock production is small and therefore Egypt plays a relatively minor role in the world trade of phosphate rock. The largest known deposit is at Abu Tartur in the Kharga region of the Western Desert. An estimated deposit of 987 million tons of phosphate rock has been identified, and an unknown large quantity of rock is thought to exist in parts of the area not yet surveyed. Of the reserves surveyed, 525 million tons of rock at a grade of 28 to 30 percent P205 can be mined. These reserves, however, lie about 100 meters below the surface and will have to be mined underground. In order to exploit the rock, its iron content must be reduced; it contains 4.3 to 4.9 percent iron in the form of pyrites and 0.7 to 1.7 percent in the form of aluminum oxide."2 It is assumed that reduction is possible, and thus production is expected to commence in 1980-81. Another deposit exists at West Sebeya, where a mine owned by the Abu Zaabal Fertilizer Company is producing 100,000 tons of rock a year. In the same area, there is another deposit of 150 million tons of 20 to 21 percent P205 grade rock with an iron content of 2 to 3 percent. By 1980-81, an annual production of 120,000 tons of P205 is expected. At East Sebeya, 150,000 to 200,000 tons of rock are being pro- duced for local consumption. The P205 content of this rock is 28 to 29 percent. A new project currently under construction is expected to reach an output of 1 million tons of rock in two stages. The first stage is to produce 300,000 tons of 30 percent P205 rock in 1980-81, and the second stage will produce the rest. Several medium grade (20 to 30 percent P20,) deposits of com- mercial significance lie in the Eastern Desert. The deposits near the Safaya port have been mined for many years, especially for exports. By 1980, all mines in this area are expected to be depleted. Further south are the Qusayr deposits, which have also been mined for several years. When these mines are also depleted (around 1980), 12. The pyrites exist in the form of impurities that cannot be used as a raw material in sulfuric acid production because of separation problems. 170 THE FERTILIZER SECTOR IN EGYPT production will shift to Abu Shigela, where the reserves are esti- mated to be equivalent to about 40 million tons of upgraded 30 percent P205 rock. This rock will be used mainly for exports. South of Safaya, at Hamrawein, is another extensive deposit of 1.3 million tons of rock. It is to be beneficiated to 600,000 tons of 32 to 33 per- cent P205 to supply the domestic and export needs of Egypt. Natural gas reserves In conjunction with Ente Nazionale Idrocarburi (ENI) of Italy, the Egyptian General Petroleum Corporation (EGPC) has developed a natural gas field (92.8 percent methane) at Abu Madi, about 40 kilometers north of Talkha. The identified recoverable reserves are estimated to be between 30 billion and 40 billion cubic meters. The anticipated daily consumption is 26 million cubic meters, which should give the identified reserves a life of at least thirty years. Addi- tional reserves may be identified as the drilling continues. The gas is now used for the Talkha I and II plants and for a local power station and township. Two other major gas discoveries have been made in the Delta area. The first field, at Abu Kir, is located in the sea about ten miles offshore. It has reserves that are currently estimated to be around 22 billion cubic meters, although more reserves are likely to be proven in this area in the near future. The gas has a methane content (by volume) of 94 percent and, like the Abu Madi gas, is free of sulfur. This gas will be used for the ammonia-urea complex due in Abu Kir in 1978-79. The second gas field, at Abu Gharahib, has known recoverable reserves of 40 billion cubic meters. This gas is expected to be piped to Helwan and Cairo. The Transport of Fertilizers Egyptian fertilizers are either produced domestically or imported. They must then be transported from plants or ports of landing to the main stations in the various governorates; this is the first activity in the distribution system required to feed the sectional stores or the stores of the agricultural cooperative societies. The various products are transported to the governorates as follows: . the calcium ammonium nitrate produced at the KIMA plant at Aswan covers the needs of the Upper Egypt governorates ex- cluding Giza; * the calcium ammonium nitrate produced at the El Nasr plant THE TRANSPORT OF FERTILIZERS 171 at Helwan covers the needs of Giza, Fayoum, Menoufia, and Kalubia governorates; * the single superphosphate produced at the Abu Zaabal factory covers the needs of Kalubia, Suez, Fayoum, Sharkia, and parts of Dakahalia and Menoufia governorates; * the single superphosphate produced at Kafr El Zayaat is trans- ported to Alexandria, Behera, Gharbia, Kafr El Sheikh, Dami- etta, and parts of Dakahalia and Menoufia governorates; * the single superphosphate from the Assiout plant covers the needs of the Upper Egypt governorates excluding Giza; * all imported nitrogenous fertilizers are distributed to Lower Egypt excluding the governorates of Menoufia and Kalubia; * all imported potash fertilizers are distributed to all the relevant governorates of Egypt. The rule that guides decisions about the distribution of fertilizers is simple: the output produced at any plant is transported to an agri- cultural area on the sole basis of its proximity to that plant. Ferti- lizers are transported either by lorries, by railways, or by river trans- port, as is shown in table 8. Clearly, the predominant mode of trans- Table 8. The Transport of Fertilizers by Governorates Covered and Mode of Transport Type offertilizer Governorates covered Mode of transport Imported Nitrogenous fertilizers Lower Egypt excluding Lorries Menoufia and Kalubia Potash fertilizers All relevant governorates Lorries and railways Domestically produced CAN from KIMA plant Upper Egypt excluding Inland waterways at Aswan Giza and railways CAN from El Nasr plant Giza, Fayoum, Menoufia, Lorries at Helwan and Kalubia ssp from Abu Zaabal Kalubia, Suez, Giza, Lorries plant Fayoum, Sharkia, and parts of Dakahalia and Menoufia ssp from Kafr Alexandria, Behera, Lorries and El Zayaat plant Gharbia, Kafr El Sheikh, railways Damietta, and parts of Dakahalia and Menoufia ssp from Assiout plant Upper Egypt excluding Lorries and Giza railways 172 THE FERTILIZER SECTOR IN EGYPT port is by lorries. In 1972, 93.3 percent of all fertilizer material trans- ported was shipped by lorries, 5 percent by rail, and only 1.7 percent by river transport. The main reasons why lorries dominate as the mode of transport of fertilizers are that they are fast and are considered to be the safest mode, thereby minimizing losses. In addition, they present a direct mode of transport from the factory or port to the governorates; thus, loading and unloading need to be undertaken only once rather than twice as in the case of river and rail transport. Rail transport is currently handicapped by the shortage of wagons in Egypt. Al- though river transport is the cheapest mode, there are few ports and few facilities for loading and unloading. The Prices of Fertilizers Fertilizer prices in Egypt have been controlled by the government since 1960. These prices are determined by Ministerial Orders of the Ministry of Industry in collaboration with the Ministry of Agricul- ture and the Agricultural Prices Balancing Fund. Fertilizer prices are fixed at all levels, and the retail prices are unified across the country. No distinction is made between domestically produced and imported fertilizers. The pricing policy for domestically produced fertilizers Basically, there are three levels of prices: the "ex factory" price, the wholesale price, and the retail price. The ex factory price is the price paid by the Agricultural Coopera- tive Credit Organization-that is, the Credit Bank-to the local producer of the fertilizer. This price varies from factory to factory and thus assures a differential return for the same product: for example, table 9 illustrates the different prices for single superphos- phate (15 percent, P205). The second stage is the wholesale price, which is the price of de- livered fertilizers at the central stores of the governorates. The ferti- lizers are delivered by the Credit Bank to the central stores that belong to the various cooperative societies. The commercial margin at this stage varies from 0.3 percent of the consumer price for am- monium sulfate to 6.5 percent for calcium ammonium nitrate. In THE PRICES OF FERTILIZERS 173 Table 9. Average Total Cost, Ex Factory Price, and Net Return for Selected Fertilizers, 1971-72 Average total Ex factory cost price (Egyptian pounds (Egyptian pounds Net return Plant and product per ton)a per ton) (percent) Kafr El Zayaat: ssp 15 percent 10.536 12.998 23.0 Assiout: ssp 15 percent 10.744 12.725 18.0 Abu Zaabal: ssp 15 percent 12.529 12.914 3.0 Aswan: CAN 31 percent 30.396 36.024 18.5 Helwan: CAN 33.5 percent 69.553 34.144 -50.9 Helwan: ammonium sulfate 20.6 percent 33.746 27.076 -6.67 Source: The General Organization of Chemical Industries. a. Includes raw materials and direct costs, wages, depreciation, indirect costs, mar- keting and administrative costs. addition to this margin, since 1965 a treasury tax has been imposed on fertilizer sales. It is a specific tax of 1 Egyptian pound for each ton of fertilizer product; it thus charges lower concentrate fertilizers a higher tax rate. This tax is added to the wholesale price. The final stage is the retail price, which is divided into two com- ponents. The first component is the retail price to consumers (or farmers) who are members of the cooperative societies. The second component is the retail price to nonmembers. The price paid by the nonmembers is 5.26 percent above that paid by the members. The levels of prices for the various domestically produced ferti- lizers in 1971-72 are shown in table 10. The pricing policy for imported fertilizers The price of imported fertilizers delivered to the stores of the Credit Bank in the governorates is made up of the import price (c.i.f.); the transport cost to the stores; customs duties; the cost of unloading, storing, and packaging; and the commission of the Credit Bank. Since the prices of the domestic and imported ferti- lizers are equalized, the difference between the retail price and the delivered price of the imported fertilizers accrues to the Agricultural Table 10. Prices and Margins of Distribution of the Locally Produced Fertilizers, 1971-72 (Egyptian pounds per ton) Price of Retail sales delivery at Whole- price to the Retail sales Ex factory Transport the central sale cooperative Retail price to Type of fertilizer sales price cost Package Taxes stores margin members margin nonmembers Superphosphate: 15 percent Powdered,in jutesack 10.950 1.800 - 1.000 13.750 l.i17 14.867 0.783 15.650 Powdered, in polyethylene sack 11.550 1.800 - 1.000 14.350 1.185 15.535 0.765 16.300 Powdered, in jute and polyethylenesack 11.450 1.800 - 1.000 14.250 1.140 15.340 0.810 16.200 Granular, in jute sack 11.050 1.800 - 1.000 13.850 0.900 14.750 1.000 15.740 Granular, in polyethylene sack 11.650 1.800 - 1.000 14.450 1.130 15.580 0.830 16.400 Granular, in jute and polyethylene sack 11.550 1.800 - 1.000 14.350 1.130 15.480 0.830 16.300 Calcium nitrate: 15.5 percent 20.800 1.800 - 1.000 33.600 1.150 24.750 1.250 26.000 Ammonium sulfate: 20.6percent 21.700 1.800 3.000 1.000 27.500 0.100 27.600 1.400 29.000 Calcium ammonium nitrate: 26percent 29.502 1.800 - 1.000 32.333 2.333 34.635 1.500 36.135 Calcium ammonium nitrate: 31 percent 35.300 2.000 - 1.000 38.300 1.871 40.171 2.629 42.800 Calcium ammonium nitrate: 33.5 percent 39.300 1.800 - 1.000 42.100 1.400 43.500 2.240 45.740 -= not applicable. Source: General Organization for Industrialization. THE PRICES OF FERTILIZERS 175 [able 11. Prices and Margins of Distribution for Imported Fertilizers, 1970-71 'Egyptian pounds per ton) Price of im- ported fertil- Retail sales Price izer at the price to the (cost and stores of the cooperative Retail price to Type of fertilizer freight) Credit Bank members nonmembers Jrea: 46percent 20.930 30.170 60.800 64.000 Ammonium sulfate: 20.6 percent 7.780 13.880 27.600 29.000 Ammonium nitrate: 33.5 percent 17.490 24.311 43.600 45.840 'otassium chloride: 60 percent - 22.700 24.750 26.000 ialcium ammonium nitrate: 26 percent 16.619 20.782 33.300 35.800 -4itro-ammonium sulfate: 26 percent 18.368 21.541 33.200 35.000 botassium sulfate: 20.5 percent - 32.700 24.750 26.000 "mmonium sulfate: 20.6percent 23.179 27.870 27.600 29.000 iiompound fertilizers 32.800 45.873 41.800 44.000 .ingle superphosphate 10.320 17.120 12.875 13.500 -n= not applicable. Source: Agricultural Prices Balancing Fund. Prices Balancing Fund. Table 11 indicates such prices for imports in 1970-71. In the case of some fertilizers, such as potassium sulfate and single superphosphate, the retail price is below the delivered price. Such products are subsidized from the Balancing Fund. The prices of different fertilizers The prices of nitrogenous fertilizers have been determined and changed by a series of six Ministerial Orders since 1960, as shown in table 12. The last order, which was in 1968, determined the prices of calcium ammonium nitrate 31 percent at 39.71 Egyptian pounds a ton and calcium ammonium nitrate 33.5 percent at 43.6 Egyptian pounds a ton. Phosphatic fertilizer prices have been determined by nine Minis- Table 12. Cooperative Sales Prices of Nitrogenous Fertilizers as Determined by Ministerial Orders, 1960-72 (Egyptian pounds per ton) Calcium Calcium Calcium Calcium Ammo- Ammo- ammo- ammo- ammo- ammo- nium Minis- Calcium nium nium nium nium nium nitro- terial Chilean nitrate: sulfate: nitrate: nitrate: nitrate: nitrate: sulfate: Urea: Compound Order sodium 15.5 20.5 20.5 26 31 33.5 26 46 fertilizers Year number nitrate percent percent percent percent percent percent percent percent 25-10-0 1960 245 24.700 23.750 26.600 26.600 33.300 - - 34.200 - - 1961 - 23.750 23.750 26.600 - - - - 34.200 - - 1962 542 23.750 23.750 26.600 26.600 - - - 32.300 - - 1963 - 23.750 23.750 26.600 26.600 - - - 32.300 - - 1964 1,316 - 23.750 26.600 26.600 33.300 - - 32.300 60.800 40.800 1965 437 - 24.750 27.600 27.600 33.300 - - 32.300 60.800 - 1966 - - 24.750 27.600 27.600 33.300 - - 32.300 60.800 - 1967 - - 24.750 27.600 27.600 33.300 - - 32.300 60.800 - 1968 99 - 24.750 27.600 27.600 33.300 39.710 - 32.300 60.800 - 1968 134 - - - - - - 43.600 - - - 1969 - - 24.750 27.600 27.600 33.300 39.710 43.600 32.300 60.800 - 1970 - - 24.750 27.600 27.600 33.300 39.710 43.600 32.300 60.800 - 1971 - - 24.750 27.600 27.600 33.300 39.710 43.600 32.300 60.800 - 1972 - - 24.750 27.600 27.600 33.300 39.710 43.600 32.300 60.800 - - = not applicable. Note: Sales price includes treasury tax beginning in 1965. Source: General Organization of Chemical Industries. Table 13. Cooperative Sales Prices of Phosphatic Fertilizers as Determined by Ministerial Orders, 1960-72 (Egyptian pounds per ton) Superphosphate: Superphosphate: Ministerial Superphosphate: Superphosphate: granulated, in powdered, in Order powdered, in granulated, in jute and poly- jute and poly- Year number Concentration of P,05 jute sack polyethylene sack ethylene sack ethyilene sack 1960 1,087 18 percent 13.300 - - - - 15 percent 10.515 - - - 642 15 percent 10.450 - - 1961-62 - 18 percent 13.300 - - - 1963 - 15 percent 10.450 - - - '4 1963 458 15 percent 11.875 - - - 1964 1,276 16 percent 12.160 - - - - 15 percent 11.875 - - - 1,316 17 percent 12.920 - - - 1965-66 - 16 percent 12.160 - - - 1967 - 17 percent 12.920 - - - - 15 percent 11.875 - - - 1968 143 15 percent 11.875 - - - 1969-70 596 15 percent 11.875 12.920 12.4785 t2.274 1971 330 15 percent 13.9175 14.630 14.535 15.440 1972 40 15 percent 14.8675 15.580 15.485 15.390 - = not applicable. Note: Sales price includes Treasury tax. Source: General Organization of Chemical Industries. 178 THE FERTILIZER SECTOR IN EGYPT terial Orders issues since 1960, as shown in table 13. The last order, which was issued in 1972, determined the prices of powdered and granulated single superphosphate in jute and polyethylene bags at an average of 15 Egyptian pounds a ton. The potash fertilizer price has been set by only two Ministerial Orders at 23.75 Egyptian pounds a ton for potassium sulfate and potassium chloride; the price level has remained unchanged since 1960. These fertilizers are not produced domestically and are thus imported; no treasury tax has been imposed on them. The Import of Fertilizers Egypt has always imported fertilizers to supplement its domestic production. Over time, as the consumption pattern changed, so did the import pattern. At the turn of the century, Egypt imported only sodium nitrate (from Chile); since then, it has imported different types of nitrogenous, phosphatic, and potash fertilizers. During World Wars I and II, trade blockades reduced imports substan- tially. By the second war, Egypt had become a heavy user of ferti- lizers, and thus the war had a major impact on the country's agri- cultural productivity. After that period, as fertilizer use increased, imports increased significantly despite large expansions in domestic production. At present, fertilizer imports are confined to nitrogenous and potash fertilizers; phosphatic fertilizer imports were stopped in 1966-67. Since no raw materials for potash fertilizers are available locally, imports will continue to be the sole source of supply. Import policy Egypt has not had a well-defined policy for importing fertilizers. Import targets are included in the five-year fertilizer plan, but they do not materialize and are subject to continuous change. The com- mon practice has been to adopt annual plans that are influenced by several factors including variations in domestic output, consumption requirements, and available foreign exchange that can be allocated to fertilizer imports. Estimates of import quantities are based on the difference between domestic output and consumption and on the availability of foreign exchange. These estimates are worked out by the General Organiza- tion for the Agricultural Prices Balancing Fund (previously known THE IMPORT OF FERTILIZERS 179 as the Fertilizer Subsidizing Fund). The Fund then submits its esti- mates to the Treasury for allocation of the necessary foreign ex- change. The Treasury considers these estimates in the light of its foreign exchange budget and determines the import quota, which is classified according to type of payment (free currency, payment agreements, credit facilities ). The Fund then coordinates the trans- actions between the different public trade agencies, which submit their preliminary tenders concerning specific quantities of different types of fertilizers, prices, and payment conditions. A special com- mittee-consisting of representatives of the Fund, the Credit Bank, trade companies, and representatives of foreign agencies-selects the most suitable offers according to price considerations and terms of shipment. The payment conditions and credits are put at the trade companies' disposal for implementation. Thereafter, the companies assume full responsibility for obtaining the fertilizers with certain specifications and according to a time schedule. These imports are required to be shipped by Egyptian vessels. In the past few years, imported types of fertilizers were confined to ammonium sulfate (21 percent), calcium ammonium nitrate (20.5 percent, 26 percent, and 33 percent), urea (46 percent), and potassium chloride (60 percent). Recent trends have shown an emphasis on imports of higher analysis fertilizers. As a result of the present import policy, the prices paid for ferti- lizers of the same type and concentration differ, even in the same year, according to the exporting country and the payment arrange- ments. This policy is also responsible for excess imports that have caused bottlenecks in unloading, transportation, and distribution in some years and shortages of imported quantities in others. The import sources are numerous. In the past ten years, the main countries exporting fertilizers to Egypt have been Italy, Rumania, Belgium, Kuwait, Poland, the Netherlands, Bulgaria, the Soviet Union, Switzerland, France, the German Democratic Republic, the Federal Republic of Germany, and Lebanon. Most of the nitroge- nous fertilizer imports-about 75 percent on the average-have been from free-exchange countries. This percentage increased steadily from about 68 percent in 1959-60 to 95 percent in 1969-70, and then decreased to about 58 percent in 1971-72. Most of the phos- phatic fertilizer imports were also from free-exchange countries; these imports ceased in 1966-67. Yugoslavia, which was a major source of the phosphatic fertilizer imports in 1961-62, was the only socialist country to export these fertilizers to Egypt. Potash ferti- 180 THE FERTILIZER SECTOR IN EGYPT lizer imports started in 1964-65 at a very low level, and from only three countries: France, Italy, and the German Democratic Re- public. East Germany has been the sole source of potash imports in the past few years. Conclusion This chapter has provided a brief, conventional description of the fertilizer sector in Egypt. Although the chapter has given some insight into the past growth and the present structure of the ferti- lizer sector and has described the role the government plays in pric- ing and importing fertilizer materials, neither the type of information provided nor the organization of that information are immediately useful for a planning model. In some respects, considerably greater detail is required; in others, entire categories of data are unneces- sary. This will become clear when several planning models of the Egyptian fertilizer sector are presented in the next few chapters. 11 An Analysis of the Egyptian Fertilizer Sector in 1975 THE STRUCTURE OF THE EGYPTIAN FERTILIZER SECTOR in the year 1975 is analyzed in this chapter.' For that purpose, a simple mathematical model of the sector, closely following the model of an existing industry developed in chapter 8 above, will be formulated. Before the analytical framework is described and the numerical information given, however, the scope and purpose of the analysis will be summarized. Initially, a model that is excessively restrictive in terms of alternative choices will be formulated; several of these restrictions will be successively dropped or relaxed in the course of the analysis so that their individual impact can be assessed. The starting point of the analysis is actual recorded use of fertilizer material by type in each of the twenty governorates of Egypt in 1975.2 By defini- tion, the supply of these fertilizers must have originated either from domestic production facilities in 1975 or from imports. The model is 1. The date 1975 refers to the "fertilizer year," which runs from July 1974 to June 1975. 2. Potassic fertilizers are not included in this analysis because their use in Egypt is limited, and no scope for their domestic production appears to exist. 181 182 AN ANALYSIS OF THE EGYPTIAN FERTILIZER SECTOR IN 1975 free to select the least-cost supply pattern-either from alternative domestic supply sources for the same product, or by importing. Domestic supplies are, however, subject to capacity constraints that are initially defined in terms of the actual production levels achieved in 1975. Fertilizer materials can be transported from supply sources to the various marketing centers (assumed to be the major town in each governorate) by waterway, road, or rail; in accordance with the actual situation, however, it is first assumed that all final products are transported by lorry, and that all raw materials are transported by barge and, if necessary, lorry. With this specification, the model can be used to select the least-cost supply and shipment pattern for fertilizers in 1975; the model will determine whether a fertilizer should be imported or produced domestically, given the existing production facilities and their location. If this determination were the only objective, however, the planning model would hardly be required, because these questions can be answered by inspecting the original data; that is, an analysis of marginal production cost in domestic plants, import price, and transport cost would be sufficient to arrive at the same answers. The value of the model will be in its ability to carry out a number of experiments by slightly varying the restrictions. In a short period of time, it can obtain a great number of detailed numerical results, presented in a systematic and uniform way, thus facilitating com- parisons. Then, the following refinements can be made to the model: drop the restriction on interplant shipments of intermediate products, relax the capacity constraints to allow 100 percent capacity utilization, and investigate the implications of greater flexibility in fertilizer use by specifying nutrient requirements rather than requirements by fertilizer type. The static analysis of these questions will indicate how the operation of the fertilizer sector can be improved in the short run. Given the simplified nature of the model, conclusions along these lines are only tentative and indicative; all require further, more detailed analysis. In this regard, the model serves as a device to generate particular options that deserve further study, not as a device that produces firm recommendations for action. The Structure of the 1975 Model In this section, a detailed description is given of the 1975 model of the Egyptian fertilizer sector. THE STRUCTURE OF THE 1975 MODEL 183 The demand for fertilizers Egypt is divided into twenty governorates; these have been con- sidered as separate demand regions in this and subsequent models in order to capture the importance of spatial aspects of the structure of the fertilizer sector. The major city in each governorate is designated as the marketing center for that demand region, and it serves as the reference point for calculating transport distances between sources of fertilizer supply and the demand region. The twenty governorates and their marketing centers are the following: Governorate Marketing center Alexandria Alexandria Behera Demanhur Gharbia Tanta Kafr El Sheikh Kafr El Sheikh Dakahlia El Mansura Damietta Damietta Sharkia Zagazig Ismailia Ismailia Suez Suez Menoufia Shibin El Kom Kalubia Benha Giza Giza Beni Suef Beni Suef Fayoum El Fayoum Minia El Minia Assiout Assiout New Valley El Kharga Sohag Sohag Quena Quena Aswan Aswan The map indicates the geographical location of these governorates and marketing centers. An estimate of the level of fertilizer use by type is given for each demand region in table 14. This table, therefore, also gives a compre- hensive listing of all relevant fertilizer types in 1975. The basic condition to be fulfilled in this model is that the sum of shipments of fertilizer material to each demand region, from domestic supply sources and imports, equals fertilizer use in each region. In this initial version of the model, no room is left for choice; that is, the Table 14. Consumption of Fertilizers by Demand Region and Fertilizer Type, 1974-75 (thousands of tons) Calcium Calcium Calcium Single ammonium ammonium ammonium super- Com- Com- Di- nitrate: nitrate: nitrate: Ammonium phosphate: pound 1: pound 2: ammonium Demand region 26 percent 31 percent 33.5percent sulfate Urea 15.5percent 25-5.5-0 30-10-0 phosphate Alexandria 0 0 5 3 1 8 0 0.0 0.0 Behera I 0 25 90 35 64 1 0.1 0.1 Gharbia 0 0 17 60 28 57 1 0.2 0.1 Kafr El Sheikh 1 0 10 45 22 25 2 0.2 0.0 Dakahlia 1 0 26 60 20 52 1 0.0 0.0 Damietta 0 0 2 15 8 5 0 0.0 0.0 Sharkia 1 0 31 50 28 43 1 0.1 0.0 Ismailia 0 0 4 6 2 4 0 0.0 0.0 Suez 0 0 1 0 0 1 0 0.0 0.0 Menoufia 0 0 24 21 30 33 2 0.1 0.1 Kalubia 0 0 25 16 7 22 1 0.0 0.1 Giza 0 0 40 6 2 14 1 0.1 0.0 Beni Suef 1 0 15 1 20 13 3 0.0 0.0 Fayoum 1 0 20 6 20 17 1 0.0 0.0 Minia 2 15 35 1 41 50 3 0.2 0.1 Assiout 1 20 26 1 27 35 5 0.1 0.0 New Valley 0 0 0 0 1 1 0 0.0 0.0 Sohag 0 65 3 0 7 20 1 0.0 0.0 Quena 0 95 2 0 3 8 0 0.0 0.0 Aswan 0 40 0 0 0 8 0 0.0 0.0 Total 9 235 311 381 302 480 23 1.1 0.5 |~~~~~h Fertilizer} AWYAdAwn 2 Th FertilizerPat PrpSed oCpcitysor N '.