Rvrptnrl Na. 1 21117-( HA China Investment Strategies for China's Coal CInd Electricity Delivery System Match i1, 1995 f r. iisl um wt ( )I ( r. itionis I )ivisin n ( 1ii.ni .I1f(I Mtv)fig(li.i I )D paln.1llf II I I .s1 i\si. l .ir(i i'.u iji R'gr(l di Cl OIli ( I (mioi ri Rt-tsii Ii ( enter SF I'L1. inning (01 11nhI%s 1(31n I p Re;ubI(l il (o China Document of the World Bank Currency Equivalents Currency Unit = Yuan (Y) = 100 fen Exchange Rate 1980 1990 1992 1993 1994 US$1 = Y 1.5 US5i = Y 4.7 US$1 = Y 5.5 US$1 = Y 5.8 US$1 = Y 8.7 Fiscal Year January - December Weights and Measures dwt = deadweight ton kV = kilovolt g gram kw = kilowatt km = kilometer MW = megawatt GW gigawatt (101 kw) (1 (r kw) std ton = standard ton kwh = kilowatt hour (5,500 kcal per kg) kcal = kilocalorie tkm = ton-kilometer kg = kilogram TWh = terawatt-hour (I 9 kwh) Abbreviations AC - Alternating Current CIECC - China International Engineering Consulting Corporation COSCO - China Ocean Shipping Company CTS - Coal Transport Study DC - direct current El - Economic Institute ERC - Economic Research Center ER[ - Energy Research Institute FYP - five-year plan GDP - gross domesti; product GIS - Geographic Information System GNP - gross national product ICIT - Institute for Comprehensive Transportation MOC - Ministry of Communications MOCL - Ministry of Coal MOEP - Ministry of Electric Power MOR - Ministry of Railways NIC - National Investment Company PCBC - People's Construction Bank of China SDBC - State Development Bank of China SPC - State Planning Commission TSP - total suspended particulates UNDP - United Nations Development Programme Five-Year Plans 5FYP = 1976-80 6FYP = 1981-85 7FYP = 1986-90 8FYP = 1991-95 9FYP = 1996-2000 IOFYP = 2001-2005 i China Investment Strategies for China's Coal and Electricity Delivery System iii China Investment Strategies for China's Coal and Electricity Delivery System TABLE OF CONTENTS Page No. CONTRIBUTIONS ...................................... ix EXECUTIVE SUMMARY .................................. xi Background ...................................... xi The Analysis Method ................................ xiii Main Conclusions and Recommendatiors by Sector .... ........ xiv Region-by-Region Summanry ........................... xix Past and Future Policy lmplications .... .................. xx 1. DEMAND AND SHORTAGES OF COAL AND ELECTRICy ... ... I Progress in Transport and Energy Development .............. I Coal and Electricity Shortages .......................... I Intensity of Energy and Transport Use .................... 2 Investment in Transport and Energy ...................... 2 Trends in Coal and Electricity Demand .................... 3 Coal and Electricity Demand Assumptions for this Study ... ..... 4 2. OVERVIEW OF THE COAL-ELECTrICITY SUPPLY CHAN: PROBLEMS AND OPPORTUNITES ..................... 7 Coal Production ................................... 7 Coal Benericiation .................................. 8 Coal Allocation and Pricing ............................ 8 Coal Transport .................................... 9 Coal Consumption and Conservation ...................... 11 Coal for Electricity ................................. 11 Environmental hnpact of Coal .......................... 13 3. PLANNING AND INVESTNENT FOR CHINA'S TRANSPORT AND ENERGY SECToRS ................................ 15 The Planing System ................................ 15 The Financing System ... ............................ 15 Problems in the Planning and Financing System .... .......... 16 The Changing Role of the State Planning Commission and its Analytical Needs .......................... 17 The CTS Network Optimization Model ............. ....... 18 iv Table of Contents 4. ANALYSIS OF INVEsTMENT STRATEGIES .................. 21 Overview .................2.... ....... .... 21 Prospects for Energy Shortages ......................... 22 Transporttion Sector Policy ImpUlations .................. 22 Nontransport Sector Policy Implications .................... 29 5. REGiONAL ANALYSIS ................................ 39 Northeast China ................................... 39 Central China ..................................... 39 Eastern China ..................................... 40 Southeastern Coastal China ............................ 40 Southwest Energy Base ............................... 40 Northwest China ............................ 41 6. PAST ANm FuTuRE POLICY LMPLICATIONS ................ 43 Policy Recommendations and Impacts of this Study .... ........ 43 Implications for the World Bank ........................ 45 Policy Issues to Be Addressed in the Future ................. 46 The Role of Intersectoral Modeling in the Market Economy ... .... 47 Action Plan for FIuture Use of the Coal Transport Study ... ...... 48 NOTES. .............................................. 48 ANWEE I Organization of the Coal Transport Study 2 Figures on Transport and Energy Annex 2 Flgures 2.1 Total Commercial Energy Flow in China 2.2 Investment in Transportation Relative to GNP and Traffic 2.3 Investment Structure of the Energy Sector, by Type of Energy 2.4 Coal Consumotion and Production Trends 2.5 Raw Coal Production, by Type of Administration, 1965-89 2.6 Maior Interregional Coal Flows, 1989-90 2.7 'Investment Structure of the Transportation Sector, by Mode 2.8 Investment Structure of the Electricity Sector 3 Background on Transport and Energy: Assorted Tables 3.1 China: Freight Traffic by Mode 3.2 International Trends in Commercial Energy Intensity 3.3 Internatonal Comparison of Reliance on Coal 3.4 Ilternz tonal Comparison of Transport Investment as a Percentage of GNP 3.5 Coal Supply and Demand Balance Sheet 3.6 Comparison of Coal Reserves-Soviet Union, United States, and China 3.7 Coal Production by Type of Coal, 1987 and 1991 3.8 Sulfur Content of Chinese Coal Reserves 3.9 Coal Output and Reserves, by Region, 1989 3.10 Interregional Coal Flows: 1980 and 1990 Table of Contents v 3.11 Steam Coal Washing 3.12 Plan vs. Market Prices for Steam Coal, 1989 3.13 E-volution of Coal Prices on the Free Market, 198690 3.14 Increases in Plan Coal Prices 3.15 Coal Traffic by Railway and Waterway 3.16 Indicators of Railway Asset Utilization: A Comparison between China, the Soviet Union, the United States, India, and Brazil 3.17 Combined Rail-Water Routes from Shanxi to East China 3.18 Share of Electricity in Total and Industrial Final Net Consumption of Energy in Selected Countries 3.19 Estimated Annual Emissions from Coal Use, by Sector 3.20 Ambient Concentrations in Major Chinese and Foreign Cities 3.21 Ambient Air Quality Standards 4 Planning and Investment for Coal, Transport, and Electricity Annex 4 Figures 4.1 Administrative Control System for Energy and Transportation 4.2 Planning System for Energy and Transportation S The CIS Analysis System Annex S Table 6 Figures and Maps on the CTS Analysis System Annex 6 Figures 6. 1 Generalized Network Diagram of the Coal-Electricity Delivery System in the Optimization Model 6.2 The CTS Coal Transportation Network with Mine Nodes 6.3 The CTS Coal Transportation Network with Demand Nodes 6.4 The CTS Electricity Transmission Network 6.5 CTS Data Flow 7 Description of 1993 Scenario Assumptions Annex 7 Tables 7.1 Comparison of Potential Capacity Increases: CTS Model versus Experts' Estimates 8 CTS Base Case (9 Percent GNP Growth): Calibration and Selected Results Annex 8 Tables 8.1 Comparison of Case 93-2 Outputs with Base Case Targets 8.2 National Totals of Coal Production 8.3 Regional Subtotals of Coal Production 8.4 Regional Breakdown of Coal Washing 8.5 Regional Coal Balance 86 Coal Transportation Q-D Table, in 2000 8.7 Electricity Transmission O-D Table 8.8 Power Plant Capacity 8.9 Ash and Sulfur Content of Delivered Coal, by Province 9 Sumnmary of Results for All 1993 Scenarios 9.1 Sunuary Solution Results 9.2 Coal Production Sector Results 9.3 Electricity Production Sector Solution Results 9.4 Electricity Transmission Solution Results 9.5 Transportation Sector Solution Results 9.6 Enviromnment Sector 10 Figures and Maps on Results of the Analysis vi Table of Contents Annex 10 Figures and Maps (Figures for Medium Demand are from Case 93-2; Figures for High Demand are from Case 93-9) 10.1 Coal Shortages, 2000, Medium Demand 10.2 Electricity Shortages, 2000, Medium Demand 10.3 Coal Shortages, 2000, High Demand 10.4 Electricity Shortages, 2000, High Demand 10.5 Rail Bottlenecks, 2000, Medium Demand 10.6 Rail Bottlenecks, 2000, Hign Demand 10.7 New Transport Projects Built 10.8 Port Bottlenecks, 2000, Medium Demand 10.9 Port Bottlenecks, 2000, High Demand 10.10 Rail Coal Flow Map, 2000, Medium Demand 10.11 Optimal Interregional Coal Flows, 2000, Medium Demand 10.12 Optimal Coal Allocation by Heat Content, from Coal Base in the Year 2000 (6 percent GNP growth, Case 92-1) 10.13 Steam Coal Washing by Province, 2000, Medium Demand 10.14 Electricity Flow Map, 2000, Medium Demand 10.15 Electricity Flow Map, 2000, High Demand 10.16 Sulfur in Delivered Coal by Province, 1995-2005, Medium Demand 10.17 Ash in Delivered Coal by Province, 1995-2005, Medium Demand 10.18 Multiobjective Tradeoff Curve for Cost and Pollution 11 Description of 1992 Scenario Assumptions 12 Summary of Results for All 1992 Scenarios Annex 12 Tables 12.1 Summary Solution Results 12.2 Coal Production Sector Results 12.3 Electricity Production Sector Solution Results 12.4 Electricity Transmission Solution Results 12.5 Environment Sector 13 Shortage Cost Data Assumptions Annex 13 Table 13.1 Market Price of Coal in 1989 14 Coal Sector Data Assumptions Annex 14 Table 14.1 Coal Production Costs in the CTS Model, by Region and Ownership 15 Transport Sector Data Assumptions Annex 15 Table 1U.1 Shipping Costs in the CTS Model 16 Electricity Sector Data Assumptions 17 Environment Sector Data Assumptions 18 Methodology for Estimating Benefits by Comparing Scenario Costs: Steam Coal Washing, Electricity Transmission, Shenmu-Huanghua Railway Completion 19 Chinese Evaluation of the CTS (December 1991) 20 List of Participants at CTS Report Meeting (October 1993) 21 Background on Edelman Award 22 CTS hnplementation Conference and Franz Edelman Award Ceremony (October 1994) Table of Contents vii .IAP. China: Provincial Coal Production and Consumption in 1992 (IBRD 26594) China: Provincial Electricity Production and Consumption in 1993 (IBRD 26596) China: Railway Network (IBRD 26857) China: Power Network (IBRD 26595) ix CONTRIBUTIONS This report is a joint effort of the The World Bank team members State Planning Commission's (SPC's) comprised Chan Juemin (RMC), Peter Cook Economic Research Center and the Bank's (CTS Training Director), Zhou Dadi (who study team following ten missions by Bank later joined the CTS Team), Terry Friesz, staff and foreign experts to Beijing and a Michael Kuby (CTS Technical Leader), nine-month training mission and two shorter Susan Neuman, Xianliang Wang, Thawat missions by the Chinese to Washington, Wetanatada (Task Manager), and Huikang D.C. (see Annex 1). It builds on the model Xu (RMC). development and testing in Phase I of the The report was written primarily by Coal Transport Study (1989-91), and Michael Kuby (CTS Technical Leader) and incorporates new material from Phase II Thawat Watanatada (Task Manager) and was (July-October 1993). Financing was received based on the results of the analysis. Other from the Japanese Policy and Human written contributions came from Zhou Dadi, Resources Development Fund and the Xie Zhijun, Cao Wei, and Sun Xufei. Also, United Nations Development Programme. Yves Albouy, Sadhan Chattopadhya, antd The Coal Transport Study (CTS) Katleen Stephenson of the World Bank team of the Economic Research Center contributed background materials on ERC) was organized into two groups: the electricity, coal production, and coal Modeling Group and the Policy Group. The utilization, respectively. Zafar Khan of the Modeling Group members included Shi World Bank made valuable suggestions Qingqi (Study Director), Sun Xufei relating to financing. A supplemental report (Modeling Leader, Phase I and II; Deputy describing the CaS Analysis System was Team Leader, Phase II), Zbang Chuntai written primarily by Zhang Chuntai and (Deputy Leader, Training Phase; Computing Michael Kuby, with zontributions from Leader, Phase 1), Zhou Dadi (Leader, Susan Neuman and Xie Zhijun. Rebecca Training Team), Cao Wei, and Xie Zhijun. Kary assisted with editing and formatting the Part-time members included Rong Qiang, report. Maps were made by Eric Khandagle, Wang Xuesheng, and Gao Shenhuai. The Denise Bergeron, Jodie Cabezas, Jose Policy Group consisted of Xu Zhen (Deputy Clavecillas, Jeff Lecksell, and Yung Koo of Director, ERC) and Liu Liru (Director, She Bank's Cartography Division, and by lCIl. Two special panels were set up to Barbara Trapido of Arizona State assist the MIodeling and Policy Groups. The University. Alice Moy and Tess Ortega were CaS Technical Panel consisted of Zhou invaluable dealing with administrative Fengqi (Director, ERI), Shi Qingqi, Zcu m&aen, and Bavani Krishnanrti helped to Yuan, Lin Fatang, and Yu Xiaodong (all of check and produce the final version of the El). The CTsS Consulting Panel was made report. up of Xu Zhen, Yang Zhenjia, Liang The Transport Operations Division Xiufeng, and Zhou Fengqi, and Liu Liru. Chief is Richard Scurfield, the Lead Other invaluable support was provided by Economist is Eliana Cardoso, and the Ga Ziyu and Gui Shiyong (Vice Chairmen Director is Nicholas Hope. Clell Harral was of SPC), Wei Liqun (General Secretary of instrumental in conceiving the study and the SPC), Huang Fanzhang (Vice Director getting the original financing. Over the last of ERC), Thou Cai Yu (lDirector, Economic several years, the study was supported by Institute), and Wu Youding (Mimistry of Bernard Montfort and Daud Ahmad as Railways, Chief Engineer, Railway Divisiorn Chiefs, Paul Stott as Acting Chief, Investment Study). Shahid Yusuf as Lead Economist, Shahid Javed Burki as Direrctr, and Zafer Ecevit as x Contributions Acting Director. Pieter Bottelier, as Chief of cover report was reviewed in March 1994 the Resident Mission in China, provided by Eliana Cardoso (EA2DR), Zafer Ecevit support to the various missions. (EA2DR), Phil Anderson CTWUTD), Lou In China, the CTS has been Thompson (TWUTD), Richard Scurfield, reviewed in three conferences. The first, in Hernan Levy, and Toshiro Tsutsumi December 1991, was chaired by Guo (EA2TP), Jeffrey Hammer (EAPVP), Vinod Hongtao, Chairman, Central Advisory Thomas (EAPVP), Robert Taylor (EAIE), Commission, along with Sun Shangqing, Shigeru Kataoka (EA2IE), Richard Vice President, Development Research Newfarmer (EA2IE), Robert Burns Center, State Council, and Gui Shiyong, (SAZIN), and C. Hugh Bannister (Intelligent Vice Chairman, State Planning Commission, Energy Systems Pty. Ltd., Australia). among others (see Annex 19). A second Comments from several reviewers were conference was held in October 1993 to adapted for use within the report. introduce the latest results, and was chaired Training of Chinese team members by Wei Liqun, General Secretary of SPC, was coordinated by Peter Cook. Participants and attended by Xu Zhen, Huang Fanzhang, included David Bernstein, Enrique Ling Zoo, and Meng Guang, Deputy Fernandez, Terry Friesz, David Hirshfeld, Directors of Economic Research Center, Tu Jerome Kreuser, Michael Kuby, Susan Zuming, Director of Energy Department of Neuman, Karen Polenske, Samuel Ratick, SPC, and Lan Shiliang, Director of the Edgar Sibley, Howard Simkowitz, and Scott (Long-term) Planning Departnent of SPC, Sitzer (U.S. Department of Energy). among others (see Annex 20). A third conference in October 1994 included an Awards for the Coal Transport Study award ceremony for the Franz Edelman Prize Finalists and a discussion of In addition to the Robert McNamara implementation issues for the recommended Fellowship won by a CTS team member, the strategies. It was attended b' Gui Shiyong, study has been recognized in other ways. It former Vice Chairnan of the SPC and has been awarded a Finalist's Award as one Deputy President of the Administrative of the six finalists in the 23rd Annual Franz Institute of China, She Jianming, Vice Edelmnan Competition in Management Chairman of the SPC, Wu Jingru, Director Science Achievement by the College on the of the Electricity Bureau of the State Practice of Management Science of the Development Bank, and Li Qun. Director of Institute for Management Science. This is the Policy Bureau of the State Development the highest international honor for Bank, awong others (see Annex 22). management science projects that have been In the Bank, the CTS inception implemented and that have produced report was reviewed in November, 1989 by verifiable benefits for a client organization C1ell Harral and Yves Albouy, both (the SPC). The study also wDn the 1993 formerly with the China and Mongolia Applied Geography Citation Awaid given by Country Department, and Terry Friesz the Association of American Geographers. (consultant). The draft CTS Phase I Final Details about the Edelman competition and Report was reviewed in July 1991 by past winners are contained in Annex 21. Kathleen Stephenson (AS3IE), Fernando This is the first time a project by China has Montes-Negret (AS3CO), Renato Schulz been selected as a finalist. It is a (ASTIN), Hernan Levy (EA2TP), and confirmation to the Bank and the Chinese Gavan McDonell (University of New South about the efficacy and appropriateness of the Wales, Australia). The white cover report large-scale modeling effort they have was reviewed in April 1993 by Hernan Levy undertaken. It also symbolizes greater (EA2TP), Robert Taylor (EA2IE), Shigeru openness on the part of China. Renewed Kataoka (EA2IE), Shahid Yusuf (EAZTP), impetus for continuing and expanding the and Robert Burns (SA2IN). The yellow study may be one result of the recognition. xi EXECUTIVE SUMMARY . "This report develops and and port bottlenecks. By rationalizing analyzes investment strategies and associated railway coal flows and by supplementing the policies for China's coal and electricity railway system with coal washing, long- delivery system, while also taking into distance electricity transmission, shipping, account the air pollution effects of the hydropower, coal imports, and energy different strategies. Financed primarily by conservation, the kind of debilitating energy the Japan Policy and Human Resources shortages experienced in the late 1980s can Development (PHRD) Fund, the Coal probably be avoided. However, if economic Transport Study (CTS) was a joint effort by growth continues in the double-digit range, the World Bank and the Economic Research severe shortages could result from Center (ERC-the policy research arm of insufficient long-distance energy-moving the State Planning Commission (SPC), capacity. Either way, the environmental China's long-term economic planning impact of the increased coal usage will be agency. This report is based mainly on extreme. The overall outlook has been Phase II of the CTS, conducted from July to improved by the Government's recent October 1993. actions to free most coal prices, raise ii. The study's original objective electricity prices and railway tariffs, focused on coal transport, but because of the decentralize many investment decisions, and interrelationships of investments in the coal, accelerate the construction of several transport, and electricity sectors, the important coal-hauling railways. However, objective gradually expanded to cover the to achieve the necessary volume and entire coal-electricity delivery system. The efficiency of energy deliveiy and utilization, study has produced two tangible fruits: an the Govermnent should continue to remove analysis system to assist the SPC with institutional barriers and subsidies, and decision-making for the coal-electricity develop and enforce market-based delivery system, and a set of policy analyses mechanisms for reducing air pollution. and recommendations proposed for the Chinese Government. Methodologically, the Background CTS developed a cost-minimizing, multisectoral, national-scale network model, iv. Shortages. China is the only similar to those used in western industrial country in the world to produce or consume countries, but well-grounded in China's real more than 1 billion tons of coal in a year, a production-transport-consumption situation milestone which they passed in 1990 and and the real options for its future expansion. which is expected to double by 2010. China The computer analysis is complemented by is also m.ore dependent on coal than any historical and policy analysis of: the causes other country, relying on it for 73 percent of of the problems; the Government and market commercial energy requirements. The forces acting on these sectors; and the policy upsurge of economiz activity since 1979 has changes needed to stimulate the put substantial pressure on China's energy recommended invesmnents. and transport systems. Since the mid-1980s, iii. The main conclusion of this China's economic growth has been study is that if economic growth during the periodically hampered by shortages of either rest of the 1990s continues in the forecast coal, electricity, or both, due to a shortage 8-9 percent range, delivery of enough coal of power-generating capacity and the and electricity should be possible to satisfy inability to transport enough coal from projected demands, despite prevalent rail where it is produced to where it is needed. xil Executive Summary V. Currently, coal shortages in GNP. The slow growth of electricity most areas have abated and are now production during the 1980s reflects estimated at onlY 3 percent of demand. suppressed rather than slacking demand Howevor. there remains a shortfall of peak because China's power sector was being electricity of about 20 percent. Also, coal, planned to satisfy 6 percent growth. As which accounts for 42 percent of rail freight growth surpassed the 6 percent level, the tonnage, continues to saturate the capacities power sector was not able to rapidly of most of th. major coal-arrying railways. readjust. On the other hand, the moderation Bottleneck links on the railway network, of coal demand growth for nonelectricity where traffic in at least one direction reaches uses is thought to be real because there is 95 percent of capacity or more, leapt from less evidence of shortages. This phenomenon 7 percent of the network in 1985 to 37 is thought to be caused by increased coal percent in 1989. As a result, rationing is prices, structural change in the economy, used to control access to congested railway technical progress, the production of higher corridors. Underlying the railway congestion quality (and therefore higher value) goods are underinvestment in transportation with little additional energy consumption, infrastructure and fast-growing transport increased road transport, and the gradual demand. replacement of wasteful residential coal use vi. Demand. Coal transport by more efficient heating and cooking demand is derived from the need to satisfy methods. demands for anthracite, coking, and ix. Theofficial GNP growth target industrial steam coal, as well as steam coal for public infrastructure planning purposes for electricity. Coal flows are the product of was 6 percent until 1992 when the forecast the tradeoffs between coal origins, coal was raised to 8 to 9 percent growth, more in types, transport modes, and other sources of line with past growth. In this study, future electric power (and their respective prices). GNP growth was forecast at 9 percent, b'zt Coal and electricity demands have been low (7.5 percent) and high (10.5 percent) rsing with total and per capita gross growth rates were also tested. The following national product (GNP), thougb not as fast. table at the top of the next page summarizes From 1980 to 1990 the GNP grew at an the assumptions of the low, medium, and annual rate of around 9 percent, and has high energy demand forecasts used in this since increased to around 13 percent. Coal study. production grew by 5.3 percent during the X. Geography. Coal was once a 1980s, while electricity production grew by widely available resource in China, having 7.6 percent, both proportionately less than traditionally been mined in over half the the growth rate of GNP. counties in China. However, in the major vii. During the 1980s, coal prices industrial provinces of East, Northeast, and were partially decontrolled, and electricity Southeast China, economically recoverable prices raised; coal prices are now market- reserves are rapidly being depleted. The determined and consumer electricity prices eastern half of the country must rely now provide for fill recovery of investment increasingly on supplies from surplus- and interest. From 1980 to 1992, coal prices producing areas in the energy base of North increased by a total of 122 percent, and real and Northwest China, centered in Shanxi electricity prices by 75 percent. Real freight and its neighboring provinces: Henan, railway tariffs have been raised by 91 Shaanxi, Ningxia, and eastern Nei Mongol. percen since 1990 and now approximate In all, this region contains 80 percent of the long-run marginal costs. nation's economically recoverable coal viii. Chinese experts expect that the reserves. The great distances between the low elasticity of coal production relative to locations of supply and demand add to the GNP will continue, but that electricity already great pressure on the transport demand will return to keeping pace with system. As a result, at the peak of the Executive Summary xiii Forecasted Demands for 2000 vs. Historical Data Forecast for 2000 1221 LQw Medium Hih Electricity demard (billion kwh) 678 1,258 1,444 1,661 Nonelectricity coal consumption 786 885 906 926 (million tons of standard coal) Approx. total coal production required 1,087 1,510 1,640 1,780 (million tons of standard coal-5,500 kcal per kg) 1988-89 shortages, market prices fbr coal in pollution benefits. Hence, in developing Shanghai were reportedly seven times higher long-term strategic plans, the Government than in Shanxi. Thus, the centerpiece of must evaluate these measures on an Chinas's strategy is to increase the already economic basis while taking into account large coal production and transport capacity timing and environmental factors. from Shanxi and the nearby provinces of Henan, Shaanxi, Ningxia, and Nei Mongol to the rest of the country. The Analysis Method xi. Opportunities in the Coal- Electricity Supply Chain. Although xii. To perform its coordinating additional railway capacity is one of the role in the overlapping transport, energy, most important ways to solve tansport and environmment sectors, the main analytical bottlenecks, other options can reduce the task facing the Chinese Government is to demand for coal transport, such as coal integrate the strategic guidance provided to washing, minemouth power plants, long- the different sectors. These three sectors distance transmission of electricity, and have not always been well coordinated and substitution of nuclear power and hydro- have suffered from underinvestment because power, especially from water-rich Southwest of (a) the rigid institutional decision-making China. These alternative investment framework; (b) the outdated methods and strategies can be considered the "demand tools used for policy analysis; and (c) the side" of the transport equation because they difficult transition to the new market-based reduce the number of tons that need to be economy. Government organizations and transported and/or the distance over which enterprises urgently need modern systematic the coal must be transported. For the energy tools suitable to China's real situation to equation, however, these strategies can be assist with policy analysis, decision-making, considered the supply side because they are and information supply. In Phiae I of the different ways of satisfying the same final CTS, completed in December, 1991, a energy demand. Most oI these options decision support system to assist the involve significant outlays of capital. Some Government in this multisectoral analysis options can be developed significandy faster was developed and tested. The mixed-integer than railways. The coal washing and programming model minimizes the tota hydropower options also have sigiificant air discounted cost of delivering coal and xiv Executive Summary electricity subject to demand and capacity is already breathtaking-China is likely to constraints and optional budget and experience more severe railway congestion, environmental constraints. At the heart of with combined shortages of coal and the model are multimodal transport and electricity exceeding 125 million tons per electricity transmission networks. Forecasts year. Even a policy of unlimited coal of demands for electricity and nonelectricity imports to coastal regions would still leave coal for the years 1995, 2000, and 2005 are a significant amount of coal and electricity inputs to the model, along with such inputs demand unsatisfied in the interior regions. as costs, capacities, new project options, Assuming greater than 10 percent growth, technological factors, and discount rates. energy shortages that would occur above and The main outputs are how much of which beyond a 9 percent growth situation are investments to build where and when; estimated at US$7 billion for the year 2000 optimal distribution patterns for coal and alone. This shortfall, expressed in 1993 electricity; and performance measures such prices at official exchange rates, is about as costs, pollution levels, shortages, three-fourths of 1 percent of the projected transport bottlenecks, and shadow prices. Chinese GNP. (Economic losses noted here The analysis system is designed to are valued in terms of replacement by coal complement, not replace, other analytical imports.) tools used for more refined economic or xiv. Railways. Rail flows are logistical analysis in the individual secors. predominantly from mines in the west to cities and ports in the east and northeast, Main Conc!usions and and from north to central. In any Recommnendations by Sector conceivable future scenario, all existing and planned railway capacity for coal being xiii. Impact of GCrowth. If the shipped out of the energy base would be economy grows at up to 9 percent per year, utilized. If growth exceeds 10 percent, more the analysis results suggest that it should be investment must be channeled to the possible, within existing constraints, to railways. The study results were used as a satisfy nearly all the coal and electricity basis to support the Govemrnment's recent demands by 2000. This can be achieved decision to accelerate the construction of the without creating major shortages and without second west-east line out of the energy base importing coal except in a few problem from Shenmu to the new port being built at regions, assuming that additional energy Huanghua. delivery strategies are carried out in xv. China in recent years has anticipation of that growth. However, if concentraed their railway investment more growth exceeds 9 percent by one or two on adding incremental capacity to existing percentage points, the planned railway lines than on building new lines, and as a network would likely be overwhelmed, result has one of the highest densities of despite railway services being priced traffic per Ikm of track in the world. As approxim:iately at long-run marginal costs. planned, China should continue expanding GNP growth significantly higher than the capacity of existing lines by 9 percent is not unrealistic given that during multiple-tracking, diesel traction, 1991-93 the economy grew at about 11 electrification, additional sidings and percent per year. In that case, even conversion to heavier axle loads. Such accelerated construction of substitute improvements generally offer lower measures, such as long-distance transmission investment cost per unit of capacity and and coal washing, would likely not be able relatively faster completion than new to alleviate the pressure on the railways. construction. This manner of capacity Unless extraordinary measures are taken to expansion, however, is not sufficient to accelerate railway planning, financing, and provide the needed additional throughput construction beyond the urrent pace-which capacity to rlve the problem of energy base Executive Summary xv bottlenecks: brand new lines are desperately efficiently. Since rail tarifts have already needed. Of course, these new lines should approached long-run marginal costs, also take advantage of the low unit cost congestion pricing on bottleneck links should offered by unit train and heavy axle load be investigated for bringing in additional technologies. revenues to supply new capital investment, xvi. The study made two specific as well as for rationalizing demand for recommendations about new lines in addition railway transport. The efficiency of coal to the Shenmu-Huanghua ]ine, both of which distribution can be improved by shipping have since been acted on by the Chinese coal with higher calorific value over greater Government: a new line from Shenmu to distances. The estimated economic savings Xian (now under construction), and a third from Shanxi would be about US$25 million west-east line from Shexian to Handan to annua1ly compared with allocating coal Jinan (now planned). Heavy haul without regard to calorific value. Congestion technology, which the MOR is implementing pricing should also have the effect of to reduce the unit capital cost of capacity encouraging shippers to differentiate expansion, is an important part of the between types of coal. Congestion pricing railway strategy. It is not expected to have should also encourage users to consume an effect, however, until after 2000, except more coal locally without using the railway on the Datong-Qinhuangdao and Shenmu system, which would contribute to a more Huanghua mine-to-port lines. Because efficient coal distribution pattern. However, railway projects take a long time to plan and the inflationary impact must also be construct, many provincial governments are considered in investigating the use of tking matiers int their own hands by congestion pricing. Congestion pricing is not rapidly developing local coal and electricity uncommon internationally: railways in the projects to satisfy their immediate demand United States typically charge higher rates in for electricity. the congested Northeast corridor. xvii. Systemwide coordination is xix. The second area of policy another key to MOR's investment strategy. reform should be to reduce the The MOR must consider interdependencies administrative allocation of railway capacity. between different links and modes in Currently, 60 percent of capacity is allocated planning the expansion of its network. For by planning instead of by market bidding. A instance, bottlenecks leading out of the case in point is that reallocation of railway energy base may possibly cause capacity from passengers to coal appears to underutilization of the new port capacity have only limited potential to solve logistical being built at Huanghua, or on the existing problems without creating substantial Beijing-Qinhuangdao line. Similarly, unsatisfied passenger demands. The third bottlenecks on railways and ports en route to policy issue is service. The MOR is moving northeast China may cause underutilization toward offering not just "raw capacity" but of railway projects in the interior of the "quality capacity" for container transport northeast. and dedicated passenger transport. Coal xviii. The efficiency of the railway shippers may also be willing to pay more for system depends not only on decisions made fast and reliable service, which could by MOR, but on the responses by railway perhaps be provided by auctioning off users. Five policy reforms that would segments of capacity to long-distance, promote market efficiency are (1) higher railbased transport services companies that prices; (2) less allocation of capacity; (3) could offer rapid unit train services. This more services offered; (4) more competition; measure would also introduce more and (5) belter and more available competition, the fourth area for policy information. First and foremeost, the MOR reform. Price and entry deregulation in the must set prices high enough that shippers rail-water transport market would also foster will have the incentive to use railways competition. Fifth, more and better xvi Executive Summary information is needed at the annual coal harbors and to develop a Panamax-class or ordering conference for coal buyers to sLupercollier fleet must take into account the consider rail and water prices and services, physical conditions at each port. While as well as coal quality. Chinese 9-meter ships are small compared to xx. Waterways. An important world standards, they are competitive with strategy for relieving pressure on the railway railways over long distances, and, in some system is to shift more coal flows onto the cases, they are the only possible direct coastal and inland waterways. Setting up supply route to some oceanside or riverside high-efficiency rail-port transport corridors factories and power plants that are not out of the coal base would create a powerfil connected to the railway system. incentive for major power and steel plants to xxiii. The recent removal of all price locate along the eastern and southern sea subsidies to railways for purchasing lanes. In the medium demand case for 2000, materials and energy supplies, and the nearly 140 million tons of coal (not potential use of market mechanisms (for including exports) would be shipped an example, congestion pricing) instead of average of almost 1,500 km along the coast, rationing to allocate scarce railway and port wbich represents a doubling of the tonnage, capacity, should help rationalize shipping 20 percent longer distance, and more than patterns. Because of the number of add-on doubling the fleet size compared with 1990. charges involved in securing port and ship The inland waterways also play a key role, access, ports are effectively charging carrying 22 million tons shipped an average congestion pricing already. The of 660 knm, which requires a 75 percent recommendee central Government role in increase in the barge fleet by 2000. the waterway sector is to work with the xxi. Constraints, however, will local port authorities to evaluate, and, if continue to limit the scope for the waterway economically justified, to provide deeper option, particularly railway bottlenecks en harbors for multicommodity ports, to route to the loading ports along the northern remove distortions from the pricing system, coast and not enough port capacity, and to decentralize more of the shipping especially at receiving ports. Shanghai will industry. Entry deregulation would need at least 25 million tons of new encourage price competition, investment in unloading capacity, while numerous other capacity, and improvements in vessel ports south of Shanghai and on the Yangtze technologies, information systems, and River will need up to 10 million tons of new service in general. Waterways are not likely capacity. Because of the shortaga of loading to fulfill their fill potential as a provider of and unloading capacity, coal is at present coal transport unless the issues related to sometimes transshipped at general cargo tariff structure, harbor depths, average ship ports lacking the specialized bulk equipment size, and limited mine-to-port railway necessary for efficient handling. capacity are addressed. xxii. Most harbors at receiving ports xxiv. Coal Production. The analysis are only deep enough to accommodate 9- results confirm the continuation of the meter draught, 20,000 dwt vessels or 9- current trend to shift more coal production meter shallow-draught 35,000 dwt vessels. to the Noih China energy base (an increase Of the main ports on the southern coast, from 43 percent of the nation's total in 1989 only Ningbo can accommodate 12-meter to around 52 percent in 2000). The results 50,000 dwt ships. Xiamen is now being also suggest that coastal regions maintain deepened to 12-meters (in a Bank-financed their absolute levels of coal production. This project), while the harbors of Wenzhou and westward shift represen-ts a sayings of more Fuzhou are only now being deepened to the than US$1 billion in coal mine investment minimal depth of 9 meters. None can match costs through 2000 relative to the alternative the 14-meter depth of Qinhuangdao, China's of mainuining the existing shares of coal major loading port. Options to deepen production across regions, which oDuld Executive Summary xvii happen if railway capacity from the energy environmental measures and enforcement base is not expanded adequately. Imports have historically been weak. Third, the appear to be economically feasible, but are proper legal framework is necessary for logistically and financially feasible mainly in willing buyers and sellers to enter into coastal cities. These regional shifts and longterm contracts that will guarantee a import increases occur during a time when reliable market for the washery's products subsidies to producers and consumers are and a reliable supply for boilers designed to being phased out, coal mines are burn washed coal. Fourth, as long as the experiencing competitive price pressures on macroeconom.y remains overheated, users their products and inflationary pressures on cannot afford to be as choosy about the their inputs, mines are generally losing quality of their coal. money, and millions of mine workers may xxvii. Power Generation Because become unemployed. there are only a limited number of xxv. Coal Washing. The results inexpensive hydropower sites near load suggest that steam coal washing should be centers, hydropower's share of total increased from the current level of 7 percent capacity, currently 24 percent, may even fall to roughly 16 to 19 percent, and to nearly slightly, though in certain regions it can double that if ash and sulfur reduction at end economically generate the majority of the use points is a policy goal. This appears to electrici-y. Thermal power is likely to be be a robust finding because steam coal China's least-cost alternative for generating washing never dipped below that level in most of the electricity in most regions. any scenario analyzed, including those in Because of the many different services which washing costs were raised. The provided by a dam, the above conclusions beniefits of steam coal washing include reflect only the competition between thermal reducing transport costs for long-distance and hydropower on an energy cost basis. flows, lessening local bottleneck effects, xxviii. Transmission. Electricity achieving environmental goals, and reducing transmission from the regions that have a ash disposal and boiler maintenance costs. surplus of new coal or hydropower capacity The net present value (NPV) of economic should be increased North China has the savings in transport, ash disposal, and boiler opportunity to economically increase its maintenance costs (net of washing and share of electricity generation from less than incremental mining costs) is estimated at 18 percent in 1989 to nearly 20 percent by US$3.8 billion over 15 years (in 1993 2000 prior to completion of the Three prices). Based partly on the strength of these Gorges power plant. Nationally, three major results, the Govermnent recently announced sets of flows stand out consistently in a guidelines asking large, state-owned mines variety of scenarios: (a) from the coal base which export coal to other provinces to wash to the Beijing-Tianjin-Tangshan area, East their steam coal before shipping. China, Central China, and even South xxvi. The success of this policy will China; (b) from eastern Inner Mongolia to depend on four main factors. First, market- the Northeast; and (c) from the hydropower detennined prices for coal and rational base in Guangxi and Guizhou provinces to prices for railway transport services will Guangdong. Construction of six 500 kV make the benefits of steam coal washing long-distance intergrid transmission lines is more transparent and spread more equitably estimated to save China roughly US$1.5 among the parties in the supply-transport- billion in combined transport and energy user chain. Second, strong enforcement of operating costs through 2000, compared environmental regulations on air pollution with the alternative strategies that would has been the key to establishing demand for have to be pursued otherwise. This study's washed coal, as demonstrated in results have been used to support China's Organisation for Economic Co-operation and recent decision to move forward with several Development (OECD) countries. In China, transmission projects. Skyrocketing power xviii Executive Summary demand, railway bottlenecks, and shorter total suspended particulates. To achieve construction lead times are helping to progress on the environmental side, the overcome years of disinterest in transmission Chinese Government will have to structure stemming from institutional, technological, its environmental policies to force firms to financial, and economic barriers. internalize the social costs of pollution while xxix. In recent years, several giving them as much latitude as possible on provinces were starting to move forward the means of response. Pollution taxes may with transmission projects with other be the most economically effective measure, provinces in a decentralized fashion. To help followed by tradeable pollution permits. In ensure smooth coordination of the regional the US, tradeable emissions allowances for and provincial grids, a regulation was passed sulfur introduced in the 1990s were priced in early 1994 that all transmission lines are by the market four times lower than the to be managed by the central Government, government's conservative estimate, saving while the market will control the power some $10 billion per year. Vigorous plants. By 2000, China will tie the regional enforcement of either measure is probably grids together into one national grid. Long- more important than which is ultimately term contracting to ensure reliable supply, adopted. demand, and financing between the local, XmXii. The level of taxation or the provincial, and central Government parties amount of credits should be high enough to involved is the key to this kind of force polluters to pay for the environmental interregional arrangement. damage they cause, yet the level must also XXX. Environmental Tradeoffs. consider what is economically feasible. As a Although evaluating the cost-effectiveness of partial answer to the question of feasibility, environmental protection measures was not the analysis found that reductions of 10 the primary purpose of this study, it was percent in the ash and sulfur content of the important to capture the fact that some coal to each province can be achieved for strategies for coal and electricity delivery only 3 percent more in cost by using better are less environmentally harmful than coal, washing more coal, and substituting others. For instance, using higher quality more hydropower for thermal power. coal, wasiiing more coal, and substituting Further reductions of 20 to 30 percent hydropower for thermal power in many (which would be needed to keep absolute cases can pay for themselves in terms of ash and sulfur levels from worsening) would reduced costs and reduced pressure on the be more expensive, costing 10 to 20 percent railway system in addition to reducing more. This increase would be a result of the pollution. In fact, increasing steam coal mounting costs for scrubbers, hydropower, washing up to a threshold of nearly and the need for energy conservation. 