- ., C y o CitiesandTowns ~ ~ ~ ~ ~ ~ ~~ ' S Y SYIA j Maln Roads L_IRQ. I IRailwavsL B A IODN ; Kil-mt-i 0 50 too 150 200 jRApRUABBI t S 050 100 OF EGY m alo 185 186 AN ANALYSIS OF THE EGYPTIAN FERTILIZER SECTOR IN 1975 recorded composition of fertilizer use has to be reproduced exactly. Later, this condition will be relaxed and the only stipulation will be that the required quantity of nitrogen and phosphate nutrient is supplied, regardless of the material used. The initial conditions can be stated mathematically as follows: (11.1) ~IE Xa + Vcy > 4j, c E CF iel je6J where x,ij = the shipment of fertilizer c from domestic plant i to demand regionj, where E means that such shipments for all plants in the set I should be summed; v,j = the imports of fertilizer c for demand region j; d,1 = the actual use of fertilizer c in demand region j. The statements on the right refer to the set coverage: CF denotes the set of fertilizers; j e J means that constraint (11.1 ) needs to be speci- fied for every demand region. The supply offertilizers The fertilizers used in 1975 were supplied from domestic production facilities or from imports. To model the possible supply from imports for the purpose of this planning model (in which fertilizer types used are strictly specified), one simply needs to specify the import possi- bility for each fertilizer product actually used in 1975 and the location of all relevant ports. In this case, all imports are assumed to have en- tered Egypt through Alexandria. To model the possible supply from domestic production facilities, one needs to specify the location of production facilities, the production capacities of all productive units at each site, and the technological coefficients that describe the transformation of raw materials to intermediates and of either or both into particular final products. The technological coefficients can be broken into two categories: input-output coefficients giving the detailed material input requirements per unit of output of the various processes in the model, and the capacity required in each productive unit as a ratio of the design capacity. As was described in chapter 8, it is often possible to distinguish a category of miscellaneous inputs that can be combined into a cost figure per ton of output, and that do not need to be specified separately. The main criterion for inclusion THE STRUCTURE OF THE 1975 MODEL 187 into this category is that the input in question can be expected to be available in the relevant quantity range at a given price. In table 15, the location of plants in operation in 1975 is given, together with the design capacity of each productive unit and the process in use during that year. Input-output coefficients are given in table 16. Labor requirements in intermediate and final product manufacture are not included among the basic data for reasons that will be explained shortly. Modeling the supply side The basic technological relationships that describe the supply side in the model are conveniently specified in three sets of material balance constraints-for final products, for intermediate products, and for raw materials and miscellaneous inputs-and in a set of capacity constraints. MATERIAL BALANCE CONSTRAINTS ON FINAL PRODUCTS (11.2) ~ E ap zpi 2E Xii ce CF ptP jeJ i e I The right-hand side of this constraint corresponds closely to one of the terms in constraint (11. I ) and represents the sum of the shipments of product c (a fertilizer type) from plant location i to all marketing centersj. The expression on the left-hand side denotes the production of final product c by all processes p at plant location i. The typical process p that provides final products can be assigned a coefficient a, = 1.0. MATERIAL BALANCE CONSTRAINTS ON INTERMEDIATE PRODUCTS (1 1.3 )E a,p zpi > ° c e CI peP i eI This constraint states that the production of intermediate product c at plant location i with process p should at least equal total use of intermediate product c at plant location i. CI is the set of intermediate products. Since final products, and some intermediate products, require intermediate products in their production, at least one of the input- output coefficients associated with process p in constraint (11.3 ) will have a negative sign. For the constraint to hold, therefore, at least one process p in (11.3) will need to have a positive sign denoting the Table 15. Design Capacity of Plants by Plant Location and Productive Unita (tons per day) Ammonium Calcium sulfate: Single ammonium Sulfuric Sulfuric Nitric acid: Ammonia: by-product super- nitrate: acid: acid: ammonia water elec- Ammonia: of coke-oven phosphate: 31 percent and Plant location sulfur pyrites oxidation trolysis coke-oven gas plant den method 33.5 percent Aswan - - 800 450 - - - 1,100 Os °o Helwan - - 282 - 172b 24 - 364 Kafr El Zayaat 200 50 - - - - 600 - Assiout 250 - - - - - 600 - Abu Zaabal 242 227 - - - - 600 -= not applicable. Source: Plant managers. a. The design capacity of a productive unit is the production level that in principle can be reached. In reality, this production level is rarely achieved, and in most cases it was found that actual production in Egypt in 1975 was lower than design capacity. b. At Helwan, ammonia is used to produce nitric acid; ammonium sulfate is a direct by-product of the coke-oven plant associated with the steel works at Helwan. Table 16. Input-Output Coefficientsa (tons of input per ton of output, unless otherwise noted) Ammo- Calcium nium Single Calcium ammo- Nitric Ammonia: sulfate: super- ammo- nium Sulfuric Sulfuric acid: water Ammonia: by-product phosphate: nium nitrate: acid: acid: ammonia electrol- coke-oven of coke- den nitrate: 33.5 Inputs sulfur pyrites oxidation ysis gas oven plant method 31 percent percent Major inputs Electricityb -12.0 - - - - - Coke-oven gas0 - -2.0 - -0.62 Phosphate rock: 30 percent P205 - - '0 Pyrites - -0.826 - - - - - - - Sulfur -0.334 - - - - - - - - Sulfuric acid 1.0 1.0 - - - -0.76 -0.41 Ammonia - - -0.292 1.0 1.0 -0.26 - -0.20 -0.21 Nitric acid - - 1.0 - - - - -0.71 -0.76 Ammonium sulfate - - - - - 1.0 - - - Single superphosphate - - - - - - 1.0 - Calcium ammonium nitrate: 31 percent - - - - - - - 1.0 - Calcium ammonium nitrate: 33.5 percent - - - - - - - 1.00 Limestone - - - - - - - -0.12 -0.04 (Table continues on the following page) Table 16 (continued) Ammo- Calcium nium Single Calcium ammo- Nitric Ammonia: sulfate: super- ammo- nium Sulfuric Sulfuric acid: water Ammonia: by-product phosphate: nium nitrate: acid: acid: ammonia electrol- coke-oven of coke- den nitrate: 33.5 Inputs sulfur pyrites oxidation ysis gas oven plant method 31 percent percent Miscellaneous inputs Electricityd -50.0 -75.0 -231.0 -1,960.0 -19,0 -14.0 - _ - Blast-furnace gas0 - - - -609.0 - - - - - Bagsf - - - - -22.0 -22.0 - -23.0 -23.0 Cooling watero -20.0 -20.0 -0.6 -700.0 -17.0 -6.0 - -49.0 -49.0 Steam - - - -4.0 - - - -0.4 -0.5 - c not applicable. a. Material input requirements per unit of output of the various processes in the model. b. Megawatt hours per ton of output. c. Thousand cubic meters per ton of output. d. Kilowatt hours per ton of output. a. Cubic meters per ton of output. f. Units per ton of output. THE STRUCTURE OF THE 1975 MODEL 191 production of an intermediate product. The latter statement applies only in the case of the particular specification of the intermediate product material balance constraint. Once interplant shipments, or imports of intermediates, are allowed, it is no longer necessary for a particular plant to produce its own intermediates. Imports of inter- mediate products did not occur in 1975, and they are highly unlikely to be efficient in the near future; therefore, they are not considered in the context of the static model. Interplant shipments did not occur in 1975 either; the case for such shipments is less clear-cut, however, and will therefore be analyzed below. MATERIAL BALANCE CONSTRAINTS ON RAW MATERIALS AND MISCELLANEOUS INPUTS (11.4) E a,p zpi + uci + Yci > 0 ce CR peP i e I The use of raw material or miscellaneous material input c in process p at plant location i is at least equal to the sum of the amounts of input c purchased domestically and imported from abroad. In this set of constraints, the coefficient a will usually be negative. CR is the set of raw materials and miscellaneous material inputs. Miscellaneous inputs normally are not imported. The coefficient ac, for miscellaneous inputs can be either expressed in physical terms (for example, cubic meters of cooling water per ton of output) or lumped together in a value term (dollars per ton of output). In this case, the former version is employed to facilitate using the model to check the consistency of data at the plant level; these data include specific requirements for some of the individual miscellaneous inputs by plant. CAPACITY CONSTRAINTS (11.5) Ebmp zpi 0 c e C i E I j e J pe P The objective function The preceding constraints state that a given pattern of fertilizer use by demand region and by fertilizer type should be met by either imports or domestic production. The latter requires that raw ma- terials be transformed to intermediate and final products-this is governed by well-defined input-output coefficients-and that installed capacity be utilized at specific rates. Only some raw materials can be imported, and no imports or interplant shipments of intermediate products can take place. Some fertilizer types (CAN 26 percent, urea, diammonium phosphate, and two compounds) had to be imported because no domestic production facilities existed in 1975; others (CAN 31 percent, CAN 33.5 percent, single superphosphate, and ammonium sulfate) were either purchased from domestic sources or imported. The selection from among the various alternatives specified is based on cost information incorporated in the objective function, in which a unit cost is specified for each of the activities in the model. The objective function includes the following cost categories, the sum of which is to be minimized: (11.7) min t = X,; + ox + THE STRUCTURE OF THE 1975 MODEL 193 where recurrent cost, including the cost of domestic raw materials; domestic transport cost; ,= import cost. Two cost categories do not need to be considered in this exercise. First, capital cost does not enter the picture, because it can be con- sidered as sunk cost (that is, a cost already incurred) in 1975, reflecting decisions that were taken previously. Whether or not installed capacity should be used is a short-term decision that may affect total benefits; cost associated with the equipment itself, irrespective of whether or not it is used, is the same whatever short- term decisions are taken, and this cost category can therefore be ignored by the sectoral planner. A similar argument applies to labor cost. In Egypt, the labor force associated with a fertilizer plant is fixed, and closely monitored by the central government. Only if production is halted entirely, and the plant is scrapped, can this cost category be considered as variable. Since the plants were in operation in 1975, labor cost can be ignored in the model. Labor cost can be added separately to the other cost categories, if total production cost is desired. RECURRENT COST The recurrent cost included here is that incurred for purchasing local raw materials and miscellaneous material inputs included in the set CR. Several raw materials can be purchased locally: - Electricity, which is used in ammonia production at Aswan in the water-electrolysis process. - Coke-oven gas, which is used in ammonia production at Helwan. The gas is obtained from the coke plant attached to the fertilizer and iron and steel complexes. - Phosphate rock, which is used in the production of ssp at Abu Zaabal, Assiout, and Kafr El Zayaat. The rock is obtained primarily from the Sebeya mines and each plant pays a different delivered price for the rock. These differences arise because of transport cost, part or full ownership of the mines, and special agreements. . Limestone, which is used in the production of CAN at Helwan and Aswan. Helwan obtains its limestone from mines close to the 194 AN ANALYSIS OF THE EGYPTIAN FERTILIZER SECTOR IN 1975 Table 17. Prices of Raw Materials from Domestic Sources by Plant Location, 1975 (Egyptian pounds) Electricity Coke-oven (per Phosphate Sulfuric Limestone gas (per megawatt rock acid Plant location (per ton) cubic meter) hour) (per ton) (per ton) Helwan 1.2 16 - - 3 Aswan 1.2 - 1 - - Assiout - _ - 3.5 - Kafr El Zayaat - - - 5.0 - Abu Zaabal - - - 4.0 - - = not applicable. complex and Aswan's supply comes from quarries belonging to the complex, about forty-five kilometers from the plant. Sulfuric acid, which strictly speaking is an intermediate product produced by the fertilizer complex. This is true in all cases except one-at Helwan it is a by-product of the coke plant and is used to produce 8,000 tons of ammonium sulfate a year. For the purposes of this study, therefore, sulfuric acid is treated as a raw material, but only in the production of ammonium sulfate at Helwan. The prices assumed for raw materials from domestic sources, at the different plant sites, are given in table 17. The miscellaneous material inputs include electricity (except at Aswan), boiler feed water, cooling water, pressurized steam, and bags.3 The prices assumed to reflect the 1975 situation are given in table 18. The detailed statement in the objective function for this cost category is: (11.8) A= LUi ceOR ieI 3. The amount of high pressure steam is aggregated over the pressures used in the plant. Although this is a fairly crude approximation, the procedure should not significantly bias the results because the cost of steam is only a small part of the total operating cost. The bags are 50 kilogram polyethylene. THE STRUCTURE OF THE 1975 MODEL 195 Table 18. Prices of Miscellaneous Inputs, 1975 (Egyptian pounds) Miscellaneous input Price Electricity (per kilowatt hour) 0.007 Blast-furnace gas (per cubic meter) 0.0075 Cooling water (per cubic meter) 0.031 Pressurized steam (per ton) 1.25 Bags (per bag) 0.28 where p'. is the purchase price for each raw material and mis- cellaneous output. TRANSPORT COST Three modes of transportation are possible: road, rail, and water- way. The bulk of Egyptian fertilizer, however, is transported by lorries. Specifically, the following assumptions have been made: * all shipments of final products from plants to the marketing centers are undertaken by lorry; * all imports of final products from the port of entry to the market- ing centers are transported by lorry; . the transport of all raw materials to the plants is undertaken by barge or by barge and lorries where appropriate. The rate schedules for each mode of transportation were obtained in consultation with the General Organization for Industrialization: Rate schedule Transport mode (Egyptian pounds per ton) Road transport 0.5 + 0.0144 X distance in kilometers Waterway transport 1.0 + 0.003 X distance in kilometers Each rate schedule has both a fixed and a variable cost. The fixed cost includes the costs of loading, unloading, and handling. The distances involved in transporting the various products are presented in tables 19 and 20. Table 19 represents the road distance from each plant to each marketing center in the relevant governorate, 196 AN ANALYSIS OF THE EGYPTIAN FERTILIZER SECTOR IN 1975 Table 19. Road distances from Plants and Alexandria (for Imports) to Marketing Centers for Final Products (kilometers) Kafr El Abh Alex- Marketing center Zayaat Zaabal Helwan Assiout Aswan andria Alexandria 119 210 244 607 1,135 - Demanhur 42 50 184 547 1,065 600 Tanta 20 65 122 485 1,003 166 Kafr El Sheikh 20 105 162 525 1,043 161 El Mansura 58 138 152 515 1,033 224 Damietta 131 216 233 596 1,114 283 Zagazig 78 60 110 473 991 256 Ismailia 241 142 173 536 1,054 381 Suez 246 224 178 541 1,059 386 Shibin El Kom 33 154 109 472 990 157 Benha 66 97 76 439 957 190 Giza 133 48 9 372 890 287 Beni Suef 248 163 105 257 775 359 El Fayoum 230 145 88 308 826 341 El Minia 372 288 230 132 650 384 Assiout 504 420 362 - 518 616 El Kharga 703 619 561 199 519 815 Sohag 603 519 461 99 419 715 Quena 746 662 604 242 276 858 Aswan 1,022 938 880 518 - 1,134 as well as the distance from Alexandria, through which imports are assumed to enter Egypt. Table 20 presents the distances covered by imports of raw materials from Alexandria to the plants. These imports are transported by barge either directly to the plant or, in some cases, to a point a few kilometers away from the plant, with the remaining distance being covered by lorries. This takes into account cases-such as at Kafr El Zayaat-where the canal does not go all the way to the plant site and the products must be unloaded a few kilometers away and then transported by lorries. Transport cost is represented in the objective function in the following way: (11.9) * = L (E E cij Xeij + E gcj v,Q) + E E i Vci ceCF iel \ if jeJ c,CR itl The first block stands for the transport cost for all domestically THE STRUCTURE OF THE 1975 MODEL 197 Table 20. Distances from Alexandria to Plant Locations for Imports of Raw Materials (kilometers) Distance Distance Total Plant location by waterway by road distance Kafr El Zayaat 104 6 110 Abu Zaabal 210 0.1 210.1 Helwan 183 - 183 Assiout 583 - 583 Aswan 1,087 10 1,097 Table 21. Prices of Imports, 1975 (U.S. dollars per ton) Average c.if. import Product price Pyrites 17.5 Sulfur 55 Calcium ammonium nitrate: 26 percent 75 Calcium ammonium nitrate: 31 percent 90 Calcium ammonium nitrate: 33.5 percent 100 Urea: 46 percent 150 Ammonium sulfate: 20.1 percent 75 Diammonium phosphate: 18-46-0 175 Compound 1: 25-5.5-0 100 Compound 2: 30-10-0 130 produced and imported fertilizers; the second block stands for the transport cost associated with the domestic shipment of imported raw materials. The transport cost for domestic raw materials and miscellaneous inputs is assumed to be included in the supply price at the plant site. IMPORT COST The final component of the objective function covers the import cost of fertilizers and raw materials. Average c.i.f. import prices for 1975 are given in table 21 for all relevant products. 198 AN ANALYSIS OF THE EGYPTIAN FERTILIZER SECTOR IN 1975 The import cost in the objective function can be specified as: (11.10) = E Spcj Vcj + E Zpevi Vci c CF jEJ ceCR itl Recapitulation Before the results obtained with the 1975 model of the Egyptian fertilizer sector are considered, the more conventional formal state- ment of the model will be given. First, the symbols are categorized and defined. Symbol Defimition THE SETS AND INDEXES iel Plant locations Aswan Helwan Kafr El Zayaat Assiout Abu Zaabal jeJ Demand regions Alexandria Behera Gharbia Kafr El Sheikh Dakahlia Damietta Sharkia Ismailia Suez Menoufia Kalubia Giza Beni Suef Fayoum Minia Assiout New Valley Sohag Quena Aswan meM Productive units Sulfuric acid: sulfur Sulfuric acid: pyrites Nitric acid Ammonia: water electrolysis THE STRUCTURE OF THE 1975 MODEL 199 Symbol Definition THE SETS AND INDEXES (continued) Ammonia: coke-oven gas Calcium ammonium nitrate Ammonium sulfate Single superphosphate: 15.5 percent PEP Processes Sulfuric acid: sulfur Sulfuric acid: pyrites Nitric acid Ammonia: water electrolysis Ammonia: coke-oven gas Ammonium sulfate Single superphosphate: 15.5 percent Calcium ammonium nitrate: 31 percent Calcium ammonium nitrate: 33.5 percent ceCF Fertilizers Calcium ammonium nitrate: 26 percent Calcium ammonium nitrate: 31 percent Calcium ammonium nitrate: 33.5 percent Single superphosphate: 15.5 percent Ammonium sulfate Urea Diammonium phosphate Compound 1: 25-5.5-0 Compound 2: 30-10-0 CECI Intermediate products Ammonia Nitric acid Sulfuric acid cECR Raw materials and miscellaneous inputs Coke-oven gas Phosphate rock Limestone Sulfuric acid: Helwan Elemental sulfur Pyrites Electricity Blast-furnace gas Cooling water Steam Bags THE VARIABLES z Process level (production level) x Domestic shipment activity v Imports u Domestic raw material purchases 200 AN ANALYSTS OF THE EGYPTIAN FERTILIZER SECTOR IN 1975 Symbol Definition THE VARIABLES (continued) Total cost 4,b Domestic recurrent cost 'kx Transport cost 4,, Import cost THE PARAMETERS a Input (-) or output (+) coefficient b Capacity utilization coefficient k Initial capacity pd Domestic price pv Import price (c.i.f.) ci Actual consumption of fertilizer in 1975 A Transport cost Second, the constraints of the model are specified. MARKET REQUIREMENTS (11.11 ) >3 X1ij + Vcj 2 6 j c E CF MATERIAL BALANCE CONSTRAINTS ON FINAL PRODUCTS (11.12) I: ap Zpi 2 > X1ii c e CF peP ifJ ieI MATERIAL BALANCE CONSTRAINTS ON INTERMEDIATE PRODUCTS (11.13) E a,p zpi > ° C e Ci pep ie MATERIAL BALANCE CONSTRAINTS ON RAW MATERIALS AND MISCELLANEOUS INPUTS (11.14) 3E a,p zpi + Uci + Vci 2 0 c e CR pep ieI CAPACITY CONSTRAINTS (11.15) -Ebmp zpi < kmi E eI PeP meM NONNEGATIVITY CONSTRAINTS (11.16) xci0, V, Zp,i u > O c Ce C i e I jEJ pefP RESULTS OF THE BASIC 1975 MODEL: SOLUTION 1 201 THE OBJECTIVE FUNCTION (11.17) min + ox + DOMESTIC RECURRENT COST (11.18 ) -, z S E c c Ce CR iel TRANSPORT COST (11.19) = ( cij Xij + Ecj Vej) + E Eci VRi CccCp iel jeJ j6J ceCR itI IMPORT COST (11.20) Pcvj vcj + E EPi Vp i CICp jfJ cfCR ifl Results of the Basic 1975 Model: Solution 1 Planning models like those presented in this volume typically produce a wealth of numerical results. Some of the results are in quantitative terms, and they concern the level of production of final and intermediate products at each location and for each productive unit, the quantity of product imported, the quantity of product shipped from each point of supply (domestic production site or point of entry for imports), the number of ton-kilometers for each mode of transport, and the quantity of domestic raw materials and miscellane- ous inputs used at each site. In addition, a variety of results are obtained in value terms, relating to operating cost, cost of domestic raw materials, import cost, and domestic transport cost. In combina- tion with this financial information, shadow prices derived from the model's solution can be put to a number of uses, as will be explained later. Additional categories of results can be achieved using either the less restrictive versions of the model described in subsequent sections of this chapter, or, in particular, the dynamic models described in chapter 12. Until recently, the sheer magnitude and detail of the results were a severe obstacle to the analyst. Using modern data management sys- tems and a computer, however, the analyst can obtain the desired 202 AN ANALYSIS OF THE EGYPTIAN FERTILIZER SECTOR IN 1975 information in tabular form that is easily assimilated. Among other things, these systems permit the analyst to concentrate on well-defined areas of interest. That is obviously important in the case of multi- faceted planning problems such as in the fertilizer sector since the sector planner, the plant manager, the agronomist, the engineer, and the farmer are all likely to be interested in different aspects of the solution. In this presentation of the results, it is clearly impossible to cater to all of these diverse needs. But that would not appear to be necessary in any case, for what is of interest is the type of results obtained rather than the complete and detailed numerical results as they apply to Egypt. At the same time, it is impossible to avoid empirical results altogether. In what follows, only the numerical output that is needed to convey a flavor of the nature of the results obtained is presented. Greater detail is found in a document that was produced specifically for Egyptian planners and sector specialists at the World Bank, in the course of conducting the present study.4 Production and imports If actual production and import levels of fertilizer products in 1975 are compared with those obtained by means of the model, there appears to be close correspondence in most cases. This is not too surprising, given the restrictive conditions under which the model was solved. Several products for which no domestic production facilities existed were consumed in 1975; obviously, to meet the demand requirements, these products needed to be imported in the appropriate quantities. Whenever domestic production facilities exist, they are used rather than imports to meet demand requirements; given that neither labor nor capital cost is taken into account in the static model, and that almost all domestic production facilities have a transport cost advantage over imports, this result is also fairly straightforward. Nevertheless, some interesting discrepancies between the original data and the model's solution appeared. For example, in the case of 4. Armeane M. Choksi, Alexander Meeraus, and Ardy Stoutjesdijk, "A Planning Study of the Egyptian Fertilizer Sector," Staff Working Paper, no. 269 (Washington, D.C.: World Bank, July 1977). RESULTS OF THE BASIC 1975 MODEL: SOLUTION 1 203 the fertilizer plant at Aswan, it was found that the plant management's estimate of the production level in 1975 was substantially above that obtained with the model. Upon inspection, it appeared that the model's result was the logical consequence of the fertilizer demand estimate made by the Ministry of Agriculture; that is, the plant was producing exactly the amount of product the ministry claimed was used in 1975. Part of the discrepancy might possibly be explained by inventory build-up.5 It would appear, however, that there is a basic inconsistency in the data: either demand was underestimated or production levels were overestimated. Since the estimated production level was close to design capacity, even though extensive repairs were carried out on the plant's equipment in 1975, it would seem that the major source for the discrepancy is an erroneous production estimate at Aswan. A comparatively minor discrepancy was found in the case of the phosphatic fertilizer plant at Assiout, where again the reported production level was above that obtained with the model. At that time, the only competitor, at Abu Zaabal, was producing at full capacity, as the plant claimed. But the reported level of demand could not possibly permit both plants to be working at full capacity, unless large inventories were built up. Since Abu Zaabal had lower marginal cost of production (derived from the shadow price information in the solution) than Assiout-an advantage that was apparently not eliminated by a possible transport advantage that the latter might have had with respect to the markets-Abu Zaabal produced at full capacity in the model while Assiout did not. Some aggregate results To provide a basis for comparing subsequent versions of the model, described in the next few sections, some of the important overall results obtained thus far are summarized in table 22. As was ex- plained above, neither capital cost nor labor cost is considered in the model. Moreover, in line with what actually happened in 1975, it is assumed that no interplant shipments of intermediates took place, and that road transport dominated. 5. Exports of fertilizer material were not reported to have taken place in 1975. 204 AN ANALYSIS OF THE EGYPTIAN FERTILIZER SECTOR IN 1975 Table 22. Summary of Aggregate Results of the 1975 Model Numerical Percent Variable results distribution Demand/supply of fertilizers (thousands of tons) Domestic supply 805.7 46.3 Imports 936.1 53.7 Total fertilizer requirements 1,741.8 100.0 Cost categories (millions of Egyptian pounds) Operating cost, 8.1 13.8 Domestic raw materials costb 3. 7 6.3 Transport cost 5.9 10.0 Import cost, 41.1 69.9 Total cost 58.8 100.0 Transport requirements (millions of ton-kilometers) Waterway 19.5 5.4 Road 337.7 94.6 Total transport requirements 357. 2 100.0 a. These refer to the cost of the total use of electricity (excluding Aswan), blast-furnace gas, cooling water, steam, and bags at all plants in the sector. b. These are the costs of limestone, coal gas, electricity (at Aswan only), phosphate rock, and sulfuric acid (at Helwan only) used in the sector. c. Imports are evaluated at the official exchange rate. Introducing Interplant Shipments: Solution 2 According to available data, no interplant shipments of inter- mediate products occurred in Egypt in 1975. This section analyzes whether such interplant shipments might have been efficient in those cases where capacity constraints did not restrict them from taking place. For that purpose, the model needs to be modified in several respects. First, a plant need not produce the intermediate products it requires; it may purchase such products from other domestic plants. This possibility can be incorporated in the material balance constraint for intermediate products as follows: (11.21) E api zpi + E (Xvi, - x,ii') > 0. ce CI pip i, if INTRODUCING INTERPLANT SHIPMENTS: SOLUTION 2 205 Table 23. Interplant Distances for Shipments of Intermediate Products (kilometers) Kafr El Abu Plant location Zayaat Zaabal Helwan Assiout Aswan Kafr El Zayaat - 85 230 479 983 Abu Zaabal 85 - 61 420 924 Helwan 230 61 - 400 900 Assiout 479 420 400 - 504 Aswan 983 924 904 504 - - = not applicable. The difference between this formulation of the constraint and that in (11.3) is the second term on the left-hand side, which stands for shipments of intermediate product c from all other plant locations i' to plant location i, minus the shipments of intermediate product c from plant location i to all other plant locations i'. Second, the objective function needs to be modified so that cost considerations can be brought to bear upon the choice. Transport cost is the only component of the objective function that needs to be altered; it is now formulated as follows: (11.22) ON = X (Z E Acij x'ij + E Icj Vcj ceCF i6I jeJ jeJ + S,ii Xcii' + Epi Vci ceCI i61 i'd ceCR iel where the third term represents the transport cost for interplant shipments of intermediate product. Third, this new version of the model requires a distance table with the transport distances among the relevant plant sites and a rate structure that can be assumed for interplant shipments. The interplant distances are given in table 23. It is assumed that interplant shipments take place by rail and that the applicable rate schedule (in Egyptian pounds per ton) is: Rail transport = 0.5 + 0.03 X km. 206 AN ANALYSIS OF THE EGYPTIAN FERTILIZER SECTOR IN 1975 With these three modifications, the option of interplant shipments of intermediate products can be analyzed. The import of final products is not affected, and the sector remains equally dependent on imports from abroad for fertilizer material. The production of ammonia from coke-oven gas at Helwan, however, is sufficiently expensive to make purchase of ammonia from Aswan attractive on the transport cost assumptions. At all plants, both production and shipments of fertilizer materials remain unchanged, but relaxing the interplant shipment constraint increases ammonia production at Aswan by almost 40 percent and brings it to within 10 percent of full-capacity utilization. In terms of the total cost of meeting fertilizer demand in 1975, the cost saving associated with interplant shipments is not negligible- 1.4 million Egyptian pounds for 1975 alone, or 2.4 percent of total cost for that year. Even if it is taken into account that no adequate shipment facilities for ammonia exist at Aswan, it might be worth looking into the scope for interplant shipments to Helwan in some more detail. Once the model is converted into a dynamic planning model, however, and once other sources of nitrogenous fertilizer material enter the picture, this possibility may turn out to be less attractive. Full Capacity Utilization: Solution 3 The 1975 model incorporates a capacity constraint on all productive units in the Egyptian fertilizer industry that is close to actual produc- tion levels achieved in that year. To the extent that the objective of the static model is to reproduce what actually happened in 1975, the choice of this capacity constraint is justified. For several reasons, however, it is of interest to assume that the various productive units could have attained capacity utilization rates of 100 percent, if these were efficient. First, by comparing the total cost of meeting fertilizer requirements in 1975 at these two different levels of capacity utiliza- tion-actual and full-one can estimate the benefits that could be derived from higher levels of domestic production and compare these with the cost that would have to be incurred to remove the constraints on higher levels of capacity utilization, whether they are import restrictions on spare parts, raw materials, or the like. Second, because several plants compete for the same markets, potentially higher FULL CAPACITY UTILIZATION: SOLUTION 3 207 production levels permit some further analysis of the relative efficiency of existing plants, as measured by market penetration. To introduce this modification, the model structure itself remains basically unchanged; only the numerical value associated with constraint (11.5 ), kmi, changes. In overall terms, the solution with full capacity utilization costs 2 million Egyptian pounds (or 3.5 percent) less than the original solution, and 600,000 Egyptian pounds (or 1 percent) less than the solution incorporating interplant shipments to meet total fertilizer requirements in 1975. This gain, although not insignificant, is surprisingly small. Several reasons account for this. First, with a few exceptions, rates of capacity utilization in 1975 were not particularly low; only one plant (Helwan) produced below 85 percent of design capacity. Second, for several products, demand constraints prevent all plants from producing at full capacity, and, in fact, the solution of this version of the model permits the relatively efficient plants to be identified easily.6 Solution 3 introduces several major changes from Solution 2. In the case of phosphatic fertilizer production, only the plant at Abu Zaabal benefits from the relaxation of the capacity constraint and increases its production and sales, at the expense of both other plants producing single superphosphate. The production of sulfuric acid on the basis of imported pyrites, as opposed to elemental sulfur, is relatively expen- sive. Consequently, the relaxed capacity constraint shows higher capacity utilization of sulfuric acid production facilities based on elemental sulfur and sharply reduced imports of pyrites. This applies particularly to Kafr El Zayaat, where facilities for both processes are in use. In nitrogenous fertilizer production, an interesting switch occurs. The plant at Aswan further increases its output of ammonia and, in fact, reaches full capacity utilization. The plant ships 35 percent of total output to Helwan, where it is converted to CAN 33.5 percent. The Helwan plant's shipments of CAN 33.5 percent compete directly with Aswan's own shipments of CAN 31 percent in Middle and Upper Egypt, and result in slightly lower production levels at Aswan than in previous solutions. 