20 percent is a win-win proposition in terms Because numerous pre- and post-combustion of reducing both cost and pollution. For mitigation measures were not included in the these reasons, the enviromnental benefits analysis, these cost figures can be thought of were measured in terms of how much they as conservative upper bounds, achieved by reduce the ash and sulfur content of choosing from only a partial menu of delivered coal. technological strategies- The real costs xxxi. The analysis shows that the should turn out to be lower. absolute amount of ash and sulfur under 9 mauiii. Energy Conservation. Energy percent GNP growth would be about one- conservation has great potential to third higher than in 1990, despite the simultaneously reduce energy shortages, recommended increase in washing steam relieve pressure on railways, and protect the coal. This is a critically important finding eironment. Conservation investments were for a country that already bears the dubious not one of the options included in the distinction of some of the world's worst analysis system used for this study. A post urban air pollution from sulfur dioxide and hoc analysis, however, suggests some Executive Summary xix tentative conclusions. Energy shortages in washing coal shipped in from the energy the model are demands that either could not base; increased mining of high-cost local be satisfied for less than the cost of imported coal; and construction of, and subsequent coal (Y 300 per ton in 1990, equivalent to transmission of electricity from, minemouth US$47 in 1993 prices) or could not be power plants on the eastern Inner Mongolia satisfied at any cost for logistical reasons. brown coal reserves. Imports are not a Given that the amortized cost of many practical solution with the exception of the energy conservation investments is less than costal port of Dalian. Y 300 per ton per year, the results suggest xxxv. Central China. Central that energy conservation investments could China-especially around the cities of reduce coal demand by as much as 100 Wuhan, Changsha, and Nanchang-is the million tons given double-digit GNP growth, next most problematic area. Its coal deficit and up to 30 million tons given growth of and highly-congested railway network are up to 9 percent. To realize these goals, somewhat offset by its central location, energy efficiency improvement must which gives it many supply options. Some increase by at least 7 percent per year, possible supplementary strategies include compared with the present rate of (a) constructing a new railway south from 3.6 percent per year. These conclusions Shennu toward Wuhan or other railway were confirmed strongly by some expansion projects; (b) producing more coal preliminary analysis using an enhanced locally; (c) producing more local model (developed in the McNamara hydropower until the Three Gorges project Fellowship program) that incorporates data can be finished; (d) redistributing coal by on energy conservation options and their inland waterway; (e) transmitting electricity investment costs, efficiency improvements, from surrounding regions, especially from and potential savings. . The continuation of minemouth thermal plants in southeast reforms that expose energy users to market Shanxi via the Henan grid; and (t) discipline, that free up coal and electricity intensifying energy conservation efforts. All prices, and that open China to technology of these options could be made more imports should provide ample incentives for attractive with congestion pricing for conservation. Congestion pricing of rail bottlenecked railway links. transport and ports, along with enforced xxxvi. South and East C4ina. The pollution taxation, would further bolster southern and eastern coastal regions must these incentives. rely mainly on medium-quality local coal and on coastal shipments of high-quality coal Region-by-Region Sumunary from the energy base. Imports of high- quality coal and investments m energy xxmiv. Northeast China. This study conservation would help keep shortages and identifies problem regions where fast sulfur content under control in most eastern economic growth or other conditions could and southern cities. Also, because of the exacerbate shortages. It recommends further distance involved and the bottlenecks that investigation into energy supply options for intervene, these two regions are best each region. Principal among these is the positioned to benefit economically from Northeast, where either high demand or purchasing washed steam coal. The reduced allocation of transport capacity environmental benefits should be considered could lead to unsatisfied demand. The as an additional advantage Northeast must supplement its own supply xmxvii. However, in eastern China, with coal from the energy base, but is possible coal supply problems face cities, obstructed by two sets of bottlenecks: from such as in Anhui and the interior parts of the energy base to the Beijing area, and then Shandong, that are without the coastal or past the Great Wall to Northeast China. inland waterway alternative. Construction of TIreelargelydecentralized strategies include a railway from Handan to Jinan would be xx Executive Summary one way to address Shandong's problem. A mostly self-supporting while continuing to strategy which has just been approved, on a supply some surplus coal to northern test commercial-scale basis, is a new coal Sichuan and central China. This region can slurry pipeline from Shanxi to Shandong. also export thermal electricity to the centra East China is particularly vulnerable to China grid and hydropower to the North electricity shortages in high-demand China grid. conditions. The electricity balance of the region can be helped tremendously by Past and Future Poflcy Implica.ons building transnmission capacity from mine- mouth plants in southern Shanxi to cities in xli. Past Policy Impacts of the Shandong and Jiangsu. Study. The CTS was reviewed at a high- xxxviii. Southeastern China does not level conference for Government officials in have the inland waterway option that eastern December 1991. Following that meeting and China has, but its industrial development has the upgrading of the GNP forecast in June historically been more confined to the of 1992, Government planning agencies coastal areas. Guangdong's electricity requested additional analyses of higher balance might be supplemented by growth rates. The results of these analyses transmission of thermal power from as far provided a basis for the National People's away as southern Shanxi, or from Congress to adopt the 8 to 9 percent annual hydropower plants in Guangxi. Other GNP growth rate for public infrastructre options for Guangdong's electricity situation planning,rather than 9 to 10 percent. might be new power plants using imported Nonetheless this represents a significant oil or coal, or new nuclear plants if they can increase from the 6 percent or so growth be built economically. rate adopted during the l980s. Subsequently, mix. Southwest C7hina. In the Phase II of the study identified strategies and Southwest, isolation contributes to its lack of projects that might make it feasible to satisfy flexibility. The Southwest is largely self- growth in the 8 or 9 percent range. These reliant in coal and electricity, and can expect latest CITS results were reviewed at another coal shortagea in the neighborhood of 10 high-level Chinese conference in October percent in conditions of rapid growth. For 1993. In April 1994, the model was used to coal, its ample reserves are generally high in analyze three different demand cases for the sulfur. For electricity, the massive Three 9FYP, the results and conclusions of which Gorges project is now under construction have been used for planning purposes. Most here. The Southwest may continue to rely on recendy, a conference was held in October its vast hydropower potential, but as more 1994 to discuss the reforms that would and more projects are developed, capital encourage the inplementation of the CTM costs become a concern. A related option is recommendations. to keep more of the power generated at xlii. S o m e o f t h e s e Southwestern hydropower plants within the recommendations are consistent with long- region instead of transmitting it to Central standing Government strategy of the last 5 to and Southeast China, or for Sichuan to 10 years. Other recommendations have receive power transmitted from Shaamxi to supported recent policy shifts of the last two the north. In the medium term, prior to to three years. Still other recommendations oDmpletion of the Three Gorges power are strategies upon which the Government project, authorities may deal with the lack of either has not acted, is still considering, or flexibility by planning and building more is going more slowly. By sbowing the dire railways and by promoting energy consequences of 9 percent GNP growth on conservation. infrastructure planned for a target of xl. Norhwesr Cina. Coal and 6percent growth, and by recommending hydropower resources are abundant in this strategies for overcoming them, the CTS area. In the future, this region can remain provided a basis for the recent Government Executive Summary xxi decisions to, among others: (a) accelerate rationalize Iheir energy and transport railway construction on the Shenmu- decisions. Even after the transition phase, Huanghu, Shenmu-Xian, and Handan-Jinan the centrally-guided infrastructure lines; (b) urge steam coal washing for state- investments, such as railways and electricity owned plants; and (c) open the door for grid and, to a lesser extent, ports and dams, intergrid electricity transmission. Pursuing will still need to be coordinated with each these three strategies alone could save China other and with developments in the as much as US$7.1 billion (in 1993 U.S. decentralized sector. This role of indicative do'lars, discounted at 12 percent) over the planning is similar to the kind of issues next 15 years, as compared with the more analyzed by comprehensive network-type expensive strategies that would be necessary models in OECD countries. and greater shortages that would likely result dliv. In the near fiture, it is most if such strategies were not considered. important for the Government to analyze xiiii. Policy kssues to Be Addressed additional railway and waterway measures, in the Future. China is in the midst of an especially heavy haul rail technology, port impressive transition from a centrally- improvemens, transport congestion pricing, controlled economy to a dynamic, market- and power transmission. The Government driven one, which necessitates a change needs to continue comparing these measures from central to indicative or guidance with hydropower, and with decentralized planning. Guidance plaing requires a strategies involving coal mining, coal multisectoral and systematic method of washing, thermal power generation, frequenty and rapidly updatng public shipping, and others. The results may be investment priorities, testing plans, published to allow decentralized enterprises identifying potential shortages, and to make well-informed transport and energy generating new energy supply strategies. investment decisions based on a dear, long- The CaS decision support system should be tenr supply-demand-transport picture. With used to investigate public sector investment slight modifications to the analysis system, strategies that will consistently be cost- it could facilitate annuial coal distribution effective under a variety of future planning. With additional model circumstances. This investment evaluation development, the analysis system could also must be complemented by anaysis of policy be used to study oil and gas alternatives, measures and market mechanisms for energy conservation, carbon lioxide, and sending undistorted economic signals to the other environmental mitigation options. decentralized sector to induce them to 1 1. DEMAND AND SHORTAGES OF COAL AND ELECTRJCITY Progress in Transport coal shortages in 1987-88 were at least 30- and Energy Development 50 million to)ns (3 to 5 percent of total production) and electricity shortages at least 1.1 China has made remarkable 70-100 billion kwh (15 to 18 percent of total progress in developing its transport and production). The effects of shortages can be energy systems over the past four decades. seen in high prices for coal in the free From 1952 to 1993, the length of the market, idleness of power plants, and railway network increased from 22,900 increasing coal imports. route-km to 53,800 route-km and the length 1.3 Tran-sport capacity shortfalls of the highway network from 126,700 reached a low point in 1987-88. At that route-km to 1,083,500 route-km. Railway time, many orders for freight cars went traffic (in tkm) grew at an average annual unfilled, industrial plants in coastal cities rate of 7.7 percent (see Table 3.1, Annex were idle up to 30 percent of the time 3). During the same period, annual coal because of the lack of raw materials, and output increastd at an average rate of 7.8 some shippers were forced to truck coal percent, crude oil production increased at over distances of 1,000 km due to the lack 16.8 percent, and electricity output increased of railway freight wagons and lack of line at 12.6 percent. China now ranks third in capacity. Some industries were forced to the world in total volume of freight traffic, 'work three days and shut down four days" fourth in total volume of commercial energy due to lack of coal, according to one high- produced, and first in total coal production. level official. Coal accounts for 73 percent of China's 1.4 As a result of the scarcity of commercial energy needs, with oil (20 coal at a time when the economy was percent), hydropower (5 percent), and overheating, market prices for coal rose natural gas (2 percent) making up the sharply during 1986-89. Market prices in balance. These figures do not include areas like Shanghai and Jiangsu approached biomass, which, if commercially traded, or exceeded international levels. In coal- would add another third to the national total producing Shanxi Province, average market energy accounts (see Annex Figure 2.1). prices of coal rose from Y 55 per ton in Industry consumes 70 percent of commercial 1986 to Y 125 per ton in early 1989, while energy, followed by households (14 minemouth prices in Xuzhou rose at a faster percent), services (10 percent), and rate, from Y 90 per ton to Y 220 per ton. agriculture (5 percent). The gap in market prices between Shami and provincial producers suggests that coal Coal and Electricity Shortages shortages were caused by transport bottlenecks rather than by production 1.2 Despite the recent progress shortfalls. Bottlenecks and shortages, and the since 1980, China's economic growth has high market prices associated with them, been severely constrained by shortages of dropped from these high levels in 1990 coal and electricity, particularly in 198485 because of the economic slowdown. and 1987-88. Transport services, coal, and 1.5 Since 1992, as the economy electricity all were heavily rationed during heated up again, transport bottlenecks and this period. While shortages are not some electricity shortages have reappeared, but tiing for which the Chinese Government coal shortages have not reached the crisis (or, for that matter, any government) keeps proportions seen earlier. Chinese power an accurate accounting, it is estimated that officials currently estimate that there is a 2 1. Demand and Shortages of Coal and Electricity shortfall between peak demand and supply unit of economic output were several times of 20 percent, which has caused weekly higher than those of countries like India and blackouts in residential areas and caused Brazil (see Annex Table 3.2). While this industries to schedule round-the-clock work comparison is problematic due to shifts. Central and local authorities establish uncertainties of GNP measurement in China, quotas for allocating electricity and ration there is little doubt that their freight and new connections. The electricity shortages energy intensities are higher than those of today are caused more by lack of generating many developed countries. One reason is capacity than a lack of coal. that China relies heavily on coal as a sobrce 1.6 Prices of coal, electricity, and of energy. Most other kinds of fuel are railway services have all risen dramatically easier to transport and can be used more in recent years. From 1980 to 1992, real efficiently. Coal accounts for almost 73 coal r.rices (adjusted far inflation) have risen per cent of commercial energy a cumulative total of more than 120 percent, production-much larger than the shares for while real electricity prices have risen about most other countries (see Annex Table 3.3). 75 percent. More than 75 percent of all coal 1.9 Besides the heavy reliance on is now sold at market prices. By the end of coal, China's freight and energy intensities 199', real coal prices were fully market- can be attributed to several other factors: the determined, and consumer electricity prices small service sector; the large heavy had risen to include complete cost recovery. industry sector; the lack of preprocessing of The Ministry of Railways (MOR) now raw materials; and the low energy efficiency purchases all materials and energy at market of outdated end-use technology; and cold prices, awl real railway tariffs have been weather in the northern areas. With the raised by 86 percent from 1990 to 1992, phasing out of the allocated system, approximating long-run marginal costs. The underpricing of energy is no longer a major effect of raising these prices so substantially problen, but it will take many years to should be greater efficiency of end use in all replace the existing inefficient energy-using three sectors. equipment that was built during times of low 1.7 After the coal market was allocated prices. In addition, freight tends to opened up, prices of coal at the minemouth be hauled over longer distances than in the energy base actually fell instead of necessary, partly because of the vertical rising, despite the nearly nationwide integration of Chinese industry and partly shortage of electricity. This anomaly was because of the heavy reliance on rationing in caused by most major railway lines leading the distribution of coal and raw materi.ls. out from the energy base being heavily congested, thus preventing energy base coal Investment in Transport and Energy mines from selling all their production. It would not be accurate, however, to say that 1.10 China's transport and energy coal shortages have been eliminated: rather, systems have expanded significantly since there are surpluses in the upstream energy- the onset of reforms in 1979, but this producing areas, and sporadic shortages in expansion has been outstripped by the the downstream areas. Downstream explosive growth of demand. Since 1955, shortages were evidenced by the fact that investment in transportation, although coal stockpiles were drawn down by 32 having grown almost 15-fold, has seriously million tons in 1993. lagged behind GNP, which has increased almost 17-fold, and behind total traffic Intensity of Energy and Transport Use volume, which has grown almost 24-fold (see Annex Figure 2.2). Government- 1.8 It is widely believed that sponsored transport investnent, only 1.3 China's transport and energy intensities in percent of GNP during 1980-89, increaw ed terms of freight traffic and energy use per to 1.9 percent in 1992, but is still below iat 1. Demand and Shortages of Coal and Electricity 3 of other major countries, such as Japan, the of denand of 0.55. The elasticity of Republic of Korea, Brazil, India, and the electricity consumption with respect to GNP former Soviet Union, where transport was 0.85 for the 1980s, which is investment ranges from 2 to 4 percent (see exceptionally low for any country over a Annex Table 3.4). In short, the two reasons sustained period, especially an for China's transport shortages are explosive industrializing one such as China. The growth of demand on one side and elasticity for coal is expected to continue underinvestment in transportation on the falling in the future, but the elasticity for other. electricity is considered unreasonably low. 1.11 Energy infrastructure 1.13 The resulting lower-ihan- investment was 3.3 percent of GNP in 1992, planned elasticities reflected suppressed and has been shifting steadily away from demand as well as gains in energy coal and toward electricity (see Annex efficiency. Consumption is limited by a lack Figure 2.3). Coal's share of the energy of available capacity. New power plant investment pie fell from over 42 percent in construction during the 1980s was planned 1953 to 27 percent in 1985 and to 1 for an elasticity greater than 1.0, but GNP percent in 1990, while electricity's share growth consistently surpassed the planned rose fromi. 42 percent in 1953 to 53 percent rate by several percentage points. For the in 1985, and accelerated to 60 percent in 1980s, China planned for 6 percent GNP 1990. Despite this boost, China remains growth, but grew at 9 percent- The underinvested in the capital-intensive electricity sector could not speed up electricity subsector, in part because construction to keep pace because it was a electricity's share of total commercial energy planned sector requiring large investments requirements grew from 20 to 25 percent and long lead times. The supply of between 1979 and 1990. This trend can be electricity was constantly fbrced to play expected to continue. (The electricity share "catch-up" during this period. Further in the United States is 36 percent.) Overall evidence that the electricity elasticities investment in the coal production, coal reflect suppressed demand lies in the fact transport, and electricity sectors for the that they moved in the opposite direction of 8FYP, 9FYP, and 1OFYP is estimated to be GNP growth (and unmeasured shortages). at least US$200 billion Cin 990 prices, not From 1981 to 1985, when GNP growth including rail transport for other averaged 10 percent, the GNP elacticity of commodities or local distribution of power). electricity demand was 0.64. FrGin 1986 to 1990, during which times GNP growth Trends in Coal and Electricity Deannd averaged a much slower 7.7 percent, the elasticity was 1.16, and averaged 1.59 1.12 The demand for coal and during the slowdown of 1989-90. electricity in China has grown rapidly since 1.14 Suppressed demand or not, the 1980; growth is expected to remain strong relatively slow-growing electricity, coal and through the IOFYP. From 1980 to 1990, railway capacity must have, in some sense, coal and electricity consumption grew by 5.3 satisfied enough demand for the GNP to and 7.6 percent per year, respectively. Both grow at nearly 9 percent, or, by definition, of these figures are lower than the average the economy would not have grown so fast. GNP growth rate of 8.9 percent over the Can hypergrowth continue -without same period. The GNP elasticity for coal increasing the investments in these three demand (computed as the percent change of sectors? While a definitive answer is not coal consumption divided by the percent possible, evidence suggests that it cannot, as change in GNP) was 0.6 from 1980 to 1990. argued below. If the coal used for electricity is separated 1.15 First, what little slack there out, final coal demand grew at only 4.9 was in the coal and electricity delivery percent, which translates to a GNP elasticity system has been mostly squeezed out: more 4 1. Demand and Shortages of Coal and Electricity railway segments are at capacity, more drive the production and distribution factories are running round-the-clock, and activities on the supply side. While not an more baseload thermal plants are running at equilibrium model per se, the sensitivity of higher capacity utilization factors and the recommendations to these assumptions postponing maintenance. Second, there has may be incorporated by using low, medium, been some relatively simple, low investment and high demand projections (see Annex 5). measures to increase the capacities in related 1.17 Demands for three kinds of sectors. In coal mining, the percentage of nonelectricity coal (anthracite, coking and coal mined by township mines, with their steam) are specified in tons for 49 zones in very low investment cost and technological 1995, 2000, and 2005. Coal demand for the level, increased by 98 percent from 1983 to electricity sector is not exogenously 1989, while state mines (planned for specified, but is an indirect result of the 6 percent growth) increased only 26 percent. demand for electricity, in kwh, at 38 zones, For railways, some minor improvements to which pulls electricity from hydropower, yards and signalling and such have been able nuclear or thermal power plants. The to increase throughput. Third, there has thermal plants in turn pull coal to those been even more structural change in the plants from mines, through washeries, and Chinese economy than anticipated, meaning across the transport network. Needless to that the unanticipated portion of the GNP say, transport demand in this study is also a growth may have come largely from high derived demand: the actual coal flows are an value-added industry and services, which are endogenous result of the tradeoffs between less energy intensive. Fourth, energy types of power, coal origins, coal types, conservation has made great strides in recent transport modes, and so on. years, in part because of the rationing, price 1.18 Electricity demand in 2000 reforms, and openness to outside technology ranges from 1,258 to 1,444 to 1,661 billion and management. The reforms have made kwh in the low, medium, and high forecasts. enterprises focus on profits, and have forced Nonelectricity coal demand in 2000 ranges them to reduce waste. While the latter two from 885 to 906 to 926 million tons of trends can continue, the former two are standard coal in the low, medium, and high problematic. There may not be as much forecasts. However, since most readers do scope for these adaptions in the future as not normally separate out nonelectricity there was in the past. Also, many adaptions coal, a more meaningful estimate of demand were costly, risky, or short-sighted. For would be the model's total national coal instance, oil power plants using imported oil production in 2000 plus coal shortages plus in coastal regions is costly; putting off electricity shortages converted to equivalent power plant maintenance is risky; while amounts of coal. This works out to about to township mines are short-sighted in the about 1.51 billion tons in the low scenario, sense that they end up leaving a much 1.64 billion in the medium scenario, and higher percentage of the minable reserves in 1.78 billion tons in the high scenario. the ground rendered unobtainable because of However, it should be kept in mind that the unscientific way they exploit the deposit. these latter figures are model outputs, not inputs; more or less electricity demand could Coal and Electricty Demand have been satisfied by hydropower or Assumptions for this Study nuclear power. 1.19 The SPC energy demand 1.16 The analytical model used as forecasts used in this study are based more the basis for this study (see paragraphs 3.14- on GNP forecasts and G3NP elasticity of 3.18) is primarily a supply-side cost- demand than on price elasticity of demand; mnmizing model for investment planning demand curves simply --innot be estimated and network opdmization. Demands for coal with any degree of ceririnty for each time and electricity are exogenous inputs that period, region, or coal type in China. 1. Demand and Shortages of Coal and Electricity 5 However, price assumptions do enter into plant demand) assumes an elasticity of the crucial GNP elasticity assumption, along around 0.20, which is less than the 0.55 with several other factors. First, the SPC elasticity of the 1980s. The elasticity of total forecasts assumed that all coal prices would coal demand works out to about 0.45, which be completely market-determined and all is also lower than that experienced in the electricity prices raised to near rheir long- 1980s (0.60). Both, however, are in line run marginal costs (they were) by the end of with the long and short term downward 1994. Second, the potential for technical trends (see Annex Figure 2.4). As the improvements in energy efficiency in Chinese economy develops, a continued shift factories and steel plants is included, but away from coal and toward oil, gas, and mostly in terms of routine replacement of electricity (of which nearly 75 percent old equipment rather than installation of comes from coal) is ekpectedto continue. In highly advanced equipment. Third, the the first and second FYPs, coal accounted forecast anticipates structural change in the for nearly 95 percent of conmmercial economy. In particular, the expanding energy, which fell to around 70 percent in tertiary sector is less energy intensive than the 5FYP. While China increased its primary and secondary sectors, though more dependence on coel to 76 percent in the reliant on electricity on a percentage basis. 7FYP, this was considered to be supply- Fourth, a wealth effect can be seen in the driven rather than demand-driven, and it has increased energy use and increased failen since then to 74.2 percent. In fact, the preference for electricity by households in GNP elasticity for coal consumption for modern residences. Fifth, the growth of road 1990-92 was a low 0.16, despite an increase transport and the development of China's of 50 million tons of coal used by power oil-producing capacity will contribute to plants over those two years. While this substitution away from coal. elasticity is probably an aberration due to 1.20 In the early stages of this overreporting of coal production before study, the official Government target for 1990 and underreporting of it afterwards, it ezonomic growth was 6 percent, but during is still astonishingly low. 1992, after Deng Xiaoping's much 1.22 For comparison purposes, the publicized trip to Guangzhou, the CTS's low demand forecast of 1.51 billions Govermment raised the target to 8 to 9 tons of coal to be produced in 2000 is percent. The medium demand forecast similar to what the National People's assumes approximately 9 percent annual Congress officially endorses. On the other GNP growth through 2000, while the high hand, the Bank's Industry and Energy and low forecasts are for 10.5 and 7.5 Operations Division, China and Mongolia percent, respectively. Department, is forecasting 1.7 billion tons, 1.21 The future GNP elasticity for which lies between the CTS's medium and electricity demand is assumed to be 1.0. high scenarios. Their forecast assumes a 0.4 This assumed elasticity is higher than the elasticity for nonelecticit-y coal, 0.5 for all historical rate for the 1980s, which did not coal, 0.9 for electricity, and 9.6 percent represent satisfying all demand, as argued in GDP growth. With this much uncertainty paragraph 1.13. Anything less than 1.0 is about future energy demand, it becomes atypical of industrializing countries. The important to look at the entire range of coal demand forecast (not including power demand scenarios. 7 2. OVERVIEW OF THE COAL-ELECTRICITY SUPPLY CHAIN: PROBLEMS AND OPPORTUNITES 2.1 The process of getting coal out 2.3 Coal is mined in over half the of the ground, beneficiating it, transporting Chinese counties. In many arcas, coal is it, and burning it to produce thermal, obtained from mines close to the market or kinetic, or electrical energy, the latter of within the same region. However, in major which can be transmitted to other places, industrial centers in East, Northeast, and can be thought of as links in a chain. This Southeast China, regional and local supply chain terminates with the final production cannot satisfy demand, and consumption of coal and electricity by end economically recoverable reserves are users, and with the emission of ash and rapidly being used up. These three regions sulfur pollutants into the atmosphere. currently mine a combined total of 30.3 Measures for alleviating coal shortages and percent of coal, but have only 10.9 percent reducing pollution are available at all links of the remaining reserves, whi,Jh occur in in the chain. (Tables and figures increasingly less economic locatic-.s, depths, accompanying this chapter can be found in seam thicknesses, anl amount. The coal Annexes 2 and 3.) deficit in the East grew by 153 percent between 1982 and 1989. Coal Production 2.4 These large cc- uming regions in the eastern half of the co'ntry rely 2.2 China is the largest coal increasingly on supplies from surplus producer in the world, with total raw coal producing areas in North China, placing production of 1. 1 16 billion tons in 1992 (see pressure on the transport network. Shanxi is Annex Table 3.5). China's current coal the main province able to produce a surplus shortage certainly is not due to any lack of for shipment outside its borders; it now coal reserves. China's approximately 800 accounts for more than a quarter of national billion tons of economically recoverable coal production (see Map IEBRD 26594). Shanxi reserves are the largest in the and the nearby provinces of Inner Mongolia, world-approximately one third of the Ningxia, and Shaanxi all have large reserves world's total (see Annex Table 3.6). The of high-quality, low-cost coal. A centerpiece majority of the reserves are high-quality of China's energy and transport strategy is bituminous coal, with a carbon content of to continue to increase the already large about 60 percent, used for coking or steam production from this region (see Annex coal, depending mainly on calorific value, Tables 3.9 and 3.10). sulfur, ash, moisture, and volatility (see 2.5 However, an important Annex Tables 3.7 and 3.8). There are also ancillary strategy for relieving shortages is sizable anthracite reserves, with a carbon to continue or accelerate mining in select content generally over 80 percent, used eastern areas in order to buy time or fill mainly for residential purposes and the gaps while the transport capacity out of chemical industry. There is also some North China is expanded to reach all parts lignite, or brown coal, in the Northeast and of China. The tradeoff in cost is that Inner Mongolia, with a carbon content of production from the limited eastern coa around 30 to 40 percent, which is mainly reserves is increasingly expensive and of burned in minemouth power plants because poorer quality. of its low weight-to-energy ratio. 2.6 In 1990, the central Government operated about 600 mostly 8 2. Overview of the Coal-Electricity Supply Chain mechanized underground mines, which around I percent, which is generally low by generated 45 percent of total production. world standards, but which ranges up to The remaining 55 percent of total production 4 percent and higher in Sichuan and came from over 60,000 mines operated by Guizhou. provincial or county governments (hereafter 2.9 Mechanical coal washing is a referred to as local mines) and by townships more expensive method that removes some or villages (hereafter referred to as township of the gangue (rock and dirt mine waste), mines) (see Annex Figure 2.5). A few ash, and pyritic sulfur (not chemically provincial mines use modem mining bonded to the coal) by a coal-specific methods, but many local and township mines combination of crushers, screens, jigs, work shallow seams or outcrops with labor- flotation tanks, and even centrifuges. The intensive pick-and-shovel operations. net result is a cleaner-burning coal that Encouraged by price and investment yields more energy and less pollution per reforms, local and township mines can ton. Although coal washing offers provide the local work force with income environmentai and transport benefits, plus and local energy users witi coal without lower disposal and boiler maintenance costs, relying on the overcommitted railway some carbon is also lost in the process. In system, and with minimal investment and power plants, washed coal can be more short construction lead times. efficient than raw coal if the boiler is 2.7 The Government could decide designed for it. About 18 percent of coal to continue encouraging local and township (but 37 percent of state-produced coal) was mine growth so as to increase production washed in China in 1991 (see Annex Table rapidly and free up investment for other 3.11), compared with more than 50 percent sectors, but there are several negative side in most Western countries. All coking coal effects: dangerous working conditions; poor (127 million tons) and some anthracite for coal quality; environmental degradation; the chemical fertilizer industry (10 million and, often, wasting of reserves because the tons) are washed by necessity. Only 69 unscientific way they exploit the reserves million tons (7 percent) of steam coal were often leaves large quantities of coal further washed in the 71 existing plants. However, below unaccessible by future mining. These almost none of the plants used for washing local and township mines generally supply steam coal were designed for that purpose; local markets and transport their coal by they are mostly plants designed for washing truck. A strategy involving greater anthracite and coking coal that are no longer production from such mines would have to needed for those purposes. include provision of railway spurs to larger 2.10 Coal fines can be lost either by local mines andlor collection of coal by being sucked up the flue or falling through truck from smaller mines. the bottoni grate of industrial boilers. Coal screening is an inexpensive but effective Coal Beneficiation method for separating lump coal from coal fines, so that the fine coal can be distributed 2.8 Methods for improving the to power plant boilers that are designed to quality of the run-of-mine coal are not use pulverized coal, or so that it can be routinely practiced, but they are often made into honeycomb briquettes. About 19 technologically simple and can play a percent of coal is screened. valuable role in eliminating coal shortages. The steam coal supplied to many users, Coal Allocation and Pricing particularly small users, can be poor-containing stones, 20 to 30 percent 2.11 The dual price system for coal, ash by weight, and a high percentage of coal featuring large differences between low fines (small particles). The sulfur content in state-determined in-plan prices on the one the North China energy base averages hand and higher open market or negotiated 2. Overview of the Coal-Electricity Supply Chain 9 prices on the other (see Annex Tables 3.1Z Coal Transport to 3.14), is now extinct. Since the dual price system w3s introduced in the mid-1980s 2.13 Railways are the predominant (replacing the preceding fully-controlled mode of transport for moving coal in China system), in-plan prices have increased in (see Annex Table 3.15). There are no major real terms (20 percent in 1991). Perhaps navigable rivers emanating from the North more importantly, the percentage sold in- China energy base. However, at least part of plan was cut in half in January 1993, and the journey to Northeast and Southeast now accounts for less than one quarter of all China can be made by water, either by coal. As of the end of 1994, the rest was coastal shipping, the Grand Canal, or by the sold at free market prices, with the possible Yangtze River. About 626 million tons of exception of a few consumer categories. For coal were shipped by rail in 1990, and of instance, the MOR now buys all of its coal this, 139 million tons were transshipped to at market prices and its electricity at state-owned water transportation (see Annex nonsubsidized prices. In East China, coal Figure 2.6). However, if locally-owned prices are now close to or above shipping is included, 210 million tons moved international price levels, while in the by water. Trucks carried an estimated 161 Northeast, some producers of low-quality million tons. Trucks are used to deliver coal coal are unable to sell their entire to users near the mines, to collect coal from production. Although the dual price system scattered mines and deliver it to railheads, was phased out in 1994, its aftereffects in and to distribute coal in cities. The the form of inefficient boilers, power plants, remaining coal was consumed locally buildings, and so forth, will linger on for without being transported, and was many years. consumed at minemouth power plants. The 2.12 The allocation of coal between railways handled 341 billion tkm, state- sellers and buyers continues to be organized owned water carriers 182 billion tkm, and through an annual conference involving roads only 10 billion tkn. Since the 1950s, producers, large consumers, and central railway investment as a percentage of total Government agencies. This conference transport investment has fallen (Annex serves the function of a not very efficient Figure 2.7). commodities market, with the interjection of 2.14 Because of the increasing the transport providers into the mix. The reliance on nordtern and western coal, the quantity and quality of coal to be supplied average transport distance for railways has are negotiated between participants who increased from 530 to 545 km from 1988 to have had longstanding supply relationships. 1990. The average waterway distance (state- Central authorities arrange for the transport owned) has increased from 1,175 km to of allocated coal. Sometimes, coal supplies 1,310 km in just 2 years. Trucking is are not well matched to consumer inefficient for long-distance transport of bulk requirements because of inflexibility and coal, thus averaging just 63 km per trip. inadequate responsiveness to changing However, there were reports of trucks market needs in the coal and transport moving coal as far as 1,000 km during the allocation system (see Chapter 6, Policy worst coal shortages. Issues to be Addressed in the Future). By 2.15 Coal represents about 42 the end of 1994, mines in the coal-rich percent of the total tonnage of freight regions of Northwest, Northeast, and South- handled by the railways. China's density of west china had been mostly freed of state freight traffic on the railways is much higher distribution plans, with the exception of coal than that of the United States and India and for some power plants for which the price is nearly on par with that of the former Soviet still controlled by state or local Union (see Annex Table 3.16). In 1993, governments. freight traffic in China averaged nearly 22 million net tkm per route-km, which 10 2. Overview of the Coal-Electricity Supply Chain represents a fairly high level of operating coastal waterways between the northern and efficiency and asset utilization. Still, there is southern seaports. Extra distance added by room for improvement. the waterway leg of the trip is not an issue 2.16 The effective strategies for for many coastal cities in East China (see increasing railway capacity from the energy Annex Table 3.17). As an example, base can be divided into two categories: distances from Shanxi Province to Shanghai building more line capacity, or increasing are nearly the same by combined rail-water the throughput of existing assets. A number routes as by the shortest rail-only routes. of new lines are being constructed, most China has 15 major ports. Qinhuangdao notably a west-east railway from Shenmu in alone loaded 54 million tons in 1991, nearly Shaanxi Province to Huanghua, a major new 70 percent of the coastal coal shipping. port south of Tianjin, and a third north- Some port capacity currently goes unused south corridor from Beijing to Guangzhou. because of railway bottlenecks leading to the The MOR is also considering adding a new ports or because of the mismatched capacity dedicated passenger line from Beijing to of receiving ports. Shanghai to the existing double-track, mixed 2.19 Many investment strategies freight-passenger line. This will separate exist for expanding water transport of coal freight and passenger traffic between the two in China. Modem port facilities are not only lines and thereby boost the combined more efficient, but they decrease the throughput capacity of both lines. turnaround time of ships in ports. The use of 2.17 There is a spectrum of self-unloading ships can reduce the cost of investments for increasing throughputs on building handling facilities at a user's dock, existing lines (see Map [BRD 26857). New especially if the annual volume is low. m.arshalling yards, longer and more frequent Pusher-barge system, in which barges can be sidings, and better signaling and used on the open seas or on the inland management can incrementally increase waterways (or both, without transshipment) capacity. For larger increases in capacity, have the advantage that the pusher units are there are some possibly very cost effective separable from the barges, thus maximizing strategies such as the use of unit trains, or their at-sea utilization. replacing existing rails and wagons with 2.20 Generally speaking, larger those designed for heavy hauls. Unit trains, vessels are significantly cheaper per ton than Marshalled as a single unit that continually smaller vessels. In this study, per unit loops from origin to destination and back investment costs fall from Y 6,000 per dwt again, is being used on the Datong- for 9-meter ships to Y 4,300 for 12-meter Qinhuangdao line and is planned for the ships and to Y 3,000 for 14-meter ships. Shenmu-Huanghua line. For heavy haul With the exception of Qinhuangdao, China's technology, the MOR recently decided to ports are not deep enough to accommodate increase maximum freight car axle loads on 14-meter (100,000 dwvt) post-Panamax 10 to 20 percent of its freight cars from 21 supercolliers. Likewise, their fleet of ships to 25 tons per car over the next 10 years. consists mainly of domestically-produced This step, which can increase net tonnage shallow draft vessels 9-meter (35,000 dwt) throughput capacity of a line by up to 30 or smaller. For many years, small ships and percent, could offer significant help on shallow ports have jointly constrained the capacity-constrained lines. However, the Chinese shipping industry from changing to effects of existing, less efficient, rail larger ships like in the international coal technology can be expected to last through Wade. However, a new generation of 50,000 2000.. Finally, the options for expanding the dwt shallow draft ships is under capacity of existing lines by the largest development for 2000 and beyond. Potential amount are multi-tracking or electrification. exists for long-distance transport of coal 2.18 China has the potential to from a single, deep-water northern port expand greatly the coal transport capacity on (e.g., Qinhuangdao) to a single, deep-water 2. Overview of the Coal-Electricity Supply Chain 11 southern port (e.g., Gaolan, near Shenzhen), Reform of coal prices will encourage some which could serve as a terminus for routes of these activities. Another important by 100,000 dwt ships and as a hub for a strategy on the coal consumption side is to fleet of smaller vessels making deliveries to accelerate the trend to channel more users. However, the high cost and industrial growth toward the coal producing environmental impacts of port construction areas. Several factors are working against and dredging, and the total lack of large 14- this, such as the existence of industrial meter domestic supercolliers, have placed linkages in existing industrial areas, and a this option on the Government's back less developed technical work force in the burner, perhaps until the Hong Kong coal regions. takeover is complete. 2.24 Some Chinese experts think 2.21 Slurry pipelines are a that over the last 10 years, about 70 percent promising technology for long-distance bulk of energy conservation has been contributed transportation, but they are untested in by adjustment of industrial structures and China. Arid conditions in the north central enhancement of energy management. The area limit the potential for coal slurry pipe- latter includes employing more engineers lines, as do their relatively small capacities. responsible for energy efficiency in A coal slurry pipeline is being built by a enterprises, and setting up special offices in Sino-foreign joint venture from Yuxian in charge of energy conservation in state, Shanxi province to Weifang in Shandong provincial, and local governments. The province. Some of the proposed pipeline remaining 30 percent of energy conservation projccts plan to ship washed coal. in China camne from introducing new technology, changing processes, using new Coal Consumption and Conservation materials, and replacing outdated equipment or facilities with advanced ones. Most such 2.22 In 1991, 1.104 billion tons of projects are designed to achieve multiple coal were consumed, of which industries benefits. Expansion of production capacity, accounted for 78 percent (see Annex Figure improvement of product quality, and 2. 1). Power plants used 27 percent of all environmental protection are generally coal, steel plants used 8 percent, and other combined with energy conservation industries used 43 percent. Of the measures. Some sample surveys show that remainder, residential buildings used 15 the investment required per unit capacity of percent, commercial buildings used 1 energy saving continua'ly increases year to percent, transport used 2 percent, agriculture year, but up to now, a marginal production used 2 percent, and building construction curve or function for energy conservation and others used 3 percent An estimated 3 to has not been estimated for China. 4 percent (30 to 40 million tons) was lost 2.25 Officially, 1.23 million tons of during handling and transport. Less than coal were imported in 1992, mainly to the 2 percent, or 20 million tons, was exported. seaports of Dalian, Shanghai, and 2.23 Many strategies could slow the Guangzhou. As the economy is opened and rate of growth of coal consumption even reformed, coal imports can play an more than the current projection. Several important role in balancing supply and Bank reports suggested many promising demand in coastal areas with little avenues for energy conservation through investment or lead time. technological advancement; operating changes; district heating; co-generation; Coal for Electricdty larger-scale industrial plants; substitution of other energy sources (for example, oil, gas, 2.26 Since 1949, China has become solar, wind, hydro, nuclear, and biomass) the fourth largest producer of electricity in where appropriate; and more emphasis on the world, with a total output of 754 TWh in high value-added industries and services.' 1992. Power in the country is distributed 12 2. Overview of the Coal-Electricity Supply Chain through nine major grids (see Map IBRD On the positive side are the air pollution 26595). In 1992, total installed capacity was reduction, irrigation, flood control, and 166 GW; about 76 percent of this was possibly navigation and recreation benefits. thermal and 24 percent was hydropower. A On the negative side are the loss of miniscule percentage of the thermal capacity agricultural land and the need for relocation is fired by oil or natural gas. From the and resettlement of villages. There are 1950s to the 1970s, the thermal share of several enormous hydropower projects in the total electricity sector investment fell while 8FYP and 9FYP, including the Three that of hydropower rose, until the 1980s Gorges reservoir and power station on the when the trend was reversed sharply (Annex Yangtze River between Wuhan and Figure 2.8). Industry is by far the largest Chongqing, approved by the State Council user of electricity, accounting for 77 percent in 1992. of consumption in 1992. Approximately 96 2.29 Local energy resources are also percent of the nation's villages and 80 relied on to meet local needs, especially in percent of rural families now have access to isolated areas. In 1989, mini-hydro, small electricity. thermal, and diesel generating sets had a 2.27 The electricity sector's 8 total installed capacity of 18.7 GW and percent share of final net energy generated 60.4 TWh. In addition, China consumption is one of the lowest in the commissioned its first nonmilitary nuclear world, but is likely to go much higher as the power plant (300 MW) in the Shanghai area service and residential sectors use more (see in 1992. In Guangdong, the Daya Bay Annex Table 3.18). One prominent nuclear plant (2x900 MW) will send 70 investment strategy is to develop more percent of its power to nearby Hong Kong. minemouth thermal power plants, thus Known uranium reserves could sustain seven avoiding the transport bottlenecks. At times more nuclear power production for 30 present, about 40 percent of all thermal years, and contingency plans include several power plants are located at minemouths additional plants. (within 50 km distance). However, with a 2.30 Since the start of economic few exceptions, most electricity transmission reforms in 1979, China has made significant is intra-provincial; the self-sufficiency of progress in modernizing its power most provinces in terns of electricity generation and transmission technology. production is evident in Map 26596. Mine- Between 1979 and ¶989, fuel consumption mouth power development is limited by the in new thermal units was reduced to about availability of water in the Yellow River 400 g of standard coal per kwh, or about 36 Basin where coal reserves are concentrated. to 37 percent efficiency. The oldest, smallest Also, to compensate for transmission losses still-operating plants use 60 percent more of high-voltage transmission lines, mine- coal to produce the same amount of mouth power generation requires about 5 to electricity as the large (600 MW) efficient 10 percent more coal mining and power- coal-fired units (600 MW)-and they generating capacity. produce at least 60 percent more pollution. 2.28 China has developed only But only 12 percent of thermal plants are in 9 percent of its hydroelectric generating units of 300 MW or more, and the national potential, estimated at 1,900 TWh per year. average conversion efficiency is only 31.8 Developing more of this hydropower percent in 1992. China can now also build potential would reduce demand for coal and operate 500kV direct current (DC) transport, but hydropower and the associated transmission lines for distances of 800-1,500 transmission lines can be capital-intensive. km. Most of the potential is located in the 2.31 A dual price system for Southwest (70 percent), about 1,500 km electricity was introduced in the nid-1980s, from the major coastal cities. Hydropower and was overhauled and simplified in 1993. also has positive and negative externalities. Three principles govern the current system. 2. Overview of the Coal-Electricity Supply Chain 13 First, higher prices can be charged for with inadequate emissions control electricity from new power plants, as the technologies. Ambient air concentrations of investment for the new plants (most of total suspended particulates (TSP) and sulfur whose inputs were purchased from the free dioxide (SO2) in Chinese cities are among market) is comparatively high. Second, the the highest in the world (see Annex Table price is composed of three parts: operating 3.20), and are worsening rapidly. They can cost, return on investment, and profit to the cause serious respiratory health problems. investors. There is thus an element of cost- Acid rain, principally a result of SO2, plus pricing in the system, but the situation damages crops, forests, and aquatic ecosys- is complicated by a variety of surchatges tems. China's contribution to global carbon and miscellaneous fees and discounts relating dioxide (CO2) buildup, largely due to to fuel prices, transport, etc. on a plant-by- burning coal, is estimated at 11 percent of plant basis. Third, the price is negotiable the world total. The envirownental impacts between the local and central govermments. of coal use are not limited to air pollution. All told, average consumer prices in regions Other impacts include mine runoff and land such as East China now provide for degradation from strip mining, and disposal complete recovery of investment and of coal ash from boilers and stoves. interest. Given the national shortage of 2.34 Particulates are considered to electricity, fear of inflation is the main present a more serious environmentai reason behind the delay. problem than S02-caused acid rain, mainly 2.32 Different rates continue to be because most of China's coal has a low charged depending on whether the power average sulfur content (about 1.5 percent) comes from old or new plants, and whether and high ash content (20 to 30 percent). the consumer is a longstanding or new Most types of TSP control equipment customer. The main difference continues to (except for fabric filters) are available in be between plants built with grant financing some parts of China, but removal before 1985 and plants financed through efficiencies can be low. Sulfur scrubbers are loans that must be repaid, built since 1985. not currently in use in China, mainly Loan-financed plants also include the semi- because of their high capital costs, which autonomous plants built by Huaneng Inc., an can add up to one third of the investment independent, date-owned firm that seeks out costs of a power plant. foreign financing. Grant-financed plants now 2.35 Emissions standards for power use an improved mechanism for plants in China are low compared with the automatically adjusting prices to reflect industrialized world. For comparison, the changes in fuel costs. In some places, if standards imposed on the new Yangzhou large users need more electricity, they are thermal plant (4x600 MW) funded by the asked to pay an advance paymnent to help World Bank are 520 tons of SO2 per day, find the construction of new generating compared with 70 tons per day (tpd) for a capacity. This advance payment can be as similar plant in the United States, 19 tpd in high as Y 1,5'00 per kw of capacity, and the United Kingdom and Germany, and 100- may have to be paid as much as two years 500 tpd for the recommended World Bank before they get any electricity. Consumer standard (100 tpd for highly polluted areas, pricing for electricity remains complex, 500 tpd for unpolluted areas). The nontransparent, and inequitable. particulate limits are similarly lax: 348 mg/r3 for the Yangzhou plant, versus 50 Environmental Impact of Coal mg/r3 in the United States, United Kingdom, and Germany, and 100-150 2.33 Air pollution is a serious mg/m3 for the Bank's recommendation, problem in China. Much of it is caused by depending on whether the site is urban (100) burning coal (see Annex Table 3.19), or rural (150). especially on the part of small-scale users 14 2. Overview of the Coal-Electricity Supply Chain 2.36 In addition to these standards, air quality standards (see Annex Table China also has pollution charges on the 3.21), though how the responsibility for books, but it is not clear how much they (or achieving them is distributed is not clear. the various standards) are enforced. The multitude of standards and taxes on the Monitoring emissions is difficult, and up books, in combination with the poor air until the recent introduction of the profit quality, raises questions about their motive to factory managers, there was little enforcement. incentive to comply. They also have ambient 15 3. PLANNING AND INVESTMENT FOR CHINA'S TRANSPORT ANiD ENERGY SECTORS 3-1 China is currently reforming its Power (MOEP). They prepare blueprints for planning and investment system. The the long-term development of their Government is decentralizing decision- industries, and develop investment projects making and reorienting the economy toward in accordance with the blueprints. However, the market in a step-by-step fashion. This the two energy ministries no longer directly chapter describes the Government's role in manage enterprises or construct projects; planning and financing the energy and they are charged with overall sector transport sectors and the problems it faces, coordination and policy guidance, and with and then briefly describes the modeling tool planning of state-owned energy projects. developed as part of this study for assisting State corporations under MOEP and MOCL the Government in its planning and construct and operate these energy projects. investment tasks. Greater detail on the 3.4 In 1992, the Chinese planning and investment system can be Government further decentralized project found in Annex 4. The model is described in management and planning. Those a general fashion in Annex 5, and in construction projects whose funds, materials, diagrams in Annex 6. and marketing can be handled by a local or provincial government can now be The Planning System organized, sponsored, engineered, and administered at that level. For some projects 3.2 The SPC is China's main that involve several ministries and/or Government agency for long-term economic provinces-such as an integrated project of planning in China. The SPC draws up the coal mining, rail transport, and thermal FYPs that outline the development strategy, power generation-different parts would be targets, and key construction projects. handled by the State Coal Mining Administratively, the SPC serves directly Corporation, MOEP, and MOR, with the under the State Council (China's Cabinet), SPC acting as the overall coordinator. which manages the economy on behalf of the National People's Congress. In 1992, to The Financing System provide a better linkage between the FYP and the yearly plan, the Government adopted a two-year rolling plan with tentative targets 3.5 China's financial sector is in for the following year. The State Economic transition from one that allocates credit and Trade Commission, newly carved from directly to one that relies increasingly on the SPC, is responsible for day-to-day markets to ensure that savings flow to high implementation of the plans. Environmental return investments. This is substantial protection is supervised by the National progress given that prior to 1987, state Environmental Protection Agency, which is investment funds were given as grants and situated one level below the SPC or ministry did not have to be repaid separately from the level, but functions like a ministry. return of profits to the state treasury. The 3.3 Ministries are also directly financial sector that emerged from the early under the State Council. The four main and middle stages of reform (1979-1991) ministries responsible for the coal-electricity was composed of one central bank (IMie delivery system are the Ministry of Railways People's Bank of China, or PBC), four (MOR); the Ministry of Communications specialized banks (industry and commerce, (MIOC), which plans and builds highways, agriculture, construction, and foreign waterways, and ports; the Ministry of Coal exchange), two comprehensive banks (multi- (MOCL); and the Ministry of Electric sector lending), seven newer commercial 16 3. Planning and Investment for China's Transport and Energy Sectors banks (mainly regional in scope), and tens 3.8 Since the early 1980s, the of thousands of urban and rural credit central Govermnent has been a diminishing cooperatives. source of investment funds. The new 3.6 The situation regarding approach is to fund a project from various infrastructure financing, however, has sources, such as (a) the sponsoring ministry changed dramatically with the introduction or corporation; (b) foreign capital; (c) bank of the three policy banks, viz., the State loans; (d) tie SDBC (or PCBC); (e) local Development Bank of China (SDBC), the govermnents; (f) local agencies; (g) private Agricultural Bank of China, and the Export- corporations; and (h) shares of stock. While Import Bank of China beginning in April, provinces and cities can plan and build large 1994. (see 'China: Financial Sector projects independently of the central Reforms: Current Status and Issues", Report Government, the 10 to 30 percent that they No. 23492-CHA). Prior to this, the four typically get as directed credit can be crucial specialized banks lent money to two classes for the project's financial success. The share of projects: first, those that offered high of central Govermnent financing in the financial rates of return and were bankable; power sector, for instance, has fallen and second, those that did not satisfy sharply, from 91 percent in 1980 to 30 commercial lending criteria but were deemed percent in 1992, while provincial and local to be of high national priority for policy govenmments now provide 40 percent or reasons. Power and transport projects often more. fell into this category because of their large scale and long payback periods. The purpose of the three policy bankls is to disentangle Problems in the Planning the 'directed credit" or policy loans from and Financing System commercial lending by creating separate institutions for each. For infrastructure in 3.9 Inefficiency in the transport the coal and electricity delivery system, the and energy sectors are common problems SDBC is taking over responsibility for for the Government to try to overcome. policy lending from the People's Various institutional factors lead to Construction Bank of China (PCBC). Thus, inefficient choice of scale and technology, the PCBC will now be able to function more particularly in the use of energy and raw on commercial lines, while the SDBC will materials. Generally, larger mines, ports and not be profit-oriented and will lend at power plants are more efficient. If a local subsidized rates. government invests in production capacity in 3.7 The SDBC supposedly will another province, they lose employment and have autonomy in deciding which projects to tax base. Marketing is not advanced in finance from a list provided by the SPC. China, which, together with the lack of The SDBC will appraise projects not only adequate transportation and for financial viability but also technical telecommunications infrastructure, makes it quality (to be contracted out) and adherence difficult to generate demand for a product to state industrial policies. At its startup, the outside the local area. Local governments SDBC had 350 projects in its lending pipe- often do not have the financial resources to line, which does indicate a close relationship build larger-scale projects, or they wish to with the SPC, at least at first. The SDBC's avoid the Y 30 million cut-off level for staff is largely composed of nonbanking staff requiring clearance by the SPC. Annual absorbed from the SPC's Investment budgetary negotiations create a bias toward Department and from the former SPC- capital cost minimization without controlled investment companies. Overall, considering operating savings of more about Y 40 billion per year of investment is efficient technologies. controlled by the SPC through SDBC. 3.10 Intersectoral and interregional coordination is also a problem. A central 3. Planning and Investment for China's Transport snd Energy Sectors 17 problem in this study is investment in power 3.13 In early 1994, the State plants, steel plants, and other coal users in Council redefined the functions of the SPC locations where enough coal or the right in the new socialist market economy. The type of coal cannot be delivered because of mne-point plan can be encapsulated as four bottlenecks. However, other examples of major functions: poor coordination can be found. Price (a) strategic functions such as researching distortions can create biases, such as and developing strategies, directions, artificially low electricity prices favoring targets, national industrial policy, and electric traction over diesel for railways 10-year, 5-year, and annual guidance even for low density corridors. Railway plans; construction has not always been optimally (b) macroeconomic adjustment functions coordinated with port construction. One such as helping to adjust the supply example is the Datong-Qinhuangdao railway, and demand for capital and which was not completed in time for the monitoring total price levels (the opening of the new berths at Qinhuangdao. National Price Management Bureau is To supply the new shipping capacity with now part of the SPC); coal, a short spur from the then-endpoint of (c) coordination functions such as helping the railway was built to connect with the to establish a national, cross-region Beijing-Qinbuangdao railway. This spur market system for important became obsolete upon completion of the commodities, info r m atio n DaQing line. Another example is the use of dissemination, and helping different unrealistically low growth rates for planning sectors to work together; and purposes, which has led to underinvestment (d) fund allocation functions such as in infrastructure. reviewing, approving, and guiding 3.11 In the financing system, the key construction projects. underinvestment in infrastructure is The key theme behind all these functions is endemic. More capital must be attracted and that their role has changed from controlling efficiently used. Adt4uate investment over the economy to steering it. Its old role was the long run will hinge on continued reform at the core of the central planning of the new market institutions and opening mechanism, while the new role represents a of the system to additional sources of transition toward indicative planning." capital, both domestic and foreign. Indicative planning involves guiding decentralized enterprises and local The Changing Role of the State Planning governments and advising the central Conmnission and its Analytical Needs Government on the implications of its policies. 3.12 Before the 1980s, the central 3.14 The SPC consists of a national Government controlled all individual agency plus provincial planning agencies for projects through the SPC. During the 1980s, each of China's 30 provinces. Planning of they controlled mainly large projects over the coal and electricity delivery system is Y 30 million. Now, townships can plan and done by several departments, including finance some of their own investment Energy, Transport, Long-Term Planning, projects, and the plans SPC makes are no Investment, and Science and Technology. longer mandatory even for state-owned The national-level SPC has a policy research industries. As the long-term economic arm that provides analytical support to the planning agency of the Government, one of planning side. The ERC is an umbrella SPC's main tasks has historically been to institute that coordinates the activities of distribute state-controlled investment funds several institutes, including the Economic between regions and industries in a balanced Institute (El), the Energy Research Institute way. (ERI), and the Institute for Comprehensive Transportation (ICT). These three institutes 18 3. Planning and Investment for China's Transport and Energy Sectors contributed researchers for this study, under became clear that in addition to increasing the supervision of the ERC itself. the transport capacity, investments such as 3.15 The SPC's new policy and coal washing and hydropower may, in some planning role emphasizes working through circumstances, reduce the Aemand for coal the market economy. Thus, they have a transportation in a cost-effective and strong need for a systematic market environmentally beneficial way. The close information system, including forecasting substitutability between rail transport of coal and modeling, in order to guide market and minemouth power plants coupled with performance and steer away from chaotic, long-distance transmission lines make it uncoordinated public infrastructure necessary to evaluate these different options development. According to Mr. Wei Liqun, in a systematic and integrated fashion. General Secretary of the SPC, they have 3.18 The analysis tool co-mbines come to recognize that planning in the past features from major energy transport models has sometimes been unreasonable. There is in the United States, but has been tailored to a growing realization of the need to analyze the SPC's needs. The CT1S model can be multiple alternatives quickly. And, to considered a Decision Support System facilitate the functioning of free markets, because it supports SPC policymakers in they need to open the planning process to making decisions by providing a fast and different fields of society, including local comprehensive way to study energy delivery officials, sectoral experts, and international as a system rather than as separate parts. It investors. And, to the extent that the SPC treats the energy delivery system as links of continues to fund state-owned companies and a chain, b'-ginning with coal supply, and allocate state funds that can make or break a continuing through washing, transport, jointly-financed investment project, the SPC utilization, conversion to electricity, should try to do so in such a way as to meet transmission, utilization of electricity, and infrastructure needs, coordinate different impact on the environment (see figures in provinces and industries, promote efficiency, Annex 6). In the model, these activities take and protect the environment. place at 48 coal supply nodes, 49 coal 3.16 China's economy has been demand nodes, 58 electricity supply nodes, increasingly dynamic in recent years. The and 38 electricity demand nodes, which are GNP forecast has been raised and may soon linked by 286 railway arcs (177 of which be raised again. Policies and priorities are have investment projects), 31 ports, 3 being frequently updated as the planning and potential slurry pipelines, and 99 investment system is reformed. The SPC has transmission lines. Given any forecast of recognized these realities by adopting two- GNP growth and energy demand elasticity year rolling plans. A computer-based for 1995, 2000, and 2005, the model will decision support system with a rapid try to satisfy those demands for the lowest response time is one of the ways the SPC is cost possible. The model features a trying to adjust to changing conditions and Geographic Information System (GIS) that policies and their own newly-defined role. facilitates the visualization of data and results. The CTS Network Optimization Model 3.19 The model's primary use is to identify logistical trouble spots and broad 3.17 Because China's transport investment directions by sector and region. infrastructure problems of the mid to late The methodology allows the dominance of 1980s spanned the entire country, the complex options over one another to be original idea was to develop a decision evaluated in a rough but relatively support tool to help coordinate the expansion comprehensive way before proceeding to a of the transport network (see Annexes 5 and more refined economic analysis. Thus, the 6 and the Supplementary Volume for greater CTS model is designed to complement, not detail). However, in studying the problem it replace, other more detailed models for 3. Planning and Investment for China's Transport snd Energy Sectors 19 planning particular sectors of the coal- among different links of the transport system electricity delivery system, such as rail and and more generally among the different links electricity planning models. Policymakers of the supply chain, require an intersectoral can answer "what-ir questions and analyze planning tool that will compare and tradeoffs between economic goals, energy coordinate investment projects from MOR, supply goals, and environmental goals. MOC, MOCL, and MOEP on a level 3.20 Methodologically, this kind of playing field. comprehensive national modeling is rare 3.21 Even if the analysis system is among the Bank's borrowing countries never ased to referee this competition for today. But in this case, it is necessary funds, the systems thinking behind the because the coal-transport-electricity- model and the consistent data base environment system is so large and complex, developed for it may be useful for so interconnected, and has so many policymaking. And if it is used, there are alternatives to be analyzed. Coal mining several modes for doing so. Using only the outside the energy base, coal washing, planned set of energy and transport options, hydropower, shipping, and minemouth the potential performance in terms of power are important -relief valves" for the meeting various growth rates or pressure on the railway system. In addition, envirorunental goals can be assessed. some of these investments are faster to Alternatively, the tool can be used on the implement or have greater environmental leading edge of the planning process to benefits than others. Interrelationships generate new project ideas. 21 4. ANALYSIS OF INVESTMENT STRATEGIES Overview wiLh China's past practice. Another key assumption relates to the shortages, which The Task Ahead are necessary bec_use demand in the medium term may not be fully satisfiable at 4.1 With rapid economic growth a reasonable cost. The cost of shortages is and reform, industries and consumers alike the economic loss from unsatisfied demand, demand more energy and transport. In mid- but it also defines the buyers' willingness to 1992, the Chinese Government increased the pay in that it acts as an upper bound for the GNP growth forecast for the rest of the delivered cost of any supply chain. The decade from around 6 percent to around approach adopted for this study is to base 9 percent. In addition, the coal forecast for the shortage cost on the cost of substitutes 2000 has been raised from 1.4 billion to at for coal or electricity supply (see Annex least 1.5 billion tons, and the electricity 13). The basic idea is that the economic loss forecast from 1.1 trillion to at least 1.3 or willingness to pay, which are nearly trillion kwh. To meet its own forecasts, impossible to quantify in a semi-market China plans to produce about 400 million economy like China's, should in any case tons more coal than is currently not be higher than the cost of an available produced-or about 60 million additional substitute form of energy, such as imports. tons each year-though the study's analysis In this study, the cost of imported coal actually suggests that the increase needs to delivered to coastal ports was estimated at be closer to 500 million tons. Of the 400 to Y 300 per ton. Electricity shortage costs, 500 million additional tons, around 300 based on converting imported coal into million is needed for new thermal power electricity, range from 20 to 70 fen per plants. Without enough coal for electricity, kwh, according to an increasing step e-onomic growth could fall short of function that varies from zone to zone. 9 percent. All this needs to be accomplished 4.3 The model was provided with in a seven-year time frame. This chapter investment options exceeding what is outlines a coordinated strategy for achieving planned in the 8FYP and 9FYP (see these goals. Annexes 14-17' For some standardized sectors, like thermal power plants and coal Key Modeling Assumptions washing, potential new capacity was virtually unlimited. For site-specific sectors, 4.2 The condusions are based on new options were far more limited, but still 13 scenarios analyzed with the CTS Analysis exceed planned ;apacities. Mining options System in 1993 (see Annexes 7-10), and to are about 100 nercent greater than the a lesser extent on 16 scenarios run in 1992 planned capacitj, hydropower options are (see Annexes 11-12). GNP growth of about 50 percent greater, and transmission 9 percent was the main forecast studied, but options are about 800 percent greater. runs with 7.5 and 10.5 percent were also Railway options include a few lines that analyzed (see Chapter 1). The analysis is were not originally in the FYPs or that were based on 1990 costs, not prices, and all listed at a lower priority, some of which Chinese currency figures are reported here have since been inserted into the plans (see in 1990 yuan. The discount rate is assumed paragraph 4.10). The list of possible projects to be 12 percent. One of the key analysis is limited by what can be accomplished assumptions is that there is no coal imported realistically in seven years. from other countries except in selected 4.4 The model was calibrated to scenarios-an assumption that is consistent approximate certain aggregate targets, such 22 4. Policy Analysis of Investment Strategies as total coaL and electricity production, exceed 125 million tons. In the worst case negligible shortages in the energy base, and scenario, shortages in 2000 are nearly a reasonable split between rail and water 8 percent of demand, with an economic cost transport and between hydro and thermal to the economy of US$10 billion, or about I power. In addition, results were checked percent of China's projected GNP. If thoroughly by SPC and Bank experts (see demand surpasses the 1.6 billion tons of coal Annex 8). needed for the medium-demand scenario, the railway projects currently planned for 2000 Prospects for Energy Shortages would probably not be able to handle it, even when supplemented by several railway Nine Percent GNP Growsth projects that have been recently accelerated in the MOR's plan (see Railway Investment 4.5 If the economy grows at about Program, paragraph 4.9). Furthennore, 9 percent per year, the analysis results accelerated implementation of nonrailway suggest that it should be possible within projects, such as long-distance transmission existing constraints to satisfy nearly all the and coal washing, will not be able to coal and electricity demands without alleviate the paralyzing pressure on the importing coal, except in a few problem railways. Coal imports can help somewhat in regions (see Annex Figures 10.1 to 10.4). the coastal cities, but domestic coal that The combined shortfall of coal and would be replaced by imports still cannot be electricity (converting electricity shortages to moved easily to interior cities (see Imports, their coal equivalents) are estimated at paragraph 4.33). For 10.5 percent growth, around 2 percent of demand. The economic the number of projects being planned or losses due to these shortages (valued at the studied does not seem to be enough. cost of importing coal) is estimated at USS3 4.7 Transport and infrastructure billion (in 1993 prices) for the single year of development need to adjust much more 2000. These conclusions are based on the quickly to rising demands. The Government assumption that additional energy delivery should open up the infrastructure sectors to strategies are carried out in anticipation of market forces as much as possible to enable that growth. Electricity transmission and them to be more responsive to changing coal washing are important components of situations. However, in the rush to reverse the delivery strategy recommended by the the historical underinvestment in long- mnodel, and these only recently gamered distance energy transport as quickly as serious consideration by planners and possible, there is a risk that railways, mines, enterprises in China. The CaS's conclusions ports, and the entire electricity sector will would be less optimistic in the absence of run into some coordination problems. such alternatives. Also, enough capital must Therefore, beyond simply setting up market be available to build all the projects. A mechanisms, the central Government should 10 percent reduction in the capital budget have a role in rationalizing network could nearly double the combined coal- investments. Their role, however, should be electricity shortages (Case 92-10). a guiding rather than controlling one, in cooperation with provincial and industry Impact of Different Growth Assumptions authorities. 4.6 Withlowergrowth(7.5 percent Transportation Sector in Case 93-13), shortages could be virtally Policy Inplications eliminated, but higher growth (10.5 percent in Cases 93-9 to 93-12) would likely Network Bottlenecks overwhelm the railway network. Combined shortages of coal and electricity in 2000 if 4.8 The major issue for future coal growth tops 10.5 percent would li-kely delivery is how to move enough coal out of 4. Policy Analysis of Investment Strategies 23 the three key provinces in the enrgy base expaision for reducing energy shortages region-Shanxi, Shaanxi, and western Nei beyond 2000 were the following: Mongol. In all 28 scenarios conducted for (a) Shexian-Handan-Jinan, a new west- this study-even an extremely low (6 percent east line from the southern part of the GNP) demand case (Case 92-3), and an energy base to Shandong Province, is increased transport capacity case (Case 92- now in the railway plan for 2005; 5)-a!l the railway lines leading out of these (b) Shenmu-Yanan-Xian-Baofeng, a new three provinces operate at full capacity (see north-south line from the far western Annex Figures 10.5 and 10.6). The part of the energy base, has beeb maximum amount, about 390 million tons completed as far as Yanan; the per year, is exported from this region in construction of the Yanan-Xian every scenario. From each case, the strong section and the planning of the Xian- conclusion is that transport bottlenecks from Baofeng section should be speeded up these three provinces must be solved because so as to open a new channel out of the rest of China will rely increasingly on the energy base; and them for coal in the future. (c) a dedicated passenger line between Beijing and Shanghai, now being Railway Investment Program actively studied, would free up capacity for coal and other 4-9 Nearly all the railway commodities on the existing Beijing- expansion projects within and leading out of Shanghai line. the energy base region are economically New lines in other regions also play justified and will play important roles in important roles in handling the energy base 3-West coal transport (see Annex Figure coal once it exits the energy base, and also 10.7). The study results were used as a basis coal from other regions (see Annex Figure to support the Govermnent's recent decision 10.7). An important caveat concerning these to accelerate the construction of a major new investment recommendations is that the west-east unit train corridor from the just- analysis includes only one commodity, coal, developing mining area of Shenmu to the albeit the most important one on most lines. new port at Huanghua via Shuoxian. The model looks at a prorated capacity for However, there is some question about coal, depending on the intercommodity whether it is feasible to complete the port allocation, and an equally prorated share of and the last section of the railway by 2000. the total investment cost. For this reason, For this reason, the 9 percent GNP growth the model can answer questions about base case was run both without it (Case 93- whether a new investment is justified in I) and with it (Case 93-2). The results terms of its coal traffic, but the results indicate that completing it by 2000 would should be interpreted in conjunction with a reduce shortages in 2000 by 7 million tons. multicommodity railway investment model Under 9 percent growth, accelerating this like the Railway Investment Study (RIS) project is justified, but not at 7.5 percent model. Ideally, before adopting a final growth. Several years ago, the State Counc:l railway investment program, the planned asked MOR to accelerate its railway capacities could be plugged back into the construction program and complete all the CTS model to explore whether they will lines planned for 1995 one year early. likely be adequate under various 4.10 To accommodate 9 percent circumstances. GNP growth, several major additional 4.11 In recent years, China has con- railway projects were included in the centrated their railway investment more on analysis that were not originally in the 8FYP adding incremental capacity to existing lines or 9FYP when we began this analysis. The than on building new lines, and as a result three most important suggestions for railway has one of the highest densities of traffic per km of track in the world. As planned, China 24 4. Policy Analysis of Investment Strategies should continue expanding the capacity of seriously undersupplied by as much as 25 existing lines by multiple-tracking, diesel million tons of unutilized capacity in the 9 and electric traction, and additional sidings. percent GNP case (93-2) for 2000 because Especially important for coal is heavy haul of expected bottlenecks on the rail segment technology, which is being implemented by coming out of Shenmu. The large capacity MOR with Bank assistance to increase Beijing-Qinhuangdao line may be only about railway line capacity at a lower per-unit half used because of bottlenecks leading into investment cost. However, the benefits of Beijing and local consumption in Beijing. increasing axle loads from 21 tons to 25 tons Likewise, bottlenecks at the ports and per car will not be felt until after 2000. railways en route to northeast China may Increasing existing capacity incrementally, cause underutilization of the interior railway however, is not enough to provide the capacity allocated to coal. Finally, in most needed additional throughput capacity to of the early model runs analyzed in 1991-92, solve the problem of energy base the railway capacity on lines south and east bottlenecks; thus, our emphasis on new line of Jinan were underutilized, which argued construction. Of course, these new lines strongly for addition of a new railway line should also take advantage of the low unit from Handan to Jinan. This line was then cost offered by modem traction systems, added to the model, and has since been unit trains, and heavy axle load approved for construction by 2000. technologies. 4.14 Another possible way to 4.12 With all the planned energy increase coal flows without making base railways plus the possible ones investmnents is to raise the ratio of coal mentioned above expected to be utilized to traffic to total traffic on each link. In the their maximum capacity by 2000 or 2005, a CTS network model, this ratio is specified natural question to ask is whether even more exogenously for each link between 20 and energy base railways would be economically 90 percent (see Armex 15). It is difficult to justified. The answer is yes, but to do so by say whether coal flows or passenger flows 2000 is unrealistic. This study has included are more important for maintaining practically all the railway projects planned economic growth, since coal provides as of early 1993, except for some spurs, energy but passengers represent skills and local railways, and railways that are information. Passenger traffic is generally primarily for other commodities. This considered to be higher value than freight suggests that railway planning has not kept traffic, and the demand for it is increasing up with railway demand. Just as railways even faster than for freight. Used are a bottleneck for the energy system, selectively, the policy of substituting coal planning is a bottleneck for the railway trains for passenger trains might help to system. Many more options should be on the equalize the throughput on parallel railway table at any given time than can be inserted links where one is a bottleneck. However, into the plans. the potential for solving China's coal 4.13 The analysis points out the transport problems in this way is limited strong interdependencies among the various because passenger transport demand already links of the railway and port network (see exceeds supply on most lines. Furthermore, Figures 10.8 and 10.9 in Annex 10). Several to successfully increase the throughput of a existing and planned railway and port coal transport corridor, traffic would have to projects will likely be grossly underutilized be reallocated by a much larger percentage because of upstream and downstream on some segments than on others in order to bottlenecks, although some of these may be increase coal capacity by the same absolute mainly justified by noncoal freight and amount on all segments along the corridor. passenger traffic. An example of less than (Scenarios 924 and 92-5 illustrate the full utilization is the new coal port of futility of changing all lines by the same Huanghua, which is in danger of being percentage amount but different absolute 4. Policy Analysis of Investment Strategies 25 amount.) TMe real issue is not whether to (a) Calorific content of coal should be give up passenger traffic capacity for coal taken into account when deciding on transport, but how much additional capacity origin-destination pairings and routes. to provide for both. Annex Figure 10.12 shows that coal with a greater energy content is Coal Distribution shipped over longer, more expensive routes. It is estimated that the 4.15 The cost minimization that transport cost savings from takes place in the CTS model replicates the optimizing flows of steam coal from behavior of coal buyers seeking the least- the northern energy base according to cost source and shipping route for the right their heat content are around Y 115 kind of coal. Figure 10.10 in Anrnex 10 million per year in the year 2000 (in shows the optimal railway coal flows by link 1990 prices). in the 9 percent GNP case in 2000. Figure (b) The model results indicate that as 10.11 shows the corresponding interregional much coal as possible should be coal flows and the waterway flows. The consumed in the local area in which it predominant pattern is nearly 300 million was produced (see Annex Figure tons of coal shipped from the North China 10.11). Long-distance railway energy base to other regions, supplemented throughput can be aided by supplying by consumption of coal produced within the as much coal from local sources as same multiprovince region. This pattern is possible. quite pronounced: no other region besides Congestion pricing of railway bottleneck North China ships more than 10-15 million corridors would provide an incentive for tons of coal to any other region. buyers and sellers to take these steps. 