6. If demand constraints do not prevent all plants from producing at full capacity, one may want to introduce artificially low demand constraints to gain insight into the relative efficiency of individual plants. Shadow prices on product shipments permit the same analysis. 208 AN ANALYSIS OF THE EGYPTIAN FERTILIZER SECTOR IN 1975 The Composition of Fertilizer Demand: Solution 4 Until now, the objective of the 1975 model was to find the least-cost pattern of supply to meet specific fertilizer requirements in each of the twenty demand regions, stated in terms of the quantity and types of fertilizer actually used in 1975. The choice was limited to the source of supply. Since several fertilizers are close substitutes, in the sense that they can provide the same major nutrient (or nutrients, in the case of multinutrient fertilizers), it is worth investigating whether the nutrient requirements in Egyptian agriculture are currently being met by the least-cost mix of fertilizer types.7 To carry out this investigation, a number of modifications need to be made. First, the set CQ, which is the set of nutrients considered in the study, must be introduced. The set contains two elements c': nitrogen and phosphorus. Second, the nutrient content of all fertilizers in the model, denoted by the coefficient a,c,, needs to be specified. With the help of this coefficient, it is possible to express shipments of fertilizer material in terms of nutrient content. The constraint that specifies market requirements is now written as follows: (11.23) a .. x.ij + E acc, vcj > de'. jEJ ceCF iel c,CF C e CQ The difference between this statement and constraint (11.1) is simply that now a given quantity of nutrients needs to be supplied in each region, d,j, rather than a given quantity of fertilizer material. With this modification, the model is free to select the least-cost pattern of fertilizer supply as long as the nutrient requirements are met. Fairly sizable savings could have been achieved in 1975 if a different combination of fertilizers had been used. In fact, compared with the original solution, the objective function value falls by more than 12 percent, or 7.2 million Egyptian pounds. To identify the changes that take place exclusively as a result of the modified demand con- 7. The degree to which substitution among categories of fertilizers can occur is a point of controversy among experts. Different fertilizers may contain the same major nutrient (N, P, K) or a combination of those; usually, however, they contain widely varying amounts of other elements affecting plant growth. Moreover, there may be important variations in other properties. SUMMARY 209 straint, the present solution should be compared with Solution 2, which is identical in its specification in all other respects. No major changes occur in the supply pattern of phosphatic fertilizer. Domestically produced single superphosphate remains competitive with the various multinutrient compounds that can be imported, given the model's price specification. If the compounds do not have to be used, they are no longer imported, with the exception of diammonium phosphate which appears competitive in the coastal demand regions with single superphosphate produced at Kafr El Zayaat. The imported quantity of DAP iS small, however. The more interesting changes occur in the case of nitrogenous fertilizers. In previous solutions, as in actual practice, a wide variety of fertilizer materials in this category are imported. If the model is free to choose the composition of fertilizer supply, the picture changes radically. With the exception of some CAN 26 percent destined for Alexandria, imports are limited to CAN 31 percent, and all other imported nitrogenous fertilizers disappear from the scene. Similarly, domestic production of CAN 31 percent increases to full capacity (at Aswan), and interplant shipments of ammonia from Aswan to Helwan virtually stop. On the basis of the solution, the Aswan plant appears to be a relatively efficient supplier of nitrogen nutrients in Upper Egypt; because of a transport cost disadvantage, however, it cannot compete in other parts of Egypt with the Helwan plant or with imports. As long as part of nitrogen demand is met by fertilizers other than CAN 31 percent, the plant runs into a market constraint. As soon as the constraint on fertilizer type is relaxed, the plant supplies all nitrogen requirements in Upper Egypt. The plant at Helwan is affected by these changes only to the extent that it can no longer pur- chase surplus ammonia from Aswan and has to produce its own ammonia requirements from coke-oven gas. Summary In table 24, a summary of the results is presented. From the statistics provided, it is clear that by far the greatest changes occur when the demand specification is modified. It would appear that efforts to increase capacity utilization in existing facilities cannot be expected to have substantial payoffs. On the other hand, the possi- bility of moderate interplant shipments of ammonia should be explored. Table 24. Summary of Results of the Static Model (amounts in millions of Egyptian pounds) Solution 4: Solution 3: interplant Solution 2: interplant shipments and Solution 1: interplant shipments no fertilizer original shipments and full capacity recommendations Ohjective function Amount Percent Amount Percent Amount Percent Amount Percent Transport cost 5.9 10.0 6.9 12.0 7.1 12.5 6.1 11.8 Domestic raw materials cost 3.7 6.3 2.9 5.1 3.1 5.5 4.2 8.1 Operating cost 8.1 13.8 6.5 11.3 6.7 11.8 9.2 17.8 Import cost 41.1 69.9 41.1 71.4 39.8 70.1 32.1 62.2 Total value of objective function 58.8 100.0 57.4 100.0 56.8 100.0 51.6 100.0 Note: Components may not add to totals because of rounding. 12 A Medium-Term Planning Model of the Egyptian Fertilizer Sector DYNAMIC SECTORWIDE PLANNING MODELS are used in this chapter to analyze possible development strategies for the fertilizer sector in Egypt. For that purpose, the model used in the previous chapter is modified, so that the time dimension is incorporated into the structure of the model, so that investment activities (and concomitant labor hiring) can be undertaken in the sector, and so that the scope for exports can be analyzed. To describe the mathematical structure of this dynamic model, the procedure adopted in chapter 11 is repeated: that is, each set of constraints is explained in detail, and the data input required with the new formulation is indicated. Following this description, the dynamic model is presented in summary form. A computer-readable description of all the sets and all the numerical data used in the dynamic model is given in an appendix to the case study. The Structure of the Dynamic Model In many respects, the structure of the dynamic model is similar to that of the static 1975 model. Several of the constraints differ only in the sense that a time subscript is added to variables and parameters. 211 212 PLANNING MODEL OF THE EGYPTIAN FERTILIZER SECTOR The demand for fertilizers The fertilizer demand constraint in the static model evolved from a formulation in which final demand was specified in terms of fertilizer product types with no choice allowed among fertilizer types supplying the same nutrient to one in which final demand in each region was specified in nutrients, so that the model could be used to find the least-cost combination of fertilizers to make up required supply levels. The former formulation of the demand constraints was given in (I 1.1 ); the latter, in (I 1.25 ). For the dynamic model, basically the same formulation is used: (12.1) E YE atc' Xcijt + 5Z a-, V,jt > jejt5 c' e CQ cCF i*, c CF j E J t e T where the notation has the same meaning as in chapter 11, with the addition of the subscript t referring to the time period. Thus: C!C' = the percentage of nutrient c' in fertilizer c; xsijt = the shipment of fertilizer material c from domestic plant site i to marketing center j during time period t; V'j t = the shipment of imported fertilizer material c to marketing center j during time period t; d,,j, = the required supply of nutrient c' in marketing center j during time period t. The sets are: CF = set of fertilizers, I = set of plant sites, J = set of governorates, CQ = set of nutrients, T = set of time periods. The set of fertilizers in the dynamic model is larger than that in the 1975 model, so that additional production possibilities can be analyzed. The set now includes the following products (asterisks denote additions to the set as specified for the static model): cECF = Urea Calcium nitrate* Calcium ammonium nitrate: 31 percent THE STRUCTURE OF THE DYNAMIC MODEL 213 Calcium ammonium nitrate: 33.5 percent Ammonium sulfate Monoammonium phosphate* Diammonium phosphate Single superphosphate Powdered triple superphosphate* Granular triple superphosphate* Nitrophosphate* The set of plant sites I has been expanded to include the following: ieI = Aswan Helwan Kafr El Zayaat Assiout Suez* Talkha* Abu Kir* Abu Zaabal Both the set of governorates J and the set of nutrients CQ remain unchanged. The added set of time periods T contains the following elements: tET = I = 1979-81 2 = 1982-84 3 = 1985-87. In addition to (12.1), the dynamic model includes a set of con- straints that limit the freedom of choice in the model regarding the type of fertilizers that can meet fertilizer demand, which is expressed in terms of nutrients. In the case of Egypt, this restriction is necessary particularly for urea, the use of which is projected to grow only gradually. This set of constraints has the following form: (12.2) E a.. x,ijt + ae, vcjt < t de,jt, c = urea c' = nitrogen t eT je J where n t = percentage of nitrogen demand that can be met by the use of urea. 214 PLANNING MODEL OF THE EGYPTIAN FERTILIZER SECTOR The assumptions regarding demand growth in terms of nutrients were decided upon in close consultation with Egyptian sector specialists-particularly, staff of the Ministry of Agriculture. Demand for nitrogen is expected to grow at an annual rate of 4.6 percent, whereas demand for phosphorus may grow as much as 10.4 percent a year. Representatives of the nitrogenous fertilizer industry disagreed with the projection of nitrogen consumption made by the Ministry of Agriculture, and it was difficult to reconcile that disagreement. In the sensitivity analysis presented later in this chapter, the model has been used to evaluate the implications of a higher growth rate for nitrogen nutrients. A similar disagreement arose regarding the likely speed of adoption of urea by Egyptian farmers. With two sizable ammonia/urea complexes under construction, it is evident that the industry believes that farmers will switch rapidly from the traditional, low-analysis fertilizers such as calcium ammonium nitrate to the more efficient urea. Experts of the Ministry of Agriculture, however, believe this will not occur rapidly. If constraint set (12.2) is used, explicit restrictions on the use of urea can be incorporated in the model. A compromise between these two positions was reached by postulating the following pattern. During the first time period, 1979-81, no more than 60 per- cent of all nitrogen requirements should be met by urea. In the period 1982-84, this proportion is increased to 75 percent; in the period 1985-87, 85 percent of nitrogen is supplied by urea. To measure the cost of this possibly less-than-optimal policy, these restrictions are deleted in the sensitivity analysis. The supply offertilizers On the supply side, several modifications have to be made to the 1975 model. As was indicated above, the set of fertilizers to be considered is expanded to include products not produced or imported in 1975. This permits an assessment of whether the product mix currently in use or being produced in Egypt can be improved upon. In particular, the feasibility of the following processes can be investigated: producing fertilizers such as monoammonium phosphate or di- ammonium phosphate (both high-analysis compounds with rapidly growing international markets) for domestic use as well as exports; converting the existing single superphosphate facilities at Abu Zaabal to one that produces powdered or granular triple superphosphate; and, finally, producing nitrophosphates at Suez-a suggestion made by sector specialists at the World Bank. THE STRUCTURE OF THE DYNAMIC MODEL 215 Since projects are under construction at Abu Kir and Talkha, and since the Suez site is being considered for another project, the set of plant sites I was similarly expanded. Moreover, to permit exports, the activity e was introduced; so also was the set L, in which each element I represents a port of exit (either Alexandria or Suez). The material balance constraints now have the following form. MATERIAL BALANCE CONSTRAINTS ON FINAL PRODUCTS (12.3) , aepit zpit > E xjit + E exilt c E CF pp6 ji' Ie' i e I t e- T The difference between this constraint and the corresponding one in the static model is twofold. First, the subscript t has been added to render activity levels period-specific. Second, the possibility of exports has been incorporated. Thus, the constraint now specifies that the production of fertilizer type c by all processes p at plant site i during time period t should at least equal shipments of fertilizer product c from plant site i to all domestic markets j in time period t, augmented by all exports of that fertilizer type from that plant site in that period.I MATERIAL BALANCE CONSTRAINTS ON INTERMEDIATE PRODUCTS (12.4) . acpit zpit + x i i t + v,it > E X.iij t + E ei i t c E CI pep i ~ i'r heL E i'/ i i'# i It te T This constraint states that the production of intermediate product c at plant site i in time period t by all processes p, augmented by shipments from other plants i' and imports, should at least equal shipments of intermediate product c to other plant sites and exports. Compared with the corresponding constraint in the static model, this constraint permits products to be imported, exported, and shipped between plants. MATERIAL BALANCE CONSTRAINTS ON RAW MATERIALS AND MISCELLANEOUS INPUTS (12.5) , a.pit Zpit + u.it + Vcit > 0 C e CR 1.ThPxprtp epbliy scostane;th poedreuedi describ e b 1. The export possibility is constrained; the procedure used is described below. 216 PLANNING MODEL OF THE EGYPTIAN FERTILIZER SECTOR This constraint is simply a restatement of constraint (11.4) in the static model, with the addition of the time subscript t. MATERIAL BALANCE CONSTRAINTS ON LABOR For all practical purposes, the labor force associated with fertilizer plants represents a fixed cost to the plant, much the same as equipment cost. There can be no short-term adjustments in the labor force in response to production changes. In the static model, therefore, labor cost did not need to be incorporated. In the dynamic model, in which it is assumed that similar treatment of labor will be imposed upon plants, labor costs should be modeled in the same way as investment cost, in the sense that the labor force is fixed with capacity installed. Thus, the material balance constraint for labor has the following form: (12.6) Uft - E ci ymir + PCi, hmir) > 0. c e CO r, and a variable cost, v. The fixed charge is 180,000 Egyptian pounds a THE STRUCTURE OF THE DYNAMIC MODEL 217 year, which is the amount required if a decision is made to build a productive unit. This includes the cost of supervision, administrative personnel, and overhead charges; it gives rise to economies of scale in labor. The variable cost is 0.45 Egyptian pounds a year per ton of capacity of total product. Labor costs are a function of capacity installed rather than size of the investment. CAPACITY CONSTRAINTS (12.7) Z bmpit zit < k-i + 5 [hmir + (f UJi,)] i e I peP TCT 1/ gG m eM t e T The capacity constraint expresses the condition that the number of capacity units used on productive unit m per unit of output of process p at plant site i in time period t cannot exceed the available capacity. Since the dynamic model explicitly permits the expansion of capacity at any of the sites in the set I, however, it is necessary to take this possibility into account in the capacity constraint. Accordingly, the term Ej hmir denotes the amount of capacity constructed at plant ,CT r 0 C e C i, i' e I ye J I E L mEM p E P t e T The objective function The objective function of the dynamic model is formulated in terms similar to those of the static model-that is, as a cost-minimizing function. Since both investment and upgrading activities, as well as exports can be considered in the dynamic model, the formulation of the objective function needs to be expanded. In addition, the time factor needs to be taken into account. Finally, the impact of working capital on total cost needs to be considered. In summary form, the objective function of the dynamic model is specified as follows: THE STRUCTURE OF THE DYNAMIC MODEL 221 (12.15) min = E at (4Kt + ¢ot + Oxt + kt, + OT't -Oct t6T r _F _ _ [otalTotal Total Discount capitlal recurrent discounted factor cost dur- + cost dur- L cost ing time Iurg time cost _ ~~~pegriod t ing tim _ ~~~~~~periot FTotal Total Total Total ToalworkmgTta export transport wr I import + cost dur- + apital + cost dur- reurnge ing time cost dur- .ig time during mg tlme ing time m metime period t _ pering t periodt t periodt J periodt pe iot where (12.16) at (I= +1 T=1 [Discount rate] A full derivation of the discount rate is given in chapter 8. In this particular application of the model, there are three time periods- namely, 1979-81, 1982-84, and 1985-87. The real discount rate p used in the model is 10 percent. Each of the components of the objective function can now be specified in detail. CAPITAL COST CHARGES (12.17) C = Ei E am (umi, ymir + VmiT hmi,) ,cT if I mfM + E E: aag (JgiT ugiT id g.G where (12.18) am= ±( + - 1 Capital recovery L factor I 222 PLANNING MODEL OF THE EGYPTIAN FERTILIZER SECTOR Table 25. The Investment Cost Functions for the Dynamic Model Fixed charge: Variable charge: t (millions v (tliousands of U.S. of U.S. dollars Productive unit dollars) a ton per year) Urea 16.7 52.2 Ammonium phosphate 4.5 32.4 Single superphosphate 7.5 25.5 Triple superphosphate 7.0 34.2 Sulfuric acid: elemental sulfur 8.5 33.0 Ammonia: natural gas 47.9 103.8 Phosphoric acid 10.0 57.0 Nitrophosphates 8.0 70.0 The derivation of the capital recovery factor orm is also given in chapter 8. With p equal to 10 percent, and with the useful life of capacity, Am, placed at an average of fifteen years, the capital re- covery factor is equal to 0.1315. The term (comir ymnt + vrai hmir) was likewise explained in chapter 8. To recapitulate briefly, w-it is the fixed element in the linear approxi- mation of the investment cost function; since it is a constant in the objective function, it needs to be eliminated if no capacity is con- structed. To that end, it is linked to y-i,, which is a variable that can assume only the values 0 or 1, where the former is chosen if no capacity is constructed. The variable hi, denotes the capacity scale that is constructed, whereas the coefficient Pi, is the variable portion of the investment cost function. Table 25 gives the value for woi, and P.i, for the productive units that are relevant to the dynamic model. The investment (or capital) costs used in this study are based on Tennessee Valley Authority estimates for different plant sizes and are given in constant 1975 U.S. dollars. They include U.S. battery limits, auxiliary and support facilities, spare parts, training, start-up, fifteen-day storage for ammonia and phosphoric acid, twenty-day bag storage, and power and steam generation. In addition, the cost of site preparation and off-site facilities needs to be incorporated. This is done by using a site factor of 1.3 for all plant sites in Egypt. This factor is applied to the total capital cost. The term (Z g E - E i,. Uir) refers to the cost of the upgrading activity-that is, the cost of converting the ammonia unit based on THE STRUCTURE OF THE DYNAMIC MODEL 223 coke-oven gas to one that uses natural gas as a feedstock. The estimated cost of this conversion is US$20 million, which is the value of o,i,., while o-, is the capital recovery factor and Ui,- is the 0-1 variable. DOMESTIC RECURRENT COST (12.19) t E Epcit UCtt -CfRUCO if With the exception of the addition of labor cost (labor requirements for new plants are included in the set CO), the treatment of domestic recurrent cost in the objective function of the dynamic model is similar to that in the static model. Prices similar to those in tables 17 and 18 have been used for miscellaneous inputs and raw materials; it is assumed that these prices will remain constant in real terms during the planning period. In the dynamic model, however, the num- ber of raw materials is larger, and they are available at more sites. Table 26 gives the prices that have been assumed at the various relevant sites. TRANSPORT COST (12.20) i = E E cijtXijt ceCF ieI j6j + E adit voit + L, tE h ei eeilto jeJ ierI /e1 + . (, E A) dEit c'e CQ ceCF iel ceCF jeJ t e T MATERIAL BALANCE CONSTRAINTS ON UREA CONSUMPTION3 (12.25) >2 acc' Xcjt +± aLxcc' v,t < r/t 4,i c = urea ieI C' = nitrogen jeJ t e T MATERIAL BALANCE CONSTRAINTS ON FINAL PRODUCTS (12.26) a.pit zit> > X~jt+ Ze.,lt ceCF p6P je6 teL ieI IeT 3. In chapter 13, seven scenarios or different project planning strategies are analyzed. This constraint is not included in scenario VII. 232 PLANNING MODEL OF THE EGYPTIAN FERTILIZER SECTOR MATERIAL BALANCE CONSTRAINTS ON INTERMEDIATE PRODUCTS (12.27 ) a,ia ±Zpit x0i ± v0i Ž > Xj + 2 e,i I t c E CI pEP i'la iEle i i e I i'i i/si t,e T MATERIAL BALANCE CONSTRAINTS ON RAW MATERIALS AND MISCELLANEOUS INPUTS (12.28) E a.pit*it + U,it + vcit > 0 C e CR peP ieI tf T MATERIAL BALANCE CONSTRAINTS ON LABOR (12.29 ) Ucit 2 2 (WCiT ymiT + vPir hli,) > 0 ce CO reT ,mM ieI r t~~~~~~~~~~~~~~~ t e T CAPACITY CONSTRAINTS (12.30) 2 bmpit zpit < k.i + 2 ±hmi7 + E (fmgiUgir)] i e I pNP 'T51t SeG m eM t e T MAXIMUM CAPACITY EXPANSION CONSTRAINTS (12.31) hmit < hmitymit iE I mME tf T MUTUAL EXCLUSIVITY CONSTRAINTS (12.32) E U t < 1 g G teT'te INTEGRALITY CONSTRAINTS (12.33) Yi = 0 or 1 el mME t e T (12.34) Ugit = 0 or 1 geG i T I tfE T THE MATHEMATICAL FORMULATION OF THE DYNAMIC MODEL 233 EXPORT CONSTRAINTS BY COMMODITY AND PLANT SITE4 (12.35) >2 e at < E Zcilt c e CF leL eL ifel teT AGGREGATE EXPORT CONSTRAINT5 (12.36) E E E ectit < e, t e T ceCF iel teL (12.37) 2 E E 2e0.i < e t = ceOF iel teL EXPORT CONSTRAINT BY MARKET SHARE" (12.38) 2 >2 E > cijt+ < Xt , eaist IT -1} -CF i6I ZtL ceCF itl leL (12.39) >2 E 52 e-zt±Ž > 2 E > e.ij5 eCif iel teL cWCF iel 1,L NONNEGATIVITY CONSTRAINTS (12.40) zit, xcijt, xci±t, .it, v,jt, e.cit, uct hmir ymir, c c C UgJir > O g e G i,i'El jeJ IEL meM pPE OBJECTIVE FUNCTION (12.41) min~= E o x A teT where (12.42) = 2( + ptO(51>7 4. Not utilized in scenarios III, IV, and V. 5. Constraint (12.36) not utilized in scenarios III, IV, and V; constraint (12.37) utilized in scenarios IV and V only. 6. Utilized in scenarios IV and V only. 234 PLANNING MODEL OF THE EGYPTIAN FERTILIZER SECTOR CAPITAL COST CHARGES (12.43) 4't E [ E E a, (Wnir y-i± + vnir hmir) T< t + >2 g U Wgri Ugir] iel geG where (12.44) U (1 + p ) 1 DOMESTIC RECURRENT COST (12.45) ¢kf = 2 Ei ceCRUCO i4r TRANSPORT COST (12.46) 0)x = E (E >2cijtxcit + >AcpitVcjt ~cyC iel jej jf j + Y E Aci I eci zt) + 1: 1 E Acii t Xcii, i>I ZeL c/C l e) i± Z + A pit vci, c6CR iel WORKING CAPITAL (12.47) otpt = 0.025 (P#t + c Epcvi e vit) (O ceCR iel IMPORT COST (12.48) 4)rt = >2 Pc2 V,,t + E E Pcpit Vit ceCF je, ceCR iel EXPORT REVENUE (12.49) Pe= t 2 p eiIgt ei It ceCF iel leL 13 Results of the Dynamic Analysis THE KIND OF RESULTS obtained with the model outlined in the pre- vious chapter are described in this chapter. This description is in- tended primarily to convey a flavor of the nature of these results, not to report on them in exhaustive detail.' The model has been used to address several of the basic questions facing the planners of the Egyptian fertilizer sector, but two general concerns dominate. First, with regard to domestic demand, what is the optimal, or least-cost, pattern of supply to meet specified re- gional requirements for fertilizer material over the planning period? This question involves investigating the relative cost of domestic supply patterns of raw materials, intermediates, and fertilizer prod- ucts. In this context, specific questions may have to be addressed: for example, what is the likely rate of adoption by Egyptian farmers of relatively new fertilizers such as urea, and what should be done with 1. For a more detailed examination of these results, see Armeane Choksi, Alexan- der Meeraus, and Ardy Stoutjesdijk, "A Planning Study of the Egyptian Fertilizer Sector," Staff Working Paper, no. 269 (Washington, D.C.: World Bank, July 1977). 235 236 R.ESULTS OF THE DYNAMIC ANALYSIS individual existing plants, particularly those that turn out to be relatively inefficient? The second main area of concern relates to the role of exports. In terms of raw materials, Egypt could well be in a favorable posi- tion to build up a substantial export-oriented fertilizer industry. Even though the model is not particularly well-suited to determine the most efficient export strategy, it can be used to assess the impli- cations of the different forms an export strategy may take. This chapter is structured as follows. To begin with, the set of "basic" results-referred to as the reference solution-of the dy- namic analysis are presented in summary form; these results are obtained on the basis of what are considered the most likely assump- tions, including highly restricted exports of fertilizers and a relatively slow adoption rate for urea. Then, a number of specific alternatives to the basic assumptions are investigated. First, various postulated production strategies at Suez are investigated. Second, several more optimistic export strate- gies-varying mainly in terms of the total volume of exports thought to be feasible-are analyzed. Finally, a complete import substitution policy and the complete elimination of the urea consumption restric- tions are examined. Next, a brief discussion focuses on the use of the model's results to obtain measures of domestic resource costs. The chapter concludes with a sensitivity analysis of the results. It focuses on the most impor- tant price and cost parameters in the model, as well as the demand projections that underlie the results. An appendix to this chapter presents the main numerical results in a series of summary tables. Basic Results: The Reference Solution Depending upon one's confidence in the robustness of the nu- merous assumptions underlying the estimates of the parameters of the dynamic model, one can consider the numerical results obtained with this model as more or less basic. Sectoral analysis carried out without a formal planning model can frequently offer the policy- maker only basic results. The present approach, however, has two major advantages. First and foremost, the full set of assumptions made in the analysis must be explicitly stated; this was done in the previous chapter. Second, once a basic set of solutions to the model BASIC RESULTS: THE REFERENCE SOLUTION 237 has been obtained, alternative scenarios can be generated on the basis of different assumptions, and thorough sensitivity analysis of crucial parameter estimates can be easily carried out. With a model of this type, it is standard procedure to solve the model using estimates of the parameters that the analyst considers most likely to materialize. Since many of these estimates involve projections into the future, the values eventually chosen may reflect a certain degree of compromise among the diverging opinions of planners involved in the study. These, in turn, provide a useful de- marcation of the range of values to be investigated in the subsequent sensitivity analysis. An overview An overview of the main results of the basic solution to the dy- namic model is given in table 31. There, the results are divided into nine broad categories: cost, demand, production, capacity expansion, imports, exports, shipments of final products, interplant shipments, and transportation requirements. The years specified in the table- 1980, 1983, and 1986-refer to the mid-point of each time period, and they are considered as representative for each time period. Several broad conclusions can be drawn from the information in this table. First, domestic products are fairly competitive with im- ports in the Egyptian fertilizer sector, with only relatively small quantities of final product (triple superphosphate) being required from outside sources in the second and third periods. Since Egypt is not self-sufficient in raw materials, several of these need to be im- ported. Second, exports of fertilizers are sufficiently attractive to take place up to the specified maximum level of 100,000 a year. Third, the only capacity expansion that should occur is at Abu Zaabal, where a phosphoric acid plant is expanded by 134,000 tons to permit the production of powdered triple superphosphate at that location. Finally, interplant shipments of surplus intermediate product are efficient in some cases-in particular, for ammonia from Suez to Helwan. The production offinal and intermediate products The production of final and intermediate products is detailed in table 32. The figures presented are the current costs for the relevant time period. 