4.16 Rail flows are predominantly 4.18 Generally, there are three from mines in the west to cities and ports in parties involved in coal distribution: sellers, the east and northeast, and from north to buyers, and transport providers. In China, central. Energy base coal by rail is largely buyers and sellers are becoming increasingly intercepo.d by the middle regions of China autonomous from the central Government, (Henan, Hebei, Shandong, and Hubei) even for state-owned plants. However, before South and East China are reached. because of the transport capacity shortages, Despite getting more coal by rail than the the MOR (rail), the MOC (ports), and China south and east, Central and Northeast China Ocean Shipping Company (COSCO) (ships) can expect shortages because of their lack of have needed to be third parties during any alternative routingsby waterways. The south negotiations. To facilitate this process, the and east rely almost exclusively on water- Government organizes an annual coal borne flows and locally-produced coal, ordering conference to bring all three parties which for the most part can be kept together. This is not so different from successfully off the southern China rail OECD countries, where carriers are network. Waterway flows are dominated by typically included in long-term contracts a huge flow from Qinhuangdao to Shanghai, between buyers and sellers. However, it is and lesser flows from Huanghua and other imperative that the quality of coal be ports to Guangdong, Dalian, and Hangzhou. included in the negotiations, and that an See Chapter 5 for a more detailed discussion information system be developed to aid in of the individual regions. this process. Contracts in the United States 4.17 The coal distributionresults are clearly spell out the latitude for coal type generally consistent with the current variance and the means of testing. direction of coal distribution policy in China. However, the analysis suggests two areas for improvement: 26 4. Policy Analysis of Investment Strategies Rail Tariffs passengers and container shippers appreciate more and are willing to pay more for tihe 4.19 Railway tariffs for freight were services, thus boosting the value of the raised by about 140 percent from 1990 to railway system to the economy. One way to 1993, compared with the GNP deflator at pursue this in the coal transport market 28 percent. Railway fteight tariffs now would be to auction off a segment of railway approximate long-run marginal costs. And capacity to independent service companies yet scarce railway assets are not used to who would provide rapid and reliable coal their maximum efficiency. Rail lines are at transport along some rail or rail-water capacity but the coal is still unwashed, low corridors. This earmarked track capacity in heat content, or weighted down with could be leased to the highest bidders on a water. These are partly the result of competitive basis. This would inject some historically low rail rates. service and price competition into the mostly 4.20 If the MOR were to charge monopolized world of coal transport. congestion prices above long-run marginal costs, it should encourage shippers to Water Transport improve the quality of their coal and switch to attractive waterway options. Congestion 4.23 Waterways are an important pricing should also provide an internal part of the strategy for handling China's incentive, and the revenue that the MOR massive north to south coal movements. For needs, to increase its capacity on bottleneck medium GNP growth, nearly 140 million links by implementing heavy haul tons of coal (not including nearly 25 million technology for rails and wagons, building tons of exports) would be shipped in 2000 lighter-weight wagons, and operating more from north to south, nearly double the unit trains. Congestion pricing is not amount in 1990 (compare Figure 10.11 for uncomnon internationally: railways in the 2000 with Figure 2.6 for 1990). The modal United States typically charge higher rates in split in terms of tonnage handled remains the congested Northeast corridor. close to the 1989 shares of 81-19 (rail- water). Other Policy Rerorms 4.24 Because they haul coal over long distances, waterways play an extremely 4.21 Rail prices should be relied on important role in terms of the nation's total more heavily to allocate resources. tkm. The tkm modal split is nearly 50-50, Currently, about 60 percent of China's compared with 63-37 (rail-water) in 1989. railway capacity is allocated administratively The average distance for coastal shipping to key users, be they centrally or can be increased from the 1989 average of provincially owned or foreign joint ventures, about 1,250 km to nearly 1,500 km. For Because railway transport demand far comparison, the current figure of 1,250 km exceeds supply at present tariff levels, the is less than the 1,275 km distance from mandatory traffic allocation system tends to Qinhuangdao to Shanghai, and half as long encourage wasteful uses of scarce transport as the 2,660 km from Qinlhuangdao to capacity. Reducing this percentage gradually Guangzhou. In the model results, all coal over time would apply competitive pressure shipped by water is shipped at least as far to wasteful customers. south as Shanghai, except for coal shipped 4.22 For other commodities, China to ports in ihe Northeast and on inland is moving towards providing more high- waterways. The increased distance value transport services-for example, fast represents the removal of over 200 billion and reliable container transport and tkm from the railway system, roughly dedicated passenger transport-which equivalent to over seven times the coal furnishes not only 'raw capacity', but traffic carried by the Beijing-Shanghai equality capacity." As a result, rail railway in 1989. 4. Policy Analysis of Investment Strategies 27 4.25 To achieve these gains, ports the flexibility to expand. Yet it is also true and ships require significant investment that shipping is held in check by the inability between 1990 and 2000. In addition to to get coal through to the ports. Assuming projects underway at Qinhuangdao and that the railway to the new port at Huanghua Huanghua, receiving ports in the south and can be completed by 2000 (Cases 93-2 and cast-central regions require substantial new 93-10), shipping investment jumps 20 to 30- capacity. Shanghai will need at least 25 percent. Even so, the port is only 60 percent million tons of new port capacity, while utilized, due to railway bottlenecks en route. Wenzhou, Ningbo, Xiamen, and Yuxikou 4.27 Becse of the economics of need at least 5 million tons each, and many ship size, the use of larger ships is generally other ports need up to 5 million tons each. desirable, but the potential for this is quite Port bottlenecks in the medium and high limited in China due to the geographic demand scenarios are shown in Figures 10.8 conditions at most ports. The advantage of and 10.9. Currently, some coal is larger ships is represented in this study by transhipped at general cargo or container the per unit investment costs, which fall terminals, rather than in specialized bulk from Y 6,000 per dwt for 9-meter ships to facilities that are far more efficient. The use Y 4,300 per dwt for 12-meter ships and to of port facilities built for handling other Y 3,000 per dwt for 14-meter ships. At commodities indicates the desperate need for present, however, 9-meter, 20,000 dwt ships more coal-handling capacity. make up the main part of the bulk shipping 4.26 The shipping fleet, which has fleet. The largest domestically-produced acquisition lead times of one year to five ships in widespread use are 9-meter shallow years, needs to be expanded in both number draft ships rated at 35,000 dwt. An even of ships and the size and types of ships. A bigger 9-meter 50,000 dwt design is under range between 2.4 million and 3.6 million development for production after 2000, but dwt of new fleet capacity will be needed by such a ship would still be smaller than the 2000. These totals do not include another ships used for coal transport in the rest of 200,000 to 300,000 dwt of inland waterway the world, which range from as small as barge capacity. In comparison, the current 35,000 dwt as used on the US Great Lakes, coal shipping fleet is estimated at some to 60,000 dwt Panamax ships, to 300,000 dwt of barge capacity, 2.3 million supercolliers as large a 200,000 dwt used tons of 9-meter general class ships, and over long distances between two very deep 100,000 tons of 12-meter vessels. The future ports. With the exception of Ningbo, which needs for vessel capacity, however, are can handle 12-meter draught ships, ports on uncertain because water transport must China's southern coast are generally 9- handle the overflow from the coal-saturated meters deep or less. None can match the railway system. Thus, investment in new twelve to fourteen meter barbor depth at ships varies by as much as 50 percent from Qinhuangdao, China's main northern loading one scenario to the next (see Annex Table port. 9.5). The lowest level of shipping is forecast 4.28 The decision support system for the 7.5 percent GNP growth scenario was used to analyze several port deepening (Case 93-13), in which shipping is curtailed projects. Several southern ports are not even more than proportionately. The highest nine meters deep. The model results support levels of shipping occurs in a scenario (Case the Government's current plan to deepen 93-12) in which growth was assumed to Wenzhou and Fuzhou harbors to a depth of exceed 10 percent and the shortage cost was 9 meters, and Xiamen's harbor to a depth of raised to Y 350 per ton. With users assumed 12 meters. An option was considered to to be willing to shoulder a higher delivered invest in creating a port at Gaolan, near cost, higher-ost coal and higher-cost Shenzhen, which has some naturally 14- routings come into play; it is mainly the meter deep water. Under this proposal, it waterways rather than the railways that have could handle ships of 100,000 dwt or larger, 28 4. Policy Analysis of Investment Strategies with a hub-and-spokes operation using 9- have waterfront access to navigable meter general class ships for short-distance channels. deliveries to Guangzhou and Shantou. The 4.31 While inland waterway traffic analysis results, however, do not seem to is only about one-sixth the tonnage of favor this option, mainly because of the coastal flows, and averages less than 600 relatively high investmnent, maintenance, and kn, they are nevertheless an imnportant environmental costs. A similar conclusion lifeline to the Central and Eastern was reached regarding the option to dredge regions-one that can potentially be part of Shanghai's harbor to 14 meters. expanded rapidly. In the medium demand Further investigation of these options is case with Huanghua Port finished (93-2), required, however, to obtain a definitive about 22 million tons of coal moves down conclusion, which depends on the scale of (and occasionally up) the Yangtze River. the operation, the changing composition of The Grand Canal, running from Xuzhou to the fleet, transport pricing, subsequent Nanjing, does not appear to be cost- improvements to other ports, and other competitive, or does not have access to a commodities. high quality coal supply. The major 4.29 Internodal competition in recipients of inland waterway traffic are China is still biased toward railways. For cities and power and steel complexes along many origin-destination pairs, rail and water the lower reaches of the Yangtze River-but costs are similar, but rail is priced lower. Shanghai mostly should rely on coastal There are numerous cost add-ons for shipping. The major rail-to-inland waterway combined rail-water routes. In addition to a transshipment ports, from upstream to port's transfer charge, ports charge pollution downstream, are Zhicheng, Wuhan, fees, construction fees, and policy fees. On Yuxikou, and Nanjing. All require new rail-water trips, the state energy- capacity by 2000, especially Yuxikou. Since transportation fund has to be paid these analyses were finished, the Govern- twice-once each for each mode-and ment has announced the plan to build a fifth intermediaries add to the rate charged. major Yangtze River port between Wuhan These add-ons effectively impose a and Zhicheng, where the Beijing-Guangzhou congestion surcharge onto port fees, which railway will cross the river. Because the should encourage longer-distance shipping railway bridge will be a botleneck, a new (see below). When all the legs of a journey port of 5 million ton capacity per year will are added together, typical shipping costs (in be constructed to handle some of the coal 1990 prices) for an all-rail route from flows destined for the south central region. Datong to Shanghai and a rail-water route 4.32 Further decentralization ofport via Qinhuangdao are both around Y 20 per and shipping operations may hasten the ton. ability of waterways to respond to China's 4.30 However, even withthe pricing energy transport problems. Currently, most biases, water transport is much cheaper than coastal shipping operations are centrally- rail over the longest distances. From Datong owned, though a fair number of local to Guangzhou, the cheapest all-rail rate is shipping companies ply the inland Y 36 per ton, compared with Y 26 for a waterways, and some power plant rail-water trip via Qinhuangdao (m 1990 companies, including the huge Huaneng prices). With the pricing and ship sizes as Group, are beginning to operate their own they are, coastal shipping has three main ships. Most ports are owned by the niches. First, it captures the overflow from municipal governments, but the central congested railways. Second, it offers the Govermment (the MOC) approves their long- lowest costs for the long hauls. Third, some run plans. Qinhuangdao is the only port of the users who receive their coal by water wholly owned by the central Government, do so because they have no rail spur but do while Xiamen and Tianjin are wholly decentralized. Greater decentralization of the 4. Policy Analysis of Investment Strategies 29 shipping sector is recommended. planning officials, especially at the Decentralization could not only help relieve provincial level, generally prefer the railway congestion by investing in fleet and flexibility of a multicomnuodity railroad to a port capacity, but it could also spur faster coal-only pipeline. This is different from the adoption of appropriate vessel sizes and United States, where railway rights-of-way technologies, develop superior information are the main barrier to pipeline construction. systems for keeping track of coal movements Second, technologically, electricity and coal qualities, provide better and faster authorities are concerned about dewatering service, and offer some additional inter- and reliability. A test line from Changzhi to modal competition to the state- and local- Jiaozuo may help to alleviate fears, work out owned shipping companies and the MOR. technical problems, and determine costs. Water availability in the energy base region may limit slurry's niche to particular Inports bottleneck areas outside the energy base. 4.33 Coal importation is an Nontransport Sector important strategy, but it was not permintted Policy Implications until recently- While imports are economically feasible, they may not be Coal Production logistically or financially feasible. Currently, provinces are free to import as much coal as 4.35 The study's analysis confirms desired, as long as they have the foreign the current trend of shifting coal production exchange to pay for it. At the moment, coal to the North and West as much as possible, is being imported to reduce shortages and from around 43 percent in 1989 to an satisfy special coal type requirements, but as estimated 52 percent in 2000. Only the lack the reforms become entrenched, importing of railway capacity is keeping it from being will be determined by an economic even higher. To achieve a 52 percent share comparison with domestic coal. Railway by 2000, the energy base area captures over system congestion will likely prevent most 61 percent of new coal mine capacity to be interior cities from receiving imports. built during the 1990s. Given that coal mine Especially well-situated are power plants and investment costs are about 13 to 18 percent other big coal users with ship berths along less in the energy base than in the coastal the coast that can receive coal from self- areas, the westward shift of production unloading foreign vessels. In addition, it is represents a savings of up to Y 4-6 billion mainly the coastal provinces that have the in coal mine investment costs through 2000 foreign exchange reserves necessary for relative to the alternative of maintaining the imports. Because of railway bottlenecks, it existing shares of coal production across is difficult for coal imports to penetrate the rtgions, which could result from the lack of interior of China. Imports could perhaps be railway capacity to transport coal out of the extended up the inland waterway system energy base. using oceangoing barges. 4.36 The shift of coal pioduction to the energy base puts more pressure on the Slurry Pipelines railway system. To move the additional 400 million tons of coal needed by 2000, the 4.34 Slurry pipelines have been average distance of railway movements is discussed for many years, and appear to be estimated to increase from around 520 km in cost effective in scenarios with high demand. 1989 to 640-670 kin in 2000 (Annex Table One pipeline, from Shanxi to Shandong with 9.5). However, the coastal-industrial a capacity of 15 million tons per year, has provinces must stabilize or slightly increase been given the go-ahead as of January 1994. their absolute levels of coal production to Slurry pipelines face several barriers. First, bridge the gap until the tixme when North 30 4. Policy Analysis of Investment Strategies and Northwest China can be relied on even coal. Imposing congestion surcharges on more heavily. Coal in the coastal regions is railway rates would have the effect of too expensive to contribute much additional decreasing the brezk-even distance even coal if GNP growth exceeds 10.5 percent, further. In some regions, however, a small and in fact should be cut back substantially amount of steam coal washing can be if GNP growth is 7.5 percent. The coastal justified mainly by its effect on reducing share of national coal production should bottlenecks, not by line-haul cost savings drop from around 27 percent in 1989 to over long distances. This can be seen in the around 22 percent in 2000, if the energy provinces of Liaoning, Hebei, and Anhui, base railway capacity is adequate. all of which ship washed steam coal over distances much shorter than 1,000 km (see Annex Figure 10.13). Steam Coal Washing 4.40 Overall, it is estimated that the new steam coal washing investment saves 4.37 A n a I y s i s a n d around Y 17.5 billion (in l190 prices, Recommendations. One of the most robust discounted at 12 percent), as compared with conclusions of the study is that steam coal the alternative measures that China would washing should be greatly increased from have to pursue if no new steam coal washing less than 7 percent in 1990 to roughly 16 to plants were allowed (see Annex 18). 19 percent in 2000 (Annex Table 9.2). 4.41 Key Assumptions. All scenarios Network analysis makes a compelling case assume an operating cost of Y 2 per ton and for the quadruple benefits of coal washing: an initial investment cost of Y 30 per ton for (a) reducing transport costs for long-distance steam coal, in 1990 currency (and, for flows; (b) lessening local bottleneck effects; coking coal, Y 4 per ton and Y 50 per ton, (c) reducing ash disposal and boiler respectively). As a test of the model's maintenance costs; and (d) achieving sensitivity to these costs, washing investment environ.ental goals. Steam coal washing is costs were raised 15 percent (in Case 93-5), justified economically even without valuing with the result that the steam coal washing the environmental benefits for a minimum of rate fell by only one-tenth of 1 percent. In 16 percent of all steam coal. For at least 200 addition to transport and environmental million tons of steam coal, the benefits, users of washed steam coal also environmental benefits of washing are benefit from increased combustion efficiency essentially free; it is a win-win situation. due to reduced ash disposal and boiler This conclusion holds true in all scenarios, maintenance. Another Bank report no matter whether demand is high or low, conservatively estimated the savings at about or whether transport capacity or costs are Y 2.50 per ton.' These benefits have been increased or decreased. represented in the model by reducing the 4.38 Somewhere between 16 and carbon loss rate and increasing the beat 19 percent, a threshold is crossed, and the content of washed coal- Steam coal washing additional coal washing is no longer justified is conservatively assumed to produce only purely on a cost reduction basis; it becomes one marketable product: crushed washed a win-pay situation-unless enviromnental steam coal. The rest is considered to be externalities are included in the cost waste. Coking coal washing is assumed to equation. To reduce ash and sulfur content produce a marketable low kcal middlings of coal 10 to 30 percent firther, washing as byproduct that can be burned only in much as 36 percent of steam coal could be minemouth power plants in addition to clean justified (see Annex Table 9.2 and Annex coking coal and waste. Table 12.2). 4.42 Current Status and Pricing. 4.39 The break-even distance for The recommendation that steam coal washing steam coal depends mainly on the washing be increased to 16 to 19 percent is unit transport cost and the ash content of the a major departure from the current level of 4. Policy Analysis of Investment Strategies 31 O to 7 percent. Coal washing has been coal. The idea of discriminating between a stalled for many years, and virtually no variety of coal choices will require a change plants in China have been designed and built in the mindset of coal users accustomed to a expressly for washing steam coal. Special take it or leave it system of coal allocation. cases appear to account for most of the With the economy overheating, users often approximately 60 million tons washed, must take what they can get. It is estimated which explains why estimates of current that tens of millions of tons of coal wastes, washing rates range from 0 to 7 percent. gangue, or poor quality coal were mixed For instance, excess coking coal washing into the washed coal at mines. ports, or capacity in Anhui is now used to wash low railway stations, and customers had no quality coking coal for power plants. In choice but to pay the higher price for the Chongqing, some anthracite is washed for lower quality coal. use as steam coal in power plants that are 4.44 New Guiderines A recent required by municipal law to limit their nonmandatory proposal formulated by average sulfur content to 2.5 percent. In the CIECC calls for all new large (greater than Northeast, some high ash bituminous steam 2.4 million tons per year) Govermnent- coal is washed because the power plant owned coal mines to be asked to build boilers do not accept coal with greater than washing plants. This new policy is targeted 30 percent ash. Also, about 15 million tons at the large-scale mines because they can of steam coal produced for export in a benefit the most; because their automated China-United States joint venture is washed. machinery generally creates a very high SPC experts have indicated that domestic percentage of gangue; because large state sources of funds could reasonably be mines have access to the long-distance expected to raise the percentage of coal railway capacity; and because locally-owned washing to 16 percent by 2000. mines could never afford the capital 4.43 Under central planning, equipment. The new po!icy urging coal producers were allowed to charge at least a washing is somewhat of a blunt policy 10 percent differential for washed steam instrument, because it is not dependent on coal, and upwards of 20 percent for washed distance and coal quality. However, coupled coking coal. While the price of in-plan raw with coal, rail, and electicity price reform, coal was generally below long-run marginal it is believed to be an attempt to jump-start costs, the washing price differentials were the industry and to break the cycle whereby enough to cover the incremental washing producers claim there is not enough demand costs, debt service, and a small profit. The for washed coal while consumers complain problem was not too low a price, but not of the lack of a reliable supply. In the enough incentives to convince coal medium term, is new policy may have the consumers to pay the higher price on a free effect of creating an oversupply. In the long market basis. Proper pricing of washed coal run. it would be more efficient to provide an also requires setting suitably low prices for economic framework which will send a the byproducts of washeries-middlings, strong stimulus to consumers to demand coal fines, and sludge-which are washed coal. problematic for users due to the difficulty of 4.45 Imnplementarion Issues. While handling, high ash or water content, low the benefits of steam coal washing may be kcal content, or all three. Potential buyers obvious, proper incentives for change are include regular or advanced power plants, missing. Incentives must make the benefits briquette makers, or the building materials of washing transparent to and shared by all industry. It will undoubtedly take some time actors in the supply-transport-consumption for the Chinese coal market to properly chain. Also, once they mutually agree to price the products, let alone the byproducts, produce and consume washed coal, standard, of coal washing, given that the market does well-recognized procedures for preparing not differentiate well between grades of raw and enforcing long-term contracts between 32 4. Policy Analysis of Investment Strategies these different actors is critical to eliminate existing factories and power plants may not any transaction risk. The incentives and fully achieve the increased combustion disincentives for internalizing each of the efficiency benefits and may not be willing to four categories of benefits (environmental, pay very much more for washed coal. New transport cost, transport capacity, and boiler boilers, which can be designed to take better operation) are addressed below, in turn. advantage of a higher kcal coal feed, may 4-46 In most countries, steam coal is represent a better marketing opportunity for washed as an inexpensive way to meet coal washeries. Use of new boiler environmental regulations on sulfur and ash technologies will promote greater use of pollution. The environmental benefit is washed coal. greatest for small boilers and stoves, which 4.49 In summary, the potential often may have poor, if any, emissions benefits from increased washing of steam control equipment. Adopting a -polluter coal are great. Enactment and enforcement pays" policy-whether based on pollution of some kind of polluter-pays laws will taxes or tradeable permits (see Envirornent, provide the strongest incentive for users to paragraph 4.63)-would be the single care about what kind of coal they burn. This strongest incentive to producers to wash in turn will drive the supply and distribution steam coal willingly. The need to enforce sectors. Freeing coal and electricity prices to whatever regulatory mechanism is adopted is seek their own levels, and pricing railway a key for China, where many envirorunental services to include congestion surcharges, rules, including some pollution taxes, are will help to make the benefits of coal already on the books. For instance, it has washing more transparent. Establishing a been reported that some power plants in legal framework for long-term contracting Chongqing refuse tD buy the aforementioned will lower the risk entailed by buyers and washed anthracite coal because the sellers who conmmit capital to producing or 2.5 percent sulfur regulation is not enforced. burning washed coal. If all these things are put in place, the new Government policy 4.47 Of the two transport benefits of urging coal washing for large mines may washing steam coal, only line-haul cost help to jumpstart the process. reduction exerts a transparent incentive to wash coal, and this would be strengthened Hydropower and Nuclear Power by congestion pricing of rail bottlenecks. The transport capacity-saving benefits are 4.50 The study's analysis of the not seen or felt by the coal producers or hydropower sector (Annex Table 9.3) is consumers. In buying higher-priced washed consistent with the high end of the range in coal, users inadvertently free up extra the FYPs, which call for 63-69 GW of total railway capacity, which will benefit other capacity to be built by 2000. In the event of railway customers, but not themselves. As 10 percent or higher GNP growth, or strong things now stand, with administrative efforts to reduce pollution, hydropower allocation of railway capacity, factory construction should top 70 GW. However, managers may think to themselves, "Why hydropower's share of capacity, currently wash coal when the MOR has allocated 24.4 percent, might drop slightly to around enough capacity to us to transport all the 22 percent with 9 percent GNP growth, or raw coal we need?" If coal users had to sign even to 21 percent, given double-digit a long-term contract with the railways that growth. The time lag in constructing a included a separate, itemized railway hydropower plant is a discouraging factor, capacity fee, there could be a stronger as is the lack of inexpensive hydropower. incentive to wash coal. 4.51 Nuclear power does not appear 4.4B Most existing boilers in China to be cost-effective in the study's analysis, were designed for raw coal with lower ba:;ed on 1990 investnent cost estimates calorific value than washed coal. Therefore, (Y 6,000 per kw for domestic construction, 4. Policy Analysis of Investment Strategies 33 Y 10,000 per kw for imported tecnmology). east, southeast, and south. In particular, Bringing down the investment costs for a these include four main directions: (a) from new generation of nudear plants might help the north of Shanxi to the Beijing-Tianjin- improve this situation. This alternative could Tangshan grid; (b) from the middle of be analyzed with the CTS model, if future Shanxi to Shandong province via the middle cost estimates were available. of Hebei province; (c) from southeast Sbanxi 4.52 The conclusions about hydro- to Jiangsu province and Shanghai; and (d) power above ignore the other benefits of from southern Shanxi to the Huazhong grid dams, such as flood control and irrigation. (Hubei, Henan, andJiangxi provinces. Other Similarly, the conclusions about both hydro- major flows include (e) from the eastern part power and nuclear power ignore the issue of of Nei Monggol (Tongliao, Yuanbaoshan, carbon dioxide reductions, which might have and Yinin River) to the Northeast put them both in a more favorable light. provinces; (f) from the hydropower plants in Whereas coal washing or switching to higher Guizhou and Guangxi provinces to quality coals can reduce ash and sulfur but Guangdong province; and (g) from thermal no carbon dioxide, hydropower and nuclear plants in Shaanxi to Huazhong and northern power completely eliminate all three. Sichuan province (see Annex Figures 10.14 and 10.15). What is particularly noticeable Minemouth Power and Electricity in Annex Table 9.4 is that the electricity Transmission flows from Shanxi and from hydropower plants in Guizhou and Guangxi are fairly 4.53 Analysis and Recommen- stable across all scenarios, while the flows dations. The analysis endorses an increase in from Nei Mongol and to the Huadong and minemouth power generation with Central regions are sensitive to demand transmission to areas suffering from forecast. This result suggests that transport bottlenecks. North China'sshare of transmission from Shanxi and Guizhou- generated electricity could be economincally Guangxi will be a good investment under increased from 17.8 percent in 1989 to most circumstances. On the other hand, Nei roughly 20 percent in 2000. There are five Mongol seems to be the marginal supplier, reasons for the recent strong interest in responding positively to increased electricity power transmission. First and foremost, demand, and negatively to increased bottlenecks on the railway system make it electricity shortage costs and increased impossible for railways to satisfy the available railway capacity. demand. Second, the lead time for building 4.55 Overall, it is estimated that the new transmission lines is much shorter than seven recommended new 500-kV intergrid for railways. Even though short-distance transmission lines could save China around railways are presumably competitive with Y 7 billion in discounted costs (in 1990 transmission, the issue of lead time has prices), as compared with not building such become of paramount importance to urban transmission lines, as a result of which officials. Third, transmission can be cheaper urban power authorities would have to turn than railways in mountainous regions, where to other alternatives (see Annex 18). Each towers and wire are much easier to build 500-kV line substitutes for about 10 million than railways. Fourth, the environmental tons per year of railway traffic. Construction impact of coal combustion can be shifted of these lines, which provide an alternative away from the highly populated urban areas. way to move energy long-distances from the Fifth, the local governments administering energy base, may partly account for the the coal base have taken the initiative, underutilization of port capacity at realizing that the more power plants they Huanghua and Liangyungang in the model build, the more profit they can make. results, as compared to the port forecasts. 4.54 The first set of major intergrid 4.56 Current Statsw. Numerous flows are from Shanxi Province to points actual proposals are under consideration for 34 4. Policy Analysis of Investment Strategies substantially increasing long-distance when the plant and line are finished, an transmission. In South China, hydropower incentive will exist to sell 100 percent of the projects are being considered in Yunnan, power to the local area rather than suffer the Guangxi, and Guizhou in the 2,000 to 4,000 line losses and sell only 90-some percent of MW range, which would provide 50 percent the power in the destination region. This or more of their output to Guangdong, up to risk is real, but can be handled with long- 1,500 km away. In North China, at least term legal contracts. Enforceable agreements 5,000 MW of minemouth thermal power is and an adequate legal framework will be being planned for transmission to the necessary for load center provinces to have Beijing-Tianjin area alone. Some of the enough trust to enter into such planned transmission lines woulA run from arrangements. However, this risk is minimal (a) Western Mongolia to Beijing; (b) Eastern for thermal power plants in the energy base, Mongolia to Northeast China; (c) Datong to where electricity shortages are not expected. Tianjin; (d) Changzhi to Jiangsu; (e) from 4.58 A second implementation issue mid-Shanxi to Shandong; (f) Shaanxi to is ownership. As described in Chapter 3, the Sichuan; and (g) Guangxi, Guizhou, and new model for infrastructure funding in Yunnan to Guangzhou. The CTS analysis China, especially for power plants, is the pie may have been a contributing factor in chart. Local, provincial, and central supports of these plans, in that at least some Govermnents, banks, and foreign partners capacity is built on all of these planned lines each contribute a certain percentage of the by the year 2000 in the 9 percent GNP base funding, with rights to a certain portion of case. The analysis also recommends the electrical load. However, this method of (h) Datong to Beijing; (i) Taiyuan to project financing is problematic when it Shijiazhuang and Jinan; (}) Changzhi to comes to transmission lines, which must Zhangzhou, Wuhan, Guangzhou, and function as part of an integrated electrical Huaiyin in Jiangsu; and (k) Yimin River to grid with complex functions dealing with Harbin and Shenyang (for lignite-generated load sharing and power outages. China's power). five big regional grids and nine provincial 4.57 Implementation Issues. In past grids are directly or indirectly under the decades, many structural (interprovincial or control of the MOEP. The problems posed intersectoral) and technological barriers have by tying new lines owned by pairs of delayed and discouraged long-distance power provinces into the centrally-controlled grids transmission and favored load center prompted the Government in early 1994 to generation fed by railway transport of coal. decide that all transmission lines, with a few First, one of the risks in developing exceptions, are to be owned and controlled transmission projects is that there will not be by the central Govermnent. Details any spare electricity to transmit as expected. regarding how coastal provinces will go This has happened before with a long- about co-financing such lines remain unclear dist&ace line from Wuhan to Shanghai, and at this time. has scared many load center officials away 4.59 Third, several incentives have from taking such a risk. Interestingly, this for many years encouraged provincial and phenomenon can be detected in the model electricity ministry authorities to build plants results. Transmission to the Central and in the load center provinces. Provincial Huadong regions is surprisingly less in the officiais favor local power plants that create high demand scenarios than in medium employment, taxes, and technology demand scenarios. This counterintuitive spillovers to the local area. (The penchant result can be explained by the fact that faster for self-sufficiency in electricity production growth would cause more electricity to be is evident in Map IBRD 26596). Electricity intercepted by cities in the north, leaving sector officials have historically preferred less available for intergrid transfers. If the load center power plants because the shortage in the origin region still exists investment for transportation is paid by the 4. Policy Analysis of Investment Strategies 35 MOR, whereas with a minemouth plant, the potential for energy conservation is in the electricity sector must absorb the investment neighborhood of 100 million tons. This cost for transportation. Finally, railways are estimate is based on the amount of demand more attractive to provincial officials than that cannot be satisfied at a delivered cost of transmission lines (or slurry pipelines, for Y 300 per ton. It is argued here that if the that matter) because they can haul other shortage cost in the model (Y 300 per ton) goods as well. is greater than the amortized conservation 4.60 Fourth, there have been investment cost, the shortages can be technological barriers as well. The DC line inte,preted as energy conservation potential. technology needed for distances over 1,000 Assuming a discount rate of 12 percent and km is new to China. However, one 500-kV a 10-year lifetime for conservation invest- DC line is already in operation from a ments, an annual amortized cost of Y 300 Yangtze River hydropower station in Hubei per ton is the equivalent of a financial to Shanghai. While it has encountered some investment cost of Y 1,700. Since the technical problems, others are still being shortages represent coal that cannot be planned. The study's analysis includes a delivered for under Y 300 per ton, investing 1,400 km 500-kV line from Changzhi to in energy conservation at a cost of Y 1,700 Guangdong, which is at the upper end of the per ton is justified. In comparison, SPC feasible length. Another technical difficulty experts report that a wide range of lies in the fact that, while most of the conservation investments exist for as low as regional grids use the same frequency and Y 700 per ton. Of course, once those Y 700 voltage, electricity shortages have per ton conservation investment destabilized the frequencies, making it opportunities are exhausted, the marginal difficult to phase them together. In the short cost would rise, and there is no guarantee term, one option is to connect the load that all 100 million tons' worth of energy center grids with single minemnouth or could be saved by investments of less than hydropower plants in the supply region, Y 1,700 per ton. rather than connecting the entire supply 4.62 Analysis with an expanded region grid to the entire load center grid. CTS model seems to confirm the above lTe single plail in the supply region would conclusions (see "Studying the Policy of function as a remote part of the demand Improving Energy Use Efficiency to Relieve region's grid. However, as the shortages Energy Shortages," Robert S. McNamara disappear, the stability of the grids will Fellowship Program, September 1994). This improve, and the long-distance lines now research should be considered preliminary being planned as will eventually become key because it was an individual effort rather links in a national grid. than an official Government-supported team effort. Real data on the costs, efficiency, Energy Conservation and energy-saving potential of investments in improved coal and electricity devices 4.61 This study did not explicitly were collected and added to the energy and independently consider energy supply and delivery options in the official conservation investments (see "China CTS model. The preliminary analysis Energy Cc-nservation Study," World Bank concludes that a combination of investments Report No. CHA-10813), but instead in energy production and energy interpreted the shortage cost in terms of the conservation can satisfy energy requirements comprehensive cost of energy conservation more economically, with fewer shortages in a post hoc analysis. The study estimates and less pollution, than a system based on that energy conservation could play a major energy production investments only. role in satisfying coal and electricity Improving energy efficiency (beyond what is demands in 2000, given double-digit GNP assumed in the demand forecast) will lower growth. In fact, a rough estimate of the systemwide costs by 5 to 10 percent 36 4. Policy Analysis of Investment Strategies compared with the CTS results without monitoring of emissions. One way around conservation variables. However, energy this is to charge presumptive taxes, placing conservation would require more upfront the burden on the enterprise to prove that capital investments. Half of the required they are polluting less than presumed. investment could come from shifting Another, much simpler method, is to charge investments from the production sectors, but taxes on polluting inputs, but this method the other half would have to come from does not send as clear a signal to finms and some other source besides coal and is therefore less efficient. In addition, with electricity supply funds. tradeable permits, a second information system is needed to keep track of who owns Enviromnent them. 4.65 Evidence from the United 4.63 As China's coal production States suggests that such programs can continues to break its own record year after succeed in cutting the costs of cleaner air. year, environmental conditions will worsen. The 1990 Clean Air Act established a Unconstrained by Government policy, the Government auction of tradeable allowances national total of sulfur in the coal that give the holder the right to emit a ton of (potentially emitted) can be expected to SO, per year. When the bill was being increase from around 14 million tons per debated, the industry warned that the cost year, the 1990 total, to nearly 20 million could be as high as US$1,500 per ton. The tons by 2000 (see Annex Frire 10.16). government estimated the cost more Particulate pollution, a resa: of the ash conservatively at US$600 per ton. But when content of coal, poses a more severe health the auction was held, the market set a much problem than acid rain. The total national lower price of US$150. The cleanup is ash content will increase from approximately expected to total 22 million tons in 250 million tons in 1990 to about 300 2000-considerably more than the 16 million million tons in 2000 (see Annex Figure tons required under law. Compared with the 10.17). govemment's estimate, the annual savings 4.64 To avoid this kind of are almost $10 billion. catastrophic situation, the Chinese 4.66 Much information is needed to Govenment should consider incentive-based go ahead with any of these proposals. To regulatory mechanisms that make enterprises impose a pollution tax, the social cost of a or individuals pay for the envirownental unit of pollution must be estimated. For damage they cause. Pollution taxes or tradeable permits or standards, the amount tradeable emissions permits are preferable to of permits must be decided on. For any of more blunt mechanisms such as emissions them, it is useful to have some data on what standards or mandated use of the "best level of pollution control is economically available technology." The incentive-based feasible. The results of this study can measures will generally achieve the same provide partial answers to some of these geoas at a lower cost by internalizing the information needs. costs and enabling those energy users who 4.67 Finding optimal environmental can reduce pollution more efficiently to do strategies was not the driving force behind so, thus lessening the negative impact on the this study. But since it is a coal transport entire economy. Taxes and permits also study, it is important to show quantitatively, raise revenues for the Government in a less in some fashion, the fact that some distortionary way than most taxes. investment strategies, such as coal washing Compared with pollution standards, taxes or and hydropower, have more enviromnental permits have the advantage of discouraging benefits than others. Estimating the social firms to get rid of all their polluion, not just costs of pollution was clearly beyond the the pollution above a certain level. Taxes, scope of this study. Instead, this study permits, and standards all require effective adopted the simple, concrete measurement of 4. Policy Analysis of Investment Strategies 37 the level of ash and sulfur content in the some users to pollute a lot while others who delivered coal. The methodology for can afford to do so make deep cuts. The analyzing the environment is explained results show the tradeoffs between further in Annex 5. noncommensurate economic and 4.68 The model was used to environmental goals and allows the estimate how much it would cost to reduce policymakers to apply their own valuations, by 10, 20, and 30 percent the ash and sulfuir rather than forcing them to accept a given content of the delivered coal to each estimate of the externalities. province. The results suggest that it is 4.70 However, the model economically feasible to impose some kind overestimates the cost of meeting the of polluter-pays principle in China. Annex environmental goals, since the model lacks Figure 10.18 shows the accelerating tradeoff several potentially inexpensive ways of curve between the conflicting goals of cost redutcing ash or sulfur emissions. Other than and ash-sulfur control, and demonstrates that incluo ing scrubbers, the model only analyzes the conflict is not as great as is generally measures that reduce the ash and sulfur believed, at least for small improvements. going into the boilers and stoves. The Tne first 10 percent of ash and sulfur unmodeled alternatives include energy reductions can be achieved with a relatively conservation, fabric filters, electrostatic small cost increase of about 3 percent. The precipitators, coal briquettes, fluidized bed cheapest strategy for doing so by 2000 combustion, coal gasification, taller stacks, involves building an extra 1-2 GW of new and coal screening. The tradeoff curve thus hydropower capacity above and beyond the represents an upper bound on the true costs. base case scenario. increasing scrubbers A more refined analysis is needed to see from none to almost 1 million ton of how much less expensive it might be to capacity, and increasing steam coal washing reduce ash and sulfur emissions at the end of from 200 million to 350 million tons, the the smokestack. latter spread liberally over many provinces 4.71 In particular, the most (see Annex Table 9.6 and Annex Figure important missing alternative is energy 10.13). Ash and sulfur reductions beyond conservation, which can reduce energy that point are increasingly expensive to demand (and therefore, in a sense, satisfy it) achieve, mainly because it necessitates with no pollution and no railway transport, increased use of scrubbers, hydropower, and, which is likely to cost less on an coal washing, and nuclear power, and amortized basis than the shortage cost because it might cause increased shortages. (Y 300 per ton, in 1990 prices). Given that The second 10 percent of reductions will the model willingly pays the Y 300 shortage drive costs up by another 7 to 8 percent, cost for over 128 million tons in the while the third 10 percent reduction adds 20 percent ash and sulfur reduction scenario almost 10 percent. For reference purposes, (Case 934), there appears to be a large in order to keep China's total ash and salfur scope for energy conservation investment at 1990 levels despite the growth of total under a tradeable permit system, or coal production would require about equivalent taxation system, that forces at 25 percent less ash and sulfur than in the least a 20 percent reduction. cost-minimizing solution. (For coal washing 4.72 From another perspective, a alone to keep pollution levels constant, about value of this tradeoff analysis is to 35 percent of coal would have to be demonstrate to World Bank and SPC policy washed.) makers that if investment strategies that are 4.69 This analysis has its not necessarily the absolute cheapest are advantages. First, by applying the ash and researched, alteinatives that are only slightly sulfur constraints at the provincial level, the more expenrive but have major positive model realistically represents the way that impacts on the environment may be found. pollution taxes or tradeable permits allow HydrDpower, steam coal washing, and 38 4. Policy Analysis of Investment Strategies er.ergy oonservation, in the right time, is a significant overlap in the strategies used place, and amount, can accomplish for meeting high demands and those for environmental and energy supply goals meeting environmental goals. simultaneously. Also important is that there 39 5. REGIONAL ANALYsIS Northeast China in the Northeast may have to be increased in absolute amount, despite the high cost of 5.1 The Northeast region of China developing the remaining reserves in this suffered severe coal shortages in 1988-89, region. Third, Hailar and Tongliao have and the shortages returned in 1992-93. abundant and cheap reserves of brown coal, Severe coal shortages for this area can be which should be developed and burned in expected in conditions of rapid growth, new minemouth power plants. The resulting enviromnental contcrls, or restricted power should be transmitted to Harbin, transport capacity, or some combination of Shenyang, and Changchun by new the three. Nortbeast China has substantial transmission lines. However, these measures coal mining capacity of its own, but not will likely not be enough if the economy enough to satisfy all its needs. It has grows faster than 10percent per year. significant reserves of coking coal, but the Nearly 30 percent of the country's shortages current energy supply policy is that the may occur in the Northeast, with electricity Northeast should not use its coking coal for shortages of arornd 10 percent of demand. power generation. Therefore, the Northeast Energy conservation should be pursued must supplement its own suDply potential by vigorously in this region. drawing coal from "inside the gate,' that is, from south oi the Great Wall. Three main Central China routes "through the gate" will all be saturated by 2000: the Qinhuangdao- 5.3 Conditions are ripe for Shenyang line; the Jining-Tongliao line; or shortages of coal and electicity in the by ship from Qinhuangdao to the ports at Central region, especially in the "Double Dalian, Yingkou, and Dandong. Hu' region of Hunan and Hubei Provinces, Furthermore, the strategy of increasing home to cities like Wuhan, Changsha, and railway capacity on these three routes cannot Nanchang. Prior to the completion of the by itself solve the problem of shortages in Three Gorges project, local sources of coal the Northeast because bottlenecks from the will fall short of demand, and the transport energy base area will prevent coal from network will continue to be congested. reaching the Qinhuangdao and Jining. Nor However, because this region is centrally can coal imports solve the problem, since located, it has many energy supply options they are feasible only for the coastal port of from other regions, and with the exception Dalian-not for most interior cities-because of the Wuhan area, shortages in most of congestion within the Northeast itself. scenarios are not severe. 