238 RESULTS OF THE DYNAMIC ANALYSIS Table 31. Summary of the Results of the Reference Solution (objective function value = 177.28 million Egyptian pounds)a Category 1980 1983 1986 Cost (millions of Egyptian pounds) Transport cost 6.08 6.24 6.83 Domestic raw materials cost 13.15 14.92 16.12 Operating cost 17.01 18.57 19.56 Working capital cost 0.81 0.90 0.95 Capital cost 1.21 1.21 1.21 Import cost Raw materials 2.24 2.52 2.52 Final products 0.01 1.07 4.07 Export revenue -3.50 -3.78 -3.78 Total cost 37.01 41.64 47.48 Total demand (thousands of tons) Nutrient demand Nitrogen 496 552 607 Phosphorus 112 135 158 Fertilizer demand 1,691 1,881 1,997 Production (thousands of tons) Final products 1,783 1,973 2,089 Intermediate products 1,664 1,737 1,788 Capacity expansion (thousands of tons) Phosphoric acid (at Abu Zaabal) 134 0 0 Imports (thousands of tons) Final products (TSP) 0 18 68 Raw materials 141 158 158 Exports (thousands of tons) Final products 100 100 100 Shipments of final products to governorates from following plants (thousands of tons) Aswan (CAN 33.5 percent) 317 311 344 Helwan (CAN 33.5 percent + AS) 67 67 67 Kafr El Zayaat (ssp) 69 168 167 Assiout (ssp) 178 178 178 Suez (CN) 23 1 3 Abu Kir (urea) 133 281 421 Talkha (CAN 33.5 percent + urea) 740 712 653 Abu Zaabal (Tsp) 163 163 163 Total shipments 1,691 1,881 1,997 BASIC RESULTS: THE REFERENCE SOLUTION 239 Table 31 (continued) Category 1980 1983 1986 Interplant shipments of intermediate products from following plants (thousands of tons) Aswan - - - Helwan _ _ _ Kafr El Zayaat (sulfuric acid to Abu Zaabal) 1 1 1 Assiout - - - Suez (ammonia to Helwan) 36 36 36 Abu Kir - - - Talkha Abu Zaabal Total shipments 37 37 37 Note: Details may not sum to totals because of rounding. a. Discounted total cost in 1975 Egyptian pounds for the period 1979-87. At Aswan, the calcium ammonium nitrate produced is entirely in the form of CAN 33.5 percent rather than the 31 percent variety pro- duced until now. This is because of the lower transport cost per ton of nutrient of the higher-nutrient fertilizer. In addition, excess ca- pacity is available for all products in all time periods, except for ammonia in the third time period. The unutilized capacity ranges from 2,000 tons a year for nitric acid in 1986 to 52,000 tons a year for CAN 33.5 percent in 1983. The reason for this is that the demand for CAN in the regions of Upper Egypt, where the Aswan plant com- petes with other sources of supply, is not sufficient to permit full- capacity utilization. The ammonia capacity is fully utilized only in the third time period. This constrains the production of CAN 33.5 percent to 344,000 tons a year; this level of production just satisfies the demands of the surrounding governorates in this time period. The marginal costs of all products remain the same in the first two time periods, but the rental of 2.5 Egyptian pounds a ton incurred on ammonia capacity in the third time period causes proportional increases in the marginal cost of nitric acid and CAN 33.5 percent. During periods of excess capacity, the production of any product to the full-capacity level will necessitate sales to sectors other than the fertilizer sector at a price equal to at least the relevant marginal 240 RESULTS OF THE DYNAMIC ANALYSIS Table 32. The Production of Final and Intermediate Products under the Reference Solution Maximum Unutilized possible Marginal cost Production capacity production (Egyptian Plant location Time (thousands (thousands (thousands pounds per and product period of tons) of tons) of tons) ton), Aswan Ammonia: 1980 137 11 148 12.3 water electrolysis 1983 134 14 148 12.3 1986 148 0 148 14.8 Nitric acid 1980 241 23 264 5.3 1983 237 27 264 5.3 1986 262 2 264 6.0 CAN 33.5 percentb 1980 317 46 363 15.3 1983 311 52 363 15.3 1986 344 19 363 16.4 Helwan Ammonia: coke-oven gas 1980 0 39 39 17.32 1983 0 39 39 17.32 1986 0 39 39 17.32 Nitric acid 1980 64 1 65 6.73 1983 64 1 65 6.73 1986 64 1 65 6.73 CAN 33.5 percent 1980 84 0 84 25.73 1983 84 0 84 20.64 1986 84 0 84 19.84 Ammonium sulfate 1980 8 0 8 15.58 1983 8 0 8 12.29 1986 8 0 8 11.96 Kafr El Zayaat Sulfuric acid: 1980 0 14 14 9.96 pyrites 1983 14 0 14 19.50 1986 14 0 14 18.46 Sulfuric acid: 1980 40 16 56, 9.18 sulfur 1983 56 0 56c 19.50 1986 56 0 56c 18.46 BASIC RESULTS: THE REPERENCE SOLUTION 241 Table 32 (continued) Maximum Unutilized possible Marginal cost Production capacity production (Egyptian Plant location Time (thousands (thousands (thousands pounds per and product period of tons) of tons) of tons) ton), ssp 15.5 percent 1980 94 74 168 13.53 1983 168 0 168 19.36 1986 168 0 168 19.51 Powdered TSP 1980 0 0 0 44.61 1983 0 0 0 66.79 1986 0 0 0 65.60 Assiout Sulfuric acid: sulfur 1980 73 1 74 9.45 1983 73 1 74 9.36 1986 73 1 74 9.42 ssp 15.5 percent 1980 178 0 178 12.78 1983 178 0 178 18.79 1986 178 0 178 18.93 Powdered TsP 1980 0 0 0 53.76 1983 0 0 0 78.40 1986 0 0 0 78.70 Suez Ammonia: refinery gas 1980 41 12 53d 8.86 1983 36 17 53d 8.86 1986 37 16 53d 8.86 Nitric acid 1980 15 117 132 4.24 1983 1 131 132 4.28 1986 2 130 132 4.28 Calcium nitrate 1980 23 152 175 9.61 1983 1 174 175 9.57 1986 3 172 175 9.51 Ammonium sulfate 1980 0 46 46 23.40 1983 0 46 46 20.93 1986 0 46 46 20.08 (Table continues on the following pages) 242 RESULTS OF THE DYNAMIC ANALYSIS Table 32. (continued) Maximum Unutilized possible Marginal cost Production capacity production (Egyptian Plant location Time (thousands (thousands (thousands pounds per and product period of tons) of tons) of tons) ton). Abu Kir Ammonia: natural gas 1980 92 205 297 19.48 1983 177 120 297 10.29 1986 258 39 297 10.28 Urea 1980 158 287 445 20.08 1983 305 140 445 20.07 1986 445 0 445 23.88 Talkha Ammonia: natural gas 1980 400 49 449 19.48 1983 402 47 449 19.50 1986 376 73 449 19.51 Nitric acid 1980 226 86 312 7.35 1983 207 105 312 7.35 1986 162 150 312 7.33 CAN 31 percente 1980 0 0 0 15.17 1983 0 0 0 17.93 1986 0 0 0 17.89 CAN 33.5 percent 1980 297 0 297 25.66 1983 272 25 297 18.43 1986 213 84 297 18.46 Urea 1980 468 22 490 20.08 1983 490 0 490 21.43 1986 490 0 490 25.50 Abu Zaabal Sulfuric acid: 1980 64 0 64 15.15 pyrites 1983 64 0 64 25.50 1986 64 0 64 24.50 Sulfuric acid: sulfur 1980 152 0 152 15.19 1983 152 0 152 25.50 1986 152 0 152 25.50 BASIC RESULTS: THE REFERENCE SOLUTION 243 Table 32 (continued) Maximum Unutilized possible Marginal cost Production capacity production (Egyptian Plant location Time (thousands (thousands (thousands pounds per and product period of tons) of tons) of tons) ton)s Phosphoric acide 1980 120 0 120 45.14 1983 120 0 120 70.07 1986 120 0 120 70.60 ssP 15.5 percent 1980 0 0 0 15.35 1983 0 0 0 19.57 1986 0 0 0 19.03 Powdered TSP 1980 163 6 169 41.92 1983 163 6 169 60.36 1986 163 6 169 60.80 a. These are the current (undiscounted) marginal costs of obtaining the product. b. All CAN production has shifted from the 31 percent variety to the 33.5 percent form. c. Includes interplant shipment to Abu Zaabal. d. Includes interplant shipment of 36,000 tons to Helwan in each time period. e. Capacity expansion of 134,000 tons at a cost of US$22.95 million. cost. For example, the sale of nitric acid from Aswan may be under- taken at a price of at least 5.3 Egyptian pounds a ton in 1980 and 1983 and 6.0 Egyptian pounds a ton in 1986. Sales at any price lower than this marginal cost would make this incremental production unattractive. Though this model does not explicitly consider exports of CAN 33.5 percent from Aswan to the Sudan, it is possible to assess the possibility of an export-oriented strategy to that country by inspect- ing the marginal production costs. If the fertilizer can be exported to the Sudan at an f.o.b. price that is at least equal to the marginal pro- duction costs plus transport cost to the border, such a strategy would be beneficial to Egypt. At Helwan, CAN production is restricted entirely to the 33.5 per- cent variety, for the same reason it is at Aswan. Ammonium sulfate is produced at its full-capacity level. No ammonia is produced by either the coke-oven gas process, which is already in existence, or by the natural gas technology, which could be used by converting the existing equipment, because the marginal cost of obtaining am- 244 RESULTS OF THE DYNAMIC ANALYSIS monia from Suez is lower. Thus, production of ammonia at Helwan would cause the fertilizer sector as a whole to incur a higher total cost than necessary. The production of CAN 33.5 percent at full ca- pacity of 84,000 tons a year in each time period is distributed be- tween shipments to the surrounding governorates (59,000 tons a year) and exports. At Kafr El Zayaat and Assiout, the production pattern is self- explanatory. In both cases, no powdered triple superphosphate is produced because its marginal cost is too high, and all single super- phosphate units operate at full capacity except at Kafr El Zayaat in the first time period. This plant produces 94,000 tons a year of ssp in the first time period, of which 69,000 tons are shipped to the governorates of Kafr El Sheikh and Menoufia and the remainder is exported. All of Assiout's production of 178,000 tons a year of ssp is shipped to surrounding governorates. All the sulfuric acid pro- duced in the first time period is made from sulfur because the mar- ginal production cost for this acid (9.18 Egyptian pounds a ton) is lower than for that made from pyrites (9.96 Egyptian pounds a ton). At Kafr El Zayaat, the excess capacity of 14,000 tons of sulfuric acid from pyrites may be utilized if the plant can sell this acid to the pulp and paper sector or to some other sector at a price of at least 9.96 Egyptian pounds a ton. In the second and third time periods, however, the increased demand for the final product forces all units to operate at full capacity.2 This accounts for the quantum jump in marginal costs of single superphosphate in the second period from 13.53 Egyptian pounds a ton to 19.36 Egyptian pounds a ton, since the plant now incurs a rental on its capacity of 5.83 Egyptian pounds a ton. The production pattern at Suez is determined by the urea "learn- ing curve" incorporated into the model. In the first time period, when consumption of urea is most restricted, calcium nitrate is produced at the rate of 23,000 tons a year to supply the governorates of Is- mailia, Suez, and Kalubia. No ammonium sulfate is produced be- cause of its relatively high marginal cost. As the urea restrictions are relaxed in the subsequent time periods, calcium nitrate production declines drastically (to 1,000 tons a year and 3,000 tons a year, re- spectively) to supply only the governorate of Suez. The production of ammonia and nitric acid declines, but the decline in production 2. Minor quantities of acid are shipped to Abu Zaabal from Kafr El Zayaat in all three time periods. BASIC RESULTS: THE REFERENCE SOLUTION 245 of ammonia is less than that of nitric acid because the Suez plant ships 36,000 tons of ammonia to Helwan in each time period. This implies that there is unbalanced excess capacity of ammonia (at Helwan) and calcium nitrate (at Suez). This provides a fallback position for the nitrogen fertilizer sector. In times of emergency or unanticipated demand, imports could only be justified, assuming the availability of foreign exchange, if the delivered price were less than the marginal cost of calcium nitrate plus the relevant transport cost. What is most noticeable is that, even in the first time period of maximum production, there is substantial excess capacity of the final product-152,000 tons a year. The low nutrient content of this product (15.5 percent nitrogen) makes its production and transport an uneconomical proposition under most conditions.' The production of urea at Abu Kir is obviously determined by the restrictions on urea consumption. As these restrictions are re- laxed, more urea production comes on stream. The favorable loca- tion of the plant, and the superior economics of production, make this an ideal export-based production unit.4 In each time period, the plant exports to the maximum limit of 25,000 tons a year. In all three periods, the bulk of the urea is transported to the neighboring governorate of Behera.5 The second largest shipment goes as far south as Minia. Thus, Minia's nitrogen requirements are supplied entirely by Abu Kir rather than by the relatively closer Helwan or Aswan. The high nutrient content of urea makes this supply pattern efficient, while Helwan and Aswan are restricted to supplying the surrounding governorates closest to them, thus minimizing the total cost to the sector. The production pattern of Talkha is characterized by a switch from producing CAN 31 percent to CAN 33.5 percent and urea. Again, the urea consumption restrictions determine the production pattern. In the first time period, CAN production is at its full capacity of 297,000 tons a year and urea production has some excess capacity 3. In general, if the transport cost per ton of final product is the same, transporting calcium nitrate is three times more expensive than transporting urea on a nutrient basis. In addition, calcium nitrate production involves a relatively high-cost technol- ogy. 4. The plant has been financed, however, under the condition that it does not export its product. 5. The amount of urea transported is 70,000 tons a year, 97,000 tons a year, and 121,000 tons a year in the three time periods, respectively. 246 RESULTS OF THE DYNAMIC ANALYSIS of 22,000 tons a year. In the second time period, as the urea restric- tions are relaxed, urea is produced at its full capacity of 490,000 tons a year, with CAN production declining to 272,000 tons a year.6 This decline continues through the last time period. The increasing nitrogen requirements are now met through urea production at Abu Kir, because the Talkha production of urea cannot increase any further. The ammonia and nitric acid production adjust accordingly. What is obvious is the excess capacity available for these intermediate products in all three time periods. Further, in the first time period, Talkha exports only urea and all CAN production is utilized for domestic consumption. As the increasing nitrogen requirements are met by urea in the next two time periods, however, CAN production is released for exports. Thus, in the last two time periods, exports of both CAN and urea take place. Urea production at Talkha attains full-capacity utilization before that at Abu Kir because Talkha has a more favorable location from the point of view of domestic supply. Total costs are minimized by shipping urea from Talkha rather than from Abu Kir, despite the slightly higher marginal costs in the latter two periods at Talkha. The higher marginal costs at Talkha in the second time period are explained by the rental on capacity incurred and the higher rental incurred in the third time period at Talkha compared with Abu Kir. Abu Zaabal is the only location where capacity expansion takes place. This is in the form of a phosphoric acid plant of 134,000 tons a year of product in the first time period. The total cost of this ex- pansion is US$22.95 million, of which US$17.65 million is for the plant and US$5.30 million for the off-site facilities. Assuming 90 percent capacity utilization, this new plant operates at full capacity of 120,000 tons a year from the start. In addition, Abu Zaabal no longer produces ssp 15.5 percent, but uses the phosphoric acid to produce powdered TSP using the existing capital equipment with minor alterations to the den. Sulfuric acid from both sulfur and pyrites is utilized to capacity limitations in the production of phos- phoric acid. Despite the higher marginal costs of powdered TSP relative to ssP, the production of high-analysis TSP is justified on the basis of its lower delivered price because of the lower transport cost per ton of nutrient. 6. Helwan produces at full capacity, but this is only possible because of the avail- ability of cheap ammonia from Suez. BASIC RESULTS: THE REFERENCE SOLUTION 247 Table 33. Exports of Final Products under the Reference Solution (thousands of tons) 1980 1983 1986 Plant location ssP CAN' Urea ssP CAN& Urea SSP CANa Urea Aswan 0 0 0 0 0 0 0 0 0 Helwan 0 25 0 0 25 0 0 25 0 Kafr El Zayaat 0 0 0 0 0 0 0 0 0 Assiout 25 0 0 0 0 0 0 0 0 Suez 0 0 0 0 0 0 0 0 0 Abu Kir 0 0 25 0 0 25 0 0 25 Talkha 0 0 25 0 25 25 0 25 25 Abu Zaabal 0 0 0 0 0 0 0 0 0 Total 25 25 50 0 50 50 0 50 50 a. 33.5 percent nitrogen. Exports The exports of fertilizer products in the three time periods are shown in table 33.7 In 1980, SSP, CAN 33.5, and urea are exported. In that period, P20 production is sufficient to meet all domestic demand and to earn some foreign exchange. In the second and third time periods, however, P205 domestic production falls short of demand, necessitating imports and a cessation of ssp exports. During these same two periods, urea consumption restrictions are relaxed, which permits more urea production at Talkha and Abu Kir and a reduction of CAN 33.5 percent production at Talkha. This reduction is mitigated, however, by exports of CAN 33.5 percent from Talkha, which were not possible in the first time period because of the urea restrictions. These changing restrictions, combined with increased demand for phosphorus, cause a shift in the export pattern from the first time period to the second; this pattern is then maintained in the third period. Alternative export strategies and potentials are investigated in greater detail later in this chapter. 7. The export constraints in the model permit exports from an individual plant of a maximum of 25,000 tons a year of a product and total exports from the country of 100,000 tons a year. All products shown in table 31 attain the upper bound stated in the model. 248 RESULTrS OF THE DYNAMIC ANALYSIS Table 34. Transport Requirements by Mode under the Reference Solution (millions of ton-kilometers) Mode 1975 1980 1983 1986 Barge 19.2 37.3 39.1 39.1 Rail 27.6 18.2 20.6 20.6 Road 332.0 294.9 291.9 328.4 Total 378.8 350.4 351.6 388.1 Transport requirements The amount of transport capacity utilized by the fertilizer sector over the planning period is given in table 34. The change in the barge capacity between the first and the second time periods is ex- plained by the increase in the import of raw materials from 141,000 tons a year in 1980 to 158,000 tons a year in the subsequent periods (see table 31). The change in the rail requirements is explained by the shift from exports of ssp produced at Kafr El Zayaat in the first time period to exports of CAN 33.5 percent produced at Talkha in the next two time periods. The latter plant is farther from the port of exit (Alexandria) than the former. The changes in road require- ments are a function of two variables: first, the nutrient content of the fertilizer; and, second, the nutrient demand. As the nutrient content rises, the transport requirement tends to decline; as the demand grows, however, this requirement increases. The transport sector is considered to be a major bottleneck in the distribution of fertilizers: there is a shortage of rail wagons, and not enough use is made of river transport. This study has not sought to determine the optimum mode of transport; rather, it has taken the existing modes as given. In this context, an interesting conclusion emerges from the requirements listed in table 34: The existing total transport capacity available to the fertilizer sector will be sufficient to distribute the products of the sector to the plants and governor- ates until 1984. After that a slight increase in transport capacity will be required. In 1975, the requirement was 378.8 mijlion ton-kilo- meters. This is not exceeded until the start of the third time period. The net decline in the first two time periods, as compared with 1975, is caused by three factors. First, the change to higher-analysis ferti- lizers such as CAN 33.5 percent, urea, and TSP lowers the road trans- port requirements. Second, the decline in interplant shipments lowers ALTERNATIVE SCENARIOS AT SUEZ: SCENARIOS I AND 11 249 the rail transport requirements. Finally, increasing imports of raw materials (because of increased demand for phosphatic fertilizers) increases the demand for barge capacity. Alternative Scenarios at Suez: Scenarios I and II Since the 1967 war between Egypt and Israel, no fertilizers have been produced at Suez. The calcium nitrate plant has, however, been repaired and is now operational. The question faced by plan- ners in Egypt is what should be done with this plant? It is clear from the preceding section that the optimal strategy at Suez would be to operate all the units in the first time period to produce 23,000 tons a year of calcium nitrate, but to reduce significantly the amount of this fertilizer produced in the subsequent periods while continuing to produce and ship ammonia to Helwan (see table 32). In view of the existing capacity and the various ramifications of bringing the Suez plant into operation, this section investigates the cost and impact of two alternative strategies. The first considers constructing a nitrophosphate unit that would produce nitrophosphate (20-20-0) and calcium nitrate in a fixed nitrogen ratio of 1:1. The second strategy involves operating the existing calcium nitrate plant at a minimum production rate of 150,000 tons a year of product. These two scenarios are compared with one another and with the basic reference solution presented in the previous section. The production of nitrophosphate and calcium nitrate at Suez: Scenario I A general overview of the results of the strategy of producing nitrophosphate and calcium nitrate at the Suez plant is presented in table 53 in the appendix to this chapter. Clearly, since the erection of this plant was forced into the solution, it represents a nonoptimal policy and has an associated cost. This cost is given by the difference between the value of the objective function in this scenario and that in the reference solution. This amounts to 2.53 million Egyptian pounds (1975 constant pounds), or 1.5 percent more than the refer- ence solution. This is not the cost of the plant alone, but the total cost associated with such a policy to the entire sector, including the costs of transportation, operation, raw materials, and so forth. 250 RESULTS OF THE DYNAMIC ANALYSIS In this scenario, the entire production of Aswan, Assiout, and Abu Zaabal is shipped to the governorates for domestic consump- tion, whereas production of the other plants-including the nitro- phosphate production at Suez-is for exports as well as for domestic consumption. The Suez plant exports 25,000 tons a year in the first period; after that, its annual production of 34,000 tons is used en- tirely for domestic consumption. As in the reference solution, Suez produces ammonia for use at Helwan. A comparison of the production pattern of final and intermediate products under Scenario I with the pattern under the reference solution is presented in table 35. The changes are at Aswan and Kafr El Zayaat in the first time period, at Abu Kir and Talkha in the second and third time periods, and at Suez in all three time periods. The capacity expansion in this scenario includes a phosphoric acid plant with a capacity of 134,000 tons a year at Abu Zaabal (as in the reference solution) and a nitrophosphate plant with a capacity of 88,000 tons a year at Suez. Both plants are erected in the first time period; the investment cost for the acid plant is US$22.95 million, and for the nitrophosphate plant it is US$18.41 million. These costs include the site costs. Given that the nutrient demand in each time period is the same in both cases, the new production of nitrophosphate and calcium nitrate at Suez alters the shipment patterns and therefore the produc- tion patterns of other plants. Thus, in the first time period, Suez satisfies part of the nitrogen and phosphorus requirements at Kalubia by supplying 31,000 tons a year of calcium nitrate and 9,000 tons a year of nitrophosphate. The rest of the nitrophosphate produced, 25,000 tons, is exported. This causes the Abu Zaabal plant to reduce its shipments of TSP to Kalubia by an equal amount and ship that TSP instead to Kafr El Sheikh. Thus, the production level at Abu Zaabal does not change; only the direction of its TSP shipment alters. This causes the shipment of ssp from Kafr El Zayaat to Kafr El Sheikh to be reduced by 12,000 tons a year, and the exports of ssp from Kafr El Zayaat to be reduced. The change in the production pattern at Aswan can be explained as follows: Under Scenario I, 31,000 tons a year of calcium nitrate from Suez are shipped to Ka- lubia, which causes the shipment of CAN from Helwan to Kalubia to be reduced. The Helwan CAN is now shipped to Beni Suef in- stead. Aswan, which supplied Beni Suef in the reference solution, must thus reduce its production of CAN from 317,000 tons a year to 295,000 tons. In all cases, the production of the intermediate products adjusts proportionately. ALTERNATIVE SCENARIOS AT SUEZ: SCENARIOS I AND 11 251 Table 35. The Production of Final and Intermediate Products under Scenario I (thousands of tons) 1980 1983 1986 Refer- Refer- Refer- ence ence ence Plant location solu- Sce- solu- Sce- solu- Sce- and product tion nario I tion nario I tion nario I Aswan Ammonia 137 127 134 134 148 149 Nitric acid 241 224 237 237 262 262 CAN 33.5 percent 317 295 311 311 344 344 Helwan Ammonia 0 0 0 0 0 0 Nitric acid 64 64 64 64 64 64 CAN 33.5 percent 84 84 84 84 84 84 Ammonium sulfate 8 8 8 8 8 8 Kafr El Zayaat Sulfuric acid: pyrites 0 0 14 14 14 14 Sulfuric acid: sulfur 40 28 56 56 56 56 ssP 94 65 168 168 168 168 Powdered TsP 0 0 0 0 0 0 Assiout Sulfuric acid: sulfur 73 73 73 73 73 73 ssP 178 178 178 178 178 178 Powdered TsP 0 0 0 0 0 0 Suez Ammonia 41 53 36 53 37 53 Nitric acid 15 40 1 40 2 40 Calcium nitrate 23 45 1 45 3 45 Ammonium sulfate 0 0 0 0 0 0 Nitrophosphate 0 34 0 34 0 34 Abu Kir Ammonia 92 92 177 173 258 238 Urea 158 158 305 298 445 445 Talkha Ammonia 400 400 402 388 376 359 Nitric acid 226 226 207 183 162 132 CAN 31 percent 0 0 0 0 0 0 CAN 33.5 percent 297 297 272 241 213 173 Urea 468 268 490 490 490 490 Abu Zaabal Sulfuric acid: pyrites 64 64 64 64 64 64 Sulfuric acid: sulfur 152 152 152 152 152 152 Phosphoric acid 120 120 120 120 120 120 ssp 15.5 percent 0 0 0 0 0 0 Powdered Tsp 163 163 163 163 163 163 252 RESULTS OF THE DYNAMIC ANALYSIS In the second and third time periods, as the demand for nitrogen and phosphorus grows, the 25,000 tons a year of nitrophosphate that were exported in the first time period are diverted for domestic use. The bulk of this nitrophosphate is shipped to the governorates of Ismailia, Suez, Kalubia, and Minia. The nitrogen requirements of these governorates were previously supplied by CAN from Talkha and urea from Abu Kir. Thus, under Scenario I, CAN production at Talkha and urea production at Abu Kir are reduced in the second period. The domestic availability of nitrophosphates also causes a reduction in the imports of TSP. This results in savings in import costs (for final products only) of 880,000 Egyptian pounds a year. These savings are, however, more than offset by an increase in the other cost categories, primarily capital and operating costs. Therefore, even though this strategy may make Egypt more independent of foreign supply sources, and may permit fertilizer production for Suez, the price of this incremental independence will amount to more than 2.5 million Egyptian pounds. The production of calcium nitrate at Suez: Scenario II In Scenario II, the calcium nitrate plant at Suez operates at a minimum of 150,000 tons a year in each time period. This is 60 per- cent of the rated design capacity of the plant. Since the reference solution indicated a low level of production in the first time period (23,000 tons a year), which subsequently declined to 1,000 tons a year in the second time period and to 3,000 tons a year in the third time period, forcing a production level of 150,000 tons in each year will distort the optimal patterns of production and consumption, thus involving a cost to the fertilizer sector. This cost, which is the difference between the objective function values of the two solutions, amounts to 1.86 million Egyptian pounds, or I percent of the refer- ence solution, over the planning period. This is, however, less than the cost incurred under Scenario I. In this scenario, all costs increase except capital cost and import cost (see table 54 in the appendix to this chapter). These cost in- creases can be explained by the change in the production pattern from using relatively efficient technology that produces high-analysis fertilizers to a relatively inefficient technology that produces a much lower analysis fertilizer, calcium nitrate. This low-analysis calcium nitrate necessitates higher levels of production to meet nutrient demand, which leads to higher transport cost on a nutrient basis. ALTERNATIVE SCENARIOS AT SUEZ: SCENARIOS I AND II 253 Table 36. The Production of Final and Intermediate Products under Scenario II (thousands of tons) 1980 1983 1986 Refer- Refer- Refer- ence Sce- ence Sce- ence Sce- Plant location solu- naHo solu- nario solu- nario and product tion II tion II tion ll Aswan Ammonia 137 127 134 135 148 149 Nitric acid 241 224 237 237 262 261 CAN 33.5 percent 317 295 311 311 344 344 Helwan Ammonia 0 0 0 0 0 0 Nitric acid 64 56 64 42 64 43 CAN 33.5 percent 84 74 84 56 84 56 Ammonium sulfate 8 8 8 8 8 8 Kafr El Zayaat Sulfuric acid: pyrites 0 0 14 14 14 14 Sulfuric acid: sulfur 40 30 56 56 56 56 ssP 94 69 168 168 168 168 Powdered TsP 0 0 0 0 0 0 Assiout Sulfuric acid: sulfur 73 73 73 73 73 73 ssP 178 178 178 178 178 178 Powdered TSP 0 0 0 0 0 0 Suez Ammonia 41 53 36 53 37 53 Nitric acid 15 99 1 99 2 99 Calcium nitrate 23 150 1 150 3 150 Ammonium sulfate 0 0 0 0 0 0 Nitrophosphate 0 0 0 0 0 0 Abu Kir Ammonia 92 92 177 172 258 258 Urea 158 158 305 297 445 445 Talkha Ammonia 400 407 402 389 376 359 Nitric acid 226 226 207 185 162 131 CAN 31 percent 0 0 0 0 0 0 CAN 33.5 percent 297 297 272 243 213 173 Urea 468 468 490 490 490 490 Abu Zaabal Sulfuric acid: pyrites 64 64 64 64 64 64 Sulfuric acid: sulfur 152 152 152 152 152 152 Phosphoric acid 120 120 120 120 120 120 ssp 15.5 percent 0 0 0 0 0 0 Powdered Tsp 163 163 163 163 163 163 254 RESULTS OF THE DYNAMIC ANALYSIS In addition, the use at Suez of the relatively inefficient technology that produces the required ammonia from refinery gas, and subse- quently nitric acid and calcium nitrate, leads to higher raw materials, operating, and working capital costs. In this scenario, only Helwan, Abu Kir, and Talkha produce for both foreign and domestic markets. The other plants ship all their production to the governorates. Table 36 shows the production levels at all plants. Alternative Export Strategies: Scenarios III, IV, and V In this section, three alternative export-oriented scenarios for Egypt are investigated. These scenarios are based on varying as- sumptions regarding Egypt's ability to export fertilizers. In the first case, it is assumed that no capacity expansion will be undertaken over the planning period and that the sector may export as much final product as is possible with the given capacity. In the second case, capacity expansion is permitted, but the total exports in the first time period are restricted to an upper limit (that is, a maximum) of 500,000 tons a year. In this scenario, an export market share is secured by Egypt in the first time period and is then maintained in the following periods or is increased by, at most, 10 percent year. The third strategy lowers the maximum of the first time period from 500,000 tons a year to 250,000 tons a year but permits capacity ex- pansion and retention and growth of the export market share as in the previous case. Since this section is intended to investigate the export side of the market, the details of product shipments to the various governorates is not discussed. No capacity expansion with unbounded exports: Scenario III Scenario III investigates the potential for exports of fertilizer products from Egypt with no change in the capacities of the existing plants; that is, no capacity expansion is permitted.8 Thus, this sce- 8. It is essential to constrain the model to no capacity expansion. Permitting expansion plus unlimited exports would yield new capacity in all time periods up to the maximum limit on capacity expansion. This would be a meaningless result. ALTERNATIVE EXPORT STRATEGIES: SCENARIOS III, IV, AND V 255 nario assumes that fertilizer products will be exported by Egypt up to the existing capacity limit as long as the country has a compara- tive advantage in exports-that is, as long as the marginal cost of production plus transport cost from the plant to the port of exit is less than the f.o.b. price. Given the cost-minimization objective function of the model, the products yielding the largest amount of export revenue will be selected. It may be argued that, given the assumption of a perfectly elastic demand for exports (that is, the small country assumption, since Egypt will not be able to influence world prices), there may be nonprice variables that will limit the total amount of exports from Egypt. Since this scenario is incorpo- rated into the analysis to study the production, import, and export behavior under the most optimistic conditions, there is no upper bound on the exports. Table 55 in the appendix to this chapter presents a summary of the results. The objective function value of 163.19 million Egyptian pounds indicates that this scenario is clearly superior to the reference solution, which permits capacity expansion. The objective function in the reference solution is 177.28 million Egyptian pounds. Thus, a strategy of no capacity expansion, with Egypt exporting as much as possible, would yield a saving of 14.09 million Egyptian pounds, or 8 percent of the total discounted cost under the basic assumptions, over the period under consideration. This is evident from the fact that the increases in the first two time periods in the transport, do- mestic raw materials, operating, import, and working capital costs are more than offset by the decrease in the capital charges (because capacity is not expanded) and by the increase in the export revenues. With the exception of the plant at Abu Kir, all plants produce for domestic consumption only. Table 37 indicates that all plants, ex- cept that at Suez, will operate at full capacity in all time periods. The Suez plant, which produces the lowest-analysis fertilizer (cal- cium nitrate), is the last to reach full-capacity utilization. Because capacity expansion is not permitted, the phosphoric acid plant is not built at Abu Zaabal. Thus, as table 37 indicates, all phosphate plants operate at full capacity, producing the low-analysis fertilizer ssp. This is not sufficient to meet domestic nutrient require- ments, and larger imports of TSP are needed than in the reference solution. The only product exported is urea from Abu Kir. The plant's coastal location provides it with a competitive edge over Talkha. The marginal export revenue (f.o.b. price) equals the marginal cost 256 RESULTS OF THE DYNAMIC ANALYSIS Table 37. The Production of Final and Intermediate Products under Scenario III (thousands of tons) 1980 1983 1986 Refer- Refer- Refer- ence Sce- ence Sce- ence Sce- Plant location solu- nario solu- nario solu- nario and product tion III tion III tion III Aswan Ammonia 137 149 134 148 148 148 Nitric acid 241 261 237 262 262 262 CAN 33.5 percent 317 344 311 344 344 344 Helwan Ammonia 0 0 0 0 0 0 Nitric acid 64 56 64 42 64 43 CAN 33.5 percent 84 84 84 84 84 84 Ammonium sulfate 8 8 8 8 8 8 Kafr El Zayaat Sulfuric acid: pyrites 0 14 14 14 14 14 Sulfuric acid: sulfur 40 56 56 56 56 56 ssP 94 168 168 168 168 168 Powdered TsP 0 0 0 0 0 0 Assiout Sulfuric acid: sulfur 73 73 73 73 73 73 ssP 178 178 178 178 178 178 Powdered TSP 0 0 0 0 0 0 Suez Ammonia 41 53 36 53 37 53 Nitric acid 15 101 1 108 2 116 Calcium nitrate 23 150 1 163 3 175 Ammonium sulfate 0 0 0 0 0 0 Nitrophosphate 0 0 0 0 0 0 Abu Kir Ammonia 92 258 177 258 258 258 Urea 158 445 305 445 445 445 Talkha Ammonia 400 425 402 427 376 429 Nitric acid 226 226 207 226 162 226 CAN 31 percent 0 0 0 0 0 0 cAN 33.5 percent 297 297 272 297 213 297 Urea 468 490 490 490 490 490 Abu Zaabal Sulfuric acid: pyrites 64 0 64 0 64 0 Sulfuric acid: sulfur 152 69 152 69 152 69 Phosphoric acid 120 0 120 0 120 0 ssp 15.5 percent 0 168 0 168 0 168 PowderedTrsp 163 0 163 0 163 0 ALTERNATIVE EXPORT STRATEGIES: SCENARIOS III, IV, AND V 257 Table 38. Exports of Final Products under Scenario III (thousands of tons) Plant location and product 1980 1983 1986 Abu Kir: urea 440 323 206 of 42 Egyptian pounds a ton.9 As the urea consumption restrictions relax over time, however, and as the nitrogen demand increases simultaneously, the exports of urea decline in order to meet the increasing domestic demand (see table 38). Since the marginal im- port cost is greater than the f.o.b. export price, the increase in do- mestic demand will always be met by reducing exports. The costs of establishing export markets are not captured in this model. In reality, these costs-which may include establishing for- eign trade agencies abroad, advertising, and so forth-will be quite substantial. Therefore, it may be argued that, once Egypt has estab- lished itself in the world market for urea (or any fertilizer), it is unlikely that the country would reduce its exports and thus its for- eign exchange earnings to supply the increasing domestic demand from existing capacity. This argument carries added strength given the desire of the government to develop and promote an export- oriented industry. The relevant question then becomes: if Egypt is at least to maintain its world market share, established in the first time period, is it optimal for the country to expand domestic capacity in order to meet increasing domestic demand? This question is addressed in the next two scenarios. Bounded export growth with maximum first-period exports of 500,000 tons a year: Scenario IV The second export strategy analyzes the expansion of domestic capacity in the fertilizer sector in Egypt that would permit growing exports of fertilizer products. In Scenario III, the first-period exports of fertilizer (urea) amounted to 440,000 tons a year, without ca- pacity expansion. In Scenario IV, an upper bound of 500,000 tons a year is placed on first-period exports. In addition, it is assumed that, once a market share in the exports of fertilizers is established in the 9. This includes the transport cost from Abu Kir to Alexandria. 258 RESULTS OF THE DYNAMIC ANALYSIS first time period, at least the same amount of fertilizers can be ex- ported in the subsequent time periods, and that exports can grow by, at most, 10 percent a year. Thus, the first time period has only an upper bound on total exports, whereas the second and third time periods have both an upper and a lower bound. Table 56 in the appendix to this chapter indicates that this sce- nario is superior to the reference solution. The objective function value of Scenario IV (158.09 million Egyptian pounds) is 19.19 million Egyptian pounds, or almost It percent, less than that of the reference solution. The increase in export revenues easily offsets the increase in the other cost components, despite the increased cost of capacity expansion. Exports in the first time period attain the upper bound of 500,000 tons, and they then grow at the rate of 10 percent a year. As in all other solutions involving capacity expansion, a phosphoric acid plant with a capacity of 134,000 tons a year is built at Abu Zaabal in 1980, which leads to a conversion of the superphosphate produc- tion from SSP to TSP. In addition, to accommodate the increase in domestic demand as well as the exports, an ammonia/urea complex is constructed at Suez in the second time period."' Interestingly, urea capacity is installed in both the second and the third time periods, whereas ammonia capacity is constructed only in the second period. The total cost, including site cost, of the new complex at Suez is US$214.27 million. Table 39 gives the detailed breakdown. Table 40 shows the production levels of final and intermediate products for Scenario IV and compares them to the levels for the reference solution. Most of the changes take place in the nitrogen fertilizer industry. The major exception is the increase in the produc- tion of ssp at Kafr El Zayaat in the first time period from 94,000 tons a year to 151,000 tons a year; this change is explained by the increase in the amount of exports of ssp from 25,000 tons a year to 82,000 tons a year (see table 41). The increased production of ssp necessitates an increase in the production of sulfuric acid. The acid production process using sulfur reaches its capacity limit at 56,000 tons a year, thus forcing the production of the higher-cost acid using pyrites. The production of sulfuric acid using the pyrites process 10. Abu Kir is the preferred location for this plant, but there is insufficient space to accommodate another complex. Area is not a binding constraint at Suez, which is the next most preferred location. ALTERNATIVE EXPORT STRATEGIES: SCENARIOS 111, IV, AND V 259 Table 39. Plant Investment under Scenario IV Capacity (thousands Investment Site cost Total cost Plant location Time of tons a (millions of (millions of (millions of andproduct period year) U.S. dollars) U.S. dollars) U.S. dollars) Abu Zaabal Phosphoric acid 1980 134 17.65 5.30 22.95 Suez Ammonia: natural gas 1983 431 92.64 27.79 120.43 Urea 1983 424 38.83 11.65 50.48 Urea 1986 319 33.35 10.01 43.36 Total 182.47 54.75 237.22 reaches its capacity limit in the second time period, thus necessitating an increase in the imports of pyrites, which is reflected in the im- ports and import cost categories in table 56 in the appendix. Table 41 shows the amount of exports undertaken by the plants. All plants not listed in that table produce for domestic consumption and ship their products to governorates (see table 56). The urea production at Abu Kir is directed toward both the export and the domestic markets. Interestingly, the first-period exports of urea from Abu Kir are 419,000 tons, as compared with 440,000 tons under Scenario III. Therefore, it is more profitable for the entire sector to reduce urea exports from that location and increase ssP exports. In the second and third time periods, the export market is supplied by the new urea capacity at Suez, causing a decline in the exports from Abu Kir and an increase in the domestic supply of urea from this location. As a result, calcium nitrate production declines to 1,000 tons a year in the second time period, since this fertilizer cannot compete domestically with urea. As domestic demand increases in the third time period, however, calcium nitrate production increases back to 87,500 tons a year. As table 41 indicates, Suez becomes basically an "export plant" and Abu Kir's production shifts toward domestic consumption. In the last time period, approximately 50 percent of the urea produced at Abu Kir is exported, as compared with 94 percent in the first time period. In the third time period, 94 percent of total output at Suez is exported, and this accounts for around 75 percent of Egypt's total exports. 260 RESULTS OF THE DYNAMIC ANALYSIS Table 40. The Production of Final and Intermediate Products under Scenario IV (thousands of tons) 1980 1983 1986 Refer- Refer- Refer- ence Sce- ence Sce- ence Sce- Plant location solu- nario solu- nario solu- nario and product tion IV tion IV tion IV Aswan Ammonia 137 149 134 148 148 148 Nitric acid 241 261 237 262 262 262 CAN 33.5 percent 317 344 311 344 344 344 Helwan Ammonia 0 0 0 0 0 0 Nitric acid 64 64 64 64 64 64 CAN 33.5 percent 84 84 84 84 84 84 Ammonium sulfate 8 8 8 8 8 8 Kafr El Zayaat Sulfuric acid: pyrites 0 7 14 14 14 14 Sulfuric acid: sulfur 40 56 56 56 56 56 ssP 94 151 168 168 168 168 Powdered TSP 0 0 0 0 0 0 Assiout Sulfuric acid: sulfur 73 73 73 73 73 73 ssP 178 178 178 178 178 178 Powdered TsP 0 0 0 0 0 0 Suez Ammonia: refinery gas 41 53 36 36 37 53 Ammonia: natural gas 0 0 0 222 0 388 Nitric acid 15 58 1 1 2 58 Calcium nitrate 23 88 1 1 3 88 Ammonium sulfate 0 0 0 0 0 0 Nitrophosphate 0 0 0 0 0 0 Urea 0 0 0 382 0 669 Abu Kir Ammonia 92 258 177 258 258 258 Urea 158 445 305 445 445 445 Talkha Ammonia 400 413 402 413 376 413 Nitric acid 226 226 207 226 162 226 CAN 31 percent 0 0 0 0 0 0 CAN 33.5 percent 297 297 272 297 213 297 Urea 468 490 490 490 490 490 ALTERNATIVE EXPORT STRATEGIES: SCENARIOS 111, IV, AND V 261 Table 40 (continued) 1980 1983 1986 Refer- Refer- Refer- ence Sce- ence Sce- ence Sce- Plant location solu- nario solu- nario solu- nario and product tion IV tion IV tion IV Abu Zaabal Sulfuric acid: pyrites 64 64 64 64 64 64 Sulfuric acid: sulfur 152 152 152 152 152 152 Phosphoric acid 120 120 120 120 120 120 ssP 15.5 percent 0 0 0 0 0 0 Powdered TSP 163 163 163 163 163 163 Table 41. Exports of Final Products under Scenario IV (thousands of tons) Plant location and product 1980 1983 1986 Kafr El Zayaat: ssp 15.5 percent 81 0 0 Suez: Urea 0 373 631 Abu Kir: Urea 419 277 214 Total 500 650 845 Bounded export growth with maximum first-period exports of 250,000 tons a year: Scenario V The third export strategy, Scenario V, is identical to the previous one, except that the upper bound in the first period on exports is 250,000 tons a year rather than 500,000 tons. The maximum per- missible annual growth rate is still 10 percent. On the basis of the value of the objective function, Scenario V is worse than Scenario IV, but is superior to the reference solution and Scenario III (unbounded exports, no capacity expansion). Table 57 in the appendix shows that the cost of exporting 250,000 tons a year is 3.87 million Egyptian pounds greater than the cost of ex- porting 500,000 tons over the planning period. Scenario V and the reference solution both involve the same phosphoric acid capacity 262 RESULTS OF THE DYNAMIC ANALYSIS Table 42. The Production of Final and Intermediate Products under Scenario V (thousands of tons) 1980 1983 1986 Refer- Refer- Refer- ence Sce- ence Sce- ence Sce- Plant location solu- nario solu- nario solu- nario and product tion V tion V tion V Aswan Ammonia 137 128 134 148 148 148 Nitric acid 241 226 237 262 262 262 CAN 33.5 percent 317 297 311 344 344 344 Helwan Ammonia 0 0 0 0 0 0 Nitric acid 64 64 64 64 64 64 CAN 33.5 percent 84 84 84 84 84 84 Ammonium sulfate 8 8 8 8 8 8 Kafr El Zayaat Sulfuric acid: pyrites 0 0 14 14 14 14 Sulfuric acid: sulfur 40 30 56 56 56 56 ssP 94 69 168 168 168 168 Powdered TsP 0 0 0 0 0 0 Assiout Sulfuric acid: sulfur 73 73 73 73 73 73 ssP 178 178 178 178 178 178 Powdered TsP 0 0 0 0 0 0 Suez Ammonia 41 39 36 53 37 53 Nitric acid 15 8 1 58 2 116 Calcium nitrate 23 12 1 87 3 175 Ammonium sulfate 0 0 0 0 0 0 Nitrophosphate 0 0 0 0 0 0 Abu Kir Ammonia 92 222 177 258 258 258 Urea 158 383 305 445 445 445 Talkha Ammonia 400 385 402 413 376 429 Nitric acid 226 226 207 226 162 226 CAN 31 percent 0 0 0 0 0 0 CAN 33.5 percent 297 297 272 297 213 297 Urea 468 443 490 490 490 490 ALTERNATIVE EXPORT STRATEGIES: SCENARIOS 111, IV, AND V 263 Table 42 (continued) 1980 1983 1986 Refer- Refer- Refer- ence Sce- ence Sce- ence Sce- Plant location solu- nario solu- nario solu- nario and product tion V tion V tion V Abu Zaabal Sulfuric acid: pyrites 64 64 64 64 64 64 Sulfuric acid: sulfur 152 152 152 152 152 152 Phosphoric acid 120 120 120 120 120 120 ssP 15.5 percent 0 0 0 0 0 0 Powdered TSP 163 163 163 163 163 163 Table 43. Exports of Final Products under Scenario V (thousands of tons) Plant location and product 1980 1983 1986 Abu Kir: Urea 250 297 205 Kafr El Zayaat: ssp 0 0 92 Total 250 297 297 expansion, but the reference solution involves an export upper bound (100,000 tons a year) that is smaller than that of Scenario V. The production pattern under Scenario V is presented in table 42; the exports are shown in table 43, which indicates that the only products exported are urea from Abu Kir and ssp from Kafr El Zayaat. All other production is channeled toward domestic con- sumption (see table 57). Unlike in Scenario IV, the total exports in Scenario V remain the same in the last two time periods; that is, exports grow by 10 percent a year during the first two time periods, but there is no subsequent increase. This 10 percent annual growth of exports is not large enough to justify new nitrogen capacity ex- pansion (as was the case in the previous scenario). Therefore, the increasing domestic demand and the relaxation of the urea con- sumption restrictions causes an increase in the production of calcium 264 RESULTS OF THE DYNAMIC ANALYSIS nitrate at Suez from 12,000 tons a year to 175,000 tons a year and a decline in urea exports from Abu Kir from 297,000 tons a year to 205,000 tons a year in the third period. Since total exports have a lower bound of 297,000 tons a year in the third time period, the difference of 92,000 tons is made up by exporting ssp from Kafr El Zayaat. This, however, causes a positive excess demand domes- tically for the phosphorus nutrient, which is satisfied by annual imports of 98,000 tons of TSP. Clearly, this is a somewhat unlikely pattern of trade, caused solely by the specification of a lower bound on exports. A Complete Import Substitution Policy: Scenario VI Scenario VI evaluates the costs of a policy that bans the imports of all final products: that is, a policy of substituting all final product imports with domestic production." As table 58 in the appendix indicates, the objective function value under the import substitution policy is marginally higher than that under the reference solution, amounting to 178.66 million Egyptian pounds; the difference, or the discounted cost of the policy, is there- fore 1.38 million Egyptian pounds over the planning period. It is not surprising that this difference-less than 0.8 percent of the cost of the reference solution-is so small since the imports substituted Table 44. Plant Investment under Scenario VI Capacity Investment Site cost Total cost (thousands (millions of (millions of (millions of Plant location Time of tons a U.S. dol- U.S. dol- U.S. dol- and product period year) lars) lars) lars) Abu Zaabal Phosphoric acid 1980 214 22.20 6.66 28.86 Sulfuric acid 1983 146 13.32 4.00 17.32 Total 360 35.52 10.66 46.18 11. The imports of raw materials-namely, pyrites and sulfur-are still permitted because these products are not available domestically. A COMPLETE IMPORT SUBSTITUTION POLICY: SCENARIO VI 265 Table 45. The Production of Final and Intermediate Products under Scenario VI (thousands of tons) 1980 1983 1986 Refer- Refer- Refer- ence Sce- ence Sce- ence Sce- Plant location solu- nario solu- narHo solu- nario and product tion VI tion VI tion VI Aswan Ammonia 137 137 134 134 148 149 Nitric acid 241 241 237 237 262 262 CAN 33.5 percent 317 317 311 311 344 344 Helwan Ammonia 0 0 0 0 0 0 Nitric acid 64 64 64 64 64 64 CAN 33.5 percent 84 84 84 84 84 84 Ammonium sulfate 8 8 8 8 8 8 Kafr El Zayaat Sulfuric acid: pyrites 0 0 14 0 14 0 Sulfuric acid: sulfur 40 40 56 43 56 31 ssP 94 77 168 0 168 76 Powdered Tsp 0 0 0 91 0 93 Assiout Sulfuric acid: sulfur 73 73 73 43 73 73 ssP 178 178 178 105 178 178 Powdered TSp 0 0 0 0 0 0 Suez Ammonia 41 41 36 36 37 37 Nitric acid 15 15 1 1 2 2 Calcium nitrate 23 23 1 1 3 3 Ammonium sulfate 0 0 0 0 0 0 Nitrophosphate 0 0 0 0 0 0 Abu Kir Ammonia 92 92 177 172 258 258 Urea 158 158 305 305 445 445 Talkha Ammonia 400 400 402 402 376 376 Nitric acid 226 226 207 207 162 162 CAN 31 percent 0 0 0 0 0 0 CAN 33.5 percent 297 297 272 272 213 213 Urea 468 468 490 490 490 490 Abu Zaabal Sulfuric acid: pyrites 64 64 64 62 64 64 Sulfuric acid: sulfur 152 152 152 284 152 284 Phosphoric acid 120 125 120 192 120 193 ssP 15.5 percent 0 0 0 0 0 0 Powdered Tsp 163 168 163 168 163 168 266 RESULTS OF THE DYNAMIC ANALYSIS through domestic production in the reference solution amount to only 258,000 tons over the planning horizon. In the reference solu- tion, the only imports of final product required were those of TSP in the second and third time periods (see table 31). The imports in the second time period amounted to 18,000 tons a year and those in the third time period to 68,000 tons a year. Comparing the results for the reference solution (table 31) and those for Scenario VI (table 58) shows that major changes occur only in capacity expansion and interplant shipments. In the reference solution, the capacity expansion involved a phosphoric acid plant with a capacity of 134,000 tons a year at Abu Zaabal, and the inter- plant shipments involved transporting sulfuric acid from Kafr El Zaayat to Assiout and ammonia from Suez to Helwan; Scenario VI necessitates a phosphoric acid plant with a capacity of 214,000 tons a year capacity (at a cost of US$29 million) and a sulfuric acid plant with a capacity of 146,000 tons a year in the second time pe- riod (at a cost of US$17 million). The detailed breakdown is pre- sented in table 44. The interplant shipments now involve an addi- tional shipment of phosphoric acid from Abu Zaabal to Kafr El Zayaat. The production pattern under Scenario VI is shown in table 45. The Elimination of the Urea Consumption Restrictions: Scenario VII The preceding scenarios have all assumed that the Egyptian farmer would require time to adjust to the consumption of the higher-analy- sis fertilizer urea. Thus, consumption restrictions were imposed so that no more than 60 percent of all nitrogen could be consumed in the form of urea in the first time period; this figure was increased to 75 percent and 85 percent in the second and third time periods, respectively. Since this consumption pattern is less than optimal, an opportunity cost is associated with this constraint. To determine this cost, the production-consumption pattern was determined with- out these restrictions on urea. Table 59 in the appendix shows that the discounted objective function value of Scenario VII is 172.31 million Egyptian pounds, which is 4.97 million Egyptian pounds less than the value of the reference solution. Comparing the individual cost categories of Scenario VII with those of the reference solution (see table 31 ) ELIMINATION OF THE UREA CONSUMPTION RESTRICTIONS: SCENARIO VIl 267 Table 46. The Production of Final and Intermediate Products under Scenario VII (thousands of tons) 1980 1983 1986 Refer- Refer- Refer- ence Sce- ence Sce- ence Sce- Plant location solu- nario solu- nario solu- nario and product tion VII tion VIl tion Vii Aswan Ammonia 137 149 134 148 148 148 Nitric acid 241 193 237 262 262 262 CAN 33.5 percent 317 253 311 344 344 344 Helwan Ammonia 0 0 0 0 0 0 Nitric acid 64 24 64 64 64 64 CAN 33.5 percent 84 31 84 84 84 84 Ammonium sulfate 8 8 8 8 8 8 Kafr El Zayaat Sulfuric acid: pyrites 0 0 14 14 14 14 Sulfuric acid: sulfur 40 29 56 56 56 56 ssP 94 69 168 168 168 168 Powdered TsP 0 0 0 0 0 0 Assiout Sulfuric acid: sulfur 73 73 73 73 73 73 ssP 178 178 178 178 178 178 Powdered TsP 0 0 0 0 0 0 Suez Ammonia 41 14 36 37 37 37 Nitric acid 15 0 1 2 2 2 Calcium nitrate 23 0 1 3 3 3 Ammonium sulfate 0 0 0 0 0 0 Nitrophosphate 0 0 0 0 0 0 Abu Kir Ammonia 92 258 177 258 258 258 Urea 158 445 305 445 445 445 Talkha Ammonia 400 295 402 258 376 258 Nitric acid 226 19 207 35 162 162 CAN 31 percent 0 0 0 0 0 0 CAN 33.5 percent 297 25 272 47 213 213 Urea 468 490 490 490 490 490 Abu Zaabal Sulfuric acid: pyrites 64 64 64 64 64 64 Sulfuric acid: sulfur 152 152 152 152 152 152 Phosphoric acid 120 120 120 120 120 120 ssP 15.5 percent 0 0 0 0 0 0 Powdered TSP 163 163 163 163 163 163 268 RESULTS OF THE DYNAMIC ANALYSIS Table 47. Exports of Final Products under Scenario VII (thousands of tons) Plant location and product 1980 1983 1986 Helwan CAN 33.5 percent 25 25 25 Abu Kir Urea 25 25 25 Talkha CAN 33.5 percent 25 25 25 Urea 25 25 25 Total 100 100 100 shows that the decline in the objective function is caused chiefly by the saving in transport cost (1.31 million Egyptian pounds) and in operating cost (3.64 million Egyptian pounds). These changes can be explained by the production pattern shown in table 46. Since the urea restrictions primarily affect the nitrogen fertilizer industry, all changes in production levels are restricted to that industry except for the production of ssp at Kafr El Zayaat, which drops from 94,000 tons a year to 69,000 tons a year in the first time period. This decline of 25,000 tons a year does not repre- sent a change in domestic consumption, but rather a change in the export pattern. The exports of ssp are substituted for by the exports of the more profitable CAN 33.5 percent from Talkha (see table 47). This switch is possible because eliminating the urea restriction per- mits Talkha CAN production to be diverted away from the domestic market to the export market. The most important impact of the elimination of this restriction is the production at full capacity of urea plants at both Talkha and Abu Kir in all time periods. All urea-490,000 tons a year at Talkha and 446,000 tons a year at Abu Kir-is shipped domestically (see table 59) except for exports of 25,000 tons a year from each location (see table 47). This level of urea production necessitates a reduction in the first time period in the production of the relatively more ex- pensive nitrogen products at Helwan, Aswan, Suez, and Talkha. As nitrogen demand increases in subsequent time periods, however, Helwan and Aswan produce CAN at full capacity in the second and MEASURES OF THE DOMESTIC RESOURCE COST 269 third periods, Suez produces small amounts of calcium nitrate in the second and third periods to satisfy local demand, and Talkha increases its production of CAN to 47,000 tons a year in the second period and to full capacity in the last period. Thus, nitrogen capacity in Egypt is sufficient to satisfy domestic demand up to 1987, even in the absence of restrictions on urea con- sumption. The production and consumption patterns so obtained result in a savings of about 5 million Egyptian pounds over the planning period. This implies that, if the cost of educating farmers to use urea is less than or equal to 5 million Egyptian pounds, such extension would be worth considering. Measures of the Domestic Resource Cost On the basis of the available cost data, the domestic resource cost (DRC) associated with the various scenarios can be evaluated. These measures provide estimates of the unit opportunity cost of the do- mestically owned factors of production employed directly (in the fertilizer sector) and indirectly (in the associated home goods in- dustries) as a function of the net change in Egypt's trade balance that would occur if the level of output were expanded or contracted by one unit. The DRC criterion is analogous to the "rate of return" criterion using the foreign exchange factor as the numeraire rather than the capital factor. Within the context of this study, the DRC is computed for the entire investment program over the planning horizon. Therefore, the numerator and the denominator are expressed in present value terms with the costs (both foreign exchange and domestic) discounted back to the beginning of the planning period. In addition, since this cal- culation is undertaken within the framework of a process analysis, economic cost can be defined as the value of the final and inter- mediate goods that could be produced in the sector through resource allocation and trade, which takes into account the possibility of substitution in production and consumption. Thus, such a measure of cost does not undertake the welfare loss attributable to a system that assumes an unaltered pattern of final consumption. Throughout this section, it is assumed that the foreign exchange component of the various cost categories are as follows: transport cost, 40 percent; domestic raw materials cost, 30 percent; operating 270 RESULTS OF THE DYNAMIC ANALYSIS cost, 30 percent; working capital charge, 50 percent; and capital charge, 70 percent. Three approaches are followed in calculating the domestic resource cost, thus yielding three different measures."2 The first approach (DRC0) ignores substitution in consumption, whereas the other two (DRc, and IDRC) do not. In addition, the last approach (IDRC) is, strictly speaking, an incremental (or marginal) DRC that evaluates the opportunity cost of domestic resources per unit of foreign ex- change on the incremental capacity expansion only; it thus takes the existing domestic capacity as given. The first approach (DRC.) calculates the foreign exchange cost on the assumption that all fertilizers consumed domestically are im- ported-that is, these fertilizers are valued at their c.i.f. import prices. The fertilizer consumption pattern is taken to be identical to the pattern implicit in the relevant capacity-expansion scenario. Then, the direct and indirect foreign exchange cost and the domestic cost associated with such a capacity-expansion scenario are obtained (see table 48). The foreign exchange savings or earnings so generated and the associated domestic cost are calculated, yielding a measure of the domestic resource cost (denoted DRCO). The resource cost coefficient, expressed as a ratio of the domestic resource cost to the exchange rate, is subsequently obtained for both the shadow and the official exchange rate.'" These ratios are expressed in percentage terms. This approach, however, ignores all substitution in the final consumption of fertilizers. The second approach (DRc.) calculates the significance of con- sumption substitution; it takes into account the change in the final product mix and thus the consumption brought about by new ca- pacity expansion. Table 49 compares the production and import patterns when no new capacity expansion is permitted with those when there is new capacity. The no-capacity-expansion situation permits domestic production from existing plants only, with the 12. All three measures of DRC are based on marginal costs for the existing plants. All the associated committed costs are treated as sunk costs since they involve decisions that have already been made and that are irreversible. 