5.2 ExpansionoftheHailar-Harbin 5.4 A combination of strategies is railway and use of heavy haul tenology needed to keep shortages to a miimum. through the gate are foremost among the First, a new railway from Shenmu in the recommendations. In addition, several non- coal base to Xiangfan, near Wuhan, would railway strategies are recommended. First, help accommodate growth after 2000. the model suggests that a fair amount of the Second, coal production in Henan Province, steam coal from Shanxi Province to the officially part of the Central region, should Northeast be washed prior to shipment, thus be increased. Production near Zhengzhou, maximizing the energy inflow on the three Yima, Pingdingshan, and Xinuiang should capacitated routes. Second, coal production be stepped up. Third, Henan, Shaanxi, and 40 5. Regional Analysis southern Shanxi provinces have the Jiangsu and Guangdong) are being planned. opporhtuity to transmit a great deal of mine- Gas has the advantages of little to no SO% or mouth electricity to Hubei and Hunan. TSP pollution, short construction time, and Fourth, if Sichuan hydropower plants are lower investment cost than nuclear. built up quickly. Sichuan can supply However, domestic gas reserves are minimal electricity to the Central region in the and aimed mainly at household use, and interim before the Three Gorges is finished. LNG imports by water are exceedingly Once completed, the Three Gorges can hazardous. Currently, oil and gas make up transmit a great deal of electricity to Sichuan less than 2 percent of thennal power plant province. capacity. Eastern China Southeastern Coastal China 5.5 Although the Yangtze delta 5.8 The southeastem coastal area area suffered severe coal shortages in 1987- shares some of the logistical problems faced 88, it is anticipated to be in less danger of by East China, but its industrial development coal shortages than Northeast China. More is more concentrated along the coast. than half its coal supply comes from self- Because of the long distance from the energy supply, and the vast majority of the base and the intervening needs of the central remainder comes from the energy base region, East China receives only a small part region by water transport, an option that of its coal by rail. Instead, it relies mainly was much less available during 1987-88 on water shipments and locally-produced before the completion of Qinhuangdao port coal. A new proposal for delivering coal to and the DaQing rail line serving it. this region (not reflected in the study's Landlocked eastern cities, such as Jinan and analysis) is to ship coal from Guizhou IHefei, are in greatest danger. In recognition province by rail to Nanning, and then by of this, construction of the Handan-linan inland waterway to Guangzhou by river. railway, which is in MOR's long-term plan, Some of the coal headed to Guangzhou, should be speeded up. Xiainen, Fuzhou, and Haikou should be 5.6 This region, however, is quite washed prior to shipment. susceptible to electricity shortages. 5.9 For Guangzhou, the major new Electricity transmission from Changzhi to dam and hydropower plant at Longtan, and Shanghai can play a major role, as can other electricity transmission from Changzhi to lines from the energy base to the interior Guangzhou (technically feasible by 500 kV cities of Jinan and Xuzhou. In part because DC line), can significantly improve the of the transmission options included in the electricity balance of the southeastern area. analysis, no new nuclear power capacity No new nuclear power capacity would be needs to be built in Shanghai, even in the built in Guangzhou other than what is high-demand and high shortage cost already under construction. scenarios. 5.7 This area seems particularly Southwest Energy Base vulnerable to shortages under stricter environmental regulations because of its 5.10 The Southwest energy base, poor coal quality. Coal imports will be able particularly Sichuan and Guizhou Provinces, to help the coastal cities in this region, and is mostly used for self-supply in the the three slurry pipelines on the drawing Southwest region. The region is somewhat boards, one of which has been approved, are isolated from the rest of China in terms of targeted here. In addition, some oil-burning transport and energy. Hardly any coal is power plants are being built along the coast exported to other regions in most scenarios, to take advantage of currently cheap as demand in Chengdu, Chongqing, imported oil, and two gas-fired plants (in Kumning, and Guiyang soaks up nearly all 5. Regional Analysis 41 the locally-produced coal. Sichuan should Some lignite reserves near the Sichuan- export some hydropower to the Central Yunnan border could be exploited, but at a region, and Guiyang should import some high cost to the environment. A proposed oit coal from Guangxi. Isolation contributes to pipeline from Xinjiang to Sichuan is also the region's being structrally prone to under consideration. Otherwise, they will substantial future coal shortages. While probably have to rely on expensive some regions, like East China, have large hydropower options. shortages in some scenarios, but none in others, the Southwest area can expect Northwest China shortages approaching 10 percent of demand in all high-demand or envirommental control 5.12 The northwest provinces of cases, as well as in some others. Despite the Xinjiang, Qinghai, Gansu, and Ningxia are fact that, in most scenarios, this region blessed with abundant coal and hydro generates about 20 OW-more than resources. In the present and future, this 60 percent of its electricity-from hydro, it region can continue to be mostly self- simply lacks much flexibility to adapt once sufficient in coal, and can ship some of its the reasonably affordable hydropower surplus to nearby regions, although long projects are gone. distances and intervening opportnities 5.11 To handle growth of 10 percent (Shanxi) limit the scope for exports. Some or more, the Southwest will have to rely coal can be supplied to northern Sichuan manly on energy conservation and on Province and the middle region of China hydropower plants in Sichuan, such as the Likewise, some thermal power can be Three Gorges. Further proof of its self- transmitted to the Huazhong grid in central reliance is that it is not expected to have any China, and some hydropower from the surplus electricity to trsmit to other upper Yellow River can be shipped to the regions in 2000 (pre-Three Gorges), even Huabei (north central) grid. under 9 percent GNP growth conditions. 43 6. PAsr AND FuTURE PoLIcY IMPLICATIONS Policy Recommendations or US$140 billion (both in 1990 prices) for and Impacts of this Study the 8FYP and 9FYP, and around US$100 billion more for the lOFYP. These figures 6.1 The analysis system and some include coal production, coal transport (not preliminary results were introduced to SPC including other commodities' share of leaders, planners, and researchers in a major transport investment), and electricity (not conference in December 1991, which was including local distribution). covered by Chinese television (see Annex 6.4 Some of the recommendations 19). Following this conference, SPC experts of the study are consistent with national worked closely with the modeling team to policy and trends over the last 5 to 10 years, improve the realism of the results, thus and some are consistent with (and may have building confidence in the tool within the influenced) some recent dramatic policy planning community and among Chinese shifts, while others go beyond even the policy makers. recent moves. There is a lot of gray area in 6.2 Then, after Deng Xiaoping's trying to sort the recommendations into much publicized tour of Guangdong in 1992 these three categories, but it is useful and the subsequent decision to accelerate nonetheless for the purposes of highlighting economic growth, three of the SPC's where China has come from, where it is planning bureaus (Energy, Transportation, now, and where it should be going. Among and Long-Term Planning) asked the CaS the policy recommendations in this first team to analyze whether and how China category, that is, those that confirm the could supply the additional energy. Within a trends and policy directions of the last 5 to week, the analysis systen was able to 10 years, are the following: answer the SPC's questions related to the (a) the shift of coal mining to the logistical problems and the potential for norhern energy base; shortages related to faster growth. On the (b) increased use of long-distance water basis of these results and other evidence, the transport, construction of mine-to- National People's Congress adopted a port railways, such as the DaQing relatively high but sustainable growth rate of line, and domestic production of 8 to 9 percent. larger ships; 6.3 Following this, the ClIS has (c) continued development of been used to identify new investment options hydropower, including in remote for surpassing 8 to 9 percent growth rates. areas; After studying the logistics, CI5 team (d) increased efforts at energy members and Chinese experts made conservation; and suggestions for new energy and transport (e) no heavy investment in the short term project ideas. These were input into the in more than a few commercial analysis system, which then showed how the nuclear power plants. new projects could nearly satisfy the energy 6.5 Ihe second category of study demands of 9 percent GNP growth. These recommendations have confirmed several results were presented at a joint Chinese- major recent trends and policy changes, World Bank conference in Beijing in which were announced concurrently to the October 1993, and were received by senior last stages of this study, that is, in 1992-93. Government officials as helpful and realistic The CTS results may have been one of (see Annex 20). Overall, the amount of new many factors that contributed to these investment required is about Y 650 billion policies and tends: 44 6. Past and Future Policy Implications of the CTS (a) The Y 30 billion complex consisting although other tentative findings of the Shenmu coal field, connecting strongly suggest that energy railway, and the Huanghua port has conservation is an even better way to been approved, and financing it has prevent shortages. become a national priority. The CTS (f) The first commercial-scale slurry team's analysis confirmed the urgency pipeline was approved in China in of this action, though it was December 1993. In the CTS analysis, originally thought impossibleto finish slurry pipelines were always chosen it by 2000. in the high-demand scenarios in (b) New railway lines from Handan to which they were considered. Jinan have been inserted into the 6.6 While post-1990 planning of plans, and the line from Shenmu to the coal-electricity delivery system has made Xian has begun construction. Without some dramatic steps in the right direction, a the Handan-Jinan line, the coal number of recommendations from this study supply to Shandong's port of Qingdao have not yet been officially acted on. These will dry up. Even with these constitute the third category of railways, the model finds it necessary recommendations. Some of them are to supplement theem with new believed to be under active consideration by transmission lines to Iinan and the Government The following Harbin. Similarly, the Shenmu-Xian recommendations should be accelerated in line could become a major coal the SFYP and 9FYP: crier as soon as it can be (a) Coal should be distributed according completed. When the crs analysis to its heat content. High kcal coal began in 1991, these lines were not should be shipped over the longest considered necessary to include in the distances to get the most out of the SFYP or 9FYP- railway and port capacity. Congestion (c) All new state-owned coal mines that pricing and improved information ship coal to other provinces are now systems are instruments that should encouraged to build coal-washing hasten this development. facilities. SPC planners think that (b) While the long-term concept of a 15 percent or more of steam coal national grid has been announced, could realistically be washed by very few long-distance transmission 2000, a share which is slightly less lines have been actually inserted into than recommended in the the plans. Because they are easier to nonenvironmental scenarios. plan and faster to construct than (d) The Government recently announced railways, transmission lines can play that a national electrical grid will be a key role in preventing massive created. Long-distance intergrid shortages in the short term. Some of transmission is a key element of this the more promising suggestions from study's recommended strategy. the analysis include those from Hailar (e) The coastal areas have been opened to Shenyang and Harbin; from up to coal imports; if the users can Datong to Beijing and Tianjin; from pay with foreign exchange, they can Taiyuan to Shijiazhuang and Jinan; gain their own access. In this study, from Zhengzhou to Wuhan; and from between 30 million and 140 million Changzhi to Zbengzhou, Huaiyin in tons of coal imports were fbund to be Jiangsu, and Guangzhou. cheaper than domestically-supplied (c) Steam coal washing can play an coal in 2000, depending on the future important role in regions other than assumptions about GNP, the the northern and southwest coal environment, and other factors. Thus, bases. In some provinces near the imports can prevent many shortages, coast, such as Liaoning, Anhui, and 6. Past and Future Policy Implications of the CTS 45 Hebei, steam coal washing can USS 7.1 billion (in discounted 1990 prices) economically increase the energy for the 15 year time perioU. Shortages in throughput on highly congested lines, 2000 were almost 50 percent higher in the despite the short shipping distances. "without" scenario. And, if tighter environmental regulations are announced, steam coal Implications for the World Bank washing could be part of the least- cost strategy for compliance in many 6.8 The substantive conclusions of regions. the study have also helped support the (d) To avoid worsening particulate and Bank's lending programs to China, in which sulfur pollution that is already among railway and power sector lending in 1995 is the worst in the world, the expected to approach a total of US$1.5 Govermnent should move quickly to billion. All of the power sector projects in make polluters pay for their the Bank's lending program were selected as environmental damage in one way or part of the least cost strategy in the medium another. The analysis results provide demand scenarios. Similarly, the results support for the economic feasibility support increased funding for railways of this type of policy. Strategies such which is part of the Railways VII loan as steam coal washing, conservation, package. and hydropower can benefit the 6.9 By recommending least-cost environment, reduce bottlenecks, and strategies, the modeling effort feeds into a in many cases pay for themselves. policy dialogue with the Chinese (e) The potential for energy conservation Government to identify and remove appears to be substantial, although the institutional barriers and price distortions study's analysis on this strategy is that may stand in the way of the free market preliminary. Reforming the economy, adopting these optimal solutions. For liberating energy prices, and instance, coal washing has long been held establishing the State Economic and back by a host of problems. Coal price Trade Commission's Y400 million reform, long-term contracting, energy conservation grant fund are environmental regulation and taxation of important steps in this direction, but pollution all would encourage enterprises to more is needed. produce and buy washed coal. These kinds 6.7 As a rough estimate of the total of implementation issues were the subject of eonomic imnpact of the most important the latest Bank-cosponsored CTS conference recommendations, a scenario was run in Beijing in October 1994 (see Annex 22). without four of the main strategies: The meeting focused mainly, though not (a) without accelerated exclusively, on coal mining, coal washing, construction of three railway electricity generation and transmission, and and port projects; energy conservation, for which investnent (b) without new steam coal decisions are at least pardy in the hands of washing; decentralized enterprises. (c) without the proposed intergrid 6.10 On the technical assistance transmission lines (zcluding level, this study has effectively transferred not only the seven 500 kV these modern management and analysis intergrid lines, but also the capabilities to researchers in China. There is shorter distance AC intergrid potential within the Bank to apply this lines); and methodology in other large coal-using client (d) without coal imports. countries, such as India and Poland. The cost difference between this scenario without these options and the base case with these options, was about Y 32 billion, or 46 6. Past and Future Policy Implications of the CaS Policy Issues to Be Addressed and energy conservation. The former In the Future only reduces ash and sulfur, while the latter not only eliminates them but 6.11 The energy supply problems eliminates C02 as well. Given that that led to this study will continue to the preliminary analysis showed that threaten China's dynamic growth. Keeping energy conservation can reduce costs, up with economic growth will remain a shortages, and pollution continuing challenge, one that requires simultaneously, and that it is constant evaluation of problems and indispensable for reducing CO2 projects. In particular, more railway and emissions, we recommend that the shipping project options need to be first energy conservation and CO, modules formulated, then analyzed, in a network be added to the Government's official setting. Planning is constrained when all the version of the CTS model, along with options on the drawing board for an updated energy conservation transporting coal out of the energy base are database. still inadequate to do the job. MOR and (b) Other energy sources, such as oil, are MOC would benefit greatly by ideas rising in importance; oil-fired power fonnulated in this fashion. plants using imported petroleum have 6.12 Particulate pollution and acid begun to compete with coal-fired rain from increased coal use will inevitably plants in some coastal locations. worsen in the 1990s, and China can expect Suggestions have been made to add growing pressure from abroad to reduce its oil and gas to the analysis system. greenhouse gas emissions. It is important for Already, the CTS has been adapted to the SPC to pursue the dual goals of fueling help the Institute for Comprehensive economic growth and protecting the Transportation to develop an "Energy environment at the least possible cost. Transportation Plan for the Year Analyses have shown that, up to a point, 2050." these two goals can go hand-in-band rather (c) The current method for matching coal than in opposition to each other. The CTS producers and consumers with each anaysis system can help the SPC determine other and with transport providers at where that point (e.g., washing some 15 the annual coal ordering conference is percent of steam coal) lies. both slow and inflexible. New long- 6.13 New initiatives for solving term contracts between suppliers and China's dynamic energy and transport users should not necessarily duplicate situation are being added to or considered the origin-destination patterns of the for the CaS analysis system, including the old arrangements. At the October following: 1993 CTS Report Conference, several (a) Energy conservation options are senior SPC officials suggested that vitally important because they can the analysis system be adapted for satisfy demand growth without adding facilitating the smooth functioning of any more pollution (other than in the coal/transport market. their manufacture). In 1994, the (d) More of the umnodeled pollution World Bank, through its Robert control altematives could be added to McNamara Fellowship program, the system so as to gain a better fimded a CTS team member from the understanding of how economically Energy Research Institute to add feasible are different levels of energy conservation options to the pollution control. model (see para. 4.62). Analysis of carbon dioxide was also added, which, among other things, distinguishes between coal washing 6. Past and Future Policy Implications of the CTS 47 The Role of Intersectoral Modeling the total for most projects, remains crucial in the Market Economy to their successful completion. For instance, in the United States, models are used to 6.14 All the past and future uses of guide investment in the federal highway the analysis system are still relevant as program, harbor dredging, or other cost- China evolves from a centrally planned to a sharing schemes. Third, it could an2lyze the socialist market economy. An intersectoral impact of policies under consideration by the modeling tool such as this need not and central Government. should not be used to control investmnent 6.17 Already, the CTS results have decisions from the center. It is certainly not reinforced the need for reforms, such as so in the United States, which just finished restructuring the distribution patten for a complete revision of its energy models. coal, opening the county up to more The U.S. Department of Energy publishes imports, modifying the tariff structure on many reports each year based on the rail and water transport, and forming national energy models. Utilities, railroads, interprovincial partnerships for transmission barge companies, and coal producers may and coal washing projects. thus inform their decision-making with a 6.18 In China's transition economy, comprehensive picture of demand, supply, there is often a blurred distinction between competition, trends, emerging technologies, the public and private sectors. With such a and good options. broad scope, it is inevitable that some of the 6.15 A good example of how an activities modeled in this study are (or are intersectoral model at the national level can becoming) private sector decisions by contribute in a market economy is its use for shippers, carriers, or users, such as where to environmental policymaking. When the U.S. build township coal mines, whether to ship Clean Air Act of 1990 was passed, the coal by water, and what type of coal to buy. models were used to find the best way to Others, such as railroad investment and meet the regulations in different regions. envirommental policy remain firmly in the Private corporations could refer to the model public sector. In between lie those activities results to find out whether it is better to that, while in the private sector in some free scrub or switch in their region, that is, to market countries, look to remain in the instail pollution control equipment or to public sector in Cbina for some time, such switch coal types (or wash coal). Another as electricity grid investment and example is when a tax was proposed on management. The central Government barge fuel, the effect on the competition should concentrate their planning efforts on between regions and between kinds of those parts of the economy where a energy was studied using the models. This continued Government intervention or information is published annually in dozens regulation is necessary to supplement market of reports. mecharisms. In particular, natural 6.16 The study recommends that the monopolies such as railways and power analysis system be used, as in Australia or distribution need to be regulated, while France, for indicative planning. A country externalities such as social costs due to in transition, such as China, falls halfway environmental pollution need to be forcibly between a full market economy as in the internalized by the Government. If the United States and a centrally planned Government moves also to free prices, economy as in China in the past. Under reform the legal system, and remove indicative planning, there would be several institutional barriers and subsidies, the roles for the CTS analysis system. First, it market should send the proper signals to could suggest directions for investment to firms regarding whether to wash coal, local govermnents and private ventures. transmit electricity, ship by water, and other Second, it could help guide state-controlled options. investment, which, while only a fraction of 48 6. Past and Future Policy Implications of the CTS 6.19 Even though decentralized or (d) Taking an active role in guiding the semi-autonomous decisions are beyond the CTIS team regarding policy scenarios newly-defined jurisdiction of Government and assumptions that are important to planning, it is still essential to include them high-level officials. in the model, in order for the Govermment (e) Continuing building support for the to predict decentralized responses to central CTS among ministry and provinc;ial decisions, to build a new railway line, for officials by holding conferences to example. The CTS modeling approach check data and assumptions, and to assumes that the decentralized sector is cost- disseminate results. minimizing, subject to regulatory and other (f) Publishing the findings periodically, constraints. Of course, once the scope of the perhaps in an annual series to be model extends beyond the jurisdiction of the distributed at such conferences and modeler, it is more useful for gaining elsewhere. strategic insights than for prescriptive (g) Using the CTS analysis system as planning. A successful example comes from technical input for coordinating the Australia, where the MENSA energy scale and timing of investments by systems model is used by federal and state different ministries. resource agencies. None of the agencies has (h) Preparing recommendations to the overriding directive powers in the areas of State Council, to provinces, and to infrastructure development for which they the private sector concerning have responsibility. Rather, their studies, investments and policies for coal which are often widely disseminated, serve production, preparation, to focus debate and act as a de facto transportation, and distribution, indicative planning mechanism. power generation and transmission, and environmental protection. Mr. Gui Shiyong, Vice Chairman of the Action Plan for Future Use of the Coal SPC, has stated in a letter tD the Institute of Transport Study Management Sciences, that "the CaS, using scientific methods, systematically described 6.20 A series of actions by the the deeper problems and contradictions in Government will be needed for the cas the system, ... that it is of great value for a-alysis system to fulfill its future role as a the adjustment of planning and decision- planning tool in a fast-growing, increasingly making, ...that it has had an important market-oriented economy. These actions influence on China's economy and should be should be carried out in a manner that will used more widely, -. .that it represents the reinforce the refonns currently under way in first time that model results of this kind have the planning and investnent process. The been used at this level in China, . -. land that recommended actions include the following: he] support[s] further applications of the (a) Continuing supervision of the CTS CTS, including making economic forecasts under the umbrella of the ERC, with for 2010." His full statement regarding how direct participation of energy planners the SPC views the CTS model and what the from FM and trAnsport planners from model has already accomplished for them is ICT. included in Annex 21. (b) Rolling the model over to the next five-year planning period, that is, to 2010. (c) Providing additional investment NOTES options, especially in the railway sectr, where heavy-axle investments 1. 'China: Efficiency and Environmental should be compared to extra tracking Impact of Coal Use" (Report No. 8915- and electrification packages. CHA, 1991) and 'China: Energy 6. Past amd Futur Policy Implications of the CIS 49 Conservation study' (Report No. 10813- 3. Martha M. Hamilton, Selling Pollution CHA). Rights Cuts the Cost of Cleaner Air,' The Washington Post, August 24, 1994, p. Fl, 2. 'China: Efficiency and Environmental F3. Impact of Coal Use.' 1991. Annexes Annex 1 1-1 ORGAmZATION OF THE COAL TRANSPORT STUDY 1. The Bank's counterpart for the 4. In addition to this group, there CsS is the Economi,k Research Center was a supporting Policy Team and Technical (ERC) of the SPC of China. The SPC is Team, and a Consulting Panel made up of China's highest economic planning agency. leaders. Members of these teams participated The ERC is the long-range research arm of in part of the training program consisting of the SPC, and oversees a number of more planning sessions, presentations, and a specialized research agencies that partici- number of site visits to U.S. coal, port, and pated in the CTS. Among these are the railway facilities. The Policy Team consisted Economic Institute (El), which served as the of Messrs. Xu Zhen and Yang Zhenjia original host organization for the CTS, the (Deputy Directors, ERC); Mr. Liang Institute for Comprehensive Transportation Xiufeng (El); Ms. Liu Liru (Director, ICT). (IC3), and the Energy Research Institute The Technical Team consisted of Mr. Zhou (ERI). Fergqi (Director, ERI); Mme Shi; Mr. Zou 2. Th:. Bank was responsible for Yuan, Mr. Lin Fatang, and Mr. Yu providing technical assistance to the ERC Xiaodong (all of El). The Consulting Panel including arranging for foreign experts to was made up of Messrs. Xu, Yang, Liang, serve as technical advisors to CTS team and Zhoui, and Madame Liu. organizing a nine-month training program in the United States for five core tean mem- Training Program bets. The Bank also worked jointly with the ERC in preparing detailed work plans and 5. The training program provided major reports. the modeling team members with customized 3. The Study Direcor on the applied courses in economics, benefit-cost Chinese side is Madame Shi-Qing Qi (El). analysis, network modeling, data base man- The core modeling team members who agemnent, software engineering. The training underwent U.S. training were Mr. Sun program also enabled the CTS team to Xufei, Phase 2 Modeling Leader (Huaneng develop a preliminary version of the requi- Co.); Mr. Zhang Chuntai, Phase I Modeling site CMS model, which subsequently was Leader (El); Mr. Cao Wei (ICT); and refined, calibrated, and tested. Starting on Ms. Xie Zhijun (E. Mr. Thou Dadi February 5, 1990, the CTS training program (ERI) served as a Bank consultant and was presented a series of five short courses and also part of the training program. a software development workshop as fol- lows: C:ourse Lecturer* Applied Economics (Huenneman, Fernandez, and Polenske) Systems Modeling (Friesz and Kuby) Benefit-Cost Analysis (Huenneman) Database Management (Sibley, Simkowitz, and Hirshfeld) Coal Transport Modeling (Kuby, Ratick, and representatives from U.S. Department of Energy and Brookhaven Laboratory) Software Development Workshop (Kuby, Bernstein, and Cook) 1-2 Annex 1 6. The first four courses were development workshop carried through to presented during the period February-April the end of the training period. The course 1990 in conjunction with training for the lecturers were supplemented by a number of China Railway Investment Study (RIS) team, invited experts who presented specialized while the fifth was presented frcm mid- topics of interest to the training team. April to end July 1990 and the software Millions ol Tons ot Standard Fuel (7.000 Kcal/kg) Coal Irom Net Coal 'Not ineluding solar b,omast geothermal o, nuclear Stsocks E-pons I Ind I\ rooe 4- - - _ ro Corwried ~ ~ ~ ~ ~ Cotrutin ~. OBkhei dwtr 011~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ lo _Ne 9 Commefcial~ ~ ~ ~~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~Rsieta tikale lEpolts SOURCE: 19t9 SrarisricSa Yearbook ol China l~~~~~~~~~~~~~~~~~~~- 4 2an Ih WddLln "'porial,01,t 2-2 Annex 2 Figure 2.2: Investment in Transportation Relative to GNP and Traffic 3000- 2500- -i* GNP /, 21 -a- Total Traffic 2000- --_-- Investment in National -0 o & / Fixed Assets L 1500- / .t ----o--- Investment in a cn /i / ' Transport 1000 500 - 1955 1960 1965 1970 1975 1980 1985 1990 Year Source: State Planning Commission Annex 2 2-3 Figure 2.3: Investment Structure of the Energy Sector, by Type of Energy 70 - 4- ' 60- (ID50 CD >, 40 aD30- 20 20 4-.10 r 0-0 O- 10I 1953- 1958- 1963- 1966- 1971- 1976- 1981- 1986- 1957 1962 1965 1970 1975 1980 1985 1989 Period -0- Coal 3& Electnicity +- nOil and Gas _.-- Coke andOthers Source: State Planning Commission 2-4 Annex 2 Figure 24: Coal Consumption and Production Trends 95 - 1200 million tons 90~~~~~~~~~~~~~~~~_ .1000 90- as - . */ | 1n11 COAL CONSUMPTION (as % ol 0/% coal. 600 total energy consumption) 80 - COAL PRODUCTION (rmillion tons) 400 75' l _ t v Ll200 70 I0 19531958 196319681973 19781983 19881992 Year Annex 2 2-5 Figure 2.5: Raw Coal Production, by Type of Administration, 1965-89 1200- o 1000- -m~ 800-r° _f c 600- O E 400O a 200- 0 1965 1979 1983 1989 Year % Increase 1983-1989 -M- Central Mines* 26 Local Mines* 42 Township Mlnes' 98 Total 47 Central mines refer to mines from which output is distributed by the central government (tongpei meikuang), local mines are operated by provinces, prefectures, or counties, and township mines (formerly referred to as commune and brigade mines) are operated by rural collectives (xiangjen meikuang). Source: MOCI. 2-6 Figure 26: Major Interregional Coal Flows, 1989-90 (Millions of Metric Tons) I ___ _ _ _ _ _ _ s K n.4 Rail traffic .- .-----. RUSSIA -'-4 BMaritime traffic J * Port e- --- China - statistical region e' boundary - i International boundary C S, Importsexport arQforCineefo CHINA -- -~~~~~~~~~~~~~~~~~~~~~~ CH INA . -- '~~~~~~~~~~~~~~~~~~~~~~ ............. ..... o........ -- a-- 1-2- NORTH- t WEST 35.3 4~ CHINA a -- 0vJWS SOUTH-SOUTH ; CHINA - HINA KO r~~~~~~~~~~~~~~~~~~~~~~~~~ - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ - Sources: U.S.. Energy Information Administration. China State Planniing Commission, and Interr -Imports/expor-ts are for 1989; Chinese flows are for 1990. Most flows under I million metric tc- Annex 2 2-7 Figure 2.7: Investment Structure of the Transportation Sector, by Mode 70- E 60- OD > 50 c 40- c30- CIS i'- 20- 4-' i!2 10 X 1953- 1958- 1963- 1966- 1971- 1976- 1981- 1986- 1957 1962 1965 1970 1975 1980 1985 1989 Period -- Railways _- Highways -0- Waterways -4A- Aviation -0- Pipeline Source: State Planning Commission 2-8 Annex 2 Figure 28: Investment Structure of thc Electricity Sector _70 - a) E 60 - \ ~50 Z 40- -*- 30- c) Wj 20- on10 *a** w i-- 1953- 1958- 1963- 1966- 1971- 1976- 1981- 1986- 1957 1962 1965 1970 1975 1980 1985 1989 Period --- Thermal |--- Hydroelectric I*-- Transmission -@-- Nuclear Source: State Planning Commission Annex 3 3-1 BACKGROUI-D ON TRANSPORT AND ENERGY: ASSORTED TABLES Table 3.1: China: Freight Traffic by Mode (billion ton-kIn) Domestic Civil water- Pipe- avia- OcLan Rail Road way I lines tion Total shipping 1952 60.2 1.4 11.8 - 0.002 73.4 2.8 Modal split (%) 82.0 2.0 16.0 - - 100.0 - 1977 456.9 25.1 102.1 38.7 0.076 622.9 174.1 Modal split (%) 73.4 4.0 16.4 6.2 - 100.0 - 1979 559.9 74.5 139.3 47.6 0.123 821.4 317.1 Modal split (%) 68.2 9.1 17.0 5.8 0 100.0 1989 1,039.4 337.5 349.8 62.9 0.690 1,790.2 768.9 Modal split (%) 58.1 18.9 19.5 3.5 - 100.0 - 1992 1,157.6 375.5 422.2 61.7 1.342 2,018.4 903.4 Modal split (%) A7 *0.5 - 31.0l Growth raoe % p. a. 1952-77 8.4 12.1 9.0 - 15.7 8.9 18.0 1982-92 6.6 14.7 9.5 2.1 21.1 8.1 9.1 1952-92 7.7 15.0 9.4 - - 8.6 - La Enxcludes ocean-going transport. Source: Statistcal Yearbook of China-1993, p. 467-8 3-2 Anncx 3 Table 3.2: Interational Trends in Commecial Energy Intendty (TCPERIGDP) La country 1979/80 1984/85 1987/88 China 1.50 1.20 1.10 Canada 0.85 0.77 0.74 Japan 0.36 0.29 0.28 United States 0.71 0.58 0.56 India 0.61 0.64 0.64 Korea 0.67 0.61 0.60 Poland 0.96 N.A. 0.95 Soviet Union 1.04 N.A. 1.04 /a Measured as total commercial primary energy requirements per US$i,000 of GDP. Ratios for countries other thn China come from OECD and ADB publications. All are computed using both 1980 prices and U.S. dollar exchange rates. Energy is measured in toe (tons of oil equivalent). These ratios generally do not include noncommercial energy. N.A. = not available. Source: Two publications by IEA/OECD: Energy in Non-OECD Counres, Selected issues, 1988 and Energy Policies and Programs of iFA Countries, 1987 Review; and ADB Energy Indicators, 1989. Annex 3 3-3 Table 3.3: International Comparison of Reliance on Coal (percent share) China 73.0 Europe 19.0 India 53.0 Japan 18.5 Korea 34.7 United Kingdom 31.4 Soviet Union 30.0 United States 24.0 Source: IEA, Coal Infontaion 1989, ADB Energy Indicators, 1989 and Statical Yearbook of Cuina, 1994. Table 3.4: International Comparison of Transport Investment as a Percentage of GNP Country Pexiod Transport investment in GNP Japan 1964-73 3.5-3.8 Korea 1979-81 2.00 Brazil 1979-81 2.36 India 1975 2.37 China 1981-92 1.34 Source: Final report. 3-4 Annex 3 Table 3.5: Coal Supply and Demand Balance Sheet (million tons) 1985 1986 1987 1988 1989 1992 1. Coal Supply, of which 874 896 930 982 1,056 1,117 a. Production 872 894 928 980 1,054 1,116 b. Import 2 2 2 2 2 1 2. Coal Demand, of which 823 870 941 1,010 1,048 1,162 c. Intermeldiate use 254 279 312 341 375 491 electricity use 164 180 203 228 252 335 d. Final use 527 545 579 616 620 607 industries use 297 311 338 361 372 388 e. Coal washing lost 35 36 37 37 38 44 f. Export 7 10 13 16 15 20 3. End-of-year stockpile 141 154 138 107 146 101 4. Change in stockpile a +39 +13 -16 -31 +37 -3 5. Statistical discrepancy b/ +12 +13 +5 +3 -29 +57 at The natonal stockpile includes govremment coal reserves, minemouth stockpiles, and users own stockpiles. b/ The statistical discrepancy may include some of the losses during handling and tansportation. Source: Official government sources. Annex 3 3-5 Table 3.6: Comparison of Coal Reserves- Soviet Union, Unied States, and Chim (bilflion tons) World Energy British Council Petroleum World total 1,791 1,174 Of which: Soviet Union 265 264 United States 268 268 China 805 184 a of which lignite 132 15 Soviet Union, United Stites, and China as % of world total 75 61 La Includes all deposits up to a depth of 900 m and a minimum seam thickness of 70 cm. Although broadly comparable, the definition of reserves in China differs from intenational standards. Source: Anrwa Energy Review 1990, p. 279, U.S. Departnent of Energy, May 1991. Table 3.7: Coal Production by Type of Coal, 1987 and 1991 (million tons and percent) 1987 % 1991 % Anthracite 192 21 214 20 Bituminous coal: 703 76 825 76 Prime coking coal (93) (10) (96) 9 Blend:4Ag coking/steam coal (362) (39) (412) 38 Other steam coal (248) (27) (306) 28 Lignite 33 3 45 4 Total 928 100 1,084 100 Source: ERC. 3-6 Annex 3 Table 3.8: Sulfor Content of Chinese Coal Reserves Percent Sulfur content of total (percent) reserves < 1 65-70 1-2 15-20 >2 10-20 Source: China General Coal Utilization Corporation, Working Paper No. 1. Table 3.9: Cal Output and Reseves, by Region, 1989 (perent) Region Raw coal output Proven reserve available for mining North China 37.2 53.3 Northeast China 14.3 2.4 East China 13.7 5.5 South Central China 14.9 3.0 Southwest China 12.1 8.9 Northwest China 7.8 26.9 Source: China General Coal Utilization Corporation, Working Paper No. 1. Annex 3 3-7 Table 3.10: Iterregona Coal Flows: 1980 and 1990 La (million tons) Net coal movement into (-) or out of (+) a %on Region 1980 1990 Change Northeast -13.4 -17.3 29 North +59.5 +157.8 265 East -29.2 -109.9 376 Central South -12.2 -27.7 227 Southwest +2.4 -12.5 -520 Northwest +5.3 +9.8 184 L& Overseas shipments not included. Sources. China Coal Industry Yearblok, 1983 and 1990. Table 3.11: Steam Coal Washing Year Million Tons 1984 31.12 1985 35.79 1989 58.08 1990 64.11 1991 68.51 Source: Energy Research Institate 3-8 Annex 3 Table 3.12: Plan vs. Market i-rke for Steam Coal, 1989 Plan Negotiated Beijing 55-70 140-150 Shenyang 45-70 80-180 Jiangsu 65-80 220-280 Shanxi 34-50 110-140 Source: Stephenson (1990), op. cit. Table 3.13: Evolution of Coal Prices on the Free Market, 1986-90 ai (yuan per ton) 1986 1987 1988 1989 1990 (Iest.) FOB Shanxi 50 80 100-120 110-140 100 FOB Xuzhou 90 100 180 220 175 FOB Shandong - - 100 200 175 LA These are indicative pices for steam coal. Shani is the main producing province in North China. Xuzhou is a mining area in the northern part of Jiangsu province, a major coal consumer in East China, which also imports coal from North China. Shandong is a coastal province in North China and is both a producer and consumer. Source: State Planning Commission. Annex 3 3-9 Table 3.14: lreas in Plan Coal Prices Period Increase Increase in Y/ton in percent 1958 2.64 20 1965 2.01 11 1979 5.07 28 1985 4.39 17 1990 10.00 16 Table 3.15: Coal Traffic by Railway and Waterway Mode Period Coal Coal Average Daily (million (billion distance wagons tons) ton-kan) (On) loaded Railways 1988 565 300 530 26,513 1989 609 316 519 28,479 1990 626 341 545 29,300 Waterways* 1988 124 146 1,175 1989 146 182 1,247 1990 139 182 1,310 3-10 Annex 3 Table 3.16: Indicators of Railway Asset Utilizaton: A Compansou between China, the Soviet Union, the United States, Iadia, and Brazil China USSR USALa India La Brazil Indicator (1993) (1987) (1987) (1987-88) (1987) Freight ton-km/route-km (mil) 22.2 26.2 6.5 3.6 8.00 Passenger-kmlroute-km (mil) 6.5 2.8 0.1 4.4 0.05 Freight ton-km/freight car owned (mil) 3.1 2.2 0.8/ 0.5 1.70 Freight tonnage/car loaded 56.1 53.6 66.6 20.1 55.0 Freight cars per train 40-50 50-75 71 n.a. 35 (est.) (est.) Freight car turnaround time (days) 4.2 6.6 18.8 11.6 12.0 Freight car turnaround distance (1an)Ld 993 1,610 2,125 1,274 750 Freight hauling distance (kmn) 761 945 1,106 776 460 Freight ton-km/locomotive owned (millions) 80.8 68.5 70.6 24.3 25.0 Freight train-krn/route-kn/dayLS 35.5 42.9 6.1 10.6 6.4 Freight train speed (inc. stops) (km/h) 30.0 31.8 n.a. 22.7 30.0 Freight train gross trailing welght (tons 2,519 3,085 4,300 2,050 2,200 Freight tonnage per in 1,300 1,800 2,390 1,053 1,200 to (est.) 1,500 (est.) La Class I Railroads oniy. /bŽ Includes cars owned by car corn d shiprs. /c 90.58 percent of all freight ton-. ometers are carried on broad-gauge rails, so most above estimates are for broad gauge only. See Indian Railways YeaiRook, 1987-88, D.23. / Data perains to 1990. Sources: China: Ministry of Railways and mission estimates. USSR: Narodnoe khozyaistvo SSSR v 1987 g. (7he USSR Nadonal Economy in 1987), pp. 21, 307-308, 312, 315. Various esimates from Hunter and Kaple, 1984 and Szynnyr and Dunn 1985. Indlia: Econoric Survey, 1988-89, pp. i7, 35-36, S-30; Indian Railways Yearbook 1987-8 pp. viii, 40, 0, 78-79 and 80-81. USA: Railroad Facts, 1387, pp.9, 31, 32, 35, 36, 43, 46. The World Bank, Railroad Database, June 1989. Brazil: Official Government sources. Annex 3 3-11 Table 3.17: Combined Rail-Water Routes from Shanxi to East China (kIn) Rail Water Total distance 1. Datong to Shanghai 1,840 1,840 Datong-Qinhuangdao-Shanghai 620 1,350 1,970 2. Taiyuan to Shanghai 1,500 1,500 Taiyuan-Qingdao-Shanghai 920 750 1,670 3. Taiyuan-Nanjing 1,190 1,190 Taiyuan-Wuhan-Nanjing 1,180 800 1,980 3-12 Annex 3 Table 3.15: Share of Electricity In Total and Indusbial Thluld Net Consumption of Energy in Selected Countries L (percent) 1980 com- 1985 com- 1988 com- mercial enegy mercial en=ry mnrcial ener= Total Industry Total Industry Total Industry Developing Countries China 6.1 7.7 6.5 8.4 7.7 N.A. Argentina 10.1 18.6 11.3 18.7 11.4 20.9 Brazil 15.5 21.0 20.1 29.5 20.3 28.8 Mexico 7.6 11.3 8.7 10.8 9.8 12.9 India 11.2 12.6 13.5 14.4 14.9 N.A. S. Korea 8.1 12.5 10.1 15.7 11.2 16.8 Developed Countries United States 13.2 14.2 15.1 16.6 15.5 16.8 Canada 16.9 20.2 21.0 24.7 22.5 25.0 Japan 17.8 22.2 20.2 25.3 20.5 26.3 France 12.6 15.2 15.9 18.6 17.2 21.0 G-nany, F.R. 13.7 18.2 15.3 20.9 16.0 22.8 Italy 12.9 18.6 14.1 20.9 14.9 21.7 UnJited Kingdom 14.6 16.5 14.9 17.5 15.3 18.7 N.A. = not available. La Final net consumption refers to energy consumed by end users, net of conversion, transmission, and other losses. These figures are not the percentage of total energy production that is used for generating electricity, but rather the percentage of final energy consumption that is electricity. Source: OECD (1989) World Energ Studstics and Balances, 1971-1987. OECD (1990) World Energy S&usics and Blances, 1985-1988. OECD (1987) Energy Blances of Developing Countries 19704985. OECD (1988) Energy Balaces of Developing Countries 1985-1986. OECD (1990) Energy Balances of Developing Countries 1987-1988. All figures were originally reported in mtoe. Annex 3 3-13 Table 3.19: Estimated Annual Emissions from Coal Use, by Sector (million tons) up _SO2 .C027 Sources 1985 1988 1985 1988 1988 Industrial boilers and plants 7.7 9.3 6.3 9.7 300 Electric power boilers 7.0 8.0 3.5 5.7 153 Residential/commercial 1.2 2.4 2.9 4.6 129 Railroads 1.0 1.0 0.5 0.5 15 Total Coal-Related Emissions 16.9 20.7 13.2 20.5 597 Total Emissions 23 N.A. 15 N.A. N.A. N.A. = not available. Note: 1985 data come from Wage Hanchen and Zhao Dianwu, 'Air Pollution Control and Energy Use in China' in Proceedings of the hinese-American Symposium on Energy Markets and the Future of Energy Demand, Berkeley: Lawrence Berkeley Laboratory, USC, November 1988. 