13. Excluding the multiple exchange rates generated by the "own imports" scheme in Egypt, three exchange rates can be observed: the official rate (US$1 = 0.4 Egyptian pounds), the parallel market rate (US$1 = 0.59 Egyptian pounds), and the black market rate (US$1 = 0.70 Egyptian pounds). The shadow exchange rate used here (US$1 = 0.65 Egyptian pounds) therefore appears to be reasonable. Table 48. The Domestic and Foreign Exchange Cost Components of the Various Scenarios (millions of currency units) Reference solution Scenario I Scenario II Scenario III Foreign Foreign Foreign Foreign exchange Domestic exchange Domestic exchange Domestic exchange Domestic cost cost cost cost cost cost cost cost (U.S. (Egyptian (U.S. (Egyptian (U.S. (Egyptian (U.S. (Egyptian Cost category dollars) pounds) dollars) pounds) dollars) pounds) dollars) pounds) Transport cost 16.76 16.34 16.57 16.15 17.24 16.81 16.94 16.52 Domestic raw materials cost 28.78 43.65 28.71 43.54 28.70 43.53 32.48 49.26 Operating cost 36.11 54.77 36.98 56.08 36.98 56.09 40.15 60.89 Working capital cost 2.91 18.88 2.94 1.91 2.94 19.08 3.15 2.05 Capital cost 9.11 25.36 16.41 4.57 9.11 2.54 0.00 0.00 Import cost 40.63 0.00 34.84 0.00 40.03 0.00 91.75 0.00 Export revenue -39.46 -0.00 -41.34 -0.00 -40.77 -0.00 -156.23 -0.00 Totalcost 94.83 119.18 95.08 122.40 94.82 120.87 28.24 128.72 (Table continues on the following page) Table 48 (continued) Scenario IV Scenario V Scenario VI Scenario VII Foreign Foreign Foreign Foreign exchange Domestic exchange Domestic exchange Domestic exchange Domestic cost cost cost cost cost cost cost cost (U.S. (Egyptian (U.S. (Egyptian (U.S. (Egyptian (U.S. (Egyptian Cost category dollars) pounds) dollars) pounds) dollars) pounds) dollars) pounds) Transport cost 16.99 16.57 16.28 15.87 17.21 16.78 15.96 15.57 Domestic raw materials cost 40.72 61.77 32.34 49.05 29.15 44.21 29.11 44.15 Operating cost 47.38 71.87 39.85 60.44 36.52 55.40 34.43 52.22 Working capital cost 3.88 2.52 3.18 2.07 2.96 1.92 2.85 1.85 Capital cost 51.88 14.45 9.11 2.54 15.39 4.29 9.11 2.54 Import cost 41.38 0.00 44.98 0.00 28.90 0.00 40.29 0.00 Export revenue -289.14 -0.00 -120.66 -0.00 -439.46 -0.00 -40.77 -0.00 Totalcost -76.91& 167.17 25.08 129.96 90.66 122.59 90.97 116.31 Note: All costs are discounted to the beginning of the planning period. a. Represents foreign exchange earnings. Table 49. Fertilizer Production and Consumption during the Planning Period, 1979-87 (thousands of tons) No capacity expansion New capacity expansion Domestic Total Domestic Total Product production Imports Exports consumption production Imports Exports consumption Calcium nitrate 192 0 0 192 81 0 0 81 Ammonium sulfate 72 0 0 72 72 0 0 72 Z.4 CAN 33.5 percent 6,045 0 450 6,495 6,024 0 375 6,399 Urea 7,077 0 450 7,527 7,077 0 450 7,527 ssP 4,632 0 0 4,632 2,898 0 75 2,973 Powdered TsP 0 0 0 0 1,464 0 0 1,464 Granular TsP 0 1,134 0 1,134 0 258 0 258 Total 18,018 1,134 900 20,052 17,616 258 900 18,774 a. The reference solution. Table 50. The Domestic Resource Cost for the Various Scenarios Objective function value (millions of dis- Percentage of shadow Percentage of official counted Domestic resource cost exchange rate exchange rate U.S. Scenario dollars) DRCo DRC, IDRC DRCo DRC, IDRC DRCo DRC, IDRC Reference solution 177.28 0.1324 0.1255 0.1094 20.37 19.31 21.57 33.10 31.38 27.35 Scenario I 179.81 0.1362 0.1288 0.1544 20.96 19.81 23.75 34.06 32.20 38.60 ScenarioII 179.14 0.1354 0.1273 0.1338 20.83 19.59 20.59 33.85 31.83 33.45 Scenarioll 163.19 0.1262 0.1267 0.1260 19.41 19.49 19.38 31.55 31.67 31.50 ScenarioIV 158.09 0.1559 0.1491 0.2308 23.99 22.94 35.50 38.99 37.28 57.69 ScenarioV 161.96 0.1350 0.1275 0.1321 20.77 19.62 20.32 33.74 31.88 33.02 Scenario VI 178.66 0.1355 0.1286 0.1497 20.85 19.78 23.03 33.88 32.15 37.43 Scenario VII 172.31 0.1307 0.1220 0.0641 20.12 18.77 9.86 32.69 30.50 16.03 Note: Basis for comparison: foreign exchange cost: DRCo = US$995.07 million for reference solution only, DRC, = US$1,044.18 million, IDRC US$163.70 million; domestic cost: IDRC = 111.65 million Egyptian pounds. MEASURES OF THE DOMESTIC RESOURCE COST 275 excess demand being satisfied by imports. As table 49 indicates, the major change in consumption that takes place if capacity expansion is permitted is the substitution of powdered triple superphosphate for some single superphosphate and imports. The higher-analysis TSP contributes significantly to meeting the demand for phosphorus nutrients, thus resulting in the reduction of domestic consumption of ssp and imports. Substitution in consumption also takes place on the nitrogen side, but to a much lesser extent; the change in the export pattern permits the substitution of CAN 33.5 percent for calcium nitrate. These substitution effects are incorporated into the second meas- ure. In this case, the foreign exchange cost of the entire program is also calculated on the assumption that all fertilizers consumed are imported, but the consumption pattern of these products is the one that is implicit with existing production capacity. This pattern is shown in the fourth column of table 49. The foreign exchange and domestic costs resulting from the various capacity-expansion sce- narios are then obtained as before (shown in table 48). These cost components incorporate the substitution of final products resulting from the change in domestic production. The domestic resource cost is then obtained as before and is denoted by DRC, in table 50. The third measure of the domestic resource cost takes the existing capacity as given and computes a DRC for capacity expansion only. Thus, the foreign exchange and domestic costs are obtained for the no-capacity-expansion case (which includes the existing capacity) and obtains the same components for each of the capacity-expansion scenarios incorporating the substitution effect. The incremental domestic resource cost (IDRC) iS then obtained as the ratio of the incremental domestic cost to the foreign exchange savings or earn- ings. This incremental cost thus represents the domestic cost asso- ciated with capacity expansion and substitution only. Since the model is structured to produce the minimum cost, the IDRC must always be less than the shadow exchange rate. If that is not the case, it implies that a less than optimal capacity expansion has oc- curred; that is, the product is being produced domestically when in fact it should have been imported. Further, the IDRC does not refer to a single project; it may also reflect multiple projects, as in Scenario IV. It is thus a measure that reflects sectoral capacity expansion. Table 50 presents the three measures of domestic resource cost and of the resource cost coefficient as a percentage of the shadow and official exchange rate. In addition, the first column presents the 276 RESULTS OF THE DYNAMIC ANALYSIS objective function value. If domestic resource cost is considered first, the DRC measures in each case are less than both the shadow exchange rate and the official exchange rate, indicating that Egypt has a cost advantage in fertilizer production under any scenario. Further substitution in final consumption affects the DRC-that is, the values of DRCO and DRC, are not the same. In each case (except Scenario III), substitution causes a decline in the value of the DRC (that is, DRCS is less than DRCO). This is primarily because of the pro- duction of 'rsp (a high-analysis fertilizer) instead of ssp (a low- analysis fertilizer), which causes a reduction in the imports of phos- phate fertilizers. There are also substitution effects in nitrogen consumption, which have a secondary influence on the DRC-that is, the substitution of CAN 33.5 percent and urea (both high-analysis fertilizers) for calcium nitrate (a low-analysis fertilizer ).14 Scenario III (which is a no-capacity-expansion case with unbounded exports) indicates a slight increase in the DRC-from a value of 0.1262 for DRC0 to a value of 0.1267 for DRc.-because of substitution. This is because the high-analysis urea and CAN 33.5 are now exported rather than consumed and thus production and consumption of calcium nitrate become relatively larger. In addition, there is no substitution of TSP for ssp. According to the incremental DRC, in each scenario the capacity expansion or strategy consistent with the scenario shows a sectoral comparative advantage for the expansion or strategy undertaken. Thus, the IDRC coefficients range from 19.38 percent to 35.50 percent of the shadow exchange rate. If, however, the domestic resource cost incorporating the substitution effects (DRc.) is compared with the incremental value in resource cost (IDRC), the value of IDRC is less than that of the DRC. only in the case of the reference solution (opti- mal capacity expansion) and Scenario III (unbounded exports). The DRC, measure incorporates existing capacity plus capacity ex- pansion, whereas the IDRC measure incorporates only capacity ex- pansion. Thus, it is possible to say something about the relative efficiency of the new and the existing capacity. In the two cases in which the IDRC iS less than the DRC, (reference solution and Scenario III), the difference indicates that the oppor- 14. This substitution leads to a higher foreign exchange cost of fertilizers, based on import prices, thus tending to lower the resource cost per unit of international value added. MEASURES OF THE DOMESTIC RESOURCE COST 277 tunity cost of earning or saving one dollar of foreign exchange asso- ciated with no capacity expansion would be greater than that asso- ciated with capacity expansion. In other words, capacity expansion (or Scenario III) has lowered the cost of earning foreign exchange.'" In the case of the reference solution, the expansion of phosphoric acid capacity-and the resulting changes in production and con- sumption-has led to a lower cost to the system as a whole. The fact that Scenario III's IDRC iS lower than its DRC8 (0.1262 as com- pared with 0.1267) is not surprising, since this scenario permits unlimited exports. Thus, the benefits of such a policy (that is, exports in excess of 100,000 tons a year) must be positive. Meanwhile, Scenario VII indicates that restrictions on the consumption of urea involve a cost to society (an IDRC of 0.0641 compared with a DRCs of 0.1220). For all the other scenarios, in which the IDRC iS greater than the DRC., the implication is that the capacity expansion or the policy associated with the relevant scenario has raised the opportunity cost of earning or saving foreign exchange, relative to a no-capacity- expansion policy. Therefore, in the case of Scenario II, an expansion of phosphoric acid and nitrophosphates will raise the cost of earning foreign exchange relative to a do-nothing strategy. The same may be said of Scenario III, which incorporates a phosphoric acid plant with minimum production of calcium nitrate at Suez, and Scenario VI, which indicates that a complete import substitution policy in- volves an economic cost to the sector (an IDRC of 0.1497 as com- pared with a DRCs of 0.1286). The objective function value of the no-capacity-expansion case is 184.92 million Egyptian pounds, which is higher than the objective function values of the various scenarios shown in table 50. Thus, all the solutions presented are superior to the no-capacity-expansion case, and standard project appraisal methods could conceivably lead to the acceptance of each program, if presented without alter- natives. The same point can be made regarding the DRC ratios, because in each case the ratio is less than the exchange rate. Thus, 15. If the foreign exchange cost of fertilizer valued at c.i.f. prices is greater than the foreign exchange cost of a no-capacity-expansion situation-which is the case in this study (US$1,044 million compared with US$164 million)-and if the IDRC iS less than DRC,, then the sectoral opportunity cost associated with capacity expansion will be less than that associated with no expansion, and vice versa. 278 RESULTS OF THE DYNAMIC ANALYSIS Table 51. The Ranking of the Scenarios According to Objective Function Value and Domestic Resource Cost Domestic resource cost Objective Without With function substitution substitution Incremental Scenario value (DRCo) (DRC,) DRC (IDRC) Reference solution 5 3 2 2 Scenario I 8 7 7 7 Scenario II 7 5 4 5 Scenario III 3 1 3 3 Scenario IV 1 8 8 8 Scenario V 2 4 5 4 Scenario VI 6 6 6 6 Scenario VII 4 2 1 1 an acceptable value of the DRC ratio is also consistent with a less than optimal pattern of investment. Finally, as the ranking of the various scenarios in table 51 shows, there is no monotonic relationship between the objective function value and any of the three measures of DRC; that is, the ranking of scenarios is not invariant with the criterion (or DRC measure) chosen. Thus, a DRC ratio cannot be used as a guide to search for the best solution, because a lower DRC ratio may be associated with a higher cost investment program. For example, the lowest DRCS and IDRC are associated with Scenario VII, which is ranked fourth from the point of view of the objective function, and the highest DRC value (for all three measures) is associated with Scenario IV, which has the lowest objective function value.'6 This discrepancy in ranking is not surprising, however, since the four criteria are essentially different: namely, total (domestic plus foreign) cost minimization, minimization of the domestic oppor- tunity cost of foreign exchange with and without substitution effects, and minimization of the incremental opportunity cost of foreign exchange. There is no reason to expect that the rankings based on 16. This point is also made in Ardy Stoutjesdijk and Larry Westphal, eds., Industrial Investment Analysis Under Increasing Returns (forthcoming). A SENSITIVITY ANALYSIS OF THE RESULTS 279 total (domestic plus foreign) cost minimization will be identical to those based on the minimization of domestic costs per unit of inter- national value added. Two further points regarding the ranking of scenarios are of in- terest. First, rankings based on domestic resource cost with and without substitution are not the same; that is, if one chooses to use a DRC measure to rank projects, ignoring the substitution effects can be misleading. Second, rankings based on domestic resource cost with substitution effects are very similar to those based on the incre- mental DRC. In fact, there is only one difference; namely, Scenarios II and V reverse their rank order between these two measures. Thus, the rank correlation between these measures would be high. A Sensitivity Analysis of the Results A detailed sensitivity, or break-even, analysis of the results has been precluded by the high cost of computer runs. Several parameters have been examined, however, to provide an answer to the question: At what value of this parameter does the basic reference solution change? In effect, this is a single-parameter, break-even analysis that indicates the value of, say, the discount rate (the opportunity cost of capital) at which capacity expansion is not optimal, or the level of import prices at which imports become inefficient (as in the case of TSP) or efficient (as in the case of urea, CAN 33.5 percent, and ssp). That is, one can evaluate the break-even point alone rather than several solutions at varying discount rates or varying import prices. This provides a range within which the optimal solu- tion remains feasible. Such an analysis is performed for the oppor- tunity cost of capital; the investment cost; the exchange rate; the import prices for TSP, urea, CAN 33.5 percent, and ssp; and the trans- port cost for the reference solution. Finally, a brief description is given of the computer runs associated with alternative demand projections. The opportunity cost of capital The opportunity cost of capital assumed in this study is 10 per- cent. Given the usual problems associated with this parameter in most developing countries (LDCS), however, planners face an im- portant question: If this value were an underestimate of the true 280 RESULTs OF THE DYNAMIC ANALYSIS cost of capital, does this affect the planned capacity expansion and the resultant change in the consumption and production patterns? In other words, how high must this cost of capital be for no capac- ity expansion to occur? Based on the assumption that the life of the plant will be fifteen years, this value is found to be 31.4 percent."7 A longer plant life leads to a higher value of break-even cost of capital and vice versa, as is seen below: Break-even Plant life cost of capi- years tal (percent) 5 18.6 10 29.6 15 31.4 20 31.8 25 31.9 Plant lives of more than fifteen years affect the break-even value of the opportunity cost only marginally; that is, a twenty-year plant life leads to a value of 31.8 percent and a twenty-five-year plant life to 31.9 percent. Compared with the value of 31.4 percent for a fifteen- year plant life, this is not a significant change. Plant lives of fewer than fifteen years lead to slightly larger changes; for example, a ten- year plant life leads to a break-even value of 29.6 percent for the cost of capital, and a five-year plant life to 18.6 percent. Thus, as plant life increases, the marginal changes in the break-even cost of capital decline. This analysis shows that, even if one chooses to make a pessimistic assumption on plant life, the phosphoric acid plant would be very robust over a large range of the cost of capital. Given this large range of flexibility in the choice of the cost of capital, one can state with some confidence that the erection of the phosphoric acid plant, and the accompanying changes in the production structure at Abu Zaabal from ssp to powdered TSP, should take place. Investment costs For any new capacity expansion that takes place, a relevant ques- tion is how sensitive is the erection of the new plant to the capital 17. This is determined by solving for the value of the cost of capital that equates the discounted net cost of the reference solution (capacity expansion) to the dis- counted net cost of the no-capacity-expansion case. An iterative procedure was used because of the nonlinearities introduced by the discount rate and the capital recovery factor, both of which are functions of the opportunity cost of capital. A SENSITIVITY ANALYSIS OF THE RESULTS 281 cost used in the study? Since the only new capacity expansion in this study is the phosphoric acid plant, it is necessary to investigate only how high the capital cost of this plant could be before its con- struction became inefficient. This value is US$43.58 million-that is, two-and-a-half times as much as assumed in the study."8 The conversion of ssp to powdered TSP at Abu Zaabal has, how- ever, been assumed to be a costless operation. This is not strictly true, although the cost involved is marginal (the estimate is between US$10,000 and US$20,000) and the conversion is simple enough to be performed locally, so that no foreign exchange need be involved. The range is large enough to accommodate errors in capital cost estimates or changes in the real value of the capital cost of the phos- phoric acid plant and to make the necessary alterations to the ssp den to obtain powdered TSP. The exchange rate The major components of total cost affected by the exchange rate are the capital cost, the import price of raw materials and final product, and the foreign exchange component of the domestic cost. The basic issue of domestic production versus imports is thus af- fected by the relative incentive provided to either production or imports by the exchange rate. Since the foreign exchange cost is a fraction of the total cost, the incentive to import the final product will decline relative to producing the product domestically as the exchange rate increases; on the contrary, the relative incentive to produce domestically will decrease as the exchange rate declines. The foreign exchange rate at which there will be no expansion of domestic capacity (that is, only imports) can be evaluated by cal- culating the various domestic resource costs, which measure com- parative advantage. For the reference solution, the incremental domestic resource cost is 0.1094. This indicates that capacity expan- sion of the phosphoric acid plant (and the associated substitution in production and consumption) would be profitable as long as the exchange rate is greater than US$1 = 0.1094 Egyptian pounds, which is well below the official and shadow exchange rates. Thus, 18. This is determined by obtaining the incremental annual capital charge that is equivalent to the savings generated by the capacity-expansion solution over the no-capacity-expansion case. Since the site factor (1.3) and the capital recovery factor (0.1314) are known, the increment in the total capital can be obtained. 282 RESULTh OF THE DYNAMIC ANALYSIS the break-even point that would reverse the optimal strategy is far too low to be of any concern."9 The import prices The import prices for the final products used in this study are the long-term prices expected to prevail over the planning horizon. In the reference solution, there would be imports of TSP in the second and third time periods (18,000 tons a year and 68,000 tons a year, respectively), whereas there would be no imports of SSP, CAN 33.5 percent, or urea. In this section, two questions are raised. First, how high could the price of TSP rise before imports become infeasible?20 Second, how low could the prices of ssP, CAN 33.5 percent, and urea fall before imports become feasible?2' in answer to the first question, the c.i.f. import price of TSP would have to rise to US$170 a ton before imports would be eliminated. Compared with the expected long-term price of US$150 a ton, there is about 14 percent flexibility before the break-even point is reached. In response to the second question, imports of ssp would be effi- cient if the c.i.f. import price dropped to US$22.87 a ton. There could, however, be imports of ssr in the second time period if the price fell to US$37.90 a ton and in the third time period if the price fell to US$38.25 a ton. Compared with the c.i.f. import price of US$50 a ton, this indicates that a drop of 20 percent in the price would be necessary to make imports attractive. For CAN 33.5 per- cent, the break-even prices are US$65.73 a ton in the first period and US$47.58 a ton in the second and third periods.22 For urea, the 19. This is particularly true in the case of Egypt, where the exchange rate is likely to be "sticking downwards" for some time to come. If labor were treated as a variable cost, however, this break-even exchange rate would be US$1 = 0.42 Egyptian pounds, which is above the official exchange rate. 20. This question is answered by selecting a representative governorate (Behera) that has all its P206 demand met by imports in the two time periods in the reference solution and the plant (Abu Zaabal) that would supply that governorate with the cheapest source of P205 if imports were not permitted. 21. In the case of ssP, CAN 33.5 percent, and urea, the governorate is chosen so that the difference in the transport cost associated with shipments from the plant and the port are minimized and so that the plant is the cheapest source of supply of the nutrient. 22. The decline in these prices can be explained by the decline in the capacity rental at Talkha. A SENSITIVITY ANALYSIS OF THE RESULTS 283 Table 52. Comparison of Break-even Import Prices and Import Price Used in the Planning Modelfor Selected Products (U.S. dollars per ton) Break-even import price Import price (c.i.f.) used in Product 1980 1983 1986 the model SSP 22.87 37.90 38.25 50 CAN 33.5 percent 65.73 47.58 47.58 120 Urea 50.20 50.20 59.70 150 break-even prices are US$50.20 a ton in the first two time periods and US$59.70 a ton in the last period. As table 52 shows, the highest break-even import price for CAN 33.5 percent and urea is still con- siderably lower than the c.i.f. import price used in the study. Since import prices are unlikely to fall to such low levels, Egypt may be expected to retain its comparative advantage in these products. Transport cost No analysis is performed on the transport cost of final products, because the major changes expected to result from an increase in this cost would be an increase in total cost and an alteration in the shipment pattern. Such an analysis would require several high-cost computer runs, and these were not performed because the cost of obtaining the information did not seem to outweigh the associated benefits. The transport cost of intermediate products is, however, a differ- ent case. The reference solution contains interplant shipments of ammonia from Suez to Helwan and no shipments of phosphoric acid from Abu Zaabal. Thus, two questions can again be raised. First, what increase in the transport cost of ammonia will lead to the production of ammonia at Helwan? Second, what decrease in the transport cost of phosphoric acid will lead to the shipment of the acid from Abu Zaabal to Kafr El Zayaat and to the subsequent production of TSP there? Two possibilities exist at Helwan since ammonia can be produced either by using coke gas or by upgrading the system to use natural gas. Ammonia from coke-oven gas production would be feasible if the transport cost of ammonia increased by at least 61.59 Egyptian 284 RESULTS OF THE DYNAMIC ANALYSIS pounds a ton.2" Since this amount is so large, one can be reasonably certain that shipment from Suez to Helwan would be preferable to ammonia production at Helwan from coke-oven gas. For ammonia production from natural gas, the necessary increase in the transport cost would be significantly smaller-3.52 Egyptian pounds a ton. This figure, however, ignores the variable cost component of the investment cost required for ammonia production from natural gas, and it is thus an underestimate. Consequently, the ammonia shipment activity appears to be stable. If, however, there are significant uncer- tainties in transport cost, then converting to production of ammonia from natural gas at Helwan should be investigated. The actual transport cost of phosphoric acid from Abu Zaabal to Kafr El Zayaat is 6.05 Egyptian pounds a ton. To permit interplant shipment of this acid, the transport cost would have to decline to 3.20 Egyptian pounds a ton, almost a 50 percent decrease. Therefore, as long as the transport cost remains higher than 3.20 Egyptian pounds a ton, the shipment activity plus the conversion of ssp to TSP at Kafr El Zayaat remains a less than optimal policy. Alternative demand projections The demand projections provided by the Ministry of Agriculture appear to be subject to some uncertainty. The annual growth rate utilized in this study was 4.6 percent for nitrogen and 10.43 percent for phosphorus. An alternative annual growth rate suggested by the ministry was used for nitrogen consumption-6.9 percent. This was thought to be an expected maximum value, so higher growth rates were not considered. Given this growth rate, the model was rerun to determine the changes. It was found that no capacity expansion was necessary for the nitrogenous fertilizer industry over the entire planning period. This is not surprising in view of the excess capacity at Talkha and Suez. Thus, the higher growth rate causes the calcium nitrate plant at Suez and the calcium ammonium nitrate plant at Talkha I to operate nearly at full-capacity utilization. No additional computer runs were carried out for the phosphate fertilizer growth rate, because the ministry and the Industrial Projects Department of the World Bank seemed to agree that the 10.43 percent annual growth was the expected maximum. Given the nature 23. This is obtained from the reduced cost of coke-oven gas ammonia production. CONCLUSIONS 285 of the system, and since phosphorus consumption accounts for less than 15 percent of total nutrient consumption, it may be said that a higher growth rate would necessitate more imports, though no attempt has been made to determine the growth rate that would lead to capacity expansion over and above the phosphoric acid plant at Abu Zaabal. Conclusions Only the major results of the case study have been described in this chapter; nevertheless, it is clear that the planning methodology used in this study yields results that are specific enough to be mean- ingful at an operational level. Moreover, the information contained in the model's solution permits an analysis of the robustness of particular project proposals in the light of possible errors in crucial parameters. Finally, this approach can be used to determine the implications of specific scenarios regarding domestic production patterns, imports, and exports. Because of these properties, this planning methodology is a powerful tool of industrial investment analysis, giving results that provide an improved basis for the subse- quent phase in the project cycle-that is, feasibility studies. The major empirical conclusions of the case study are the follow- ing. First, Egypt is completely self-sufficient in nitrogen fertilizer with the existing capacity up to 1987. Triple superphosphate will need to be imported, however, in the 1982-84 period at the rate of 18,000 tons a year and in the 1985-87 period at the rate of 68,000 tons a year. Second, the nitrogen fertilizer industry should produce calcium ammonium nitrate with 33.5 percent nutrient content at Aswan, Helwan, and Talkha I, thus moving away from the 31 percent va- riety. Third, it is not economical to produce ammonia at Helwan using coke-oven gas; neither is it advisable to convert to a production process using natural gas. It is more economical to produce the required ammonia at Suez and ship it to Helwan over the entire planning period. Thus, Egypt needs to build the required ammonia movement facilities. Fourth, calcium nitrate production at Suez is optimal at the level of 23,000 tons a year in the 1979-81 period and at only 1,000 tons a 286 RESULTS OF THE DYNAMIC ANALYSIS year and 2,000 tons a year in the two subsequent time periods. In practical terms, since the plant cannot operate at such low levels, these figures imply that it must shut down. The erection of a nitro- phosphate complex producing a mixture of an equal-weight nitro- phosphate (20-20-0) and calcium nitrate is not optimal and will involve an additional cost of 2.5 million Egyptian pounds to the fertilizer sector over the planning period 1979-87. A minimum level (150,000 tons a year) of calcium nitrate production at Suez also entails an additional cost to the sector of 1.9 million Egyptian pounds over the planning period. Thus, if some activity other than the opti- mal production pattern mentioned above must be undertaken at Suez, calcium nitrate production is superior to a combined nitro- phosphate/calcium nitrate production activity. Fifth, phosphorus demand can be adequately taken care of by the erection of a phosphoric acid (54 percent P205) plant with a capac- ity of 134,000 tons a year at Abu Zaabal in the first time period (1979-81) at a cost of US$23 million. The erection of this plant must be accompanied by a subsequent change in the production of the final product at Abu Zaabal from single superphosphate to powdered triple superphosphate.24 This acid plant is robust (in the sense that it appears in all capacity-expansion scenarios) and has a high rate of return (31.4 percent) associated with it. This is one of the most important conclusions of the study. Sixth, the production of phosphoric acid and powdered TSP at Abu Zaabal and the continued production of single superphosphate at Kafr El Zayaat and Assiout will satisfy domestic demand for phosphorus up to 1981. From 1982 to 1987, imports of TSP will be necessary. A policy of complete import substitution (or complete self-sufficiency) will cost the sector an additional 1.38 million Egyp- tian pounds over the planning period. (This is less than the cost to the sector of calcium nitrate production at Suez.) Such a policy entails the erection of a larger phosphoric acid plant (with a capac- ity of 214,000 tons a year) at Abu Zaabal in the first time period (1979-81) at a cost of about US$29 million, and the subsequent erection of a sulfuric acid plant in the second time period (1982-84) at the same location with a capacity of 146,000 tons a year at a cost of US$17 million. This must be accompanied by a conversion to production of powdered TSP at Abu Zaabal in the first time period and at Kafr El Zayaat in the second time period. 24. The cost of this conversion is marginal, about US$15,000. CONCLUSIONS 287 Seventh, in all scenarios considered, an export-oriented strategy is beneficial to the sector, since Egypt possesses considerable com- parative advantage. An export strategy without any capacity expan- sion will result in a savings of 14 million Egyptian pounds over the period under consideration, if Egypt can capture the required export market for urea.25 As domestic demand increases, however, this will result in Egypt's relinquishing the captured export market to its competitors. In order to capture and to preserve an export market of 500,000 tons a year of final product, a phosphoric acid plant at Abu Zaabal and an ammonia urea complex (of 743,000 tons a year of urea) at Suez would have to be erected. Other loca- tions not considered might be equally preferred; they would, how- ever, need to have easy access to natural gas and good port facilities. The urea capacity expansion would have to be phased between the second and third time periods. The total investment cost for the acid plant and the new complex at Suez would amount to US$237 million. Such a strategy would result in a savings to the sector of 19 million Egyptian pounds. Capturing and maintaining a smaller export market of 250,000 tons a year would result in a smaller sav- ings of 15 million Egyptian pounds and would entail the erection of only the phosphoric acid plant at Abu Zaabal. Eighth, considerable concern has been voiced over the transport sector being the major bottleneck in the distribution of fertilizer. The study indicates that the total transport capacity used by the sector in 1975 is sufficient to transport fertilizers up to 1984. Only in the last time period (1985-87) is the 1975 transport capacity of 378 million ton-kilometers exceeded. The conversion to the higher- analysis fertilizers-urea and TSP-permits the desired distribution of the fertilizers to all the governorates since more nutrients are transported per ton of final product. Therefore, the necessary dis- tribution of fertilizers can be accomplished with the 1975 transport capacity and, apart from replacement costs, major new investments will not be required to transport fertilizers until 1985. Finally, a cost saving of 5 million Egyptian pounds can be achieved over the plan period if the farmers are prepared to use all the urea that is produced, instead of the traditional fertilizer. Therefore, educating the farmers could be beneficial, as long as the associated costs are less than 5 million Egyptian pounds. 25. The savings refer to the difference in the discounted total cost between the optimal-capacity-expansion case and the case incorporating the export strategy. 288 RESULTS OF THE DYNAMIC ANALYSIS Appendix. Summary Tables of the Results of Scenarios I-VII In this appendix, tables 53 to 59 present the summary results of Scenarios I-VII. These tables correspond to table 31, which gives the summary results of the reference solution to the dynamic planning model of the Egyptian fertilizer sector. Table 53. Summary of the Results of Scenario I (objective function value = 179.81 million Egyptian pounds), Category 1980 1983 1986 Cost (millions of Egyptian pounds) Transport cost 5.83 6.29 6.90 Domestic raw materials cost 13.16 14.86 16.06 Operating cost 17.33 19.11 20.05 Working capital cost 0.82 0.91 0.97 Capital cost 2.17 2.17 2.17 Import cost Raw materials 2.23 2.52 2.52 Final products 0.00 0.18 3.19 Export revenue -3.90 -3.78 -3.78 Total cost 37.63 42.27 48.08 Total demand (thousands of tons) Nutrient demand 608 687 765 Nitrogen 496 552 607 Phosphorus 112 135 158 Fertilizer demand 1,695 1,922 2,033 Production (thousands of tons) Final products 1,795 2,022 2,133 Intermediate products 1,663 1,753 1,796 Capacity expansion (thousands of tons) Phosphoric acid (at Abu Zaabal) 134 0 0 Nitrophosphate/cN (at Suez) 88 0 0 Imports (thousands of tons) Final products (TsP) 0 3 53 Raw materials 137 158 158 Exports (thousands of tons) Final products 100 100 100 APPENDIX. SUMMARY TABLES OF THE RESULTS OF SCENARIOS I-Vll 289 Table 53 (continued) Category 1980 1983 1986 Shipments of final products to governorates from following plants (thousands of tons) Aswan (CAN 33.5 percent) 295 311 344 Helwan (CAN 33.5 percent + As) 67 67 67 Kafr El Zayaat (ssp) 58 168 168 Assiout (ssp) 178 178 178 Suez (NP + CN) 54 79 79 Abu Kir (urea) 133 273 421 Talkha (CAN 33.5 percent + urea) 740 681 613 Abu Zaabal (Tsp) 163 163 163 Total shipments 1,695 1,922 2,033 Interplant shipments of intermediate products from following plants (thousands of tons) Aswan _ _ _ Helwan Kafr El Zayaat (sulfuric acid to Abu Zaabal) 1 1 1 Assiout _ _ _ Suez (ammonia to Helwan) 36 36 36 Abu Kir - - - Talkha Abu Zaabal - - - Total shipments 37 37 37 Note: Detail may not sum to totals because of rounding. a. Discounted total cost in 1975 Egyptian pounds for the period 1979-87. 290 RESULTS OF THE DYNAMIC ANALYSIS Table 54. Summary of the Results of Scenario 11 (objective function value = 179.14 million Egyptian pounds), Category 1980 1983 1986 Cost (millions of Egyptian pounds) Transport cost 6.21 6.46 7.04 Domestic raw materials cost 13.25 14.79 15.99 Operating cost 17.45 19.03 19.96 Working capital cost 0.82 0.91 0.96 Capital cost 1.21 1.21 1.21 Import cost Raw materials 2.24 2.52 2.52 Final products 0.00 1.07 4.07 Export revenue -3.78 -3.78 -3.78 Total cost 37.39 42.19 47.97 Total demand (thousands of tons) Nutrient demand 608 687 765 Nitrogen 496 552 607 Phosphorus 112 135 158 Fertilizer demand 1,695 1,922 2,033 Production (thousands of tons) Final products 1,860 2,065 2,176 Intermediate products 1,723 1,791 1,833 Capacity expansion (thousands of tons) Phosphoric acid (at Abu Zaabal) 134 0 0 Imports (thousands of tons) Final products (TsP) 0 18 68 Raw materials 138 158 158 Exports (thousands of tons) Final products 100 100 100 Shipments of final products to governorates from following plants (thousands of tons) Aswan (CAN 33.5 percent) 295 311 344 Helwan (CAN 33.5 percent + AS) 57 39 39 Kafr El Zayaat (ssp) 69 168 168 Assiout (ssp) 178 178 178 Suez (cN) 150 150 150 Abu Kir (urea) 133 272 421 Talkha (CAN 33.5 percent + urea) 715 683 613 Abu Zaabal (TsP) 163 163 163 Total shipments 1, 760 1,965 2,076 APPENDIX. SUMMARY TABLES OF THE RESULTS OF SCENARIOS I-VII 291 Table 54 (continued) Category 1980 1983 1986 Interplant shipments of intermediate products from following plants (thousands of tons) Aswan _ _ _ Helwan Kafr El Zayaat (sulfuric acid to Abu Zaabal) 1 1 1 Assiout Suez (ammonia to Helwan) 24 24 24 Abu Kir - - - Talkha (ammonia to Helwan) 7 - - Abu Zaabal - - - Total shipments 32 25 25 Note: Detail may not sum to totals because of rounding. a. Discounted total cost in 1975 Egyptian pounds for the period 1979-87. 292 RESULTS OF THE DYNAMIC ANALYSIS Table 55. Summary of the Results of Scenario III (objective function value = 163.19 million Egyptian pounds)a Category 1980 1983 1986 Cost (millions of Egyptian pounds) Transport cost 6.13 6.41 6.79 Domestic raw materials cost 16.28 16.32 16.37 Operating cost 20.09 20.17 20.26 Working capital cost 0.95 0.95 0.95 Capital cost 0.00 0.00 0.00 Import cost Raw materials 1.53 1.53 1.53 Final products 4.54 7.54 10.55 Export revenue -18.49 -13.55 -8.64 Total cost 31.03 39.37 47.81 Total demand (thousands of tons) Nutrient demand 608 687 765 Nitrogen 496 552 607 Phosphorus 112 135 158 Fertilizer demand 1,896 2,024 2,153 Production (thousands of tons) Final products 2,336 2,346 2,358 Intermediate products 1,748 1,757 1,767 Capacity expansion (thousands of tons) None 0 0 0 Imports (thousands of tons) Final products (Tsp) 76 126 176 Raw materials 77 77 77 Exports (thousands of tons) Final products 440 323 206 Shipments of final products to governorates from following plants (thousands of tons) Aswan (CAN 33.5 percent) 344 344 344 Helwan (CAN 33.5 percent + AS) 92 92 92 Kafr El Zayaat (ssp) 168 168 168 Assiout (ssp) 178 178 178 Suez (CN) 153 163 175 Abu Kir (urea) 5 123 240 Talkha (CAN 33.5 percent + urea) 787 787 787 Abu Zaabal (ssp) 168 168 168 Total shipments 1,896 2,024 2,153 APPENDIX. SUMMARY TABLES OF THE RESULTS OF SCENARIOS I-VII 293 Table 55 (continued) Category 1980 1983 1986 Interplant shipments of intermediate products from following plants (thousands of tons) Aswan _ _ _ Helwan Kafr El Zayaat - - - Assiout _ _ _ Suez (ammonia to Helwan) 24 21 19 Abu Kir - - - Talkha (ammonia to Helwan) 12 15 17 Abu Zaabal - - - Total shipments 36 36 36 Note: Detail may not sum to totals because of rounding. a. Discounted total cost in 1975 Egyptian pounds for the period 1979-87. 294 RESULTS OF THE DYNAMIC ANALYSIS Table 56. Summary of the Results of Scenario IV (objective function value = 158.09 million Egyptian pounds)$ Category 1980 1983 1986 Cost (millions of Egyptian pounds) Transport cost 6.21 6.23 6.96 Domestic raw materials cost 16.91 21.55 25.28 Operating cost 20.89 24.51 28.02 Working capital cost 1.01 1.21 1.40 Capital cost 1.21 10.21 12.48 Import cost Raw materials 2.47 2.52 2.52 Final products 0.00 1.07 4.07 Export revenue -19.39 -27.30 -35.49 Total cost 29.29 30.81 45.24 Total demand (thousands of tons) Nutrient demand 608 687 765 Nitrogen 496 552 607 Phosphorus 112 135 158 Fertilizer demand 1,748 1,911 2,089 Production (thousands of tons) Final products 2,248 2,561 2,934 Intermediate products 1,953 2,108 2,348 Capacity expansion (thousands of tons) Phosphoric acid (at Abu Zaabal) 134 0 0 Ammonia (at Suez) 0 431 0 Urea (at Suez) 0 424 319 Imports (thousands of tons) Final products (TsP) 0 18 68 Raw materials 152 158 158 Exports (thousands of tons) Final products 500 650 845 Shipments of final products to governorates from following plants (thousands of tons) Aswan (CAN 33.5 percent) 344 344 344 Helwan (CAN 33.5 percent + As) 92 92 92 Kafr El Zayaat (ssp) 69 168 168 Assiout (ssp) 178 178 178 Suez (cN + urea) 87 9 125 Abu Kir (urea) 27 169 231 Talkha (CAN 33.5 percent + urea) 787 787 787 Abu Zaabal (TsP) 163 163 163 Total shipments 1,748 1,911 2,089 APPENDIX. SUMMARY TABLES OF THE RESULTS OF SCENARIOS I-VII 295 Table 56 (continued) Category 1980 1983 1986 Interplant shipments of intermediate products from following plants (thousands of tons) Aswan _ _ _ Helwan Kafr El Zayaat (sulfuric acid to Abu Zaabal) 1 1 1 Assiout Suez (ammonia to Helwan) 36 36 36 Abu Kir - - - Talkha Abu Zaabal Total shipments 37 37 37 Note: Detail may not sum to totals because of rounding. a. Discounted total cost in 1975 Egyptian pounds for the period 1979-87. 296 RESULTS OF THE DYNAMIC ANALYSIS Table 57. Summary of the Results of Scenario V (objective function value = 161.96 million Egyptian pounds), Category 1980 1983 1986 Cost (millions of Egyptian pounds) Transport cost 5.69 6.10 6.97 Domestic raw materials cost 15.09 16.96 17.31 Operating cost 18.34 21.01 21.66 Working capital cost 0.89 1.01 1.04 Capital cost 1.21 1.21 1.21 Import cost Raw materials 2.24 2.52 2.52 Final products 0.00 1.06 5.86 Export revenue -10.50 -12.48 -10.69 Total cost 32.96 37.39 45.87 Total demand (thousands of tons) Nutrient demand 608 687 765 Nitrogen 496 552 607 Phosphorus 112 135 158 Fertilizer demand 1,685 1,968 2,056 Production (thousands of tons) Final products 1,935 2,265 2,353 Intermediate products 1,737 1,960 2,035 Capacity expansion (thousands of tons) Phosphoric acid (at Abu Zaabal) 134 0 0 Imports (thousands of tons) Final products (Tsp) 0 18 98 Raw materials 138 158 158 Exports (thousands of tons) Final products 250 297 297 Shipments of final products to governorates from following plants (thousands of tons) Aswan (CAN 33.5 percent) 297 344 344 Helwan (CAN 33.5 percent + As) 92 92 92 Kafr El Zayaat (ssp) 69 168 77 Assiout (ssp) 178 178 178 Suez (cN) 12 87 175 Abu Kir (urea) 133 148 240 Talkha (CAN 33.5 percent + urea) 740 787 787 Abu Zaabal (Trsp) 163 163 163 Total shipments 1,685 1,968 2,056 APPENDIX. SUMMARY TABLES OF THE RESULTS OF SCENARIOS I-VII 297 Table 57 (continued) Category 1980 1983 1986 Interplant shipments of intermediate products from following plants (thousands of tons) Aswan _ _ _ Helwan Kafr El Zayaat (sulfuric acid to Abu Zaabal) 1 1 1 Assiout - - - Suez (ammonia to Helwan) 36 36 19 Abu Kir - - - Talkha Abu Zaabal Total shipments 37 37 37 Note: Detail may not add to totals because of rounding. a. Discounted total cost in 1975 Egyptian pounds for the period 1979-87. 298 RESULTS OF THE DYNAMIC ANALYSIS Table 58. Summary of the Results of Scenario VI (objective function value = 178.66 million Egyptian pounds)a Category 1980 1983 1986 Cost (millions of Egyptian pounds) Transport cost 6.12 6.45 7.16 Domestic raw materials cost 13.15 15.08 16.69 Operating cost 17.01 18.48 20.52 Working capital cost 0.81 0.91 1.01 Capital cost 1.52 2.43 2.43 Import cost Raw materials 2.32 2.76 3.22 Final products 0.00 0.00 0.00 Export revenue -3.50 -3.78 -3.78 Total cost 37.43 42.34 47.25 Total demand (thousands of tons) Nutrient demand 608 667 765 Nitrogen 496 552 607 Phosphorus 112 135 158 Fertilizer demand 1,680 1,737 2,002 Production (thousands of tons) Final products 1,780 1,831 2,102 Intermediate products 1,669 1,838 1,954 Capacity expansion (thousands of tons) Phosphoric acid (at Abu Zaabal) 214 0 0 Sulfuric acid (at Abu Zaabal) 0 146 0 Imports (thousands of tons) Final products 0 0 0 Raw materials 141 160 182 Exports (thousands of tons) Final products 100 100 100 Shipments of final products to governorates from following plants (thousands of tons) Aswan (CAN 33.5 percent) 317 311 344 Helwan (CAN 33.5 percent + AS) 67 67 67 Kafr El Zayaat (ssp) 52 0 76 Kafr El Zayaat (TsP) 0 91 93 Assiout (ssp) 178 105 178 Suez (CN) 23 1 3 Abu Kir (urea) 133 281 421 Talkha (CAN 33.5 percent) 297 247 188 Talkha (urea) 443 465 465 Abu Zaabal (Tsp) 168 168 168 Total shipments 1,680 1,737 2,002 APPENDIX. SUMMARY TABLES OF THE RESULTS OF SCENARIOS I-VII 299 Table 58 (continued) Category 1980 1983 1986 Interplant shipments of intermediate products from following plants (thousands of tons) Aswan _ _ _ Helwan Kafr El Zayaat (sulfuric acid to Abu Zaabal) 8 0 0 Assiout - - - Suez (ammonia to Helwan) 36 36 36 Abu Kir - - - Talkha Abu Zaabal (phosphoric acid to Kafr El Zayaat) - 67 68 Total shipments 44 103 104 Note: Detail may not add to totals because of rounding. a. Discounted total cost in 1975 Egyptian pounds for the period 1979-87. 300 RESULTS OF THE DYNAMIC ANALYSIS Table 59. Summary of the Results of Scenario VII (objective function value = 172.31 million Egyptian pounds), Category 1980 1983 1986 Cost (millions of Egyptian pounds) Transport cost 5.45 6.24 6.70 Domestic raw materials cost 13.65 14.79 16.12 Operating cost 15.56 17.90 19.56 Working capital cost 0.79 0.88 0.96 Capital cost 1.21 1.21 1.21 Import cost Raw materials 2.24 2.52 2.52 Final products 0.00 1.07 4.07 Export revenue -3.78 -3.78 -3.78 Total cost 35.12 40.83 47.36 Total demand (thousands of tons) Nutrient demand 608 667 765 Nitrogen 496 552 607 Phosphorus 112 135 158 Fertilizer demand 1,563 1,830 1,997 Production (thousands of tons) Final products 1,663 1,930 2,096 Intermediate products 1,350 1,591 4,730 Capacity expansion (thousands of tons) Phosphoric acid (at Abu Zaabal) 134 0 0 Imports (thousands of tons) Final products (TrsP) 0 18 68 Raw materials 138 158 158 Shipments of final products to governorates from following plants (thousands of tons) Aswan (CAN 33.5 percent) 253 344 344 Helwan (CAN 33.5 percent + AS) 14 67 67 Kafr El Zayaat (ssp) 69 168 168 Assiout (ssp) 178 178 178 Suez (CN) 0 3 3 Abu Kir (urea) 421 421 421 Talkha (CAN 33.5 percent + urea) 465 487 653 Abu Zaabal (TsP) 163 163 163 Total shipments 1,563 1,830 1,997 APPENDIX. SUMMARY TABLES OF THE RESULTS OF SCENARIOS I-VII 301 Table 59 (continued) Category 1980 1983 1986 Interplant shipments of intermediate products from following plants (thousands of tons) Aswan - - - Helwan Kafr El Zayaat (sulfuric acid to Abu Zaabal) 1 1 1 Assiout - - - Suez (ammonia to Helwan) 14 36 36 Abu Kir - - - Talkha Abu Zaabal Total shipments 15 37 37 Note: Detail may not add to totals because of rounding. a. Discounted total cost in 1975 Egyptian pounds for the period 1979-87. Appendix Computer-Readable Representation of the Model GAMS 0 E G Y P T I A N FERTILIZER MODEL 01/30/80 16.44.20 PAGE SET DEFINITIONS 4 SET I PLANT LOCATIONS / 6 ASWAN I HELWAN 8 ASSIOUT 9 KAFR-EL-ZT IT1 SUEZ 12 TALKI(A 13 ADU-KIE 14 ABU-ZAABAL 15 / 16 17 J DEMAND REGIONS / 18 19 ALEXANDRIA ALEXANDRIA 20 BEDERA DATANHUR 21 GHARDIA TANTA 22 KAFR-EL-SH KANE EL-SHEIKH 23 DAKAHLIA EL MAASJRA 24 25 DADIIETTA DANlIETTA 26 SHARKIA ZAGAZ1G 27 ISMAILIA ISMAILIA 02 SDOD SUEZ 46. 2 MENZOUFIA SHIBIN El. KOII 3D 31 KALUBIA BENHA 32 G1ZA GIZA 33 BENI-SUEF BENI-SUEF 34 FAYOUM EL-FAYODU 35 MINIA EL-IIINiA 36 37 ASSIOUT ASSODIUT 38 NEW-VALLEY EL KHARGA 39 SOHAG SOHAG 40 QUENA QURNA 41 ASWAN ASWAN 42 43 L EXPORT PORTS OF EDIT 44 45 ALEXANDRIA 46 SUEZ 47 48 04 PRODUCTIVE UNITS 49 50 SULF-A-S SULFURIC ACLD: DULFUR 51 SULF-A-P DULFURIC ACID: PYRITES 52 NITE-ACID NITNIC ACID 53 AMM-ELEC AMMONIA: WATER ELECTROLYSIS N4 AUM-C-GAS AMMONIA: COKE GAS 53 56 AMM-N-GAS AMMONIA: NATURAL GAS 57 A0-0R-GAS A0MONIA: REPINERY IAS w - - - - H o Fo o < H H po o o U o - O e < J o O n < O o ,,> , X D X g Q H . 5 < H X F X = w - p, g Z ' ...... w S w e e w w e < - o a e e iz H > g b > w < H H H H X fH H P Es Dl o H H H H H X z D > D w ° >, Z Z Z S < e e S H Z H H H < EX ww rA SK zy =E Es WX Www > ° w >=EEW e° o e S w H w e rv S S H < °e w U z X w X X b a D X X > w O = z 2 X ffi e D. X ffi H U v w < z 2 Z 1h X O < o _l o M w < U U n > M < 14 < ° ° °Z D Z S X X e 8 z > t S -> z v U < W 1:4 D > N < <: U 5 ° Z e Q S < w D Z : sL, . , ..... w J H Z z z w , 0d e t z w w O z U t Z H w w w o o .. z w b _ :: z H C4 F z Z J a O : 2 _ Z Z Z Z o.  , DS z9D zD SZ e a: w X z H : e w H H H H z S S O; S S Z  H ; 4 > S S S S D; X < S Z < H e e H w S a: w _ < 2; 4 S O X z. O < < w 9 Z w w Z < < < < o U o O S S Q W > H Z D <: z N Es Q U Cl z < S > _ D H H W w e w ° w < < ° Z w ° z W o = _1 o z Z I I o W o o o U o o w a un 3 e H Z Z v O O o X Zzo zSa = O = < < < Z ' <-< a - >0 0 Z: ' £ W47_a)D e 1 ;5 M S w w H °e D D H Z Z S Z Z Z Z Z Z >4 ffi > w w w S D. S D D z w Z Z z Z S > > Z N Q < D < w > z z w w z <: c: < < Z c> U U 5: D _ O W H H Z D <: Z Z =< :o D U O s O < _ N 1; 0 v W > c O C) w ew D z :s C _ cs n / rl sD > on ov o _ N n :r ur Eo > o: oo o _ N m < w Q b w v o _ N o < w fio S w oa o _ es1 < w hD x z o o _ W = = = = oa ¢. D. JE ffi o z w .^ O U z A o o _ H uo Q o w w w _ ve H w u, , z o o _ I I I I N v _ U H t1 o o N w w + w :^ - 4 Eq +  o os o =u v1 1 z ffi r _ .^ w _ _ _ | _ 8 > w ur _  o _ 1 4 C1 b U w w um > o < c) O o um z ° w < J J H O H O = N g z u w O n o:: W = W W .^ O O O D D D W Z _ rs - g- g O H X W 14 W Z W o < ^ z F W a w X o o > S > w e w z U = > > a o w r G H z z _ . vo O _ W W o w o en > o a o 1l < N X + w > S :4 S _ K H a P H W ^ Q g - o W z a z 9 w + < _ .e o o _ D O D Z N 0 4 _ w 4 o K - n a a 20 Xw D z * @ * * * * * * z O z z _ = v 1 W O D s zD O D O W U Cl w w o _ W X 1 N < o < w e w 1l !1 " I SI H K 1< w z O _  ]S X X r) W rs _ _ D :4 < xD e sD _ re O _ W O un O uz Q :^ 0 W z a a X z z < _ N e 7 N _ _ N N e u z a c> Z a] > > X O = O a z e 9 z < < + b z w w < ^ w  H  p vs 1; o un H ^ b g o v H " H C;n r o o um 1 o 3 N un _ W m: . U . W z O g so _  o o O O < e o U .n we z > <+ E S g S f g nno z oc) z z w a o < w S Z Z Z Z z < <: w w w N | E ffi z z a z w <: e o = e > U O c> U < = a z > eq z o O :1 Q o 308 00 00 00 ,-0000 000, ~ 01000 000 0.0000 00 ..0 ....0 309 OAMS 0 E G Y P T I A N FERTILIZER MODEL 01/30/80 16.44.20 PACE 7 TRANSPORTATION DATA 329 TABLE RAIL INTERPLANT RAIL DISTANCES (KES) 330 331 ADU-KIR KAFT-CL-ZT TALKHA ABD-ZAABAL HELWAN 332 333 KAFR-EL-ZT 128 334 TALKRPA 196 58 335 ABU-ZAARAL 219 85 138 336 HELWAN 253 142 155 57 337 SDEZ 371 246 298 2.4 178 338 ASSIOUT 616 504 518 420 362 339 ASWAN 1134 1022 1036 938 880 340 EXPORT-PTS 16 104 199 210 166 341 342 + SUEZ ASSIOUT ASWAN 343 344 ASSIOUT 527 345 ASWAN 1045 518 346 EXPORT-PTE 10 583 1087 340 348 349 TABLE IMPD IMPORT DISTANCES (KMS) 354 351 BARGE ROAD 352 353 ABU-XIR 16 354 RAFR-EL-ZT 104 6 355 TALKIIA 199 356 ABU-ZAABAL 210 .1 357 HELWAN 183 358 SUBZ 346 50 359 ASSIOUT 583 360 ASRAN 1087 1T 361 362 363 PARAMIETERS TUF TRANSPORT COST (LE PUR TON): FINAL PRODUCTS 364 MUFV TRANSPORT COOT (LE PER TON): IMPORTED FINAL PRODUCTS 365 MUE TRANSPORT COST (LE PER TON): EXPORTS 366 4U0I TRANSPORT COST (LB POR TON): INTERPLANT REIPMENT 367 MUR TRANSPORT COST (LB PER TON): IMPORTED RAR MATERIALS 368 369 RAIL(1,IP) = RAIL(lI,IP) RAIL(IP,I); 370 ROAD(J,"IMPORT-PTB") = MIN(ROAD(J,"ABU-KIR"),ROAD(J,"SUEZ"))i 301 372 MUF(I,J) -4 .5 A .R144IROAD(J,1 )I5ROAD(J,I); 373 MUFV(J) - 4 .5 + .0144*ROADIJ,"IMPORT-PTS') (3ROAD(J,IMPORT-PTS")4 374 MUE(I) = ( .5 + .0300MRAIL("EXPORT-PTS",I) (9RATL("EXPRET-PTS',I)2 375 MU1(I1,7 - (3.5+ .0300*RAIL(I,IPI )$RAIL(:,IP); 306 MUR(I) I ( 1.0 * .0030^1MPD(",RARGE") (SIPS(I,"RARUO"3 377 A ( .5 + .0144*IRPD(I,"ROAD' ) )ISMPD(I, ROAD - .e Q q X > o col - cs d m . so Q Ln tn r D , ; - l _ I .0 £ O ez S N D l , l O O 0s = = a: | _ 1/L 1_ C1 .. 1 7 Z e ,,, , <., O ,,, e., = N _ aD < Uf I I I I ,95 . . _ I I I Cw N1 RA O ' ' N 4 O O . , , ,,: Z _ S < , I O O Z J H .. 1 2 a, o _ N . O °' < O N X ; X r ,,, 5 , o 8 z , ,, , _ = b o In < , _ = A, W Z' H D O iD s O 0 1 I 9 O N . . . _ I I I ;s vl < O Q N 01 H ; Z > ^ N . N _ N X Z l., _ .s o < £ N O G rc .D r o N uw < so . _ a N < S s H . . _ I I I I eo z < z I I as o < < J o 0 N CX =} > P o X o o o g o o i X o o O Q O O .t V O O U. ; ° o ° U 9 < 3 ° O O D 9 A A -I .: H <|: H < £ 0 _ 1t D 1£ < 1,: < .£: H uO o I ffi J X H < < D I In I z I P >1 n =S z EH Z = + I I I I _ N O b H S E + I D EH I I > z O = N J Z Z Z Z = X X Du Z H C7 r O O Ei = :s X z I = a H :d o o I a: = o J o ° z J = < O X O O O o < D  O Y 3 5 = X N ffi Z .t al p H z O X 5 5 v Y ffi < 1L vl o cLs n < Ur So rs c0 > O  N _ 4 n D N X 0> o o O O < O tD o o o--N        N N zE X X < O 1 5 X O H X 1z1 > F =: Y a X H 0 5 _ p so, <& Y 1 5 M Q n , <: o X v _ -  s < 8 7 o s Ko S *- v n U v ') 5 * ' * ' H Y c; V y z Z z V 5 t 5 0 Z < > X O C :5 0 (s (s N N z X X X Dl X- ° X, z u-s O > _ _ _ _ n 4 J J w 4 >: S q b un O O C O um O O O on O O 0 5 LD Q _ _ ur - o _ N [ co _ 1: _ r uE ffi a1 n X V > 5  S O o e ul Q 1z1 H < ffi a O o O N N X 5 X s#D x O o vl; vl 5 ::; < > V r P 5 > X X H m - ^ ^ D vn o I X I = n 5 0 2: < H ¢ Y *r B <: S u <: Y I :- Z SC 5 N X X X m I D 5 D D J 5 V 5 ie 5 1 z; >1 0 C we s:s rs D V V u <: 5: D 0 > H 5 n: .z1 .t = X D 5 U: Ch Y m 5 5 5 X OsoOwoon<4n>0 oNo_n :D = < D a al = < ur : ^ ^ z o < 2 0 0 0 a o ^ o = xo .0 ;0 0 N a r 2 _ X a Y O e = < < a S Z : q 2 0E s0 < Y 2 Y S = D _ J O a a _ q n O O H U H < a N O g O ,<°) o > cD D = q Z X 2 o > I = < a < a > 5 = + < ' ' O = n _ a N 0 2 a I < S _ 0 0E v ?' < 1H 0 xD ° J < Y b' > :_ = 4 a 2 Z _ _ I < I H = a s E ' = > P X < ,, ' __ -z ce = < O S I X a C z < O U 2 N < X 0N 1k: 2 U O O N <.a O X <: :R S < N > Q O :.1 a a 1l z Y a ,> > < 2 a i _ w a o z! < b a I < 2 2 > 2 o 2 ro < < z C; 2 0 a, o N - X > - s., a > ,_7 . 2 < ., a . a X a > > 2 X < H 2 M H 2 H I I I H < = < 2 z - y < 5 s a a a < :D N J D J X a 3 :; 1 X 3 S 2 1 3 2    aa a U < Es < t z < < < H = ul < < :. < :U > o 314 v c o - E o o o s - .... O _ _ _ Q > z z zr y z T: C Z :R 14 bS m o C 0 z z z s s C C - y w z I I C C v v A.1 Q z _ v > S s s Y cs v   eq Z s s s o H < N v o _ Q b O Z Z Z _ C Ez > z w z y y v v C o C w W ,; o : S b o - w 9 O X H .S 11 zs s F v w s > H P w D o s z y 0 k7 0 + s s Y ; 8 . o o g v x o U c c ^ _ v n X _ ss 9 z > c c U c o U > O v v o > a v s a 8 3 3 s N :c H > -: a c V = 8 o o O Z ,, V V > V Z X S  S s £ C A: > _ o Z z C a o 3 > S I I I .v co w A _ oQ 7 N o o X c s v ai ) > < 8 < O v ; xD Z < q ,, D O .... ....... ..... .......... I o = S Ss 9 I : = X on K Ct X w C oa oe cE . < z w X n X m c1 o r o va a C s Oc C z ^ z z o J w o .. O  , d H o la _ X Y X > H w u: . _) () _ z # z C ec U n S C CS C F S Z + ^ ^ ^ :4 X Z C - _ _ - H C O Y Z X w H C X m o O O bn vq O O . z .^ ^ O J X S o = o + + + w S n Ko; N n X rs g S a X v o o M o w w ,, .x S v 3 s _ _ _ 8 X o z Z x w o 1 03 0 C C w .^ < C W Y z S C D: n R z cx m O a w c v n =.s Q 1 v  v > J o X U o O X - _ _ E S v 1l > F S X ZW O T z I C e < c ,- e e C CS e f W 14 v { < I Y S I w S V V V :: Y X O J =, E VS S L > 3 w <. > 14 S S S E > - e 9 w O o 315 e o o o > > > > >< :: r Z : * io:w>>N X X 7^X aNe eeeec P H > H F H e H N > >, F 4 X W >, >F >e O O O O O O H O o o o o o o S o > > > > > > r o ° ° ° ° l=z E Z = M o o o D Cs ,,S, ,£ y A, SL, Sx, Cl. :L, 1l, C. o o o o :L A N X X X Z ----- --- e c e e e e H H > H b b H _ _ _ _ .. 05 > > > :S > :S O O O O O O O H z = a X J X .1 a P = nH D:MCsf a °°°° ° °°9 zZ° °°°°° b g a e A X L X Di Di Dw Z1 _ _ _ _ _ _ _ _ e x Z o O o O O O O > a a s>J-Xums z z ci z_ _--__ _ > e N ; J a a a X J D O :_ u O O O O O O O C) P H 0: a Z X S O P es 0050000 n NU1£ J <: XCI U . 11 H 11  Z ;: ° °-° ° ° ° ° 2 D D e e e e X .) O . J X _ Ul ______ e0 n S Y H P W X X z H e e Z > a S > > > r = - > Cc E  > z ^>o oOZ > zO ^0> oz=Su zbe X u > 7 > > % X o z ^ H o = Z z > 3 = e o z s <: u: S U X b X X . b. _ >>> X se z° s nnSz ee eZ O> z E X ezo t N x w sone e ox > U > >w fz>s sa uz a - ^ X evow sz=s z > o zZOZz zz x =8rme >ffi°z>> O x z F > H = ld X] C O z E> o 2 H F H1 P p1 U] W 11 Z Z Z. + : _ sZ->> e> ZzZZ; rezEv° Q ^a e- zzaz szzz < zFe X uzzzz zz>s eozee omozx e U e z X ->0> z zeeee Je<, mx>> 0>3Ss = - o > - z s e ex- e m 8 > U 0 $ Neeee > > > X - u >>9> >>>>> evEz>> >eee >zJ HO>>z zzzzz H _ > .< u ossEE oxeee e>w eY=g >>> Y=U< EoFE8 88888 - e e F g YOOzz xoew ozoYozx 0_=_, =eve sezuc eeee  > S e x z = E t X s+Swwfi ma=Xs = ve== z X e Nxx>> ===> ;v>>>> O_>= =muxx -£e ecee + + e n _ > H a o . . w e e n e o o > S = a Ir o r O £ vD £ 40 xD .o 0 0 0 £ 0 £ £ O 0 YD £ o xD sC .D £ .D o £ o £ 0 ND .o o £ O SD .D 0 £ .D .o 0 40 £ £ xD 0 XD o 316 o o - - - - - 1 A X C) H S ' . ^ _ UX O 04 _ _ _ D H Eq O C _ 7 .,!: A 9 O : q a, t) =: ^ ^ H - S _ _ u H _ > _ S X - U a A s C ; = X H .w W = o 11 ^ H . o s :. ^ ^ - 1 S D ^ _ A .^ _ 12 4: H W ^ 14 O. 'D _ H CC C> ^ - S + ^ W S U U a b D 4 S | W W D ::. W Z > .^ S Z v ._ a. 11 9 - n s s _ _ E It H [< H - P _ t_, _ S .