3-14 Annex 3 Table 3.20: Ambient Concentrations in Major Chinese and Foreign Ciies TSP (ug/In3) SO2 (yp/m3) Annual Maximum Annual Maximum City La average Sk reading L average Lb reading LQ Beijing 321-504 965 73-249 566 Chongqing 620 710 300-400 N.A. Guangzhou 118-325 640 7-143 342 Shanghai 192-360 854 13-115 238 Shenyang 304-598 1,546 78-213 623 Taiyuan 1,070 N.A. 250 N.A. Xian 495-708 1,310 124-170 317 Bangkok 198-243 386-741 18 48 Delhi 291-453 831-1,062 28-68 97-197 New York City 44-62 87-121 65 116 Tokyo 52 143 25-34 58-68 Warsaw 58-67 213-248 33-46 180-205 N.A. = not available. La Data for Chinese cities are for 1988. Data for foreign cities are for 1982-85. 1k For those cities where there were data firom more than one monitoring station, a range is given. k The 98th percentile of daily values; it represents the level below which fall 98 percent of daily measurements in a year and the threshold, above which lies 2 percent of the measurements (7 days of the year). It provides information on the most polluted days of the year. Ranges represent more than one monitoring station. Source: World Resources Institute, World Resources, 1988/89. Monitoring data were submitted by national governments to the Global Environmental Monitoring System (GEMS), maintained by WHO and the UN Environmental Program. Emission levels can be affected by seasonal patterns Ln emissions and meteorological conditions. As a result, annual means and 98th percentiles are most meaningful if they include measurements taken during all seasons of the year. Since the analysis did not always include seasonal data, it should be treated with caution. Also, data provided by NEPA and various local EPBs. Annex 3 3-15 Table 3.21: Ambient Air Quality Standards Period Class I Class I Clm T (Mg/m3 = micrograms per cubic meter) TSP Daily average 150 300 500 Max. at any time 300 1,000 1,500 TSP (< 10 microns) Daily average 50 150 250 Max. at any time 150 500 700 SO2 Annual average 20 60 100 Daily average 50 150 250 Max. at any time 150 500 700 Source: National Environmental Protection Agency (NEPA), China. . Annex 4 4-1 PLANNING AND INVESTMENT FOR COAL, TRANsPoRT, AND ELECTCITY Overview targets, and key capital construction projects to be pursued over the period. The State 1. China is currendy reforming the Planning Commission (SPC) assists the State planning and investment system. They are Council in preparing, the FYPs. The newly- decentralizing their decision-making and formed State Economic and Trade reorienting the economy toward the market Commission (SETC) has taken over the in a step-by-step fashion. responsibility of day-to-day implementation 2. Investment planning in China is of the FYPs from the SPC. Formerly known carried out in two parallel processes: (a) as the State Economic Commission, the formulation of aggregate plans for capital SETC was merged into the SPC in 1988 and investment; and (b) planning and then separated back out in 1992, with management of individual projects. The state additional responsibilities relating to trade government continues to coordinate the and reinvestment. Finally, there are gross allocation of investment funds between ministries in charge of finance and planning different sectors and provinces, and to for energy and transportation. macromanage the economy to balance Eavironmental protection is supervised by consumption, savings, and investment. the National Environmental Protection Before the 1980s, the central government Agency, which is situated one level below controlled all individual projects. During the the SPC and ministry level, but functions 1980s, they controlled mainly the large like a ministry. projects. Now, even townships can plan and 5. Paralleling the SPC and miistries at finance some of their own investment the lower levels, each province has its own project. Often, the state government is but provincial planning commission (PPC) and one of several partners. departments of finance, energy and 3. Government in China comprises the transportation (Figure 4.1). central level and three to four lower levels: 6. The SPC (the World Bank's (a) provinces, autonomous regions, or major counterpart agency in this study) is the long- municipalities (Beijing, Tlanj in, Shangbai term planning agency of the central only) (b) counties, and (c) townships. government in charge of new investments. However, sixteen major cities (e.g., There are several plannin departments in Chengdu, Guangzhou, Shenyang) are now charge of finance, energy, and transport "directly managed" by the SPC rather than planning (which, by the way, have used the the province in which they are located. preliminary results of this study). The SPC's Long-Term Planning Bureau prepares FYPs Organizations and their Functions and ten year strategic plans. The SPC's National Economic Comprehensive Planning 4. Figure 4.1 shows the administrative Bureau translates the FYPs into yearly plans control system for economic planning of the that set macroeconomic targets such a. transport ' energy sectors. The State money supply and financial budgets and Council, at the top of the economic planning quotas for new investment and equipment hierarchy, is responsible for the five-year replacement among the ministries and plans (FYP) for social and economic devel- provinces. To provide a better linkage opment and for obtaining approval of these between the FYP and the yearly plan, a 2- Flans from the National People's Congress. year rolling plan was adopted 92, with The FYP outlines the development strategy, tentative planning targets fue .: ilowing 4-2 Annex 4 year. The SPC coordinates the ministries, 9. In 1992, the Chinese government the state corporations, the provinces, and further decentralized project management those cities whose plans are dlrectly 2nd planning. Those construction projects managed by the SPC. whose finds, materials, and marketing can 7. In addition, the SPC approves all be handled by a local or provincial "key" capital construction projects costing govermments can now be be organized, more than Y 30 million to build (Y 50 sponsored, engineered, and administered at million for transport projects). The projects that level. For some projects that involve exceeding Y 200 million must also be several ministries andlor provinces, such as cleared by the State Council. Projects an integrated project of coal mining, rail costing less than Y 30 million (or Y 50 transport, and thermal power generation, million for transport projects) are approved different parts would be handled by the State by the ministries, provinces, and directly- Coal Mining Corporation (SCMC), MOEP, managed cities according to their and MOR, with the SPC acting as the responsibilities and the investment quotas set overall coordinator. by the SPC. 10. Those projects that do need support from upper level governments must follow Planning Process a seven step planning process: (a) the sponsoring ministry or province carries out 8. Each ministry and province in China forecasting and programming of its has its own institutions for planning and investment projects; (b) the sponsor programming, surveyirig, engineering discusses its project with the SPC (or PPC) design, and research and development, as and identifies priorities; (c) the sponsor shown in Figure 4.2. These institutions draw submits a preliminary project proposal to the up a blueprint of long-term development for SPC (or PPC); (d) CIECC (or provincial their particular industry or region and equivalent) appraises the proposal and the prepare investment projects in accordance SPC (or PPC) approves it; (e) the sponsor with the blueprint. The Ministry of Railways submits to the SPC (or PPC) a feasibility (MOR) is in charge of both planning and study (for joirt ventures) or specification building the railway networr. The Ministry (for domestic projects) containing detailed of Communrications (MOC) plans and builds information about marketing conditions, the highway network, waterways, and ports financing, material goods supply, at the national level. The Ministry of Energy technology, production capacity, economic (MOE) was broken up in May, 1993 into a costs and benefits, and condition of Ministry of Coal (MOCL) and a Ministry of infrastructure; (t) CIECC (or provincial Electric Power (MOEP) and the existing equivalent) appraises the study or ine speci- China National Petroleum Corporation. fication and SPC or the State Council (or These two ministries are no longer invohved PPC) approves it; and (g) SPC (or PPC) in enterprise management or project inserts the project into the yearly plan for construction; they are charged with overall execution. sectr coordinadon and policy guidance, and for planning state-owned energy projects. Financing Various sta corporations under MOEP and MOCL construct, finance and supervise 11. Since the natio-al economr, Ls facing these energy projects. Provincial a long-term shortage of capital resources, governmeats are in charge of planning and the number of projects that can be constructing provincially-run energy undertaken will depend on how much money projects, coordinated by MOEP and MOCL, is available and can be borrowed. Thus, and highway and waterway systems, projects must be carefully chosen on the coordinated by MOC. basis of economic priority so that savings may flow to high reun investments. Annex 4 4-3 12. The transition of China's financial 15. Policy loans generally meet two sector from one that allocates credit directly criteria: they are mandatory for the banks to to one that relies increasingly on markets finance, anct they may not satisfy the bank's can be divided into three periods. In the first commercial lending criteria. Five types of phase (1979 to 1986), the government began loans fall into this category. First (and of to bir-ak up the monobank system. The primary concern here) are infrastructure People's Bank of China (or PBC) was investment loans that are essentially finan- refocused on controlling tie money supply, cially viable but are large scale and have bank reserves, and other traditional central long payback periods, such as power and bank roles. PBC's commercial lending was transport projects in the FYPs. The oLier spun off to ihe Industrial and Commerce four categories include loans for: Bank of China (ICBC), which joined three technological renovation of fixed assets other specialized banks including the based on the FYPs; rural development and People's Construction Bank of China (or food security programs; subsidized social PCBC, mainly for infrastructure lending); sectors such as health and education; and the Bank of China (mainly for foreign subsidized working capital to priority exchange transactions); and the Agricultural enterprises. The latter includes strategic Bank of China. industries of national importance, export- 13. During the second phase (1987 to oriented industries, and state-owned 1991), two comprehensive multisector banks enterprises of major regional or national (the Bank of Communications (BOCOM) importance that are structurally loss-making and the China Insurance Trust lavestrent (including coal mines). Company (CMC) Industrial Bank) were 16. In an effort to enable the specialized created, and the specialization barriers banks to function like commercial banks, the between banks begap to be eliminated. The lending for policy purposes is being third phase (since 1991) was characterized separated out from the four specialized by further market development and banks and shifted to three policy banks. The transformation, including the establishment first policy bank to open its doors was the of stock exchanges in Shanghai and State Development Bank of China (SDBC) Shenzhen. In addition to these institutions, in April 1994, followed shortly thereafter by there are now seven commercial bank (ma- the Export-Import Bank of China and the inly regional in scope) and some 65,000 Agricultural Development Bank of China. near-bank institutions (urban and rural credit The focus here will be on the SDBC because cooperatives). Finally, there has been a it is responsible for transport and energy proliferation of nonbank financial policy loans. The SDBC is not profit- institutions, including almost 400 trade and oriented, and would lend at subsidized rates. investment corporations, 90 security Funding for the SDPC in 1994 came from companies, 40 finaice and financial leasing three sources: (a) allocation of Y S billion companies, three insurance companies, and from the state budget; (b) inflows of Y 2 100 branches of international banks. billion from earlier SPC-fumded projects; 14. The specialized banls served two and (c) domestic bond issues of Y 65 bil- functions until 1994. On the one hand, they lion. engaged in business banking using a 17. A short history of some now-defiuct growing amount of discretionary funds institutions may help the reader to which they may lend to projects which get understand the roots of the SDBC and also the highest rates of return. However, they to appreciate how far financial sector reform also acted as an implementation arm of has had to come. Before 1987, state central Gvernment policy. Within the last investment funds were given as grants and year, the Government has initiated a new did no: have to be repaid. The funds were reform to deal with the latter group of so- derived from taxes and fri-n the profits of called 'policy loans" or 'directed credit." state-owned enterprises-a system which 4-4 Annex 4 blurred the lines between the Government, typical example might be fimded by any the Zlxaance sector, and the enterprises number of sources, including: (a) the themselves. After 1987, the Government sponsoring ministry or corporation; (b) established a state investment fund with foreign capital; (c) bank loans; (d) the these monies, from which low-interest loans SDBC (or PCBC); (e) local governments; (f) were extended by specialized banks such as local agencies; (g) private corporations; and the PCBC. To manage this fund, the (h) shares ot stock. Government in 1988 set up six National 21. While provinces and cities can plan Investment Companies (NICs) under direct and build large projects independently of the control of the SPC. 'Trhe NICs would then central government, the 10 to 30 percent choose the projects and find the partners for that they typicz-ly get from the cental the SPC. The NICs no longer exist, their Government can be crucial for the project's staff having been absorbed into the new financial success. The share of central SDBC, along with staff from the SPC's Government financing in the power sector, Investment Department. One potential for instance, has fallen sharply, from problem facing the new SDBC is that much 91 percent in 1980 to 30 percent in 1992, of the staff comes from a nonbanking while provinces and local govermments now background. provide 40 percent. About 70 percent of 18. The SDBC will appraise projects not total investment is now allocated through the only for financial viability but also technical banking system, with about Y 40 billion per quality (to be contracted out) and adherence year being controlled by the SPC. to state industrial policies. The exact process 22- Whereas the SPC oversees most new for choosing projects under the new system investnent in China, the State Economic and is not clear. Before the new SDBC, the Trade Commission (SETC) oversees China International Engineering Consulting equipment renewal, rehabilitation, and Corporation (CIECC)-also under the innovation. hIowever, the SETC has fewer SPC-would appraise projects. The project funds at their disposal, because most of the would then have been forwarded to the SPC depreciation fees are retained by local and State Council for approval, which would governments and enterprises. All large-scale have passed it on to the PCBC for financing. equipment replacement projects must be Occasionally a project already approved by approved by the SETC or its local branches, the NICs and the CIECC was struck down though the planning process is much more because of the PCBC's appraisal, but for the streamlined than fur new investments. The most part the PCBC implemented the central SETC also can fumd demnstration projects government's investment se-ategy. for new technologies, and they have a Y 400 19. The SDBC supposedly will have million fund for energy conservation grants. autonomy in deciding which projects to finance from a list provided by the SPC. At its startup, the SDBC had 350 projects in its The Now Nearly Extinct lending pipeline, which does indicate a tight Dual Price System relationship with the SPC, at least at first. Prior to the formation of the SDBC, about 23. 'The dual price system for coal, Y 40 billion per year of investment was con- featuring large differences betwepn low trolled by the SPC. state-determined "in-plan" prices on the one 20. In any case, the SDBC is not the hand and higher "open market" or sole funding source for transport and energy -negotiated" prices on the other, is now investment finds, which since early 1980s virtually extinct. Since the dual price system has come from various sources. The new was introduced in the mid-1980s (replacing flexible "method" for financing is cost- the preceding fully-controlled system), in- sharing, symbolized by a pie chart showing plan prices have increased in real terms (20 the ownership shares of different entities. A percent in 1991). Perhaps more importantly, A Annex 4 4 the percentage sold in-plan was cut in half in Analytical Support January 1993, and now accounts for less than one quarter of all coal. The rest is to be 26. The State Council, SPC, ministries, sold at free market prices by the end of and provinces have their own research 1994, with the possible exception of a few centers and institutions to provide analytical consumer categories. For instance, the MOR support. For the SPC, the Economic now buys all of its coal and electricity at Research Center (ERC) is an umbrella market prices. In East China, coal prices are institute which coordinates the activities of now close to or above international price several institutes, including the Economic levels, while in the Northeast, some Institute 'El), the Energy Research Institute producers of low-quality coal are unable to (ERI), and the Institute for Comprehensive sell their entire production. Transportation (ICT), all of which provided 24. Under the dual system, in-plan coal researchers for this study. Analytical results was priced well below incremental provided by these institutes and by otLers production costs or border prices. Usually, are one consideration among many used by ownership was the dominant factor in the decision makers, and not necessarily the determining whether an enterprise's inputs domniant factor. Past experience and or outputs are sold in or out of the plan. For established patterns tend to be the factors coal producers, most of the coal produced most heavily relied upon in decision making, by state-owned mines was until recently and models tend to be used as a supporting allocateL under the plan, although reference. production in excess of quota can be sold with a 50 percent surcharge. Township Existing Constraints and Outlook for the mines sold most of their coal at market 1990s prices, with province-owned mines selling under both systems. For coal consumers, all 27. The task for SPC is to achieve some the enterprises owned by the townships and balance in the distribution of investment part of those owned by the local gov- resources among different industries and ernments were supplied by out-of-plan regions, recognizing demand and supply resources. All the enterprises owned by the realities and the availability of foreign and central Government could access the in-plan domestic financing sources. Most ministries resources, although in most cases, the in- and provinces each year ask foir an increase plan supply to the centrally-owned in their invesunent quotas from the cenitral enterprises were not enough to meet their government, citing rising transport and requirements. energy demand. Naturally, some of the 25. Central and local governments still projects put forward for state approval are control most coal reserves of high quality, more affordable and more badly needed tnan as well as the main transportation, such as others. There is a tendency for some project railways, major ports, and large ships. The sponsors to underestimate project costs in township mines can access only the order to improve the chance of getting the remaining coal reserves and shorter distance project approved, or to provide insufficient transportation modes. Because the ability to allowance for price increases. To identify sell in the market is dependent upon high-priority projects, the ministries and securing access to transport, the local and provinces discuss their proposals in a contin- township mines in recent years have had uing dialogue with the SPC, often submitting little choice but to sell their excess projects for approval several times. A major production beyond local needs to the state improvement in the investment planning mining bureaus at in-plan prices. For all process since the begining of economic these reasons, the in-plan and out-of-plan reforms in 1979 has been the introduction of systems were nowhere near independent of economic benefits as a key criterion for each other. project selection. .4-6 Annex 4 28. Reform of the investment and scale projects. Enterprises and local goverm- banking systems is proceeding step-by-step. ments find it easier to invest in smaller Inevitably, difficulties and distortions arise plants costing below the Y 30 million cut-off when reforms in different sectors are not level for requiring clearance or financial coordinated. The banking system still does contribution by the SPC. The design of not control the total volume of investment, provincial and county coal mines and coal nor is it free to choose which projects to washing plants are often not optimized in finance. Profit rates for infrastructure order to stay within a certain threshold of projects tend to be lower than for free scale and cost. market consumer goods because of price distortions, ownership problems, special Other Problems privileges for Special Economic Zones and joint ventures, and subsidized housing, 31. The administrative nature of the medicine and insurance. The Chinese plannng and investment allocation system rationale for step-by-step reform is that, continues to emphasize supply targets rather until these distortions are eliminated, a than efficiency, profitability, quality, and totally free banking system would tend to technical innovation. Annual budgetary underinvest in infrastructure. In mid-1993, negotiations create a bias towards capital some free market measures in banking were cost minimization without considering rescinded. operating savings; this tends to discourage upgrading investments or acquisition of Reasons for Inelicient Scale of Projects more efficient technologies. 32. Price distortions can lead to 29. Institutional factors encourage local undesirable side effects. Distortions in the government investnents to be small scale, prices of inputs to rail transport, especially resulting in inefficiencies, particularly in the the low price of electricity, may result in use of energy and raw materials. If a local nonoptimal choice of railway technology government invests in production capacity in (e.g., electric traction instead of diesel another province, they lose that tax base. traction). The low prices of co; also dis- Local governments, in pursuit of high courages investment in coal mining. employment, try to produce as many of their 33. Uneven and inflexible foreign needs as possible in local plants, and also try exchange allocation often hampers imports to keep the processing plants for locally- of intermediate or final tedcnologies which produced raw materials in their jurisdiction. could foster more rapid modernization of Also, markeing is not advanced in China, domestic equipment in a number of maing it difficult to generate demand for a industries-including coal, cement, and product outside of the local area. fertilizer. The lack of foreign exchange may 30. Local governments often do not have force enterprises to use local materials and the financial resources to build larger scale equipment which have low initial costs but projects. In some cases, the slow and sometimes do not last long enough to be cumbersome system of state investment economical. approval also encourages inefficient small- Amnex 4 4-7 Figure 4 1: Adeinistr tive Controt Sy t- for ErsHr *nd Tr fqwtXartion f State Ptanin Consso >qChina International Engineerfng | CSPC) ConsuLting Corporation CCIECC) State ~ Ecnmc and N uaneng Energy Investmont TradeCommision SETC)Corporation | Ministry of Fiiname (HOF) State Development Bank of |Chfna Prople's ~ ~ ~ ~ ~ ~ ~ ~~to Bank offhn lfig ofb Cnrai Luays) (N Chinae Buem o Rilay |State Council q- l _-ministry of communications Minitry f Col (RCL)|stae Col MiingCorporation toca col iing Corporation Sinisty of tectrc EL triciy Admnistrativ B ureaus Pouer tRCEP) for~~~~~~~~~~~~~ ie n dor.d) -China MatimoLa Petroleua Corporatfen China Nuclear Energy CENTRMAL LIEVEL Corporation National E mvironmental _ ~~~~~~~~Protection Agency LOCAL LEVEL PoicPtning Commissfon Province CovernoentDeparmn of Energy Department of Coormuications Department of Finance_ 4-8 Annex 4 Figure 4.2: Ptanniua Syutm for Eneww and Traupartatien H Pl-nPing DepDrtment ofvo oR Planing Department of -APlanning oDeprtment of XECan State Cost 14fnfn Corp. LocaL Cosl gining Corp.C State Pta ngun DPoanning nepfrtment of Cal ssien_ CSPC) Ptamiing Department of 140C Centrat Levet LoraL LeveL~~~~~~~~~~ Ptamingr Division of ProvinciaL Department of communicgtions { | _ P(anning ~~~~~Division of Provir;ial D,epartment _ of Energy Annex 5 5-I TUE CTS ANALYsis SYSTEM 1. This annex provides an overview of can analyze the effect of hypothetical rail or the CTS analysis system, a summary of port projects, cost differentials, import strengths and limitations of the model, and restrictions, budget constraints, a theoretical analysis of supply and demand environmental goals, or other policy in linear programming models. The technical variables on such outcomes as bottlenecks, aspects of the analysis system, such as the flows, shortages, costs, capital budgets, and model formulation, data base structures, and environmental impacts. software system, are described in Volume 4. The task of simultaneously II. optimizing all of these interrelated energy and transport activities is not doable without a large model. And yet, the CaS model is Model Overview designed to complement other more detailed models for planning particular sectors of the Purpose and Scope coal-electricity delivery system, such as rail planning models or electricity system 2. In studying China's coal and programming models. electricity shortages, it became clear that in adidiiion to increasing the transport capacity, Methodology investmnents such as coal wasting and hydropower may, in some circumstances, 5. The CTS model is a mixed-integer reduce the demand for coal transportation in Oinear) program that minimizes the total a cost-effective and environmentally cost of delivering coal and electricity subject beneficial way. Likewise, there is a close to capacity constraints and optional budget substitutability between rail transport of coal and environmental constaints. It combines and minemouth power plants coupled with features from the major coal and electricity long distance transmitssion lines. Therefore, planning models used in the United States, it is necessary to evaluate these different but has been tailored to the SPC's needs. options in a systematic and integrated The cost function includes operating costs fashion. and nnnualized investment costs for 3. The CaS analysis systen was transportation, mining, and the other designed as a strategic-level Decision sectors, as well as a penalty cost for coal Support System (DSS) to help the SPC in and electricity shortages. Given any forecast this task. It is a tool to support SPC of GNP growth and energy demand policymakers in making decisions by elasticity for 1995, 2000, and 2005, the providing a fast and comprehensive way to model will try to satisfy those demands for study energy delivery as a system rather the lowest cost possible. than as separate parts. Its primary use is to 6. The CaS analysis system models all identify investment priorities by sector, of these activities simultaneously by treating corridor, or region. A fully automated the energy delivery system as links of a software system allows the user to easily chain (see Figure 6.1, Annex 6). The create new scenarios modifying key nontransport activities are represented as a assumptions. With this DSS, policymakers generalized network, which is well-suited can answer "what-if' questions and analyze for modeling energy systems in which the tradeoffs between economic goals, energy output of one process is the input to another. supply goals, and environmental goals. It The chain begins with coal mining by three 5-2 Annexc 5 levels of technology and four broad types of locations of bottlenecks and shortages, and coal. Anthracite and low ash steam coal cam data on systemwide costs and amounts of be transported in raw form, coking coal ash and sulfur in the delivered coal. must be washed, and high ash steam coal has the option to be washed. Coal washing Level of Detail is modeled as a flow that converts one ton of raw coal to a lesser amount of better coal. 9. Temporally, the model covers three The raw or beneficiated output is then time periods: currently these are the final transported over a multimodal transport years of the 8th, 9th, or 10th five-year plans network to demand nodes, where shortages (1995, 2000, 2005). Demand must be met in occur if not enough is delivered. All users each time period. Investments made in any of each type of coal, that is, anthracite, time period carry over into later time coking, and steam coal, are aggregated periods, while some preexisting capacity is together at each node. retired each period. 7. Next (in the second half of Figure 10. Spatially, the Chinese economy is 6.1, Annex 6), coal competes with hydro divided into 48 coal supply nodes, 49 coal and nuclear power. Thermal power demand nodes, 58 electricity supply nodes, generation is modeled as an activity where and 38 electricity demand nodes. The the tons of coal in kilocalories are converted transport network is represented by 286 to a lesser amount of kilowatt-hours of railway arcs (177 of which have investment electricity. Electricity from all sources then projects), 31 ports, 3 potential slurry flows over a transmission network, with pipelines, and 99 transmission lines (see some losses, to electricity demand nodes, Figures 6.2-6.4, Annex 6). where shortages can occur in any of five 11. Technologically, there are four basic end-user sectors, each with a different types of coal (low and high ash steam, willingness to pay. Demand drives the anthracite, coking), but within each type, the model, pulling coal and electricity across ash, sulfiur, and kilocalories per ton are this entire system. Finally, ash and sulfur uniquely specified for each zone. Four content of the delivered coal (from both classes of coal users (steam coal for sides of Figure 6.1) are tallied after electricity, steam coal for industry, accounting for optional coal washing and anthracite, coking) and five types of scrubbers. electricity users (rural, urban, light and heavy industry, and agriculture) are the Inputs and Outputs driving forces in the model. There are four types of power plants (baseload thermal, 8. The CTS model uses fbur kinds of middlings thermal, hydro, and nuclear), with inputs. Project data define which investment the hydro plants being uniquely defined options are to be considered. Technical data based on local conditions. Four types of specify capacities, wash-out rates, electricity transmission lines (220 iV AC, 330 kV AC, loss factors, and electricity conversion 500 kV AC, and 500 kV DC) are used to factors. Economic data set investment and transmit electricity both within and between operating costs. Poliy assumptions are grids. demands and shortage ;osts, and, if desired, environrmental or importation restrictions, or Environmental Analysis budget constraints. The model's primary outputs are (a) optimal type, location, scale, 12. The CTS model facilitates and timing of new investment projects; (b) multicriteria analysis by allowing constraints optimal coal and electricity distribution to be placed upon the total tonnage of ash patterns; (c) optimal use of existing mining, and/or sulfur content of the delivered coal to washing, transport, generating, and each province in each of the time periods. transmission capacity; and (d) predicted The analyst can disable these constraints, in Annex 5 5-3 which case they simply count the totals. and there is always the danger that Alternatively, the analyst can force something that has been left out cannot be reductions of any given percent or to any adequately treated by scenarios. Fourth, for given target (or a range of percents or those activities that are not left out of the targets) in order to find out how much more model, complexities are ignored and details it would cost to do so and wihat is the least are suppressed. The way that the cas model cost way of doing so. in this fashion, the simplifies China's reality in each of these user can trace out a Pareto efficient fronteir four ways determines its strengths and between the conflicting goals of minimizing weaknesses. cost and minimizing delivered ash and sulfur content (see Figure 10.12, Annex 10). The Strengths solutions found in this way have the property that, given the options available in 15. The strength of the analysis system the model (see Limitations, below), there are is its use as a spatially-based strategic-level no other solutions that are both lower cost planning tool. The fact that the mining, and lower in ash and sulfur content. This washing, transportation, and consumption of type of analysis recognizes the difficulty of coal are optimized simultaneously with the putting an economic valuelcost on ash and generation, transmission, and consumption sulfur content, and instead leaves it up to the of electricity, and that all of this is done in reader to determine which point on the a spatially disaggregated way in the context tradeoff curve is best, considering also of environmental concerns over a fifteen various factors not included in the analysis. year time horizon is certainly the main strength of the modeling system. However, User Friendliness in modeling, a broader scope does not always mean a better model, because 13. The analysis system runs on a 486 realistic detail and performance are usually IBM-compatible computer and is completely sacrificed by ir.cluding so much. So why is menu-driven. It is automated so that SPC the broad scope considered to be a strength planners can easily generate data inputs for of the CsS model? Because logistically and scenarios with modified assumptions about economically, these investment activities are costs, demands, capacities, routes, imports so closely interconnected that there are too and exports, interest rates, shortage costs, or many tradeoffs for them to be realistically environmental policies. It automatically compared outside of such a model. Take processes the results and outputs them in coal washing vs new transport infrastructure summary tables and maps (see Figure 6.5, vs new minemouth power planm and long- Annex 6). distance transmission lines. All of these investments can help alleviate the shortage of electricity in an urban area. But their Strengths and Limitations comparative economics depend on coal type, distance, budget availability, their 14. Models have been defined as contributions to sulfur and ash pollution, and "simplified representations of reality." There losses in washing, generation, and are four ways in which any model simplifies transmission, etc. Perhaps for a given origin the system under study. First, the model and destination, their economics can be simplifies reality by limiting the choices compared without a network optimization available for investment and routing model, but in China there is a multitude of decisions. Second, the model simplifies origin, destination, route, and coal type relationships by linearizing a nonlinear combinations to choose from. relationship or ignoring a feedback loop. 16. Given that the comparative Third, some activities must be left out of the economics dictates a broadly-defined model that really are related to the system, strategic network model, the next question 5-4 Annex 5 is, how can this be accomplished without best way that multiple modes and coal types sacrificing too much realism? In the CTS, could have been handled in a large network, the level of detail is pitched to allow for the use of paths is limiting precisely because spatial disaggregation of supply, demand, the number of paths is limited. The number and transportation, while still including of logistically-possible paths from just a related economic activities and multiple time single origin to a single city can be periods. By not disaggregating these enormous. Despite the use of a least-cost activities to the level of individual plants, path generator, it is inevitable that some mines, and rail spurs, enough "room" is left valuable path options will not be included. within the model to allow the electricity This has the tendency to force as much coal sector and three time periods to be modeled. as possible on this smaller set of cheapest 17. The use of path variables is both a paths, which is no' a Lad result in and of strength and a weakness. If one can live itself. However, when arcs become with a limitation on the number of transport capacitated, coal has to flow on whichever routing options (see weaknesses), the use of remaining paths do not use the congested path variables eliminates the need for mass arcs. balance constraints at every transportation 20. A similar limitation of this sort is junction point, which would have had to caused by the use of transport packages. have been by coal type, and would have Although a necessary simplification in terms made the number of constraints much too of model tractability, it limits the flexibility large to be included. With path variables, it of the model to freely optimize investment is possible to uniquely specify the heat, spending in the transport sector. For sulfur, and ash of each kind of coal at every instance, if the first phase of an origin. Plus, the use of path variables infrastructure package is economically enables the policy ipakers to determine the justified but the second is not, the model is source(s) of each user's coal. forced to make an all-or-nothing choice on 18. Another strength of the model is that both investments together. In summary, the mixed-integer programming is model results are really only as good as the straightforward, well-known, and understood paths and packages that are included as by planners and economists in China and at inputs. the World Bank. There are no heuristics 21. The second type of methodological used here with unknown performance levels. simplification is caused by linearity. Many Also, the model solution is the result of a economic phenomena tend to be highly single run of a single model, so there are no nonlinear. Some examples from this model questions about suboptimality that are often would include: congestion costs which raised by decomposing a model into several increase exponentially as links approach parts or by using a family of linked models. their capacities; shortage costs per ton or Certain problems dictate the use of these kWh which also tend to increase at an other techniques, but for every heuristic increasing rate; economies of scale which used, the "black box" gets more and more tend to produce flattening total cost opaque. In this case it must be considered a functions; environmental impacts which tend plus that such a complex problem has been to increase suddenly as thresholds are made amenable to such a straightforward surpassed; and demand which is a nonlinear solution methodology. function of price. In the CTS, congestion costs are avoided by imposing 'brick wall" Limitations capacities on arcs; shortage costs are linearized based on the international price of 19. The first and foremost limitation is coal; economies of utilization (not of scale) the limited choice set, and the most limiting are approximated by a total cost function case of this is the use of the path variable with a fixed-charge intercept and fiatte methodology. While path variables are the slope; and environmental impacts are Annex 5 5-5 ignored in favor of ash and sulfur contents point on the load curve). Some accuracy is which are more quantifiable units. (Demand, inevitably lost in this approximation. A a more complex and central matter, is second major approximation by the cas is discussed in detail in the third section of this that peak demands are pooled for the entire Annex.) These are all standard linearization grid instead of being satisfied at each node. techniques, but they must be recognized as This implicitly assumes that electricity approximations. In the future, it is possible management within a grid does not need to to enhance the model by using a step be planned at the national level. Another function for coal shortage costs, and thus example, outside of the electricity sector, is eliminate the shortage upper bounds. that the CTS model ignores other coal 22. The third type of limitation is caused characteristics besides heat, sulfur, and ash. by drawing an artificial line around the system and, by necessity, having to designate some activities as being part of the Demand in the CTS system's environment rather than part of the Linear Programming Model system itself. Despite t' : model's already broad scope, the CTS model is still quite Exogenously-Specified Demands, No limited by the fact that many energy and Explicit Prices environmental activities could not have been included as endogenous variables. Most 24. From the outset, it must be limiting is the omission of technologies that acknowledged that the CaS model is reduce ash emissions, because it leaves the primarily a supply-side model. It is not an model with only indirect methods of equilibrium model that solves for- the reducing ash (e.g., changing coal types, welfare-maximizing supplies, demand and substituting hydropower, washing coal), prices. The strategies that come out of the which is inconsistent with the way sulfur is model are, of course, price sensitive, and treated in the model. The social costs of obviously, prices have to be determined by pollution are also well beyo:ad the scope of an equilibrium argument. This model is not this model. It must be recognized that designed for determining prices. In fact, the reducing ash and sulfur content of the model deals only with economic costs, not delivered coal is not the true ultimate goal of prices. However, price assumptions can be environmental policy in China. Likewise, (and have been) made in forecasting leaving out ash disposal and boiler demands. maintenance activities makes it harder to 25. In the CTS's linear programming evaluate the benefits of steam coal washing. methodology, demands for coal and Energy conservation, so valuable for electricity are exogenous inputs that drive reducing shortages and cleaning the the production and distribution activities on environment, must also be modeled the supply side. Demands for three kinds of exogenously, though it is being added to the nonelectricity coal (anthracite, coking and model (along with C02 constraints) in a steam) must be specified for each of 49 spinoff research project funded by a World zones in 1995, 2000, and 2005. Coal Bank McNamara Fellowship. demand is for tons of standard coal (5,500 23. The fourth type of limitation is kcal per kg), which can be satisfied by less oversimplifying or ignoring complex than one ton of better quality coal or more matters. Oversimplification is primarily than one ton of lesser quality coal. visible in the electricity sector of the CTS Likewise, demand for electricity, in kwh, is model. For instance, the CaS model does specified at 38 demand zones in each time not explicitly use load curves for electricity period, but is subdivided into separate demand. Instead, the CTS satisfies total demands for the agricultural, urban, rural, demand for kWh (the area under the load light industry, and heavy industry curve) and peak demand for kW (the highest components, each of which is separately 5-6 Annex 5 forecasted. Coal demand for the electricity the coastal regions, washes more coal, sector is not exogenously specified, but is an substitutes more expensive hyrdopower for indirect result of the demand for electricity, coal, ships more coal by water, takes more which pu!ls electricity from hydropower, round-about routes, and otherwise attempts nuclear or thermal power plants. The to satisfy that demand at the lowest cost. thermal plants in turn pull coal to those This process continues up until the marginal plants from mines, through washeries, and delivered cost reaches the shortage cost across the transport network. Transport specified for the coal type, zone, time demand in this study is thus a derived period, and user type (in the case of demand: the actual coal flows are an electricity). After that, further increases in endogenous result of competition between the demand forecast would lead only to types of power, coal origins, coal types, higher shortages. transport modes, etc. IUsefulness of a Cost-Minimizing Model Low, Medium, and High Demand for China's Emerging Market Economy Scenarios 28. Models such as these are useful for 26. Obviously, there is a great deal of decision support for a whole variety of uncertainty surrounding any forecast of decisions that the SPC may be grappling future coal and electricity demand, with, both policy decisions and investment especially as prices are deregulated and as decisions. In this multisectoral physical one looks further and further into the future. system, there is a complex set of logistical The major way of dealing with this interrelationships that cannot be understood uncertainty in this model is to run separate in any other way. Twhat is not to say that the scenarios with low, medium, and high many minds that together make up the demand forecasts. Ideally, an even more market will not eventually allocate resources theoretically appealing way to use this kind efficiently. But the methodology affords an of model would be to iterate between a opportunity to evaluate various complex general equilibrium model and production- options more completely than the current distribution models like the CTS. The transition market economy can evaluate procedure would be to get price and demand them. The methodology allows the forecasts for coal, electricity, oil and gas dominance of complex options over one from the equilibrium model; pass them to another to be evaluated in a rough way the CaS model or similar oil and gas before proceeding to a more refined models; solve the various production- economic analysis. Furthermore, distribution models; pass delivered cost and externalities such az environmental effects shortage information back to the equilibrium may not be comprehended by the system of model; convert delivered cost to p-ice in prices in China for many years, if ever. some fashion; restimate prices and demands; 29. In China's socialist market econiomy, and iterate until consistent solutions are -much economic decision making has been reached. This approach is being taken in the decentralized. lust because a variable is in US Department of Energy's new National the model does not mean that the Energy Modeling System. government has to centrally control it. Of 27. In the absence of such an course, many of these decisions are internal equilibrium model, the next best thing is to to firms. The variables are included in the run the CTS model with a range of different model so the relationships and tradeoffs demand assumptions. As the demand can be analyzed. With the model, the key forecast is raised from low to medium to Chinese decisonmakers at least have a better high over three different scenarios, the sense of where profit-maximizing (cost- average delivered cost of energy is driven up minimizing) firms might build thermal plants as the model mines more expensive coal in and where they might build hydro than if Annex 5 5-7 those essentially decentralized decisions had import coal in the coastal areas to alleviate been left out of the model. the coal shortage. Presently, some provinces 30. A cost-minimizing model need not like Guangdong have followed this distract attention away from the actual proposal.' This is a pro-reform impact of market policy instruments the government is the model. likely to be able to use. In fact, it can do 31. Also, might not this model help exactly the opposite. For instance, despite convince firms to investigate these the fact that most industrial coal users strategies? In fact, other Bank studies have already pay at least the long-run marginal analyzed coal washing and transmission. costs for transport, and therefore the market Although their results provided many useful signal that would encourage firms to wash insights, these studies had to assume an coal are present, steam coal washing has not origin, a destination, a coal type with a taken off in China. There is no well- particular ash, sulfur and heat content, a developed market for washed coal yet. We distance, a terrain, a mode, etc. If any one cannot yet rely too much on China's not- of those factors differs, the analysis no fully developed markets. In the US, 50% of longer applies. So, how convincing would coal is washed, and even a larger percentage those analyses be to a Chinese power or in Europe. The main reason why market mining company located elsewhere? On the forces have not yet caused an increase in other hand, the CTS mode! not only coal washing is that the transaction costs considers all of those factors explicitly, but (institutional barriers) for washing might be also takes into account whether, iii the grand too high and the environmental policies for network scheme of things, there is likely to internalizing social costs of pollution are not be enough transport capacity. If a Chinese in place. Therefore, these model results add entrepreneur saw these results in which (a) fuel to the policy dialogue regarding there are likely to be bottleneckcs on the lowering those transaction costs, removing routes to his or her city, and (b) that steam those institutional barriers, and internalizing coal could not only be washed and externalities. To cite another example, Mr. transported to his or her city for a cost less Gui Shiyong, Vice Chairman of the SPC (in than the cost of importing coal, but (c) could a letter to the management science prize pay for itself in terms of weight reduction, commission) wrote that the "CTS proposed might not that be useful information? for the fir.st time that it is reasonable to Annex 6 6-1 FIGURES AND MAPS ON THE CTS ANALYSIS SYSTEM 6.1: Generalized Network Diagram of the Coal-Electricity Delivery System in the Optimization Model COAL COAL COAL - - _COALENIOMTACNRL NODES N OD TRANSPORT PATHS NODEMASD (PROVINCES) Now New ~~~Steam E Coal 1 ______To_Sleam Coa 4 New Slal \ Generaing Nodes Now Sttea CoSI/ To Middlings Coal Now ~ ~ ~ ~ ~ ~ ~ ff Genemafng Nodes E)dsfl ~ ~ (tns w StaSam Existing Towsh --I~~~~ State remnovedS5 Now and carbon t Sf 1on Esng(tons) (tns s.c.) I Coldng sulhrr . Angtrae Consumpton _ u,urV ns EAstingons(tons) Wsn, (tons s.c.) PRh (tonC) EiAstin /I (lons)~ ~ a cabo Codn - A«alConsu)mptkm - S. v 1~~~~~~~~~~~~~~at (tn 1 I N, I 6.1- Gnerlized Network Diagram oftel (conetrinud)t ENVIROuNpMRolS Ft cENERATIO MODLtKSES COAL (ttoOnss ,S,) ll_ AHlSR).