^ ^ ^ H ; 9 H H W H P S _ , - - ffi ' C a. a. ,4 :,, 11 S ^ ^ ,.n ^ _ S _ O z _ ^ O - - - _ ^ - C X S ^ I _ VM _ W w: O. X H > p A tz F1 _z 1£ _ 91 _ > H H O ^ H I O ^ > N H N _ _ A _ = < _ W D P - - - H O : o. H S ^ D ' Z _ ^ > ^ f  O O W _ u D D 11 _ S D D _ D _ D D . D W W ° DW D O W W .r W ,il W W _ W :t : ^': _ - _ _ H H H _ v H . _ _ 9 w . > z a. O U i ^ S 5 _ 1; _ _ _ _ _ _ _ F1 H O _ _ < _ > > s m s ; - - x 8 X U O D  S 5 = 0 5 S W aI S 14 g os 0N O o o :t 08 e o o o o o o o o o o  _ _ _ _ _ _ _  g 317 o o + o o ^ . ^ - _ _ Ev o) ^ * _, S D U 4  _ _ > , ^ 8 ^ ^ ^ * ^ ,= X A, _ _ _ _ = _ . _ ^ b _ ^ ^ P [., Pr7 e. o v , £, K Q - £ H x X _ O, _ _ . D -: S N X :h < .e S _  _ S _ _ _ + X X S 8 W X X D , > la _ Ff O o g_ ^ 2D 5D > O O > 1X. >. .. M vl A _ 4 < _ w t 9 w EW 19 X H H D D D D D X U O u D x b X D O D U :i 07 td - D D D _ Ce _ _ _ _ t.] z v S >, m X + S CE: > N Cd :Y 0: 4 , 1C1 n D D D O 1X1 >1 X s S ^ n rr + + 11 n + u S ^ . y X N X X X X 17 _ ;1: _ D u: b z D Y 0^ o1 2 < 1e X S X [; C4 X 2 > ffi N b oo * . . + . . 2 . . , , , * 5, O A _ A A A :E P n _ _ _ _ _ _ O D o < 4: .e <: wt < o 14 X W s > N r. r rv S N N s r ro r s r rs 2 318 o - oooooo'000 0 0 000 0 0 000 000A 00000 0 00 0000 0 0 0o 0 0 0000000 0 00 0000 0 0 0 000 000000000 0 00 0000 0 00000000AA000000000 A00 A A000A000 00 00000000000000000000 00000000000 0000000000 000000 A OAOOO 000000000000 0000 OAAAAOOAOO 000000 0 0 0 0 0 0 0 0 0 00000000000000000000 00000000000000000000000000000 0 0 0 0 0 0 00 0 0 0 0000 000000000000000000A0A0000000A00 00000 000A0000 00 0 0000000000 0000000000000 00000000000000000000000000 0 0 0000000000000000000000000000000000000000000000000 000000000 0 0 00000 0 0 0 0 0 0 0 0 0 0 0 0 00000 0 0 0 0 0000 00 0 0 0 0 0000 0 00 0 0 0 0 0 00 00000 00 0 0 0 0 0 0 0 0 <0 00000000 0 00 0 000000000F-000 0000 0 0 0000000 00000 o0000 000 0 0 0o-00000 o'00000 0000000000000 ooo-oo'00000 000 00 0 0000000000000 0000000000000 0000000000000 0 OO0OAA 00 000 0000 0' 0000000 00000000 0 319 o V) - ox U N CC O oo n U! _ s N > r D _ vo N a s a o S n e z n o a; ; , e o _ (,, o X xo X z a a X xo * z z S < 8 8 a o < _ e N N < o e v oX o a , 7 a _ _ o z > e x _ X - > N ur) o _ o U 2 a a o N sr z X ° 8 a - ° e. _ o a a n ; s o a n < o < U m Z ) sD eo en o S X N 14 S a a a r > so X cs < < o > N e = O s D U Q U U _ a a o; >7 \ c > U X W es avl>_xos o oro RD annnaEo z z mam xoo mazz z N C,, d, pz x a a a a b o s WD mero>rYz rs zoa joKDxo.oca Z>an a a a a _Mxamo nsvs 06Qz0O^Ntso-0N<-so0vVvvvvvvvv >X-0-SvN v zovononanv ezve > N N :R W x N = N X N N N a annca a a SO>NL mO\VV - t44X^¢¢z;;OUN ozso>ezovvobavazvennnnnnan ooNobam vavzasooooro nove C 9 + x C C ,,, X N N 0 X a a a cl. e 0 o v m oE ur _ _ N N N _ r) < < vr N or un _ 0 0 s IA O a o Ct e. o o  oy o < o o 0 _ N  _ _ _ _ v o - 5: v os _ < r _ v e v oe v K v N / e r r 0 o cz O o > O £ 1_ N rs rv r N r N n os o < < rn N N S q N N N s v v e ^ v r S n v fm v bI v 31 > _ v v Uh ur v v v v v r. N n X Y t 1 v 1_ n so rs o A ut rs > rs rs r um 1 N v v7 X N N N N CS N e N :S 0 O V oe _ e X X 0 N X on V e 0 X X 0 0 nOf X Z S Z Z Z Z os 0 0 RY {> g2: Y C: Z 0 M V M oS X 1: M C 0 X 0 X a: sr 0 0 Z 0 c > a a 8aaaaaaa a 0 0 <: X X : Lz w z=> b HXHX> aaDaaaaXzM00£Y7:nM > MXMYccccXIY:XX mO X X X <: <<<<<<<<< o 0 S H rd J :N X _n a 14 a X W 0 0 0 ,a  a o a > n D 0 S a >2 S: 0 0 0 0 0 ' 0 ' C v a0 0 O >  z n n M X J = nC v :1: z -;: = m: Z Z v w Z a a X > P. ci X :4 A :- X X X ffi 0 0 M 0 320 w <_sae- os_swon W V N rs W s :s * v * 9 _ W W w a a w _>-orw v N C o N < o _ N U1 v ; O N < O O X S IA < W ° a >0>4o W >sSNeavvanseva w 4;oF ¢_X^a_ea__wwoor¢XuQo wov>NssosN=nno>sna n cnevoree>semeaxv>>Xen0 V Q t_ b N I_ rv > W W n 14 Z aa w > W o o :d V < UT W V n N O ,,, O _ O 0 0s o N V N V V t_ _ b S ° ° o' O OY 6 C W W m 5 Lt _ P I_ 3 , ,_s W < W .e _ - > , e o woOevxo>_w-a-LQewo e e az _ Ha_ z N_ovo_nooNooc_oDwowOoO aw zz O So rt SvvNvoaNsabOvsQvvvw aw Cw C a ; wUw s O O whw ;9>Z ° wF w Q azQz ;s ar w w w wwwwz_ WHaS w B** > w°° cw a awp Fzz_° g > taz at SaPH^X __ SwHE_a oaoPsawon IwH- Hw mZ w otwawwHaa? W W W 1 fh 16 H = O > > :g X :/ N a < 321 os ~ ~ ~ ~ ~ ~ ~ ~ 00 o ~ ~ ~ ~ ~ ~ ~ ~ 0FFFF : ~ ~ ~ ~ ~ ~ ~ 0 00 0E o ~ ~ ~ ~ ~ ~ ~ F.0 - 0 0 r ~ ~ ~ ~ ~ ~ 0 00 0 0 - ~ ~ ~ ~ ~ ~ ~ ~ 0 0o0.0-o..-0 0 0 o ~ ~ ~ ~ ~ -F-- 0 0 0 ~~~~WWW0 0020 0 0F0 0 0 ' o 000 0 F- 0 WWo 00 0F-F-0 ,2, 0000IY00a ° P F---FFF 00 wXwwwwweS F-- ° 2 °0 100 0Z 0 0WoO < - - 2 W 0Z0Z00WF---- 000000 ,,4Z 0 F- 0, Z -FFFO . ..... F-... . 000 00 00... 0 0v 0°00 0° ,, o0 0000000o;o 00 . ~ ~~ O00 0 00 C00 OOW9H A : : F, -4 000 00 00 0 9 : e ww>~vvSv 0 OF-F- 04~a c00000:M2 0 20 zF4l > o 0- H 000 >ZHZ O0 D0 00 00 00 0 0 0 0 O 0 0 0000D 00 = 2r _ 02 e N 2 X 0.0, 00 0 0 0 00 00,4 U~> z ° 00W 008X < < 0 0, 000000000000292 0000,4000'22S°>> 0 00 ,4 000 0000=Z F- 0 0 00 0 0 0 00000000H^H ,4.0^a 2H0 2z U 000000n ns ia 00 00 00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0, 00 00 F-0 0 0 ¢ 0 0 0 0 0 oF.0 0 0 0 0 0 0 0 0 F--0° Wo 0 F- 00 0 00 0 0 00 0 00W0 0 0 00 00- 00 0 00 0 00 0 00 00 a0000o0000,4 w - OF 000F 000w=z 002 00000000F-000,400000NS OO O 0 r 00 0 0 0,40wX ~ YJ 2 2 rz A z 000 0, 00 ,4 00 ,4 00 , 00 . 000 0 e o x xrz W 00 000 0, 00 00 00 000 0 0 0003220 o ~ ~ ~ 0 0 0 0 0 0 0.0000 0 0000 0Z00 0 0 o ~ ~ ~ o zo 0 F o ~ ~ Z ~ '0 0 0 c ~ 0 ~ o 0 0 *t 0004000000Y A 0000o00000000p, 000 0, 0000 ooooO OOOO D0000000 00ZC0C OOO ^^ ^ 0 0000000000OO 0O 0 00000: 000000000000000000E_E- b: X X £ X lY ___323 Index Abu Zaabal Fertilizer Co., 163, 167-68, Atmospheric evaporation process, 30 169 Autothermal process of ammonia syn- Agglomeration granulation, 64 thesis, 22 Agricultural model link with fertilizer sector model, 144-45 Agricultural Prices Balancing Fund, Basic slag, 146 172-75, 178-79 Beneficiation, of phosphate rock, 42-43 Ammonia, 17-24, 39, 163-66, 243-45, Biuret, 30 266, 283, 285 Bones, calcium preparation from, 54 Ammoniated superphosphate, 67-68 Break-even analysis, 279-85 Ammonium chloride, 37 Brine, solar evaporation of, 58 Ammonium p htrate, 33-36 Bulk-blending, 14, 63-64, 71 Ammonium phosphate, 65-67: urea By-products: choice of, 77-78; sales of, (APU), 67 104-05, 109, 118, 132; use in other Ammonium sulfate, 31-33, 161-65, 168, sectors, 145-46 243, 244 Ammonium sulfate-nitrate, 36-37 Anhydrous ammonia, 17, 39 Application rates for fertilizer, 156-57 Calcium ammonium nitrate (CAN), 161- APU. See Ammonium phosphate urea 65, 175-76, 192, 207, 209, 224, 239, Aqua ammonia, 39 243-52, 268-69, 279, 282-85 Arcanite, 61-62 Calcium nitrate, 38, 68, 160-62, 245-50, Assiout, 168 254-59, 269, 276, 285-86 Aswan High Dam, 154, 158, 160-64 CAN. See Calcium ammonium nitrate 325 326 INDEX Capacity constraints: in complete Copper liquor, 23 model, 127; in dynamic model, 120- Cost analysis, 85-86 21; in dynamic Egyptian model, Cost including insurance and freight 217-19, 232; in existing industry (c.i.f.), 85 model, 94, 95, 98, 103, 108, 191-92, Costs: of capital, 117, 130, 193, 221-23, 200; in static capacity model, 113-14, 279-81; of conversion, 218, 222-23; 116 domestic input, 108, 117, 131; domes- Capacity expansion, 83-84, 109-18, 129, tic resource, 269-79; of fertilizers in 218-19, 232, 254-65 Egypt, 172-78; import, 102, 109, 118, Capacity planning models: dynamic, 132, 197-98, 201, 225, 234; in 1975 118-21; static, 109-18 model, 192-98, 201; of production, Capacity scaling, 81-83 92-93; recurrent, 193-95, 201, 223, Capacity utilization, 206-07 234. See also Transport costs Capital constraints, 137-38 Credit Bank, Egyptian, 172-75 Capital costs, 117, 130, 193, 221-23, CRF. See Capital recovery factor 279-81. See also Investment costs Cropping pattern in Egypt, 158-59 Capital recovery factor (cRF), 112-13,222 Crystallization: for ammonia nitrate Carbamate-recycle process, 28-29 production, 35; in prilled urea produc- Carbon dioxide, 146 tion, 30-31 Carbon monoxide, 22-23 Carnallite, 58-60 Chamber process, 46 Choksi, Armeane M., 202n, 235n DAP. See Diammonium phosphate c.i.f. (cost including insurance and Dasgupta, P., 7n freight), 85 Decisionmaking problems. See Planning Coal gasification, 22 problems Coke production, 164-65 Delivered cost, 3 Coke-oven gas, 283-84 Demand for fertilizers in Egypt: alterna- Compacting of fertilizers, 64-65 tive projections of, 284-85; applica- Complete investment model, 122-49: tion rates and, 156-57; composition alternative constraints on, 134-44; as of, 208-09; cropping pattern and, a dynamic capacity model, 118-21; 158-59; in dynamic model, 212-14, linkage with agricultural model, 144- 231; in 1974-75, 183-86; projected, 45; linkage with other sectors, 145-48; 157-60 objective function in, 143-44; size of, Demand growth, 79-80, 83-84 132-34; specification of, 123-32 Demand regions, 78-79 Complex fertilizers, 63-71 Design capacity, 188 Compound fertilizers, 63-71. See also Diammonium phosphate (DAP), 15, 65, Multinutrient fertilizers 66, 91, 168, 209 Concentrated fertilizers, 67 Dicalcium phosphate, 54-55 Constraints: in modeling, 107-08; Discount factor, 130 unique to a particular problem, Discount rates, in modeling, 119 134-43. See also Capacity constraints; Distances, and transport cost, 195-97, Material balance constraints; Non- 205, 225-26 negativity constraints Domestic input cost, 108, 117, 131 Consumption. See Demand Domestic labor constraints, 138-39 Contact process of sulfur production, Domestic recurrent cost, 223, 234 45-46 Domestic resource cost (DRc), 269-79 Conversion costs, 218, 222-23 Dorr Oliver process, 66 INDEX 327 DRC. See Domestic resource cost Existing industry model, 92-109: by- Dry-mixed fertilizers, 63-64, 71 product sales in, 104-05; constraints Dry-mixing, 14 in, 107-09; fertilizer recommenda- Duplex process, 56 tions in, 105-06; foreign trade in, Dynamic capacity model, 118-21. See 100-04; interplant shipments in, 99- also Complete investment model 100; mathematical structure of, 107- Dynamic Egyptian model, 211-34: de- 09; multi-product, 93-94; process mand in, 212-14; mathematical for- model in, 95-99; production costs in, mulation of, 227-34; objective func- 92-93; productive units in, 94-95; tion in, 220-27; results of, 235-87; product substitution in, 105-06; sym- structure of, 211-27; supplyin, 214-29 bols used in, 106-07 Dynamic Egyptian model results, 235- Exogenous variables. See Parameters 87: basic results (reference solution), Expansion of capacity. See Capacity 236-49; conclusions from, 285-87; expansion on domestic resource cost, 269-79; Export bounds, 104, 108, 117, 254-64 on export strategies, 254-64; on im- Export constraints, 128, 134-37, 219-20, port substitution policy, 264-66; on 233 production strategies at Suez, 249-54; Export markets, establishment of, 257 Scenarios I through VII, 249-69; Export modeling, 100-04, 247 sensitivity analysis of, 279-85; on Export revenue, 102, 109, 118, 132, 226, urea restriction, 266-69 234 Export strategies: conclusions on, 287; with bounded export growth, 257-64; Economies of scale: in fertilizer indus- with unbounded exports, 254-57 try, 4, 5; investment planning and, 72-73, 81; in static capacity plan- ning model, 109-13; transport costs Feddan, 156n and, 82-83 Feedstocks, choice among, 74, 85-86 Egyptian Chemical Industries, 163-64 Fertilizer sector planning: in Egypt, Egyptian fertilizer sector, 153-287: con- 153-80; in Egypt in 1975, 181-210; clusions of case study on, 285-87; de- formulation of investment program mand in, 154-60; domestic resource and, 80-87; introduction to, 3-8; cost in, 269-79; dynamic analysis of, models for, 88-149, 211-34; problems 211-34; dynamic analysis results on, in, 72-87; results of analysis of, 235- 235-87; exports by 247, 254-64; im- 87; specification and, 74-80 ports by, 178-80; in 1975, 181-210; Fertilizer Subsidizing Fund, 172-75, overview of, 153-80; prices in, 172-78; 178-79 production in, 160-68, 237-46; raw Fertilizers: application rate for, 156-57; materials for, 168-70; supply and, bulk-blended, 63-64, 71; classifica- 186-92, 214-19; transport in, 170-72 tion by plant nutrients in, 11-12; Egyptian General Petroleum Corpora- combination of, 208-09; compound, tion, 170 63-71; concentrated, 67; conclusions El Nasr companies, 160-68 on Egyptian production of, 285-87; End products, 74-75 delivered cost of, 3; description of, Ente Nazionale Idrocarburi (ENI), 170 12-15; Egyptian demand for, 154-60, Evaporation: solar, 58; for urea pro- 183-86, 208-09, 212-14, 231, 284-85; duction, 30 Egyptian exportation of, 247, 254-64; Exchange constraints, 139-40 Egyptian importation of, 178-80; Exchange rate, 269-78, 281-82 Egyptian prices for, 172-78; Egyptian 328 INDEX Fertilizers (continued) Hargreaves process, 62 production of, 160-70, 235-87; Egyp- Hassan, R., 158n tian supply of, 186-92; Egyptian Helwan, 164-65 transport of, 170-72; Egyptian use by High-analysis fertilizers, 12 type, 183-86; end products of, 74-75; Hignett, T. P., 70n grades of, 12; high-analysis 12; least- cost, 208; low-analysis, 12; multi- nutrient, 14-15, 63-71; nitrogenous, 13, 16-40; nutrients in, 10-12; phos- IDRC. See Incremental domestic resource phatic, 13-14, 41-56; potassic, 14, cost 57-62; recommended use of, 105-08, Import constraints, 140-42 117, 128; straight, 12; terminology Import cost, 102, 109, 118, 132, 197-98, used with, 10-15 201, 225, 234 Final products: constraints on, 126, Import substitution policy, 264-66 140-41; in dynamic Egyptian model, Imports: Egyptian, 178-80; modeling 215, 231; modeling results on produc- of, 100-04; prices of, 282-83 tion of, 237-46; in 1975 model, 187 Incremental domestic resource cost Fison's "Miniphos" process, 66 (IDRC), 270-77 Fixed nitrogen, 16-17 Indexes. See Sets and indexes Flotation, froth, 60 Input cost, 101, 108, 117, 131 Foreign exchange constraints, 139-40 Input-output coefficients, 189-90 Foreign exchange cost, 269-78, 281-82 Integrality constraint, 219, 232 Foreign trade, 100-04. See also Exports; Integrated-loop process for urea manu- Imports facture, 29 Frank, Charles Jr., 8 Intermediate products: constraints on, Frasch process, 44, 46 96-98, 103, 107, 116, 126, 142, 147; Froth flotation, 60 in dynamic Egyptian model, 215, 232; Fuel oil use in ammonia manufacture, 18 modeling results on, 237-46; in 1975 Furnace phosphoric acid, 47, 49 model, 187, 191, 200; shipment of, 99- 100, 204-06 Interplant shipments, 99-100, 206-07 Gas purification, 23-24 Investment constraints, 117, 137-38 Gestation lag, 79 Investment cost function, 222 Governorates, 185 Investment costs, 112-14, 222, 280-81. Grades of fertilizers, 12 See also Capital costs Graining for ammonium nitrate pro- Investment planning: capacity expan- duction, 35-36 sion and, 83-84; capacity scaling and, Granulation: agglomeration, 64; in 81-83; complete model for, 122-49; ammonium nitrate production, 36; of Egyptian fertilizer sector, 153-287; by compacting, 64-65; of compound feedstock choice and, 85-86; intro- fertilizers, 63-65, 71; melt, 64; of duction to, 3-8; models for, 88-149, multinutrient fertilizers, 14; in prilled 181-234; problems in, 72-87; process urea production, 31; by prilling, 64; selection in, 86; product choice in, processes for, 64-65; slurry, 64, 66; of 86-87; program formulation in, 80- triple superphosphate, 53-54 87; results of Egyptian analysis on, 235-87; site selection and, 80-81; specification and, 74-80; transport Haber-Bosch process, 19 problems and, 84-85 Harberger, A. C., 7n Ion exchange, 62 INDEX 329 Kafr El Zayaat, 166-67 Maximun capacity expansion con- Kendrick, David, 88n, llln, 133n, 144n straints, 218-19, 232 KIMA (Egyptian Chemical Industries), Meeraus, Alexander, 8, 88n, 133n, 202n, 163-64 235n Koppers-Totzek process, 22 Melt granulation, 64 Melt process of ammonium nitrate production, 35 Labor constraints, 98, 103, 108, 116, Methanation, 23 126, 138-39, 142 Methodology, 3-149: basic approach to, Labor cost, 193, 216-17, 232 5-7; of planning models, 88-149; Langbeinite, 58-61 previous work on, 7-8; on production Least-cost mix of fertilizers, 208 of fertilizers, 16-71; problems of Leunasalpeter, 36 sector planning and, 72-87 Linear approximation in modeling, 110- Micronutrients, 11-12 11 Mineral flotation, 60n Linear programming, and size of model, Mining: of phosphate rock, 42; of po- 132-34 tassium chloride, 58-59; shaft, 58-59; Little, I. M. D., 7n solution, 59 Low-analysis fertilizers, 12 "Miniphos," 66 Lurgi process, 22 Mirrlees, J. A., 7n Miscellaneous material inputs: con- straints on, 103, 108, 116, 127; defined, Macronutrients, 11 96; in dynamic Egyptian model, 215- "Make-buy" choice, 4 16, 232; in 1975 model, 191, 200; Manne, Alan S., 8 prices of, 194-95 Mannheim furnace process, 62 Mixed fertilizers. See Multinutrient fer- MAP. See Monoammonium phosphate tilizers Marglin, S., 7n Mixed-integer models, 153-54 Market location, 140 Models for investment planning: agri- Market requirements: in complete cultural, 144-45; alternative specifi- model, 128; in existing industry cation of constraints in, 134-43; by- model, 94, 95, 98, 104-08; in 1975 prodtuct sales in, 104-05; capacity model, 200; in static capacity model, constraints in, 120-21; complete, 122- 117 49; domestic labor constraints in, 138- Marketing centers: of Egypt, 185; 39; dynamic capacity, 118-21; dyna- selection of, 78-79 mic Egyptian, 211-34; economies of Markets, export, 257 scale in, 109-13; of Egyptian fertilizer Markowitz, H. M., 8 sector in 1975, 181-210; of an existing Material balance constraints: in com- industry, 92-109; export constraints plete model, 126-27; in dynamic in, 134-37; fertilizer recommenda- Egyptian model, 215-16, 231-32; in tions in, 105-06; foreign exchange existing industry model, 95, 97-98, constraints in, 139-40; import con- 102-03, 107-08; on fertilizers, 97, straints in, 140-42; imports and ex- 102, 107; on intermediate commodi- ports (foreign trade) in, 100-04; ties, 97-98, 103, 107; in linkages with interplant shipments in, 99-100; in- other models, 146-47; in 1975 model, vestment constraints in, 137-38; limi- 187, 191, 200; on raw materials and tations on, 67; linking with other labor, 98, 103, 108; in static capacity models, 144-45; linking with other model, 116 sectors, 145-48; mixed-integer, 153- 330 INDEX Models for investment planning (cont.) nitrate, 36-37; calcium nitrate, 38; 54; multiproduct, 93-94; objective description of, 13; Egyptian imports function in, 143-44; process, 95-99; of, 179; Egyptian prices for, 175-76; production costs in, 92-93; produc- Egyptian production of, 160-66; tive units in, 94-95; product substi- Egyptian self-sufficiency in, 285; tution in, 105-06; regional preference Egyptian use of, 155, 158-59; model- constraints in, 142-43; size of, 132-34; ing of demand for, 209; nitrogen specification of, 122-32; specification solutions, 38-39; nitric acid, 24-25; of constraints in, 134-43; static capa- production of, 16-40; sodium nitrate, city, 109-18; symbols used in, 106-07, 38; urea, 25-31; ureaform, 39-40 115-16, 124-25; time element in, Nitrophosphates, 68-70, 249-52 118-21; transport costs in, 89-92 Nonbinding constraints, 140-42 Monoammonium phosphate (MAP), 15, Nonnegativity constraints: in complete 65, 66, 168 model, 129, 148; in dynamic Egyp- Montansalpeter, 36 tian model, 220, 233; in existing in- Multinutrient fertilizers, 63-71: ammo- dustry model, 94, 95, 98, 104, 108; in iated superphosphate, 67-68; am- 1975 Egyptian model, 192, 200; in monium phosphate, 65-67; bulk- static capacity model, 114, 117 blending of, 71; granulation process Norsk Hydro, 66 for, 64-65, 71; nitrophosphates, 68- Nutrient ratio, 12 70; potassium nitrate, 70-71; pro- Nutrients, plant, 11-12 duction of, 14-15, 63-71 Multiproduct model, 93-94 Muriate of potash, 58 Objective function: in capacity model, Mutual exclusivity constraints, 218-19, 114; in complete model, 130, 143-44, 232 148; domestic resource cost and, 278; in dynamic Egyptian model, 220-27, 233; in dynamic model, 121; in 1975 Naphtha, 18, 21 Egyptian model, 192-98, 201; in sta- Natural gas: Egyptian reserves of, 170; tic capacity model, 108, 117 possible use in Egypt, 163; steam re- Odda process, 68 forming of, 21; use in ammonia man- Once-through process for urea manu- ufacture, 17-18, 284 facture, 27-28 Nili season, 158 Opportunity cost, 279-80 1975 Egyptian model, 181-210: objec- Other sectors, fertilizer sector's link tive function in, 192-98; results of, with, 145-48 201-10; structure of, 182-201; supply side in, 187-92; symbols used in, 198-201 Parameters in modeling, listed, 107, Nitric acid, 24-25, 239, 243, 245 115-116, 123, 125, 200, 230-32 Nitrochalk, 36 Partial oxidation of hydrocarbons, 21-22 Nitrogen: consumption in Egypt, 214; Partial-recycle process, 28 fixed, 16-17; liquid, 23-24; solutions, "Phos Al" process, 66 38-39 Phosphate rock, 41-43, 50, 169-70 Nitrogenous fertilizers: ammonia, 17- Phosphatic fertilizers: basic slag, 55-56; 24; ammonium chloride, 37; ammo- description of, 13-14; dicalcium phos- nium nitrate, 33-36; ammonium phate, 54-55; Egyptian imports of, sulfate, 31-33; ammonium sulfate- 179; Egyptian prices for, 175, 177; INDEX 331 Egyptian production of, 166-68; 175-78; for domestic fertilizers, 172- Egyptian use of, 155-56, 159; model- 74; for imported fertilizers, 175-78; ing of demand for, 209, phosphoric for miscellaneous inputs, 191, 194-95, acid for, 47-50; phosphate rock, 200; for raw materials, 194 41-43, 50, 169-70; production of, Prilled ammonium nitrate, 34-35 41-56; sulfuric acid for, 43-46; super- Prilled urea, 30-31 phosphate, 51-52, 244-48, 281-86; Prilling, 64 thermal, 55-56; triple superphosphate, Primary nutrients, 11-12 52-54, 237, 244-46, 252, 258, 264, 266, Problems in planning. See Planning 274, 276, 281-87 problems Phosphoric acid, 47-50, 168, 255, 258, Process choice, 75-77, 86 261, 266, 277, 281, 286 Process model, 95-99 Phosphorus, 41 Product choice, 74-75, 86-87 Planning horizon, 79 Product substitution in modeling, 105- Planning models. See Models for in- 06 vestment planning Production cost in modeling, 92-93 Planning problems, 72-87: by-products Production of fertilizers: conclusions of choice, 77-78; capacity expansion this study on, 285-87; dynamic Egyp- and, 83-84; capacity scale and, 81-83; tian model results on, 237-46; in demand regions and, 78-79; factors Egypt, 160-70; materials used in, in specifying the problem, 74-80; 11-15; modeling of, 95-99; of multi- feedstock choice, 74, 85-86; formula- nutrients, 63-71; nitrogenous, 16-40; tion of investment program, 80-87; phosphatic, 41-56; potassic, 56-62; overview of, 72-74; process choice, strategies for, 249-54 75-77, 86; product choice, 74-75, Productive units, 94-95 86-87; program formulation and, Programming: linear, 132-34; mixed 80-87; site selection, 77, 80-81; time integer, 153-54; 0-1 (zero to one) periods and, 79-80; transport alterna- variables in, 111-13, 133-34 tives, 78, 84-85 Project planning. See Specification in Planning process, introduction to, 3-8 project planning Plant nutrients, 11-12 Plants for fertilizer production: design capacity of, 188; investment in, 259, Raw materials: constraints on, 98, 103, 264; life of, 280; sites for, 77, 80-81, 108, 116, 127, 142: cost of, 193-95; 213, 219-20 in dynamic Egyptian model, 215-16, Polyhalite, 61 232; in Egypt, 168-70; in 1975 model, Potash, 57-62. See also Potassic fertili- 191, 200; plant location and, 224 zers Recommended use constraints, 100-06, Potassic fertilizers: description of, 14; 108, 117, 128 Egyptian imports of, 179-80; Egyp- Recrystallization refining method, 59-60 tian price for, 178; Egyptian use of, Recurrent costs, 193-95, 201, 223, 234 156, 160; potash ores, 57-58; potas- Reference solution, 236-49: exports, sium chloride, 58-60; potassium sul- 247; overview of, 237; production, fate, 61-62; production of, 57-62 237-46; transport, 248-49 Potassium chloride, 58-60 Regional preference constraints, 142-43 Potassium nitrate, 70-71 Resource cost, 269-79 Potassium sulfate, 61-62 Results of dynamic Egyptian model. See Prices in Egypt: for different fertilizers, Dynamic Egyptian model results 332 INDEX Revenue: from by-product sales, 104-05, plant sites, 77; of products, 74-75; of 109, 118, 132; from exports, 102, 109, time period, 79-80; of transport, 78 118, 132, 226, 234 Squire, Lyn, 7n Road distances, 225 ssp. See Single superphosphate "Run-of-pile" (Rop) process, 52 Static capacity model, 109-18: con- straints and sets in, 116-18; economies of scale in, 109-13; mathematical structure of, 116-18; structure of, Sauchelli, Vincent, lOn 113-14; symbols used in, 115-16 Scaling of capacity, 81-83. See also Steam reforming of natural gas, 21 Economies of scale Stoutjesdijk, Ardy, 8, 88n, llln, 133n, Scenario 1, 249-52 144n, 202n, 235n, 278n Scenario II, 252-54, 277, 279 Straight fertilizer, 12 Scenario III, 254-57, 276, 277 Suez plant, 161-63, 249-54 Scenario IV, 257-61, 274, 278 Sulfur, crude, 44 Scenario V, 261-64, 279 Sulfuric acid, 43-46, 163, 168, 244, 246, Scenario VI, 264-66, 277 258, 266 Scenario VII, 266-69, 277, 278 Superphosphates: ammoniated, 67-68; Scottish Agricultural Industries, 66 production in Egypt, 166-68; single Seasons in Egyptian agriculture, 158 or normal, 51-52, 244-48, 281-86; Secondary nutrients, 11-12 triple, 52-54, 244-46, 281-87 Sector planning problems. See Planning Superphosphoric acid, 47, 49-50 problems Supply: of fertilizers in Egypt, 214-19; Sen, A., 7n of fertilizers in Egypt in 1975, 186-92; Sensitivity analysis, 279-85 modeling of, 187-92 Sets and indexes, in modeling, 106, 115, Sylvinite, 58, 60 123-24, 198-99, 227-30 Symbols used in this study: for complete Shaft mining, 59 model, 123-25; for dynamic Egyptian Sheldrick, William T., 63n model, 227-31; for existing industry Shipment of products, 99-100, 206-07. model, 106-07; for 1975 Egyptian See also Transport model, 198-200; for static capacity Sinai oil fields, 163 model, 115-16 Single superphosphate (ssp), 51-52, 244- 48, 258, 268, 274, 276, 281-86 Site selection, 77, 80-81, 213, 219-20 Size of model, 132-34 Talkha, 165-66 Slag, 55-56, 146 Tennessee Valley Authority (TVA), 69, Slurry granulation, 64, 66 222 Sodium chloride, 59 Terminology of this study, 10-15 Sodium nitrate, 38 Thermal acid, 47 Solar evaporation of brines, 58 Thermal phosphate fertilizers, 55-56 Solution mining, 59 Thermo-urea process, 29 Solution and recrystallization refining Thomas slag, 146 method, 59-60 Time in modeling, 118-21 Specification in project planning, 74-80: Time period, planning, 79-80 of by-products, 77-78; demand Total-recycle process for urea manu- growth and, 79-80; of demand re- facture, 28 gions, 78-79; of feedstocks, 74; of Trace elements (micronutrients), 11-12 INDEX 333 Transport: alternatives in modeling, 78, tion of farmers in use of, 287; elimina- 84-85; conclusions on, 287; in Egypt, tion of restrictions on consumption of, 170-72; modeling of, 89-92; modeling 266-69; export of, 257, 263, 287; results on, 248-49. See also Transport prilled, 30-31; production of, 25-31, cost 259; restrictions on production of, Transport cost: in complete model, 131; 244-45 distances and, 195-97; in dynamic Ureaform, 39-40 Egyptian model, 223-26, 234; econo- mies of scale and, 82-83; in existing industry model, 109; interplant ship- Vacuum evaporation of urea, 30 ments and, 205; in modeling, 89-92, van der Tak, Herman, 7n 102; in 1975 model, 195-97, 201; Variables used in modeling: lists of, 107, overview of, 283-84; in static capa- 115, 124-25, 199-200, 230; 0-1 (zero city model, 118 to one), 111-13, 133-34 Triple superphosphate (TsP), 52-54, 168, Vietorisz, Thomas, 8 237, 244, 246, 252, 258, 264, 266, 274, 276,281-87 Tsp. See Triple superphosphate Washing of phosphate rock, 42-43 TVA. See Tennessee Valley Authority Westphal, Larry E., 8n, 278n Wet-process acid, 47, 49 Working capital, 225, 234 Upgrading costs, 218, 222-23 Urea: advantages of, 69; concentrated fertilizers from, 67; demand in dyna- 0-1 (zero to one) variables, 111-13, 133- mic Egyptian model, 214, 231; educa- 34 I II I comprehensive dynamic investment plan- ning model that can be used to draW up an investment program for the sector in its broad outlines, taking into account in- ternational trade and domestic transport, as well as alternative sectoral supply patterns. A case study describes the application of the methodology to the Egyptian fertilizer sector. Several of the models discussed in the first part of the book are used to address a number of typical problems encountered by sector planners in Egypt. All three authors are staff members of the World Bank. Armeane M. Choksi is an economist in the Country Programs De- partment of the East Asia and Pacific Regional Office, Alexander Meeraus is an economist in the Development Research Center, and Ardy J. Stoutjesdijk is senior adviser to the Development Economics Department. Alexander Meeraus and Ardy J. Stoutjesdijk are the editors of the series, THE PLANNING OF INVESTMENT PROGRAMS. Cover design by Carol Crosby Black -w The World Bank The Planning of Investment Programs edited l) Alexander Meeraus and Ardy J Stoutlesedl = ItOOKS IN I ilIS Nl W St RILS describe a systematic approach to investment analysis in the public and private sectors and Its application to specifhc industries and to ° multicountry investment Each volume is designed to provide a useful tool for = the practical planner as well as the student of development economics The series -- is entirely self-contained, requiring no prior knowledge of either mathematical co programming or the specific industry under consideration Ongoing research at the World Bank has studied the importance for investment analysis of interdependence among economic activities, with reference to par- ticular industrial sectors Two main conclusions emerged First, in the presence of significant economies of scale, interdependence is important enough to war- { rant explicit recognition in the investment analysis phase Second, computer 3 technology now permits a far more comprehensive and systematic analysis of 3 sectorwide investment problems than was ever possible before Project evaluation COD need no longer focus on the rate of return for only a single variant of an invest- ment program The use of mathematical programming and modern computers now enables the analyst to examine many variants of a given project or groups of co interdependent projects THE PLANNING OF INVESTMENT PROGRAMS makes this methodology available to a wider audience The introductory volume, Tlhe Planning af Industrial Investment Programs A Methodology by David A Kendrick and Ardy J Stoutjesdijk, out- - lined the properties of newly developed models and assessed their use in solving various kinds of problems Thib second volume deals with the formulation of = sectorwide investment programs in the fertilizer industry, and will be followed by similar treatments of the forestry and steel industries, a book on planning multi- CO country investment programs, and a user's guide to the methodology C- The Johns Hopkins University Press BALTIMORE AND LONDON e ISBN O-8018-2138-X __