* poWER P>s p r odu_~~~~~~~~~idfn | Xl . _ Strll coa llw JI|II1 F 51m M dd ngs C1a3 s S. NO' Big InterUGrid U)ks S ulfu ,, \I l tewldubl W 3 ovediJ< _ ftsris° Ini rancn| Ash ubbens) % a %=_cneslnr¢t \~~~~~~~~~~~~~~S-l =nm I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Inu F.K N-t ALIMPORT | /~~~~~~~~~~~~~~~~ I 0 C | L I U L II O 4 Y < I Z I~~~10NG t V ^ \ ,.t~~LITANG I%1. (/ 0 200 400 600 800 II5AIK'OU - apprx. 317m. m _rl / ,,!IAILAIt ...IAIL ARI £ hNu I " - < ,/ < ~~~~~~~~~~~~~~~~~~~~IIAERLIN J. .IILN(Y N \ ,z, ~ZIIONG%VCI lia A \ t2btANtIOU f tt At - £ < \ < I B\ t IN^4 C E NtS U >\ ;;2~ ~~N EXPOROT| g t-1~~~~~~~~~iNZI IMPOIIT ,0 ~~ ) / Y - ~UI { NC ,ILNG ANG k Q Lz ;IzlJ> 9 4 ~~ \ t I 9 NNI NG, o71iLI zf u 03 200 400 600 800 I * apprx. 317 mi. 6-6 Annex 6 6.4: The CTS Electricity Transmission Network 0 dt~~~~~~~~~~~~ j -~~~~~~~~~~~~~~~~ a_ CD CD L~~~~~~~~~~~~~C C.)~~~~~~~~~~~~~~~~~~C U3 ~ ~ z~~~~U 2 1~~~~~\ /~~~~~~~~~~~~~~~~~~~~W Inp'a Data hI MlthPlro Fi Reatainal Dats htormediato Dnta Outpu II I Agg*egaon, Noa-oma wd I Procssing |Dataembr K Dala Mad~~~~~~~~~Pro~~atx \ > F*adf \ i | X 1 lRead ro ModW SdchiUc (wP FoITNaSUn) Annex 7 7-1 DESCRIPON OF 1993 SCENARIO ASsumTONS Scenario Assumptions 1. Thirteen scenarios were run mental reductions are imposed. Major from July-October, 1993. They can be investment projects are optimized. divided into two groups, according to the handling of the investment decisions for Case 93-2-Huanghua Railway Base Case major projects (i.e., transport projects, large dams, and 500 kV DC power lines, modeled 3. Same as base case, except it with 0-I variables). In Case 93-1 (the base assumes that the railway from Shijiazhuang case) and Case 93-10, the investment deci- (continuing from Shenmu) to Huanghua sions for major projects are optimized. (new port) can be completed by 2000, These cases are valuable for policy purposes whereas the base case assumed (accurately because they endogenously compare trans- so, at the time) that the railway could not port investments with the nontransport reach the port by 2000. substitutes, like hydopower and coal 4. The main result: shortages in washing, under various combinations of 2000 fall by 7 million tons (18 percent), and conditions. (However, it should be noted systemwide costs fall by 0.22 percent; 6 that the yes-or-no decisions that the model million more tons of coal washed in 2000. makes about transport projects gives no consideration to other commodities, i.e., the Case 93-3-Ash and Sulfur Reduced by decisions are based on coal's share of the 10 Percent projects' investnent cost and capacity.) In the other cases, the major projects were 5. Constraints were added to fixed at their values from Case 93-1 (1 =to forcibly reduce the national total of ash and be built, O=not to be built). In essence, this sulfur content by 10 percent below Case latter group of cases can tell us how well the 93-1 (unconstrained) levels. Major set of major projects optimized for the investment projects are the same as in Case 9 percent GNP forecast would be able to 93-1. meet China's energy supply needs under 6. The main result: systemwide various conditions. The exception to these costs rise by 3.12 percent compared with two groups is Case 93-2, which differs from base case. the base case only in assuming that the railway from Shijiazhuang (continuing from Case 93-4-Ash and Sulfur Reduced by Shenmu) to Huanghua (new port) can be 20 percent completed by 2000. 7. Same as Case 93-3, but Case 93-1-Bae Case (9 percent GNP, no reduced by 20 percent. policy changes) S. The main result: systemwide costs rise by an additional 7.5 percent com- 2. Coal demand in the year 2000 pared with Case 93-3. is 906 million tons of standard coal per year, which represents an assumed elasticity Case 93-5-15 Percent Increase in Coal of 0.3 (percentage growth in coal Washing Investment Costs demandlpercentage growth in GNP). Elec- tricity demand in 2000 is 1.44 trillion kwh, 9. All coal washing investment assuming an elasticity of 1.0. No environ- costs are increased by 15 percent. Major 7-2 Annex 7 investment projects are the same as in Case the 9 Percent Case. Major investment 93-1. This case determines the cost projects are the same as in Case 93-1. sensitivity of the coal washing strategy. 18. The main result: coal and 10. The main result: the steam coal electricity shortages in 2000 are a combined washing rate falls only 0.1 percent from 138 million tons. Hydropower and coal 16.3 to 16.2 percent. washing are increased slightly, but intergrid transmission must fall, as less electricity Case 93-6-15 Percent Increase in Elec- production is considered surplus available to tricity Transmission Investment Costs be transmitted to other grids. 11. All electricity transmission Case 93-10-10.5 Percent GNP Case (with investment costs are increased by 15 per- reoptimized transport capacity) cent. Major investment projects are the same as in Case 93-1. This case determines the 19. Same as Case 93-9, except cost sensitivity of the long-distance transmis- major project investment variables are sion strategy. reoptimized using a 3-stage heuristic 12. The main result: transmission optimization process yielding an all 0-1 but from Nanning falls by 22 percent, most probably suboptimal solution. others remain stable. 20. The main result: the only difference between the railway projects in Case 93-7-Steam Coal Washing Benefit this case and those in the base case is the Case addition of the Shijiazhuang-Huanghua railway line; it was assumed that construc- 13. Upper bounds were added to tion could be accelerated given faster forcibly prevent any new steam coal washing growth. All the railway projects that can capacity from being built. Major investment increase railway throughput were already projects are the same as in Case 93-1. This chosen in the 9 percent GNP case. As with case is used for estimating the benefit of Cases 93-1 and 9302, the addition of the steam coal washing investments, by Huanghua port by 2000 results in 7 million comparing the total cost with that of Case tons fewer shortages. 93-I. 14. The main result: systemwide Case 93-11-Triple New Railway Capacity costs increase by almost 2 percent. Case Case 93-S-Large Transmission Lines 21. A question that can be asked Benefilt Case about these runs is whether the model's decisions to invest in "railway substitutes' 15. Same as Case 93-7, but for like coal washing, hydropower, and trans- new 500 kV power transmission line mission are really economically optimal capacity. from a cost-minimizing point of view, or are 16. The main result: systemwide they suboptimal investments that are forced costs increase by 0.8 percent. into the model's solution due to lack of enough railway project alternatives? This Case 93-9-10.5 Percent GNP Case (using scenario is a hypothetical case that can shed 9 percent transport capacity) some light on that key theoretical question. In this model run, all assumptions are identi- 17. GNP growth is assumed to be cal to Case 93-10, except that the 0-1 vari- rapid. Electricity demand in 2000 is ables for transport investments are allowed assumed to be 1.66 trillion kwh, 15 pcent to vary continuously between 0 and 3, higher than the 9 Percent case. Coal demand meaning that any new railway or port is 926 million tons, 2.2 percent higher than project can be tripled in size by tripling the Annex 7 7-3 investment. This case provides only a rough 26. The main result: virtually no cut answer to the question, because tripling shortages in 2000 (under I million tons coal the size is not a realistic option for some equivalent). arcs, and because this option was not given to existing railway arcs with no expansion Comparison of Investment Options plans. between the CTS Model 22. The main result: the steam coal and Five-Year Plans washing rate actually rises, since greater transport capacity ;llows more washed coal 27. The ideal way to test the from the energy base to be shipped long- intersectoral investment strategy is for the distance, implying that steam coal washing potential investment options to outnumber by pays for itself by lowering transport costs. far the investment requirements for each Hydropower is virtually unchanged from sector (mining, washing, transport, and other high demand cases, but intergrid electricity). Then the model will be able to electricity transmission falls by 40 percent, choose among them and to coordinate them. by far the lowest level of any scenario. In these runs, the model does indeed have this flexibility, but less so for the transport Case 93-12--High Demand, High Shortage sector. To a certain extent, this does create Costs a situation ia which the transport sector, which has the least flexible set of investment 23. Same as Case 93-9 (10.5 per- projects, may force suboptimal investments cent GNP), except coal and electricity to be chosen in the other sectors. shortage costs are Y 350 per ton instead of 29. In general, all of the planned Y 300, 17 percent higher. Electricity projects in the relevant sectors have been shortage costs are assumed to be higher by included as options, either as individual the same percentage. The purpose of this projects or aggregated with others, but they case is to determine the sensitivity of the have been supplemented by additional results to the shortage cost assumption. The projects to an extent that varies from sector results tell us how much of the unsatisfied to sector. Generally, two types of data demand in the high demand scenarios is a sources were used for identifying a list of logistical shortage and how much is an potential additional projects to give the economic shortage, i.e., demand could be model some flexibility in coming up with an satisfied but at a delivered cost up to optimal investment plan. First, for 17 percent higher. standardized technologies, such as coal 24. The main result: the combined washing, coal power plants, ships, and coal-electricity shortage drops from 138 scnrbbers, the model allows any amount of million tons in 2000 to 125 million tons, new capacity to be built at all possible meaning that only 13 million tons of unsatis- places, except in a few cases where water or fied demand is because of too high delivered land availability is an issue. (In the model costs. results, the only place where the water availability constraint on coal washing was Case 93-13-7.5 Percent GNP Case (using binding was in Datong.) Second, for site- 9 percent transport capacity) specific investments or those constrained by resources, such as coal mining, hydropower, 25. GNP growth is assumed to be and transmission investments, the specific slower. Electricity demand in 2000 is list of planned projects is supplemented by assumed to be 1.25 trillion kwh, 13 percent extra projects that are (a) planned for lower than the 9 Percent case. Coal demand beyond 9FYP; (b) proposed, but not is U&5 million tons, 2.4 percent lower than included in a plan; (c) suggested by CrS the 9 Percent Case. Major investment team members or by a panel of experts; (d) projects are the same as in Case 93-1. suggested by the SPC; (e) a modified scale 7-4 Annex 7 or timing of planned or proposed projects; model relative to experts' estimates of what or (f) across-the-board capacity expansion projects are most likely to be in the FYPs potential. Table 7.1 shows the flexibility in through the year 2000. investment options accorded to the CTS Table 7.1: Comparison of Potential Capacity Increases CTS Model versus Experts' Estimates Unit Experts' Estimate CS Modei Coal MiningLa million tons 406 855 Coal Washing million tons 30 No Limit New Rail Lines& route-kn 10,339 10,939 Ports/c million tons 450 203 Ships million dwt 5 No Limit Slurry Pipelines million tons 15 25 Power Plants Thermal million kw 185 No Limit Hydro million kw 62 100 Nuclear million kw 4 45 Transrmission Lines 500 kV DC km 3,000 2,340 other km n.a. 25,760 kwh n.a. No Limit Scrubbers million tons n.a. No Limit /a The experts' estimates are not really comparable because the experts' estimates were net of mine retirements, while the CTS model figures are for absolute potential increases. The experts' figures therefore underestimate the real capacity increases, especially for township mines. L& Both the CTS model and the Experts' Estimates include the Handan-Jinan railway (under long- term consideration, but not in the FYP), the Hailar-Harbin expansion, and the Shenmu-Xian line (not planned). The 600 additional k3n of railways include possible lines from Sbuichen to Chengdu and Chongqin (now planned), and a local line from Xinyi to Nantong. Not shown on this table are the additional capacity, in tkm, on existing or planned lines. & The experts' estimate of port capacity is for all commodities, not just for coal. Also, the CTS Model's potential is now known to be too small, having left off planned port expansion at user- owned ports in South China. Annex 8 8-1 CTS BASE CASE (9 PERCENT GNP GROWTH): CALIBRATION AND SELECTED RESULTS Calibration or the CTS Model which only about 100 are origins and desti- nations for coal. Therefore, the total traffic, 1. The model was calibrated over tlkm, and average railway distances are not nearly a year, and was thoroughly checked strictly comparable to reality or to the Plan. against the current or planned system on a This is especially true with regards to self- number of aggregate performance measures. supply by individual nodes. A worst case Table 8. 1 summarizes the comparison be- example of this might be Kunming, which tween the Case 93-2's model results and the represents all of Yunnan province in the targets agreed upon in advance for calibra- model. Coal flows from any part of Yunnan tion. Case 93-2 is used because, at this mo- to any other part, no matter how far, are ment, it is deemed to be more realistic than treated as a local flow that is not loaded onto Case 93-1, since the accelerated railway the rail network. Overall, the model tends construction program now calls for finishing toward self-supply whenever possible so as the Huanghua port and the Shennu-Huang- to keep the network capacity free for long hua railway by 2000. distance traffic. 2. The targets are based on past, 5. On the other hand, sometimes the planned, or desired activities on an aggre- network structure can have the opposite gate level. At this level of aggregation, the effect, where two nearby places, assigned to test case is roughly in line with expectations different network nodes, are not given the on these points. Shortages in the energy base local trucking option in the model. Some of are practically nonexistent (see Table 8.1). the structuralw shortages that appear in all Total coal and electricity production are far scenarios may be due to the way that coal higher than the targets published in accor- mines are aggregated into the Taiyuan and dance with the 6 percent GNP growth plans, Shijazhuang nodes. In China today, about 20 but are within a few percent of the SPC's million tons are transported out of the ener- new unpublished targets for 2000. The gy base by truck, and the Plan calls for thermal and hydropower shares are within about 30 million tons. Even Chinese experts 1 percent of their current levels. The biggest cannot say precisely from where to where deviation from current practice is in the rail- this coal is trucked, but some of it goes water modal split, which is significantly from Taiyuan (in the energy base) to Shijia- shifted toward water transport, which is zhuang (outside of the energy base), which consistent with the planned trend. is only about 200 km. However, in the 3. Table 8.2-8.9 show detailed results model, all coal from Taiyuan to Shijiazhu- for Case 93-2. The tables show subtotals by ang must go by rail because they are in region or by node for the major sectors in different nodes. This means that some of the the coal-electricity system. The tables are all precious 360 million tons of capacity leading produced automatically by the CTS analysis out of the energy base-all of which is used system. in every model run-is used for this short distance traffic, thus causing a structural Technical Note Regarding Shortages and shortage in all runs of the model. The model Transport Figures in the CTS builds new transmission capacity from Taiy- uan to Shijiazhuang, equivalent to 5-10 mil- 4. Some discrepancies between the lion tons of coal, but there is little doubt that model and reality may be due to the network some of the 30 million tons of shortage that structure of the CTS (or any similar model). exists in all cases is due to this network The CTS network has about 200 nodes, of problem. 8-2 Annex 8 Table 8.1: Comparison of Case 93-2 Outputs with Base Case Targets Target Base Case 93-2 Total Raw Coal Production 1.5/i 1.6 in Year 2000 (billion tons) Total Electricity Production 1.2b i.5 in Year 2000 (trillion kwh) 1.38-1.48/& Shortages in Year 2000 (percent of demand) Energy Base Area Industrial Steam Coal 0X5 0.0 Coking Coal O/d 0.0 Anthracite Coal 0/4 0.0 China Total Industrial Steam Coal n.a. 0.0 Coking Coal n.a. 0.0 Anthracite Coal n.a. 0.1 Electricity n.a. 3.9 Combined Coal-Electr. n.a. 1.8 Modal Split in Year 2000 (percent share of ton-km) Rail 63.5L& 53.0 Waterway 36.5Le 47.0 Generating Capacity in Year 2000 (percent of total) Thermal 75.4L2 77.0 Hydro 24.6Le 22.6 Nuclear 1.6ff 0.4 La = forecasted target under 8-9 percent GNP target. L2 = forecasted target under 6 percent GNP growth; new target not yet announced. J = Yangzhou Thermal Power Project Staff Appraisal Report (November 1993) ad = desired target. L = current share. lf = under construction share. n.a. = not applicable, i.e., no basis exists for forecast or reasonable desired target. Note: The energy base is defined here as the provinces of Shanxi, Shaanxi, Henan, Nei Mongol, Ningxia, and Guizhou. Annex 8 8-3 Table 8.2: National Totals of Coal Production (10,000 tons) 1995 2000 2005 Anthracite Step 1 4155.88 4145.72 4195.76 Step 2 4148.34 4531.66 4845.96 Step 3 4401.62 5344.23 7120.45 Total 12705.84 14021.61 16162.16 Steancoal Step 1 44859.84 49522.6E 57375.38 Step 2 21321.24 24231.09 27293.97 Step 3 34678.63 53104.30 62930.46 Total 100859.72 126858.05 147599.80 Cokingcoal Step 1 31.50 133.24 123.98 Step 2 2359.01 2894.85 2856.45 Step 3 14739.72 16756.80 19867.06 Total 17130.23 19784.89 22847.49 Step 1 49047.23 37.5% 53801.62 33.5% 61695.12 33.1% Step 2 27828.59 21.3% 31657.60 19.7% 34996.38 18.8% Step 3 53819.97 41.2% 75205.33 46.8% 89917.96 48.2% Tatal 130695.79 160664.55 186609.47 Step 1 = township. Step 2 = provincial. Step 3 = state. 8-4 Annex 8 Table 8.3: Regional Subtotals of Coal Production (10,000 tons) Reg. Year Anthracite Coking Coal Steam Coal All Types New Total New Total New Total New ToLal N 1995 1055.0 5236.2 603.2 5827.4 17312.9 40638.5 15971.1 51702.1 2000 1622.9 5392.9 1047.2 7036.7 33365.6 52291.6 36035.7 64761.2 2005 2325.4 6524.1 1193.9 7010.0 46795.8 62062.3 50315.1 75596.5 NW 1995 11.0 125.6 322.7 3275.7 4290.7 13490.8 4624.4 16392.1 2000 12.0 123.1 889.3 3792.3 7181.2 15338.5 9082.5 19253.9 2005 25.0 132.1 1533.0 4486.0 9147.1 16421.4 10755.1 21039.6 E 1995 359.5 1906.4 1175.4 4021.4 2136.9 11170.8 3721.7 17093.6 2000 506.1 1356.2 1492.4 4023.4 5172.5 13199.5 7171.3 19079.1 2005 940.1 2061.7 3662.4 6097.4 3S09.9 15677.6 13112.4 23836.7 SC 1993 625.3 3025.1 116.9 1332.2 3594.6 11652.7 433'.8 16010.1 2000 1263.1 4132.9 183.2 1373.4 6451.6 14236.0 7898.6 19742.3 200S 2067.0 4557.9 388.9 1541.4 9592.5 15833.7 12048.4 21933.0 SW 1995 248.3 1866.4 60.0 1683.2 3021.5 10801.6 3329.3 14351.2 2000 445.4 1950.5 3S2.7 2146.1 5242.1 11620.0 6040.2 15716.5 2005 B72.7 2253.7 405.0 2120.7 7960.2 12759.9 9237.9 17134.2 NW 1995 340.0 546.1 0.0 990.3 6691.8 13105.2 7031.8 14641.7 2000 380.6 576.1 165.4 1363.1 13431.8 191B9.2 13977.9 21123.4 2005 453.0 632.6 449.8 1592.0 13272.9 23344.9 19175.7 25569.5 Tar 1995 2639.2 12705.3 2278.2 17130.2 37098.3 100859.7 42015.6 130695.8 2000 4230.9 14021.6 4130.2 19784.9 71323.4 126858.1 80169.4 160664.6 2005 6683.2 16162.2 7683.0 22947.5 101778.4 147599.5 116144.6 186609.5 ES 1995 1547.7 6166.5 590.5 6066.0 22516.3 52545.5 24654.5 64777.9 20O0 2403.4 6779.9 1060.6 7!83.0 45497.5 71094.0 48961.5 85056.9 2005 3229.4 7736.5 1301.0 7175.4 62970.1 83916.1 67500.5 98823.0 co 1995 597.8 3093.9 1439.1 7817.7 4753.3 19933.6 6790.1 30845.1 20D0 941.0 3483.3 1870.0 8234.1 9716.2 23356.7 12527.2 35104.1 2005 1648.0 3839.2 4121.7 10401.4 14166.1 26204.3 19935.8 40494.9 N=NORTlH (HUABEl) NE =NORTlHEASr (DONGEE!) E=EAST (HAUDONO) SC=SOUTH CFNTRAL (ZHONGNAN) SW=SOUTFHWESr (XNAN) NW= NORTHWESr (XIBEI) IMP = IdPORTS TOr =NAllONAUL TOrTAJL EB=ENERGY BASE (NORTH AND SOl.mIWrFS) CO =COASTAL PROVINCES Annex 8 8-5 Table 8.4: Regional Breakdown of Coal Washing (10,000 tons) Region Period Coal Type Raw Coal Washed Coal Existing New Total Huabei 1995 7464.9 7402.0 14866.8 10753.9 Steam-2 4085.3 4954.1 9039.4 6779.6 Coking 3379.6 2447.8 5827.4 3974.3 -clean 3291.8 -mid 682.5 2000 6755.2 11549.4 18304.6 13356.6 Steam-2 3187.0 8031.0 11218.0 8413.5 Coking 3568.2 3518.4 7086.7 4943.1 -clean 4159.6 -mid 783.5 Dongbei 1995 4193.4 1888.3 6081.7 4270.7 Steam-2 1637.7 1168.3 2806.0 2104.5 Coking 2555.7 720.0 3275.7 2166.2 -clean 1724.1 -mid 442.0 2000 5158.0 2338.6 7496.6 5255.7 Steam-2 2536.0 1168.3 3704.3 2778.2 Coking 2622.0 1170.3 3792.3 2477.5 -clean 1936.8 -mid 540.7 Huadong 1995 3780.0 1048.8 4828.8 3530.2 Steam-2 190.0 617.4 807.4 605.6 Coking 3590.0 431.4 4021.4 2924.7 -clean 2294.7 -mid 630.0 2000 3606.0 1802.5 5408.5 3956.5 Steam-2 190.0 1195.1 1385.1 1038.8 Coking 3416.0 607.4 4023.4 2917.7 -clean 2331.1 -mid 586.5 Zhongnan 1995 659.1 963.0 1622.1 1193.2 Steam-2 82.1 207.8 289.9 155.8 Coking 577.0 755.2 1332.2 1037.4 -clean 976.6 -mid 60.8 2000 607.5 2100.9 2708.5 2074.3 Steam-2 13.3 1321.8 1335.1 1001.3 Coking 594.2 779.2 1373.4 1073.0 -clean 1010.1 -mid 62.9 8--6 Annex 8 Region Period Coal Type Raw Coal Washed Coal Existing New Total Xinan 1995 1365.0 1614.2 2979.2 2054.3 Stean-2 96.0 1200.0 1296.0 972.0 Coking 1269.0 414.2 1683.2 1082.3 -clean 888.6 -mid 193.7 2000 1601.8 1909.1 3510.9 2397.1 Steam-2 164.8 1200.0 1364.8 1023.6 Coking 1437.0 709.1 2146.1 1373.4 -clean 1120.0 -mid 253.5 Xibei 1995 554.0 1869.4 2423.4 1816.2 Stean-2 104.0 1329.1 1433.1 1074.8 Coking 450.0 540.3 990.3 741.4 -clean 559.4 -mid 182.0 2000 554.0 2961.2 3515.2 2662.6 Steam-2 104.0 2048.1 2152.1 1614.1 Coking 450.0 913.1 1363.1 1048.6 -clean 825.2 -mid 223.4 Total 1995 18016.4 14785.7 32; 1 23680.1 Steam-2 6195.1 9476.7 15671.9 11753.9 Coking 11821.3 5308.9 17130.2 11926.2 -clean 9735.2 -mid 2191.0 2000 18282.6 22661.6 40944.2 29702.8 Steam-2 6195.1 14964.2 21159.4 15869.5 Coking 12087.5 7697.4 [9784.9 13833.3 -clean 11382.7 -mid 2450.5 Annex 8 8-7 Table 8.5: Regional Coal Balance (10,000 tons of std coal) Region Period End User Demand Supply Shortage Huabei 1995 ANTHUSER 3961.2 3961.2 0.0 COKIUSER 2099.3 2099.3 0.0 INDUUSER 10654.3 10654.3 0.0 TOTAL 16714.9 16714.9 0.0 2000 ANTHUSER 4540.4 4540.4 0.0 COKIUSER 2467.1 2467.1 0.0 INDUUSER 11480.3 11480.3 0.0 TOTAL 18487.8 18487.8 0.0 Dongbei 1995 ANTHUSER 915.1 762.5 152.6 COKIUSER 1871-4 1871.4 0.0 INDUUSER 9732.2 9457.0 275.1 TOTAL 12518.7 12090.9 427.8 2000 ANTHUSER 1042.6 1042.6 0.0 COKIUSER 2047.7 2047.7 0.0 INDUUSER 12025.7 12025.7 0.0 TOTAL 15116.0 15116.0 0.0 Huadong 1995 ANTHUSER 3916.1 3916.1 0.0 COKIUSER 2575.7 2575.7 0.0 INDUUSER 10872.1 10134.7 737.3 TOTAL 17363.9 16626.6 737.3 2000 ANTHUSER 4097.2 4097.2 0.0 COKIUSER 3131.8 3131.8 0.0 INDUUSER 12406.6 12406.6 0.0 TOTAL 19635.6 19635.6 0.0 Zhongnan 1995 ANTHUSER 3129.2 3129.2 0.0 COKIUSER 1374.7 1374.7 0.0 INDUUSER 9873.5 9873.5 0.0 TOTAL 14377.4 14377.4 0.0 2000 ANTHUSER 3220.8 3220.8 0.0 COKIUSER 1541.5 1541.5 0.0 INDUUSER 11621.5 11621.5 0.0 TOTAL 16383.8 16383.8 0.0 Xinan 1995 ANTHUSER 1750.5 1750.5 0.0 COKIUSER 925.0 925.0 0.0 INDUUSER 5969.1 5969.1 0.0 TOTAL 8644.5 8644.5 0.0 2000 ANTHUSER 1833.4 1833.4 0.0 COkIUSER 1201.5 1201.5 0.0 INDUUSER 7653.1 7653.1 0.0 TOTAL 10688.0 10688.0 0.0 8-8 Annex 8 Region Period Coal Type Raw Coal Washed Coal Existing New Total Xibei 1995 ANTHUSER 526.5 507.1 19.5 COKIUSER 444.0 444.0 0.0 INDUUSER 5187.8 5187.8 0.0 TOTAL 6158.4 6138.9 19.5 2000 ANTHUSER 578.5 557.6 20.9 COKIUUSER 548.2 548.2 0.0 INDUUSER 6025.3 6025.3 0.0 TOTAL 7152.0 7131.2 20.9 Export 1995 ANTHUSER 0.0 0.0 0.0 COKIUSER 445.0 445.0 0.0 INDUUSER 2225.0 2225.0 0.0 TOTAL 2670.0 2670.0 0.0 2000 ANTHUSER 0.0 0.0 0.0 COKUIJSER 445.0 445.0 0.0 INDUUSER 2670.0 2670.0 0.0 TOTAL 3115.0 3115.0 0.0 Total 1995 ANTHUSER 14198.6 14026.5 172.1 COKIUSER 9735.2 9735.2 0.0 INDUUSER 54513.9 53501.4 1012.5 TOTAL 78447.7 77263.2 1184.6 2000 ANTHIUSER 15313.0 15292.i 20.9 COKIUSER 11382.7 11382.7 0.0 INDUUSER 63882.5 63882.5 0.0 TOTAL 90578.2 90557.3 20.9 Annex 8 8-9 Table 8.6: Coal Transportation O-D Table, in 2000 (10,000 tons) To: Minenode Huabei Dongbei Huadong Zhongnan Xinan Xibei Export Total From: Huabei 33656 7120 10496 7135 43 2240 60690 Dongbei 17005 17005 Huadong 17099 17099 Zhongnan 3150 15909 19059 Xinan 343 157 13897 14397 Xibei 1506 930 3916 1307 485 11976 156 20276 Imort 0 Total 35162 25055 35003 24507 14382 12019 2397 148526 8-10 Annex 8 Table 8.7: Electricity Transmission O-D Table (100 milion kwh) To: Huabei Dongbei Huadong Zhongnan Xinan Xibei Huanan Total From: Huabei TI 734.50 325.14 228.33 115.92 0.00 0.00 0.00 1403.89 T2 1115.86 593.44 288.22 115.92 0.00 0.00 168.00 2281.43 T3 1499.57 838.26 293.15 24.62 0.00 0.00 336.00 3051.59 Dongbei Tl 0.00 76.49 0.00 0.00 0.00 0.00 0.00 76.49 T2 0.00 70.97 0.00 0.00 0.00 0.00 0.00 70.97 T3 0.00 69.45 0.00 0.00 0.00 0.00 0.00 69.45 Huadong Ti 0.00 0.00 1427.54 0.00 0.00 0.00 0.00 1427.54 T2 0.00 0.00 1640.18 0.00 0.00 0.00 0.00 1640.18 T3 0.00 0.00 2508.05 0.00 0.00 0.00 0.00 2508.05 Zhongnan Ti 0.00 0.00 0.00 103.21 0.00 0.00 0.00 103.21 T2 0.00 0.00 0.00 235.68 0.00 0.00 0.00 235.68 T3 0.00 0.00 0.00 285.88 0.00 0.00 0.00 285.88 Xinan TI 0.00 0.00 0.00 0.00 44.87 0.00 0.00 44.87 12 0.00 0.00 0.00 0.00 90.21 0.00 0.00 90.21 T3 0.00 0.00 0.00 45.56 122.63 0.00 23.77 191.95 Xibei Ti 0.00 0.00 0.00 47.74 0.00 47.62 0.00 95.36 T2 0.00 0.00 0.00 47.74 0.00 56.00 0.00 103.74 T3 0.00 0.00 0.00 456.69 0.00 92.07 0.00 548.76 Huanan TI 0.00 0.00 0.00 0.00 0.00 0.00 67.07 67.07 12 0.00 0.00 0.00 0.00 0.00 0.00 67.07 67.07 T3 0.00 0.00 0.00 0.00 0.00 0.00 273.12 273.12 Total TI 734.50 401.62 1655.87 266.88 44.87 47.62 67.07 3218.43 T2 1115.86 664.41 1928.40 399.34 90.21 56.00 235.07 4489.28 T3 1499.57 907.71 2801.20 872.75 122.63 92.07 632.89 6928.80 T1=1995 12=2000 13=2005 Annex 8 8-11 Table 8.8: Power Plant Capacity Coal Hydro Nuclear Elgnode Year Exist New Total Exist New Total Exist New Total Grand Harbin Ti 693.0 438.6 1131.6 18.8 0.0 18.8 0.0 0.0 0.0 1150.4 T2 O558.3 679.8 1338.2 18.8 0.0 18.8 0.0 0.0 0.0 1356.9 T3 623.7 783.2 1406.9 18.8 0.0 18.8 0.0 0.0 0.0 1425.7 Hfailar TI 50.0 69.0 119.0 0.0 0.0 0.0 0.0 0.0 0.0 119.0 T2 47.5 486.1 533.6 0.0 0.0 0.0 0.0 0.0 0.0 533.6 T3 45.0 927.5 972.5 0.0 0.0 0.0 0.0 0.0 0.0 972-5 Changchu TI 429.0 7.5 436.5 280.4 15.0 295.4 0.0 0.0 0.0 731.9 T2 407.5 36.7 444.2 280.4 81.0 361.4 0.0 0.0 0.0 805.6 T3 386.1 398.8 784.9 280.4 96.0 376.4 0.0 0.0 0.0 1161.3 Shenyang TI 666.5 27.6 694.1 116.6 10.0 126.6 0.0 0.0 0.0 820.7 T2 633.2 332.7 965.8 116.6 20.0 136.6 0.0 0.0 0.0 1102.4 T3 599.8 883.7 1483.5 116.6 120.0 236.6 0.0 0.0 0.0 1720.2 Qinhuang TI 1.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 12 0.9 0.1 1.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 T3 0.9 0.1 1.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 Tongliao Ti 172.0 0.0 172.0 0.0 0.0 0.0 0.0 0.0 0.0 172.0 T2 163.4 263.1 426.5 0.0 0.0 0.0 0.0 0.0 0.0 426.5 T3 154.8 345.3 500.1 0.0 0.0 0.0 0.0 0.0 0.0 500.1 Dalian TI 150.0 0.0 150.0 0.0 0.0 0.0 0.0 0.0 0.0 150.0 12 142.5 97.1 239.6 0.0 0.0 0.0 0.0 0.0 0.0 239.6 T3 135.0 266.5 401.5 0.0 0.0 0.0 0.0 0.0 0.0 401.5 Yingkou Tl 60.0 0.0 60.0 0.0 0.0 0.0 0.0 0.0 0.0 60.0 T2 57.0 0.0 57.0 0.0 0.0 0.0 0.0 0.0 0.0 57.0 T3 54.0 0.0 54.0 0.0 0.0 0.0 0.0 0.0 0.0 54.0 Dandong Ti 1.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 T2 0.9 0.0 0.9 0.0 0.0 0.0 0.0 0.0 0.0 0.9 T3 0.9 0.0 0.9 0.0 0.0 0.0 0.0 0.0 0.0 0.9 Baotou Ti 276.0 84.1 360.1 3.4 0.0 3.4 0.0 0.0 0.0 363.5 T2 262.2 252.6 514.8 3.4 0.0 3.4 0.0 0.0 0.0 518.2 T3 248.4 486.1 734.5 3.4 0.0 3.4 0.0 0.0 0.0 737.9 Tangshan Ti 395.0 42.8 437.8 0.0 0.0 0.0 0.0 0.0 0.0 437.8 T2 375.2 43.6 418.8 0.0 0.0 0.0 0.0 V.0 0.0 418.8 T3 355.5 45.2 400.7 0.0 0.0 0.0 0.0 0.0 0.0 400.7 Beijing TI 230.0 0.0 230.0 27.1 0.0 27.1 0.0 0.0 0.0 257.1 T2 218.5 0.0 218.5 27.1 0.0 27.1 0.0 0.0 0.0 245.6 T3 207.0 0.0 207.0 27.1 0.0 27.1 0.0 0.0 0.0 234.1 Tianjin Tl 260.0 0.0 260.0 0.0 0.0 0.0 0.0 0.0 0.0 260.0 T2 247.0 0.0 247.0 0.0 0.0 0.0 0.0 0.0 0.0 247.0 T3 234.0 0.0 234.0 0.0 0.0 0.0 0.0 0.0 0.0 234.0 Datong Ti 420.0 227.4 647.4 0.0 0.0 0.0 0.0 0.0 0.0 647.4 T2 399.0 227.4 626.4 0.0 0.0 0.0 0.0 0.0 0.0 626.4 T3 378.0 605.8 983.8 0.0 0.0 0.0 0.0 0.0 0.0 983.8 Zhungeer Ti 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 12 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 T3 0.0 4.5 4.5 0.0 0.0 0.0 0.0 0.0 0.0 4.5 Hebaopia Ti 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 T2 0.0 422.5 422.5 0.0 0.0 0.0 0.0 0.0 0.0 422.5 T3 0.0 644.6 644.6 0.0 0.0 0.0 0.0 0.0 0.0 644.6 Taiyuan TI 276.0 870.3 1146.3 22.8 0.0 22.8 0.0 0.0 0.0 1169.0 8-12 Annex 8 T2 262.2 1000 1262.2 22.8 102.0 124.8 0.0 0.0 0.0 1387.0 T3 248.4 1000 1248.4 22.8 242.0 264.8 0.0 0.0 0.0 1513.2 Changzhi Ti 84.0 54.9 138.9 0.0 0.0 0.0 0.0 0.0 0.0 138.9 T2 79.8 554.9 634.7 0.0 0.0 0.0 0.0 0.0 0.0 634.7 T3 75.6 1000 1075.6 0.0 0.0 0.0 0.0 0.0 0.0 1075.6 Houma Ti 40.0 481.6 521.6 0.0 0.0 0.0 0.0 0.0 0.0 521.6 T2 38.0 500.0 538.0 0.0 0.0 0.0 0.0 0.0 0.0 538.0 T3 36.0 500.0 536.0 0.0 0.0 0.0 0.0 0.0 0.0 536.0 Shijiazh Ti 444.0 31.1 475.1 0.0 0.0 0.0 0.0 0.0 0.0 475.1 T2 421.8 464.2 886.0 0.0 0.0 0.0 0.0 0.0 0.0 886.0 T3 399.6 788.6 1188.2 0.0 0.0 D.0 0.0 0.0 0.0 1188.2 Jingn Ti 507.0 30.3 537.3 7.5 0.0 7.5 0.0 0.0 0.0 544.8 12 481.6 867.9 1349.5 7.5 0.0 7.5 0.0 0.0 0.0 1357.0 T3 456.3 1457.9 1914.2 7.5 0.0 7.5 0.0 0.0 0.0 1921.7 Weifang Ti 235.0 0.0 235.0 0.0 0.0 0.0 0.0 0.0 0.0 235.0 12 223.2 0.0 223.2 0.0 0.0 0.0 0.0 0.0 0.0 223.2 T3 211.5 0.0 211.5 0.0 0.0 0.0 0.0 0.0 0.0 211.5 Yanzhou Ti 265.0 0.0 265.0 0.0 0.0 0.0 0.0 0.0 0.0 265.0 12 251.8 0.0 251.8 0.0 0.0 0.0 0.0 0.0 0.0 251.8 T3 238.5 29.1 267.6 0.0 0.0 0.0 0.0 0.0 0.0 267.6 Hefei Ti 88.9 96.2 185.1 52.0 0.0 52.0 0.0 0.0 0.0 237.0 T2 84.5 364.0 448.5 52.0 0.0 52.0 0.0 0.0 0.0 500.4 T3 80.0 364.0 444.0 52.0 0.0 52.0 0.0 0.0 0.0 496.0 Huainan Ti 213.5 0.0 213.5 0.0 0.0 0.0 0.0 0.0 0.0 213.5 T2 202.8 0.0 202.8 0.0 0.0 0.0 0.0 0.0 0.0 202.8 13 192.1 388.9 581.1 0.0 0.0 0.0 0.0 0.0 0.0 581.1 Huaibei Ti 102.0 0.0 102.0 0.0 0.0 0.0 0.0 0.0 0.0 102.0 T2 96.9 57.8 154.7 0.0 0.0 0.0 0.0 0.0 0.0 154.7 T3 91.8 313.2 405.0 0.0 0.0 0.0 0.0 0.0 0.0 405.0 Xuzhou Ti 416.5 12.2 428.7 0.0 0.0 0.0 0.0 0.0 0.0 428.7 T2 395.7 12.2 407.9 0.0 0.0 0.0 0.0 0.0 0.0 407.9 T3 374.9 830.4 1205.2 0.0 0.0 0.0 0.0 0.0 0.0 1205.2 Nanjing Ti 65.0 0.0 65.0 0.0 0.0 0.0 0.0 0.0 0.0 65.0 T1 61.8 100.0 161.8 0.0 0.0 0.0 0.0 0.0 0.0 161.8 T3 58.5 100.0 158.5 0.0 0.0 0.0 0.0 0.0 0.0 158.5 Jianbi Ti 617.7 0.0 617.7 0.0 0.0 0.0 0.0 0.0 0.0 617.7 12 586.8 444.3 1031.1 0.0 0.0 0.0 0.0 0.0 0.0 1031.1 T3 555.9 444.3 1000.2 0.0 0.0 0.0 0.0 0.0 0.0 1000.2 Nantong Ti 127.0 0.0 127.0 0.0 0.0 0.0 0.0 0.0 0.0 127.0 T2 127.0 0o0 127.0 0.0 0.0 0.0 0.0 0.0 0.0 127.0 T3 127.0 133.1 260.1 0.0 0.0 0.0 0.0 0.0 0.0 260.1 Wangting TI 122.0 0.0 122.0 0.0 0.0 0.0 0.0 0.0 0.0 122.0 T2 115.9 0.0 115.9 0.0 0.0 0.0 0.0 0.0 0.0 115.9 T3 109.8 0.0 109.8 0.0 0.0 0.0 0.0 0.0 0.0 109.8 Shanghai TI 670.0 87.1 757.1 0.0 0.0 0.0 30.0 0.0 30.0 787.1 12 636.5 710.1 1346.6 0.0 0.0 0.0 30.0 0.0 30.0 1376.6 T3 603.0 988.6 1591.6 0.0 0.0 0.0 30.0 0.0 30.0 1621.6 Hangzhou Ti 117.0 24.0 141.0 233.2 30.0 263.2 0.0 0.0 0.0 404.2 T2 111.2 317.8 428.9 233.2 60.0 293.2 0.0 0.0 0.0 722.1 T3 105.3 349.2 454.5 233.2 60.0 293.2 0.0 0.0 0.0 747.7 Ningbo Ti 270.0 0.0 270.0 0.0 0.0 0.0 0.0 0.0 0.0 270.0 T2 256.5 0.0 256.5 0.0 0.0 0.0 0.0 0.0 0.0 256.5 T3 243.0 83.3 326.3 0.0 0.0 0.0 0.0 0.0 0.0 326.3 Wenzhou Ti 144.5 0.0 144.5 0.0 0.0 0.0 0.0 0.0 0.0 144.5 T2 137.3 0.0 137.3 0.0 0.0 0.0 0.0 0.0 0.0 137.3 Annex 8 8-13 T3 130.1 0.0 130.1 0.0 0.0 0.0 0.0 0.0 0.0 130.1 Fuzhou TI 106.2 0.0 106.2 317.2 45.0 362.2 0.0 0.0 0.0 468.4 T2 100.9 0.0 100.9 317.2 162.0 479.2 0.0 0.0 0.0 580.1 T3 95.6 0.0 95.6 317.2 282.0 599.2 0.0 0.0 0.0 694.8 Xiamen Ti 55.0 0.0 55.0 0.0 0.0 0.0 0.0 0.0 0.0 55.0 T2 52.2 87.2 139.4 0.0 0.0 0.0 0.0 0.0 0.0 139.4 T3 49.5 205.6 255.1 0.0 0.0 0.0 0.0 0.0 0.0 255.1 Nanchang Ti 171.0 81.1 252.1 139.5 25.0 164.5 0.0 0.0 0.0 416.6 T2 162.4 293.4 455.8 139.5 50.0 189.5 0.0 0.0 0.0 645.3 T3 153.9 534.0 687.9 139.5 50.0 189.5 0.0 0.0 0.0 877.4 Changsha Ti 237.5 189.0 426.4 349.7 50.0 399.7 0.0 0.0 0.0 826.1 T2 225.6 481.2 706.8 349.7 186.0 535.7 0.0 0.0 0.0 1242.6 T3 213.7 927.0 1140.7 349.7 296.0 645.7 0.0 0.0 0.0 1786.4 Yueyarage Ti 70.0 0.0 70.0 0.0 0.0 0.0 0.0 0.0 0.0 70.0 T2 66.5 0.0 66.5 0.0 n.0 0.0 0.0 0.0 0.0 66.5 T3 63.0 0.0 63.0 0.0 0.0 0.0 0.0 0.0 0.0 63.0 Wuhan Tl 310.0 0.0 310.0 595.0 50.0 645.0 0.0 0.0 0.0 955.0 T2 294.5 93.9 388.4 595.0 168.8 763.8 0.0 0.0 0.0 1152.2 T3 279.0 93.9 372.9 595.0 597.8 1192.8 0.0 0.0 0.0 1565.7 Zhengzho Ti 762.1 0.0 762.1 46.8 0.0 46.8 0.0 0.0 0.0 808.9 T2 724.0 805.4 1529.4 46.8 0.0 46.8 0.0 0.0 0.0 1576.2 T3 685.9 805.4 1491.3 46.8 0.0 46.8 0.0 0.0 0.0 1538.1 Guangzho Ti 820.0 0.0 820.0 282.2 35.0 317.2 60.0 0.0 60.0 1197.2 T2 779.0 208.9 987.9 282.2 70.0 352.2 60.0 0.0 60.0 1400.1 T3 656.0 208.9 864.9 282.2 70.0 352.2 60.0 0.0 60.0 1277.1 Shantou Ti 1.0 70.5 71.5 0.0 0.0 0.0 0.0 0.0 0 0 71.5 12 0.9 95.6 96.6 0.0 0.0 0.0 0.0 0.0 'J.0 96.6 T3 0.9 142.1 143.0 0.0 0.0 0.0 0.0 0.0 0.0 143.0 Haikou Ti 52.3 0.0 52.3 52.8 24.0 76.8 0.0 0.0 0.0 129.1 T2 49.7 8.9 58.6 52.8 24.0 76.8 0.0 0.0 0.0 135.3 T-3 47.0 43.8 90.9 52.8 24.0 76.8 0.0 0.0 0.0 167.6 Nanning T1 154.0 7.1 161.1 397.9 45.0 442.9 0.0 0.0 0.0 604.0 t2 146.3 48.6 194.9 397.9 194.0 591.9 0.0 0.0 0.0 786.8 T13 138.6 241.4 380.0 397.9 772.1 1170.0 0.0 0.0 0.0 1550.0 Kunrming TI 139.0 17.8 156.8 359.4 30.0 389.4 0.0 0.0 0.0 546.2 T2 132.0 17.8 149.9 359.4 195.0 554.4 0.0 0.0 0.0 704.2 T3 125.1 78.3 203.3 359.4 305.0 664.4 0.0 0.0 0.0 867.7 Guiyang TI 192.9 28.4 221.2 204.3 30.0 234.3 0.0 0.0 0.0 455.6 T2 183.2 81.9 265.1 204.3 60.0 264.3 0.0 0.0 0.0 529.5 T3 173.6 81.9 255.5 204.3 425.0 629.3 0.0 0.0 0.0 884.8 Chengdu TI 249.0 3.5 252.5 398.8 90.0 488.8 0.0 0.0 0.0 741.4 T2 236.6 3.5 240.1 398.8 634.8 1033.6 0.0 0.0 0.0 1273.7 T3 224.1 3.5 227.6 398.8 1316.0 1714.8 0.0 0.0 0.0 1942.5 Chongqin Ti 310.0 226.8 536.8 20.0 0.0 20.0 0.0 0.0 0.0 556.8 T2 294.5 230.4 524.9 20.0 68.0 88.0 0.0 0.0 0.0 612.9 T3 279.0 236.3 515.3 20.0 226.0 246.0 0.0 0.0 0.0 761.3 Xian Ti 365.6 149.9 515.4 116.8 0.0 116.8 0.0 0.0 0.0 632.3 12 347.3 441.3 788.5 116.8 0.0 116.8 0.0 0.0 0.0 905.4 T3 329.0 1650.3 1979.3 116.8 0.0 116.8 0.0 0.0 0.0 2096.1 Shenmu TI 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 T2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 T3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Yinchuan TI 105.1 65.1 170.2 27.6 0.0 27.6 0.0 0.0 0.0 197.8 T2 99.8 97.4 197.3 27.6 44.0 71.6 0.0 0.0 0.0 268.9 T3 94.6 288.5 383.1 27.6 44.0 71.6 0.0 0.0 0.0 454.7 B-14 Annex 8 Xining TI 40.4 0.0 40.4 174.9 15.0 189.9 0.0 0.0 0.0 230.2 T2 38.4 0.0 38.4 174.9 227.7 402.5 0.0 0.0 0.0 440.9 T3 36.3 0.0 36.3 174.9 416.0 590.8 0.0 0.0 0.0 627.2 Lanzhou Ti 198.5 225.1 423.6 233.5 25.0 258.5 0.0 0.0 0.0 682.1 T2 188.6 237.2 425.8 233.5 X1.3 314.8 0.0 0.0 0.0 740.6 T3 178.7 278.2 456.8 233.5 254.5 488.0 0.0 0.0 0.0 944.8 Wulumuqi TI 165.7 109.2 274.9 44.5 0.0 44.5 0.0 0.0 0.0 319.4 T2 157.4 210.1 367.6 44.5 0.0 44.5 0.0 0.0 0.0 412.0 T3 149.2 406.3 555.5 44.5 0.0 44.5 0.0 0.0 0.0 599.9 Mannshan Ti 79.5 0.0 79.5 0.0 0.0 0.0 0.0 0.0 0.0 79.5 T2 75.5 0.0 75.5 0.0 0.0 0.0 0.0 0.0 0.0 75.5 T3 71.6 0.0 71.6 0.0 0.0 0.0 0.0 0.0 0.0 71.6 Jiujiang Ti 75.0 0.0 75.0 0.0 0.0 0.0 0.0 0.0 0.0 75.0 T2 75.0 0.0 75.0 0.0 0.0 0.0 0.0 0.0 0.0 75.0 T 75.0 0.0 75.0 0.0 0.0 0.0 0.0 0.0 0.0 75.0 Total Ti 13268 3758 17026 4523 519 5042 90 0 90 22158 12 12615 11678 24293 4523 2429 6951 90 0 90 31334 T3 11880 20387 32267 4523 5596 10119 90 0 90 42476 (PCT) TI 78% 22% 90% 10% 1% 0% 77% 23% 0% 12 52% 48% 65% 35% 1% 0% 78% 22% 0% T3 37% 63% 45% 55% 1% 0% 76% 24% 0% Annex £ B-1S Table 8.9: Ash and Sulfur Content of Delivered Coal, by Province (in 10,000 tons) Province Period Ash Sulfur Beijing 1995 546.0 28.1 Beijing 2000 562.5 29.7 Beijing 2005 517.0 26.4 Tianjin 1995 437.9 27.5 Tianjin 2000 428.5 28.3 Tianjin 2005 464.3 31.5 Hebei 1995 1457.0 91.1 Hebei 2000 1744.1 123.4 Hebei 2005 1959.4 159.8 Shanxi 1995 2324.6 201.5 Shanxi 2000 2722.5 233.5 Shanxi 2005 3006.1 244.7 Neimeng 1995 832.0 60.2 Neimeng 2000 1365.9 93.4 Neimeng 2005 1867.7 120.5 Liaoning 1995 1401.4 79.4 Liaoining 2000 1925.6 98.9 Liaoning 2005 1745.2 93.2 Jilin 1995 1089.9 32.6 Jilin 2000 1082.3 34.8 JilH 2005 1201.6 35.4 Heilongj 1995 1437.7 45.8 Hcilongj 2000 1922.5 58.6 Heiongj 2005 2229.0 62.5 Shanghai 1995 419.2 37.7 Shanghai 2000 627.9 58.3 Shanghai 2005 783.1 74.2 Jiangsu 1995 1373.9 110.2 Jiangsu 2000 1492.1 111.4 Jiangsu 2005 2082.1 151.1 Zhejig 1995 465.4 36.9 Zhejiang 2000 559.5 38.7 Zhejiang 2005 621.1 47.0 Anhui 1995 601.6 39.5 4nlui 2000 719.0 51.9 Anhui 2005 1229.9 88.3 Fujian 1995 345.1 21.4 Fujian 2000 353.1 24.0 Fujian 2005 459.9 29.3 Jiangxi 1995 606.5 29.2 Jiangxi 2000 803.0 38.3 Jiangxi 2005 731.4 39.4 8-16 Annex B Shandong 1995 1124.5 133.5 Shandong 2000 1548.1 153.1 Shandong 2005 1539.4 146.6 Henan 1995 1536.6 88.6 Henan 2000 1905.1 125.5 Henan 2005 2039.7 131.4 Hubei 1995 718.2 46.9 Hubei 2000 865.7 60.8 Hubei 2005 982.3 65.1 Hunan 1995 750.5 60.4 Hunan 2000 929.5 78.3 Hunan 2005 1143.3 85.9 Guangdon 1995 760.8 53.1 Guangdon 2000 853.4 65.2 Guangdon 2005 1080.3 77.7 Guangxi 1995 430.6 24.1 Guangxi 2000 458.2 24.9 Guangxi 2005 606.3 33.3 Hainan 1995 41.9 3.3 Hainan 2000 51.7 4.0 Hainan 2005 61.9 4.7 Sichuan 1995 1802.3 197.0 Sichuan 2000 1896.5 194.0 Sichuan 2005 2119.3 220.3 Guizhou 1995 627.7 38.3 Guizhou 2000 767.1 45.6 Guizhou 2005 774.5 49.5 Yunnan 1995 574.3 26.7 Yunnan 2000 719.8 32.5 Yu.nian 2005 845.2 38.1 Shaanxi 1995 851.9 46.8 Shaanxi 2000 1065.7 59.1 Shaanxi 2005 1561.9 89.0 Gansu 1995 532.5 18.5 Gansu 2000 575.0 20.3 Gansu 2005 633.0 22.8 Qinghai 1995 58.0 3.1 Qinghai 2000 59.1 3.2 Qinghai 2005 64.7 3.5 Ningxia 1995 299.7 10.1 Ningxia 2000 358.4 12.3 Ningxia 2005 513.8 17.9 Xinjiang 1995 471.7 23.6 Xinjiang 2000 588.2 29.2 Xinjiang 2005 761.0 38.0 Export 1995 334.6 32.7 Export 2000 399.6 27.1 Annex 8 8-17 Export 2005 446.1 38.8 Total 1995 24253.9 1647.5 2000 29349.9 1958.1 2005 34070.5 2265.8 Annex 9 9-1 SUMMARY OF RESULTS FOR ALL 1993 SCENARIOS Table 9.1: Summary Solution Results Case Coal Electr. Total Total Coal Electr. Totl Demand Demand System Invest- Shtg. Shtg. Shtg. in 2000 in 2000 Disc. ment in in 2000 in 2000 in 2000 (million (bil. kwh) Coat 8-9FYP (mil. (bil. (mil. std tons) (bil. Y) (bil. Y) std tons) kwh) std tons) 93-1. 9% GNP 906 1444 897 660 0.2 67.6 39.3 93-2. 9% GNP wJ 906 1444 895 671 0.2 58.4 32.3 New Port Railway 93-3. Sulfur, 906 1444 925 654 3.4 95.6 55.9 Ash -10% 934. Sulfur, 906 1444 994 637 67.3 110.6 128.2 Ash -20% 93-5. Wash Inv 906 1444 898 660 0.2 67.6 37.3 Cost +15% 93-6. Transm mIv 906 1444 899 659 0.2 69.2 38.3 Cost +15% 93-7. No New 906 1444 914 658 3.5 86.7 51.6 Steam Coal Washing 93-8. No New 906 1444 904 664 0.2 64.1 35.5 Big Transmiss. Lines 93-9. GNP 926 1661 1151 748 80.0 105.6 138.1 10.5%, Same Rail as 93-1 93-10. GNP 926 1661 1147 759 75.7 101.7 131.6 10.5%, Rail Reoptimized 93-11. GNP 926 1661 1093 741 47.7 110.2 108.3 10.5%, Triple New Rail Capac 93-12. GNP 926 1661 1185 766 74.9 91.3 125.1 10.5%, Iigh Shortage Cost 93-13. GNP 885 1258 713 571 0.04 1.1 0.65 7.5%. Same Rail as 93-i 9-2 Annex 9 Table 9.2: Coal Production Sector Results Case Coal Yr. 2000 Coal Yr. 2000 Tonm Steam Coal Production Prod. from Coal Prod. Washed in Wahing in 2000 North and SW from Coastal 2000 Rate (million Energy Base Regions (mil. tons) in 2000 tons) (mil tons.) (muil. tons) (percent) 93-1. 9% GNP 1601 855 349 403 16.3 93-2. 9% GNP wI 1607 851 351 409 16.7 New Port Railway 93-3. Sulfur, 1576 839 341 550 28.5 Ash -10% 93-4. Sulfur, 148S 808 305 582 33.4 Ash -20% 93-5. Wash Inv 1601 855 349 403 16.2 Cost +15% 93-6. Transia Inv 1599 853 349 404 16.3 Cost +15% 93-7. No New 1589 842 349 n.a. n.a. Steam Coal Washing ___ 93-8. No New 1604 857 350 405 16.4 Big Transmiss. Lines ¶i3-9. GNP 1638 867 359 417 16.5 10.5%. Same Rail = 93-1 93-10. GNP 1643 865 361 429 17.3 10.5%. Rail Reoptimized 93 -11. GNP 1655 895 347 437 18.2 10.5%, Triple New Rail Capac. __ _ _ 93-12. GNP 1648 868 363 451 18.9 10.5 %, High Shortage Cost 93-13. GNP 1514 814 326 384 16.3 7.5 %. Same Rail as 93-1. Annex 9 9-3 Table 9.3: Electricity Production Sector Solution Resmlts Case Electricity Total Total Nuclear I Thermal-Hydro Generated in Thermal Hydro Capacity Built Shares of Eloc. 2000 Capacity Capacity Prod. in 2000 (biUion kwh) by 2000 by 2000 by 2000 (percent) (million kw) (million kw) (million kw) 93-1. 9% GNP 1507 240 70 .90 81.8- 17.8 93-2. 9% GNP w/ 1515 243 70 .90 82.1 - 17.5 New Port Railway 93-3. Sulfur, 1478 234 71 .90 80.4 - 19.1 Ash -10% 93-4. Sulfur, 1459 229 73 .90 80.4 - 19.1 Ash -20% 93-5. Wash Inv 1507 240 70 .90 81.8 - 17.8 Cost +15S% 93-6. Tranm aIv 150S 240 70 .90 81.8 - 17.8 Cost +-15% 93-7. No New 1490 236 71 .90 81.2 - 18.3 Steam Coal Washing , 93-8. No New 1511 244 69 .90 82.1 - 17.5 Big Tranamiss. Lines 93-9. GNP 1702 279 73 .90 83.2 - 16.4 10.5%, Same Rail as 93-1 93-10. GNP 1704 280 73 .90 83.2 - 16.4 10.5%. Rail Reoptimized _ 93-11. GNP 1688 277 73 .90 83.1 - 16.6 10.5%, Triple New Rail Capac. 93-12. GNP 1716 282 74 .90 83.2 - 16.4 10.5%, High Shortage Cost 93-13. GNP 1372 216 63 .90 81.9 - 17.7 7.5%, Same Rail as 93-1 9-4 Annex 9 Table 9.4: Electricity Transmission Solution Results Case Total Electricity Electricity Net Elec. Electricity Electricity Intergrid Transmis- Transmis- tricity Transmis- Transmission Electricity sion from Sion from E. Tranmis- sion from to Hunan, Transmis- Shanxi Pr. Nei sion to Nanning in Hubei, & sion in 2000 in 2000 Mongolia in Huadong big 2000 Henan in (bil. kwh) (bil. kwh) 2000 (bil. grid in 2000 (bil. kwh) 2000 kwh) (bil. k:wh) (bil. kwh) 93-1. 9% GNP 135.3 131.4 48.4 31.3 12.0 23.6 93-2. 9% GNP wl 121.3 131.4 49.1 28.8 6.7 16.4 New Port Railway 93-3. Sulfur. 127.3 125.5 40.0 41.8 10.3 18.8 Ash -10% _ 93-4. Sulfur, 102.3 119.1 24.4 26.2 10.9 26.1 Ash -20% 93-5. Wash Iv 135.4 131.4 48.3 31.4 12.1 23.6 Cost +15% _ 93-6. Transm Inv 133.9 131.5 46.1 31.3 9.3 22.8 Cost +15% _ 93-7. No New 147.8 134.4 55.5 30.1 14.1 27.5 Steam Coal Washing _ _ . 93-8. No New n.a. n.a. n.a. n.a. n.a. n.a. Big Transmiss. Lines . 93-9 GNP 115.9 135.6 51.0 19.5 14.8 19.3 10.5%, Same Rail as 93-1 93-10. GNP 102.4 142.0 45.1 16.9 12.0 13.4 10.5%, Rail Reoptimized 93-11. GNP 81.4 114.9 11.6 28.6 0 16.1 10.5%, Triple New Rail Capac. 93-12. GNP 104.8 143.1 16.6 15.6 13.2 15.4 10.5%, Hijgh Short.ige Cost 93-13. GNP 99.3 89.1 4S.4 23.4 6.9 11.7 7.5%, Same Rail as 93-1 Annex 9 9-5 Table 9.5: Transportation Sector Solution Results Case ToWa Ton-km Average Rail- Rai- Shipping Number Traffic in 2000 Transp. Water Water Fleet of Botdle- in 2000 (billion Ditance Shares of Shares of lavatmnt. necks in (million tkm) in 2000 Tonnage Tkm in by 2000 2000 tons) (rail km/ in 2000 2000 (bil. Y) waterkin (percent) (percent) 93-i. 9% GNP 952 952 671/2268 79-21 53-47 8.64 58 93-2. 9% GNP WJ 937 941 671/2310 80-20 53-47 10.43 64 New Port Railway 93-3. Sulfur, 925 996 67012806 81-19 50-50 8.29 63 Ash -10% 93-4. Sufur, 909 984 67112860 81-19 50-50 7.81 59 Ash -20% 93-5. Wash Inv 950 953 673/2210 79-21 54-46 8.64 61 Cost +15% 93-6. Transm Inv 952 953 667/2369 80-20 53-47 B.65 58 Cost + 15% 93-7. No New 970 968 663/2248 79-21 53-47 8.79 62 Steam Coal Washing 93-8. No New 961 947 670/2280 79-21 54-46 9.03 60 Big Transmiss Lines 93-9. GNP 971 962 642/2362 80-20 52-48 8.65 57 10.5%, Same Rail as 93-1 93-10. GNP 968 981 624/2662 81-19 s5-50 11.43 61 10.5%, Rail Reoptimized 93-11. GNP n.a. n.a. n.a. n.a. n.a. n.a. n.a. 10.5%, Triple New Rail Capac. _ 93-12. GNP 974 1000 629/2675 80-20 49-51 11.35 56 10.5%, Eigh Shortage Costs 93-13. GN9P 914 797 666/2939 83-17 63-37 7.94 55 7.5%, Same Rail as 93-1 9-6 Annex 9 Table 9.6: Environment Sector case Total Ash TOal Scrub. Tona of Hydro Nuclear Delivered Sulfur Capacity Steam Coal Capacity Capacity in 2000 Delivered Built Washed Built Built (mil. tons) in 2000 by 2000 in 2000 by 2000 by 2000 (mil. tons) (mil. tons) (mil. tons) (mil. kw) (mil. kw) 93-1. 9% GNP 293 19.6 0 206 70 .9 93-2. 9% GNP wI 294 19.6 0 211 70 .9 Now Port Railway 93-3. Sulfur, 266 17.B .79 353 71 .9 Ash -10% 93-4. Sulfitr, 236 15.9 1.38 387 73 .9 Ash -20% 93-5. Wash lav 293 19.6 0 205 70 .9 Cost + 15% 93-6. Transm mv 293 19.6 0 206 70 .9 Cost + 15% _ 93-7. No New 308 19.9 G n.a. 71 .9 Steam Coal Washing 93-8. No New Big 294 19.6 0 207 69 .9 Transmiss. Lines 93-9. GNP 10.5%, 300 19.9 0 213 73 .9 Same Rail as 93-1 93-10. GNP 300 19.9 0 224 73 .9 10.5%, Rail Reoptimized 93-1 1. GNP 301 20.1 0 239 73 .9 10.5%, Triple Rail Capac. 93-12. GNP 299 19.9 0 246 74 .9 10.5%, High Shortage Cost 93-13. GNP 7.5%, 277 18.8 0 194 63 .9 Same Rail as 93-1 Annex 10 10-1 FIGURES AN!) MAPS ON RESULTS OF THE ANALYSIS (All figures are for 9 Percent GNP Growth, Case 93-2, unless otherwise indicated.) *HARBIN WULUMUQI >\ fS CHANGCIIU j ANSIJAN O DAOTOU DATONG -'1 DANP ONG TIA14lNJ ALIA YINCIIUAN T U YDALIAN rAlru^ L YANTAI CX~~~~~~~~~r XINING CHANGiHI WEIFANG HOUMA I XIAN ZHENGZHO EXPORT H FEI NANTONG E SHANGHAI CHENGDU WUHAN 'HANGZIIOU * CIIONGQIN JIUJIANG NINGBO * CHANGSHA WENzHOU * GUIYANG FUZHOU KUNMING / XIAMEN * GUANCZH0 (I NANING Ralio of Shortage/Demand (percent) U ~~~~~~~0.OTo'O. I 0.1 To 5.0 0 200 400 600 800 ,§7IJAIKOU -,_apprx.317mi. HARBIN 0 WULU MUQI CHANGCHU O ANSHAN BAOTOU DATONG DANDONG *YINCIIUAN IN DLA TAIYUAN - YANTAI XINING H JINAN QDAD HIOUMiA KlAN ZIIENGZIIO EXPORT NANJIN 0 hiAA ANSIAN CIIENGDU WUHAN HANGZHOU . CIONGQIN 0 JIUJIANG NINGBO . CHANGSIIA WEN7ZiOU *GUIYANG FUZHOU Ratio of Shortag/Denmand (percent) KUNMING / *0.0 TO 0.1 AXIMEN * 0.1 To *5.0 II ./ / 10.0 0.OTot1 5.0 NANNING GUkNG HO a 50To e50.0 \ N A N N I NHAIKOLJ I apprx. 317 ml. * IIAILAR KARBIN CHANGCIiU SHENYAN BAOTOU DATONG YINGKOU ZHUNGEER Tl ANJIN YINCitUAN SIIENiU k DALI AN XINING * TAIYUAN / iAN HOUNiA YANZIIOU XIAN * ZHENGZHO 0 IIUAINAN P . HEFEI. AVANGTING CHENGDU * WUHAN *- HANGZHOU o*CHIONGQIN * YUEYANGE NINGBO ry * CH A NGSII A /wENZIiOU 0 GUIYANG FUZHOU KUNNIING (n /X'IAMEN 6 Ilk IikIILl. GUANG ZHO gelDemand (percent) 0.1 To '5.0 *5.0 To '15.0 (S AOI K 0 U 0 200 400 600 800 I H IAIKOU * -, p _3 mi. . ty ~~~~~~~~~~~~~~~~~~~~ ~ ~ ~~~~1- -f apprx. 317 mi.O HAILAR ARBIN . CHANGCHU * SHENYAN O BAOTOU YINGKOU ZIIUNGEER * BEIJING QINHUANGJ YINCIIUAN 5IENUS * DALIANS XINING a 0 J1NA / S SHJINAZI llOUMA YANZHOU XIAN S ZHENGZHO C,, * NANJING *HENU * ) SHANGHAI p X CtlENGDU * WUtiAN '*lANGZHOU * CliONGQIN YUEYANGE NINGBO * CHANGSHA WENZtIOU * GUIYANG eu KUNMIING GUANCZHO (J To t0 I ~NANNING V*. I To 45.0 a5.0 To 10 0 200 400 600 gOO Ir *apprx. 317 mi. :0 I i i Bold lines Indicate onie-way or two-way bolcicnecks. 0 200 -400 600 Sao I C ? 1. .~~~~~~~~~~~~~~~~~~~~rapprx. 31 7ml.a Annex 10 1-7 Figure 10.6: Rail Bottlenecks, 2000, Figh Demand . . .~~~~~~~~~~~~~~~~~~~. aL CD 10-8 Annex 10 Figure 10.7: New Transport Projects Built n hchojan Bw fu - - - - - - - - - - - - - - - - - - - - - -- - * - . 1: - Y3 dt . *-s . Z Ct wP O;is*- , t Sihnz H "U Sh~~~Um B _ _ g4 -n Xh-" or U-7-W" -:-.' / IJ4' r ~~~~~~ - S & ; ~~~MOOEL RE-SULTS LFGEN- m~~~~~~~~~~~~~~~~~~~~~~~~7 We _rx C '.gd \ Xhyu n . ' '_ \ .#t '' "s V.kwWV 1 '. % . q R~~~~~~ ~~~~~~~~~~~~~~~~BAJZI MA . <---tND_ N~X66 ;Sanm. o0{] Zhwq w SI' -d D~~~~~~~~ 0 1~~~~~~~~ 0 0~~~C e 4. 4.~~~~~~~~~~~~otlnc 0 2==j00 400 600 100 Er~~ ~~~~~~~~~~~~~ - *apprx. 317 rn1. 10-10 Annex 10 Figure 10.9: Port Bottlenecks, 2000, High Demand I~~~~~~~~~~~~~~~~~~~.I a. 0~~~~~~~~~~~~0C I. I.- C;0 0 0 a E. Maxidmum widih = 100 million tonLs per yearI 0 200 400 60O---0 ao 6? I' apprx. 317 mi. 1 * 10-12 Anne 10 Figure 10.11: Optimal Interregional Coal Flows, 2000, Medium Demand (Mil&ions of Metric Tons) _-4 Rail traffic * -'-4 Mantmetraffic j ~~~R U S S I A -- Maritime traffic , .t2P7 0 China - statistical region r J - boundary ,70°0 Intemational boundaty *. , NORTHEAST - 6 Case CNA WEST 139.0 3 CHINA ~ ~ ~ ;~---~ EASTA HONG ~ TIA -- .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~m Case 93-2 Annex 10 10-13 Figure 10.12: Optimal Coal Ailocadon by Heat Content, from Coal Base in the Year 2000 (6 percent GNP Growth, Case 92-1) ....................... ..................................................... ...................... .......... .07- ... *...................... ................................................ --------- Z~~~~~~~~~ ........................................................... .................. C I- 4o'- ----------------7000--000-8000--000 ~3 C- I- <2 0~~~~ ~5000-6000 6 000-7000 700 0-8000 8000 Kcal. 1kg ,^,WA#J~~~~ ~ ~ ~ ~~~~~~~~ d O W4W q A ,i A K, - r ,~* t VihasN 3u , XMJinngU&a. $mgAwHubdRtQ \ t Siahuan ,@s.f@>,, XvZhZi unani FuR Wdhi ash i t 'lCSu)Hnn) Fgn Coal 100. g andsuMaurndong washed, YOu na ') )Shauf In 50-9o Wth 10% e _ of tons *u,0 0 U redLucion"oC c [nan ' mmmmomw a 5MI' .~ ~ ~ ~ ~~lnAR Hu_ . 4in., I H AlL AR HIARDIN0 / 1CHONGQINGTONG HEN N _ ) / I \1, !f~CANGHI , V\\ ~~~~~GUANG2H0/7 ANNINGZ 1 - V MaWdrum width - 50 billion kwh per year 0 200 400 600 goo U | I~~~~~~~~~~1' * aPprx. 3 17 ml. M 0% I 0 I. MaLxmum widih = 45 bililon kwh per year O, -2,00 _40_0 l 60 ===o -~ ~ ~~~ I' a*pprx. 3 17 mi. Annex 10 10-17 Figure 10. 16: Sulfur in Delivered Coal by Province, 1995-2005, Medium Demand a~~~~~~~~~~~~~~C th-nEa i~~~~~~~~~~o l7 >~~~ __ _d5Z o ½~s / MrAfP ¶it 21E I ~ ~~~~~~ .. . .... ' ~ ~ ~ ~ ~ . A'/ 'XN , CD~~~~~~~C ~~~ r ~ ~ ~ 'N tr~~~C -b t.w! w , z ss ~~~~~~~~~~~~Halbogangv C | ~~~Xlnliang Uygur A.R. |' Llon| sVt\av,s, !; X > < > / itf ^ ] ; SGans LI > *,,-, .. ,.S^-.- S \ KAS }' J ~~~~~~cm L at t dng _ 5;v except tat the 0-1 vari- In this case, demand is at the High level. ables for transport investments are allowed The CTS Team hasopened up a number of to vary continuously between 0 and 3, mea- new options, including extra capacity (and ning that any new railway or port project cost) for some key coal-only railway arcs, can be tripled in size by tripling the invest- extra mining capacity in the corresponding ment. This case provides only a rough cut energy base nodes, three coal slurry pipe- answer to the question, because tripling the lines, and unlimited coal imports for a size is not a realistic option for some arcs, number of coastal demand zones. The SPC and because this option was not given to had previously told the CTS Team not to existing railway arcs with no expansion consider slurry pipelines and coal imports as plans. part of any base case analysis. Major project Annex 12 12-1 Summary of Results for All 1992 Scenarios Table 12.1: Summary SolUtion Results Case | Coal Elcctr. Total Total Coal Elcctr. Total Dcmaud Dcmnnd Systcm Invcst- Shtg. Shtg. Shtg. in 2000 in 2000 Disc. nent in S- in 2000 in 2000 in 2000 (nillion (bi). kwh) Cost 9FYP (bil. (mil. (buil (ma. std torn) (bil. Y) y) sd tonr) kwh) Nd tons) 92-1. Mcd. Dcm. 946 1176 771 994 2.26 62.3 36.53 92-2. High Dem. 1018 1279 931 S1l0 51.00 74.0 91.70 92-3- Low Den- 916 1151 726 952 1.20 61.2 34.86 92-4. TranEport 946 1176 795 992 12.74 63.1 47.45 Capacity -10% 92-5. Transpoir 946 1176 756 996 0.52 60.4 33.74 Capacity +-10% __-_. 92-6. Transport 946 1176 813 990 6.48 62.3 40.75 Cost +100% 92-7. Sulfur, 946 1176 795 996 12.74 70.6 51.57 Ash -10% 92-8. Sulfur, 946 1176 846 970 69.12 75.1 110.43 Ash -20% 92-9. Sulfur, 946 1176 929 9}4 t81.63 78.9 225.03 Ash -30% 92-10. Budget-10% 946 1176 781 880 21.95 84.1 68.21 92-11. No New 946 1176 780 973 10.20 62.4 44.53 Stcam Coal Washing 92-12. No New Big 94.6 1176 773 961 3.17 61.4 36.94 Tranamission Lincs 92-13. SPC Case 1018 1349 1.119 1230 105.18 65.4 141.15 (High Dem., Envionment, High Sh. Cost) 92-14. 'Combined' 946 1176 876 886 46.19 72.6 86.12 or Econ. Lirnits 92-15. High Dfmd. 1018 1279 911 1123 18.4 63.4 53.27 W/ Extra Options . 92-16. Tnple New 946 1176 743 977 0.53 59.6 33.31 Rail Capacity _ _ 12-2 Annex 12 Table 12.2: Coal Production Sector Results Cas Coal Yr. 2000 Coal Yr. 2000 Tons Wuahed Steam Coat Prduction Piod. from Coal Prod. in 2000 Washing in 2000 North and SW f(om Coastal (mil. toar) Rate (million tons) Energy Baee Regiona in 2000 (mil tons.) (mil. ton) (pcrcent) 1. Med. Demand 1498 793 334 417 18.7 2. High Demnd 1567 839 348 428 18.4 3. Low Deniaand 1456 767 328 398 18.1 4. Transport 1485 781 335 406 11.4 Capacity -10% 5. Transport 1496 799 331 395 16.8 Capacity + 10% . 6. Tranport 1489 783 336 414 19.0 Coat +100% 7. Sulfur, 1465 791 311 512 28.3 Ash -10% 8. Sulfur, 1397 769 280 549 33.3 Ash -20% 9. Sulfur, 1265 712 226 529 36.7 Ash -30%9 10. Budget -10% 1467 784 324 392 18.1 11. No New 1485 787 332 257 5.4 Stean Coal Washing _ 12. No New Big 1494 791 332 414 18.4 Transmission Lines 13. SPC Came 1538 821 340 594 32.3 (High Demand, Envrornment, High Sh. Cost) 14. 'Combined" 1441 755 331 496 28.1 or Econ- Lirnmi I 15. High Dmnnd. wt 1593 856 346 427 17.7 Extm Options 16. Triple New 1495 800 329 389 16.3 Rail Capacity I I Annex 12 12-3 Table 12.3: Electricity Production Sector Solution Results Cawe Electicity Themia Hydro Nuclear Tbhrmal-Hydro- Gunirajed in Capacity Built Capacity Built Capacity Built Nuclear Sham of 2000 by 2000 by 2000 by 2000 Elec. Prod. in 2000 (billion kwh) (million kw) (million kw) (million kw) (percent) 1. Med. Demand 1214 185 64 .90 79.8/19.710.5 2. High Danand 1313 201 69 .90 79.7/19.810.5 3. Low Dcmand 1186 181 63 .90 78.9/19.5/0.5 4. Tmzrpozt 1215 183 66 .90 78.9I20.6/0.5 Capacity -10% 5. Tansport 1215 186 64 .90 79.9/19.6/0.5 Capacity +10% 6. Tansport 1214 185 65 .90 79.5s20.0u0.5 Coat +100% 7. Sulfur, 1205 180 69 .90 78.1121.3/0.5 Ash -10% 8. Sulfur, 1198 176 70 2.56 76.7/21.B/1.5 Ash -20% 9. Sulfur, 1194 175 70 3.30 75.9/22.1/1.9 Ash 30.% 10. Budget 1187 183 60 .90 80.8/18.710.5 -10%_ 11. No Ncw 1214 184 66 .90 79.3120.2/0.3 Stearn coal Washing _ 12. No New Big 1212 187 65 .90 79.5/20.010.5 Transmission Lines 13. SPC Cawe 1397 214 72 2.34 79.5119.3/1.2 (High Dcanmd, Environent, High Sh. Cost) _ 14. wCombined" 1202 183 65 .90 79.2/20.3/0.5 or Econ. LimitB 15. High Demand 1325 204 69 .90 79.8/19.7/0.5 with Extpa ___ _ 16. Triple New 1216 186 63 .90 80.1/19.3/0.5 Rail Capacity I__ _ _ _ _ _ _ I__ _ _ _ _ _ _ _ 12-4 Annex 12 Table 12.4: Electricity Transmission Solution Results Came ToWl Electricity Electricity Nt Electr. Eletbicity Eletricity Intergrid Tramimn. Transmisn. Tnamn. Trnamnian. Tranmisn. to Electr. from Shanxi fiom E. Nei to Guizhou from Hunan, Tmnmsiisn. Pr. in 2000 Mongolia Province Naning Hubei,Hcnan in 200D (bil. kwh) in 2000 in 2000 in 2000 in 2000 (bil. (bil. kwh) (bil. kwh) (bil. kwh) (bil. kwh) kwh) 1. Mcd. Demand 284.3 66.5 15.3 +1.4 8.0 4.5 2 High D1ma 323.9 8S.6 20.6 +3.3 10.1 28.0 3. Low Demand 263.1 61.0 11.6 +0.9 8.4 2.6 4. Tmnsport 315.3 84.1 20.8 +1.4 8.0 16.7 Capcily -10% S. Transport 269.8 60.0 11.6 +1.4 8.0 2.5 Capacity +10% 6. Tnsport 293.7 74.2 15.0 +1.4 8.0 11.8 Cost +100% 7. Sulfur, 29B.8 67.1 13.4 +1.4 12.0 13-3 Ash -10% B. Sulfuir, 273.6 52.9 11.6 +1.3 11.4 22.2 Ash -20% 9. Sulfur, 257.5 14.8 7.9 +13 10.1 12.9 Ash -30% 10. Budget 245.7 48.2 12.5 +1.3 9.2 2.5 -10% i11. No New 300.2 76.3 17.6 +1.4 9.1 11.2 Steam Coal W ashing__ _ _ _ _ _ __ _ _ _ _ ___ _ _ _ _ _ _ 12. No New Big 262.9 50.8 15.2 +1.4 9.1 6.0 Trnsmission Le _ ______ 13. SPC -se 323.2 66.4 13.7 0 10.9 19.3 (High Demand, Envirmnent, High Sh. Cost) 14. "Combined' 290.4 6S.7 14.0 +1.3 13.6 13.8 or Econ. Limit. 15. High Demnd 329.9 69.5 19.8 +3.3 9.4 17.7 w/ Extra Options _ 16. Triple New 264.0 63.5 11.6 +1.4 7.4 2.4 Rail Capacity Annex 12 12-5 Table 12.5: Environment Sector Case Total Ash Total Sulfur Scrub. Toro of Hydtm Nuclear Dclivcrcd in Delivered in Capacity Steam Coal Capacity Capacity 2000 2000 (mil. Built Washod Built Built (mil. tmo) towi) by 2000 in 2000 by 2000 by 2000 _ _ _ _ ____ (mil. tan) (mi!. ton) (mil. kw) (mU. kw) 1. Med. Demand 273 18.4 0 215 64 .9 2. High Dcefand 287 19.2 0 223 69 .9 3. Low Demand 268 1_.1 0 203 63 .9 4. Tmnport 272 18.3 0 212 66 .9 Capacity -10% 5. Transport 27S 18.7 0 193 64 .9 Capacity +10% _ _ _ 6. Trunspoit 273 18.5 0 218 65 .9 Cost +100% 7. Sulfur, 249 16.9 .547 321 69 .9 Ash -10% _ . B. Sulfur, 222 15.0 1.151 361 70 2.6 Ash -20% 9. Sulfur, 194 13.1 1.266 361 70 3.3 Ash -30% 10. Budget 269 18.2 0 207 59 .9 -10% 11. No New 290 19.0 0 62 66 .9 Stcam Coal Washing 12. No New Big 273 18.3 0 211 65 .9 Transuiision Lines 13. SPC Care 250 16.8 1.257 383 72 2.3 (High Demand, Environment, High Sh. Coat) _ 14. 'Conjbined' 246 16.8 .328 317 65 .9 or Econ. Limits 15. High Dmnd wl 291 19.7 0 218 69 .9 Extra Options 16. Triple New 278 IB.7 187 63 .9 Rail Capacity I Annex 13 13-1 SHORTAGE Cosr DATA ASSuMFflONS 1. In this study, the term shortaget variables. Second, because unit "shortage costs" connotes something energy delivery costs vary from region to different from the usual meaning of "shadow region, analysing the allocation or prices" in a Bank study. Shadow price refers distribution of shortages by region will be to the economic cost, as opposed to the informative for policy analysis. Third, a financial cost, of any economic activity. In very ambitious idea originally proposed for this study, the financial costs that were interpreting the shortage variables was that adopted were determined to be mostly the model could endogenously decide when market-determined, and therefore no shadow it is more economically reasonable to have price conversion factors were applied. a shortage than to continue to deliver more Shortage costs, on the other hand, refer to energy even when it is logistically possible the penalty imposed in the CTS model to do so. In this case, shortages are neither against any shortfall in the final products of strictly nor solely interpreted as logistical coal or electricity at demand nodes. bottlenecks, but also as instances where it is 2. Theoretically, ash and sulfur more economical to (a) go without energy, emissions also have a shadow price in the (b) substitute another fuel, or (c) conserve Chinese economy, based on the social cost energy. of pollution. However, the CTS team decided that detennining the shadow prices Specification of Shortage Costs for the of ash and sulfur emissions with any CTS Model accuracy was beyond the scope of this study, and therefore environmental impact is 4. Because the CTS model is a measured in physical quantities-the ash and cost-driven model, the methodology for sulfur content of delivered coal-and is not setting the numerical values of shortage endogenously traded off against cost in the costs in the objective function will determine model. which of the above-mentioned roles the shortage variables will play. Various Shortage Cost Methodology economic meanings can be attached to the shortage costs in the CTS model, which Background to Shortage Cost Discussion determine how to specify the value of shortage costs. 3. In the CTS model, shortage 5. For the first purpose described variables and shortage costs are introduced above (getting a feasible solution to a for coal and electricity end uses. There were realistic scenario), the specific value of three original purposes for introducing shortage cost is not important, although its shortage variables in the CTS model. First, general level is. This method could be in a constrained optimization model, the described as a "penalty cost method," in that demands have to be met in every zone for it simply discourages any shortages from every user type in every time period. The occurring unless absolutely necessary. The balances between supply and demand are value of the shortage cost should be constrained by many capacities of different significantly higher than any possible activities and other technical constraints. If delivered cost of energy, and then adjusted shortages are real and not just an artificial downwards in steps. The second purpose result of the model, the model should not be mentioned above (to find the geographic forced to give us unrealistic results, which allocation pattern of shortages) is partly would be expected if the model had no achieved ir the penalty cost approach as 13-2 Annex 13 well. The third purpose mentioned above whether the users are owned by the central, (letting the model decide the tradeoff local, or township government. A second between delivering more coal and suffering pricing track has emerged outside the the shortage cost) requires us to give the planned system, but the market is still far exact value of shortage costs to each end from perfect. Because the distribution of user to find out the value of economic losses coal reserves is uneven, and transportation is if shortages occur. still a rigid constraint combined with the 6. Value-Added Approach (not barriers between administrative regions, idle adopted). From the point of view of an production capacity and severe shortages economist, the shortage cost is what will be exist simultaneously in different parts of lost from the entire economic system if China. In regions where shortages were energy is in short supply, not only from the severe, coal and electricity prices soared energy sector itself. An activity-based value- upwards, while in the regions where idle added approach was considered but rejected capacity still existed, the market coal price because of three methodological and remained lower than the planned price. theoretical problems. First, suppose that I Also, the out-of-plan market is not close to ton of coal is used to produce 1 unit of a being independent of the planned economy, particular product. One cannot assume that since in-plan producers must get their supplying an additional ton of coal will marginal supplies from the out-of-plan necessarily lead to production of an market. Thus, out-of-plan prices can additional unit of product because coal is fluctuate widely from time to time and place only one of the raw materials of input and to place. not the only one in short supply. Second, the 9. The SPC has conducted a opportunity cost due to the coal shortage is survey on market coal prices. Table 13.1 not a stable linear function of the shortage, gives a rough idea of market coal prices in but a nonlmear increasing function of the 1989 in some cities. The figures are results volume of coal shortage. Third, in the CTS processed from many different survey analysis system, all industries or activities reports. In Table 13.1, the market price is that use the same type of coal have been found to differ greatly from place to place. aggregated into a single category. But surely The extremely low price in the coal base is the economic loss for each economic activity the result of spare coal production capacity in a coal type user group is not the same. there and because some small township For these three reasons, it was decided that mines, equipped with very backwards it would not be feasible to try to estimate technology, had produced coal at a much from an activity analysis the loss to the lower cost but were unable to access the economy of coal and electricity shortages. transport facilities. These mines would sell 7. Two alternatives to the activity their products at prices lower than their analysis approach for determining energy normal price, in order to pay the salaries for shortage costs are market prices and workers and managers to prevent the mines substitution costs. The latter-specifically from closing. Clearly, such a market price import prices for coal-was eventually cannot be used as a shortage cost in the adopted for this study. model, because the model would choose to 8. Market Prices (nor adopted). not supply any coal at all. Using real market prices oE coal and 10. The electricity price system in electricity as the shortage costs for each China is similar to that for coal prices in demand node seer's at first to be a plausible that a multi-track price system has been approach, but does not hold up under introduced. The basic situation is that in an scrutiny. The dual track price system is too electricity grid, a basic price of electricity is distorted. In-plan prices do not depend on fixed, but many additional fees are added for the willingness-to-pay of the users, but on various reasons. Surcharges are added for the balances planmed between sectors and high price fuels, for construction funds of Annex 13 13-3 energy and transportation that are charged is an upper bound on the economic loss by central government (and sometimes again from shortages, in the sense that they could by the local government separately), and for avoid that loss by paying for the substitute. many miscellaneous reasons. Local 13. Estimation of Coal Substitution governments will charge high prices for amport) Cost. For coal, the intemational electricity from new power plants, as the border price of coal comes closest to investment for new plants is comparatively meeting this criterion. The CIF (cost, high and the return to the investment should insurance, and freight) price of steam coal at be recouped from the prices charged to the coastal Chinese ports was computed as users. Similarly, power plants constructed follows. First, the price of standard by the semi-autonomous HuaNeng Inc. are international steam coal (6,700 Kcal/kg) mainly funded from abroad, and are allowed FOB (freight on board) New Castle, to charge high prices to users because of the Australia, is assumed to equal US$32/ton. high cost of investment and shorter payback So, the CIF price of standard international period for foreign capital. Because of severe coal at a Chinese coastal port is electricity shortages in some coastal areas, US$32 + US$10 = US$42/ton.Multiplying some diesel or gasoline generators are even by the currency conversion factor of 8.8 loaded at peak times, which may cost as then yields Y 370/ton. Converting this price much as 40-50 fen per kwh. to the equivalent price of standard Cbinese ]1. It is not reasonable to take raw coal (5,500 Kcallkg) on the basis of these kinds of market prices for coal and heat content alone gives Y 370 electricity to be valid measures of the x 5,500f6,700 or about Y 300/ton. This CIF economic loss to society due to coal or price of standard Chinese coal at coastal electricity shortages. Taking into account the ports of Y 300 per ton is adopted as the base complexity and data difficulties of market case value for the cost of coal shortage. prices, we have concluded that it would be 14. Estimation of Electricity inappropriate to use the market price as the Substtution Cost. Since about 0.5 kg of shortage cost in the CTS model. coal can generate I kwh of electricity, one 12. Subsniturion Costs (adopted). ton of coal can generate about 2,000 kwh of Since the CTS model is primarily intended electricity in the model. In this case, if Y to choose investments for economic reasons 300 were the coal shortage cost, then I kwh (it really is a long-term strategic planning of electricity shortage should be priced at model), not because of logistical constraints, about Y 0. 16 plus the annualized investment estimating the economic values of energy cost and nonfuel operating cost of one kwh shortages cannot be avoided, no matter how of electricity, with the total value around Y difficult. Therefore, the shortage costs to be 0.2 to Y 0.3 per kwh. Using diesel power as employed in the CTS model must be based the substitute, however, yields a high on economic considerations. The approach estimate around Y 0-5 per kwh. adopted for this study is to find the cost of 15. Like any linear program, the substitutes for coal or electricity supply. The CaS model is sensitive .o the coefficients of basic idea is that the economic loss should variables. Competition for coal supplies not be higher than the cost of a substitute between electricity demand and direct coal form of energy, such as oil, gas, energy demand is a basic characteristic of the CTS conservation investmnent, or imported coal. model. During calibration, electricity For example, if users stand to lose, say, shortage cost levels were moved up and Y 500 for every ton of coal that they are down. Higher electricity shortage costs denied, but they could buy imported coal for caused there to be no electricity shortages, only Y 300 per ton, a rational user would while lower electricity shortage costs caused choose to buy the imported coal and reduce there to be no coal shortages. Electricity their losses by Y 200. In this way, the cost shortage costs in the CaS are implicitly of a substitute form of energy, if available, represented by a step function. A single 13-4 Annex 13 shortage cost is specified for each of five lower end, 21 fen per kwh, roughly user groups, but taken together, the five corresponds to the short-run marginal cost. groups constitute a step function. 17. Conclusion. Recognizing the 16. The electricity shortage costs great uncertainties inherent in estimating finally adopted were calibrated to create a shortage costs on an economic basis, the balance between coal and electricity approach adopted in the CaS is one of shortages in high demand runs. The simplicity: (a) use the same level of shortage electricity shortage costs increase step-wise costs for coal and electricity throughout the from 21 fen per kwh for rural users to 33 model, regardless of time and place; and (b) fen per kwh for light industry users. The use sensitivity analysis to find out the effect upper range, 33 fen per kwh, roughly of the shortage costs on the choice of corresponds to the long-run marginal cost transport and energy projects. (LRMC) of thermal electricity, while the Table 13.1: Market Price of Coal in 1989 (yuan per ton) \i City Coal Price City Coal Price Beijing 120 Fuzhou 150 Tianjin 170 Xiamen 150 Shijiazhuang 80 Nanchang 90 Tangshan 110 Jinan 90 Datong 35 Weifang 90 Taiyuan 42 Qingdao 120 Changzhi 37 Yantai 100 Houma 35 Zhengzhou 60 Baotou 30 Wuhan 140 Shenyang 120 Changsha 95 Dalian 130 Guangzhou 250 Dandong 120 Shantou 200 Anshan 100 Nanning 120 Changchun 120 Haikou 200 Harbin 85 Chongqing 50 Shanghai 200 Chengdu 90 Xuzhou 55 Guiyang 30 Nanjing 200 Kunming 30 Nantong 150 Xian 110 Hangzhou 250 Lanzhou 50 Ningbo 250 Xining 50 Wenzhou 200 Yinchuan 40 Hefei 150 Wulumuqi 25 \a Because market prices of coal varied very nuch in these years, prices of coal in this table are a sumnary of surveys conducted in 1989 and 1990. The necessary processing, modification, and estimation have been introduced. Annex 14 14-1 COAL SECrOR DATA AssumPoNs Coal Type Prinary Data Base resources of the region. The :sh content, sulfur content and thermal value of CW, Spatial and Technological Assumptions SW, MI are get from washing submodel. 5. Coal type data are organized by 1. Coal types are catalogued into mine node and coal type. Coal types includ- three categories: steam coal, coking coal, ed in CTS are the following: and anthracite. In the CTS, all the coking Al raw anthracite; coal produced in one zone is aggregated into Cl raw coking coal; a single type of coking coal, and likewise Si raw steam coal with high for anthracite. For steam coal, two kinds can quality (low ash content, not be defined within one zone having different washed); coal qualities. These two kinds are generally S2 raw steamn coal with low quali- high and low ash steam coal, and all steam ty (high ash content, can be coal within the zone must be aggregated into washed); one of these two categories. CW clean coking coal (after wash- 2. The coal types are identified ing); according to the users' technical require- SW clean steam coal (after wash- ments, meaning that coal is classified ac- ing); cording to how it is used rather than its Ml middlings (after washing). inherent physical properties. In order to simplify the model, coking coal in CTS is Cost Data Assumptions the coal which be used as metallurgical coking coal. Coking coal used as steam coal None. has been identified as steam coal. Generally speaking, main coking coal, fat coking coal Capacity Data Assumptions and gas coking coal are classified as coal used for coking in the CTS. Thin coking None. coal and other coking coal are put into the steamn coal category because they are gener- Other Technological Factors Assumptions ally used in steam boilers in China. All the township mine coking coal output is aggre- None. gated into steam coal except for several township mines that produce the highest Coal Mine Nodes Primary Data base quality coking coal. 3. The anthracite coal category Spatial and Technological Assumptions includes all anthracite coal mines plus some steam coal mines in certain regions where 6. In the CTS, mines around no anthracite is available and steam coal is China are aggregated into some (47 at pres- commonly used for residential and commer- ent) production zones. Each zone is repre- cial purposes. sented by a centroid which is also a node on 4. Coal quality in the CTS is the ansport network. Mines were aggregat- measured in three physical characteristics: ed into zones based on several often conflict- thermal value, sulfur content and ash con- ing considerations, such as the transportation tent. In general, coal produced in different network, province boundaries, coal type nodes should have different qualities. These localization, and model size. Generally coal quality data are based on the typical spealing, aggregation into zones should 14-2 Annex 14 enable the model to identify all significant ship mines. Mine step 2 includes those flows while at the same time reducing the mines which can access the middle level number of coal production nodes. reserves of coal with general equipments. In 7. Some important coal supply the CTS, local mines are included as mine provinces are divided into several nodes for step 2. Mine step 3 are those mines which more accurate modeling of intra- and inter- can access the deep reserves and some provincial flows, while in other cases prov- surface reserves with advanced technology, inces can be represented by one node. In the which are assumed to be all of the state coal base area, where large amounts of coal mines. While in some cases, township, are exported to other places, each production local, and state mines may employ technolo- zone reflects a subprovincial region from gy that is more typical of other levels, these which coal is generally transported by a assumptions are generally valid. different railway line. For instance, there 11. Coal mine data are organized are five nodes in Shanxi province, which has by mine node, raw coal type, and mine five typical routes out. Another reason for type. Raw coal types included following: subdividing Shanxi province is that some Al raw anthracite regions predoniinantly produce steam coal, Cl raw coking coal while others produce mostly coking coal. Si raw steam coal (low ash con- 8. Another reason fbr disaggre- tent) gating a province into several nodes is if S2 raw steam coal (high ash coal production is related with a specific content) railway project. In such cases, the node Raw mine types included following: aggregation should allow all major flows to I Mine step I be modeled. For instance, Shandong prov- 2 Mine step 2 ince is separated into two production nodes, 3 Mine step 3 Jinan and Taian because the model considers a new rail project between Jinan and Taian. Cost Data Assumptions Most other provinces take one node to represent the total coal production of each 12. Cost data for existing coal province such as Harbin as the centroid for mine. are based on statistical data from the Heilongjiang, Changchun as the centroid for local and state coal mine companies. Cost Jilin, and so on. data for new coal mnines are based on plan- 9. For mine step 1, more than ning data and experts from the local and 50,000 township mines were aggregated into state coal mine companies. The source is the the 47 nodes. For mine step 2, more than 1989 statistics book of the coal industry 2500 local mines capacity were aggregated published by China State Coal Mine Compa- according to coal type and location. For ny, October 1990 and the 1989 statistics mine step 3, according to geographic ioca- book of provincial and county local coal tion, a total of more than one hundred exist- mines published by China Local Coal Mine ing and new state mine bureaus were aggre- Company, May 1990. Data for the Eighth gated into 47 nodes. Five Year Plan and Ninth Five Year Plan 10. In each production node, there also come from these two companies. are three kinds of mine steps. Different mine 13. Different mine steps have steps have different costs and capacities. different operating costs and investment Taken together, mine steps aprroximately costs, even in the same zone. Operating cost represent an increasing step function for coal includes labor, material, power, maintenance production costs. In the CTS, mine step I and others (including management, adminis- indicates those mines that only have access tration, safety, and insurance). Depreciation to the shallow resources of coal using low is excluded from the operating cost for both tchnical level equipment. This set of labor- existing and new coal mine of three steps. intensive mines is taken as the set of town- Investment cost not only includes mine Annex 14 14-3 engineering but also includes infrastructure, Both investment and operating costs are that is, the comprehensive investment cost. higher for state mines than for township and Inflation is not considered in cost data, and local mines, and for niines in coastal areas it is assumed that the same technology for than in the coal base. Table 14.1 shows the mining will be used over the next 15 years. ranges for each category of costs and mines. Table 14.1: Coal Production Costs in the CTS Model, by Region and Ownership (yua per ton) Coal Base Coastal Operating Cost Township 12-46 40-60 Local 20-50 40-70 State 35-58 50-120 Investment Cost Township 100-140 125-150 Local 160-200 200-240 State 200-250 250-300 14. For mine step 1, operating and Capacity Data Assumption investnent costs were based on a survey of plans for 200 key counties of coal produc- 17. The sources of mine capacity tion. Township mine production costs are data statistics are the same as for the cost lower than for local mines because of lower data. Existing capacity refers to capacity administrative expenses and less support for existing at the end of 1989. the workers in terms of housing, medical, 18. Existing capacity data of mine etc. The investment cost used for mine step step I are taken from statistics of township 1 in the CTS is higher than the historical mines. New capacity data are based on the figures for the already existing township gth and 9th FYPs made by the local mine mines in order to get reliable capacity. companies. New capacity was calculated as 15. For mine step 2, operating cost the planned key year output minus the exist- data for existing mines came from the sratis- ing capacity remaining in that year. In rich tics of local mines. Several typical mines reserve areas, greater capacity than what is were identified for each node and each coal planned was given to provide some flexibili- type. These mines are relatvely bigger than ty in where and how much capacity is built. others in the node. Weighted average cost of 19. Existing capacity of mine step these mines was taken as operating cost. The 2 comes from the statistics book of the local investment cost for new step 2 mines was mine companies. New mine capacity is also based on the planning of Local Mine Com- based on the planned key year output minus pany and on experts' opinions. the existing capacity remaining in that year. 16. For mine step 3, the weighted 20. Existing capacity of mine step average operating cost of existng mines was 3 come from the statistics book of state mine takn for each node and coal type. companies. Each state mine capacity was collected according to the mine node and 14-4 Annex 14 coal type. New capacity is from the planning tent steam coal (coal type S2) are considered of new mine projects for the Eighth and for washing. Washing is optional for high Ninth Five Year Plans. In order to give the ash steam coal, but it is required for coking CTS optimization model more choices than coal. In the CTS model, coal washing is provided by the Five Year Plans, some new permitted at every coal mining node that has capacity currently slated for development the ability to produce coking coal or high after the year 2005 is also added into the ash steam coal. 2001-2005 time period new capacity. 24. It is assumed that all coking 21. There is a big difference in coal will be washed with standard technol- mine depreciation rate among the three kinds ogy, that is heavy medium separation ap- of mines. The existing capacity remaining in proach. This standard technology is assumed each key year reflects the fact that big state to produce two outputs: clean coking coal mines will retire one by one. Each big state with a standard ash content of 10 percent, mine to be retired in the next 15 years is and a usable byproduct known as "middling- treated uniquely. But the existing capacity of s." Steam coal is always assumed to be township mines is assumed to decline at washed using ajigging approach to get clean least 25 percent during each five year period steam coal with no middlings byproduct. in all nodes. The average decline rate used The thermal value, sulfur content and the for local mines is about 10 percent for each weight ratio of washed coal to input coal five year period. So there is not a constant will be calculated by the washing submodel depreciation rate applied to all mines or based on the thermal value, sulfur content, node. Therefore, the remaining capacity of and ash content of the input coal. existing mines in the years 1995, 2000, and 2005 for each mine step were calculated Cost Data Assumption exogenously and put into the primary data base directly, and the depreciation rate is 25. The source for coal washing given as 1 in primary data base. data is the 1989 statistics book of the coal industry published by China State Coal Mine Other Technological Factors Assumptions Company, October 1990 and the 1989 statis- tics book of provincial and county local coal 22. The lead time for coal mine mines published by China Local Coal Mine construction and the mine lifetimes are based Company, May 1990. This is the same data on Chinese experts' suggestions. Generally resource as for mining data. speaking, a new mine with 30,000 ton ca- 26. Since a standard washing tech- pacity per year needs two years to be set up; nology is adopted in the CTS for coking a 300,000 ton capacity new mine needs 4 coal, the same operating cost, 4 yuan per years; and a 1.2 million ton capacity new ton, was applied everywhere. Depreciation mine needs 6 years. These three kirnds of is excluded from this operating cost. The mine were taken as typical mine sizes for investment cost of coking coal washing is the three mine steps respectively. For mine also the same, 50 yuan per ton of capacity, step 1, two years is the lead time and 15 no matter where it is. These numbers are years is the lifetime. For mine step 2, four the averages for the whole country. The years is the lead time and 25 years is the same assumptions are made for washing lifetime. For mine step 3, six years is the steam coal. Because steam coal washing lead time and 50 years is the lifetime. technology is simpler than coking coal washing technology, the operating cost and Coal Washing Nodes Primary Data Base investment cost are less than fr- washed Spatial and Technological Assumption coking coal. The operating cost for jigging steam coal in the CTS is 2 yuan per ton and 23. Only coling coal (coal type Cl the invesunent cost is 30 yuan per ton of in the primary data base) and high ash con- capacity. Annex 14 14-5 Capacity Data Assumption 31. The lead time for building new washing plants and the lifetime of new 27. The data for coal washing washing plants are based on the suggestions capacity are also based on the statistics book by experts from the State Coal Mine Com- of the State Mine Company. The existing pany. The CTS assumes a three year lead capacity as of the end of 1989 was 189 time and a 25 year lifetime.Washing plant million tons annually. This total was identi- retirement within the next 15 years is ig- fied as either steam or coking coal plants nored. That means no retirement of existing and aggregated into the same zones as were washing plint in CaS from 1990 to 2005. the existing mines. Some washing nodes This assumption is reasonable because 50 have no existing washing capacity for steam percent of existing washing plants were built coal and/or coking coal. within the last 10 years. 28. The upper bound on the amount of new washing capacity that can be Coal Demand Nodes Primary Data Base built in the model depends on the water resources and other technological consider- Spatial and Technological Assumptions ations. Based largely on the opinion of Chinese experts consulted by the CTS, water 32. The data on coal demand are resources are a limitation on the extent of based on the coal demand forecast made by coal washing, especially in the coal base Energy Research Institute of SPC. The area where lack of water is a problem. In report was published in 1988 under the title Northwest China, some high sulfiir content "Coal Base Area Development Strategy in coal is difficult to wash for technical rea- 2000 and Balance of Coal Supply and De- sons. Technology constraints for this area is mand.' the major concern. 33. In the CTIS, there are presently 29. For all other nodes, the maxi- 49 coal consumption zones around the coun- mum possible washing capacity was only try. Every province has one node at least. limited by the coal production at this node, Some provinces have several nodes because which, essentially, is no limit at all. Plans there are several consumption centers in do exist for construction of washing plants these provinces and they will influence by the State Coal Mine Company, but in the transportation flow direction. For example, CTS, much more choice is given. The 8th Shanxi province has four nodes, and Jiangsu and 9th Five Year Plans call for 100 million has two. Another reason for disaggregating tons of washing capacity to be constructed, a province is because shipping and port but the total potential capacity allowed to be investment for coal transportation is one of built in the cas is much higher. the major concerns of the CaS. Usually in China, there is no transshipment from ship Other Technological Factors Assumptions to railway, so some province which have different ports will be divided into several 30. The standard weight loss fac- nodes. For instance, in Guangdong province tors and final ash contents for coal washing there are two nodes: one is Guangzhou, one are described in the write-up of the coal is Shantou; and in Fujian province, there are washing submodel (see Volume II). Howev- two nodes: one is Fuzhou, the other is er, SPC experts suggested that the washing Xiamen. submodel not be used. Instead, they suggest- 34. Coal demand in CTS is divided ed that the thermal value of washed steam into three kinds of users. Anthracite user, coal be increased in order to capture the coking coal user and steam coal user. An- increased efficiency from burning washed thracite users include sectors which use coal. Therefore, a carbon loss rate of 0 per- anthracite, those are the residential and cent was assumed for the CTS Yellow Coy- commercial sectors, and part of the metal- er model runs. lurgical, chemical and building materials 14-6 Annex 14 industries. For the 9 percent GNP growth steam coal except utilities. For the medium case, total anthracite demand is around 142 demand case, general industry demand is million tons of standard coal in 1995, 153 570 million tons in 1995, 668 million tons million tons in 2000, and 177 million tons in in 2000, 770 million tons in 2005. 2005. For the anthracite demand constraints, 37. In the medium demand case, heat content of coal is taken into account, so the total demand for coal not including that the exact amount of anthracite produced utility demand, that is, the sum of the three and consumed may vary from these figures parts separately discussed above, is 784 even if there are no shortages. million tons in 1995, 906 million tons in 35. Coking coal users are strictly 2000, and 1,046 million tons in 2005. the metallurgical industry, with the indus- try's anthracite demand subtracted out. It Shortage Cost and Shortage Upper uses washed coking coal only. Raw coking Bounds coal will not be delivered unless it is cata- loged as steam coal (see coal types, above). 38. The base figure for coal short- Only washed coking coal will be transported age costs is Y 300 per ton of standard coal, to the user. Middlings are assumed only to which corresponds to the international price be used locally, mainly by minemouth fluidi- for coal (see Annex 13). This is the figure zed bed combustion (FBC) power plants. used for steam and anthracite coal shortages For the medium demand case, metallurgical in most of the country. uses 97 million tons washed coking coal in 39. No upper bounds are imposed 1995, 114 million tons in 2000, and 132 on coal shortages in the Yellow Cover million tons in 2005. The heat content of model runs. Shortages are free to go as high coking coal is ignored in the coking coal or low as possible, as determined by the demand constraints, since metallurgical supply and demand situation. This is impor- plants require a given tonnage, not a given tant because it means that all of the invest- heat content. ments that the model builds are used to 36. Steam coal users are separated deliver energy at a reasonable cost- into two. One is utilities, the other is general none are subeconomic projects forced in by industry. Both can use steam coal 1, steam a upper bound on the size of coal shortages. coal 2 and washed steam coal. Coal demand On the electricity side, upper bounds are of utilities depends on electricity demand. necessary to make the step function work; thus, electricity demand is given exoge- however, in our computational experience, nously, but utility consumption of coal is rarely did a shortage reach the upper bound endogenous. Gieneral industry demand for on the highest cost step. steam coal includes all sectors which use Annex 15 15-1 TRANSPORT SECrOR DATA AssumrTIoNs Origin-Destination (OD) Pairs Primary impacted by shortages: Shanghai, Xiamen, Data Base and Guangzhou. Plus, a few other new options are tested in the CTS. Spatial Assumptions 4. The source of the O-D pairs data is the current annual coal distribution 1. The O-D pairs data base is plan, modified to suit the CTS model. The used by the path generator to control the O-D pairs data base has been checked and creation of path variables. The CTS model revised by experienced experts and planners does not permit coal to be shipped from from the State Planning Commission (SPC) every origin to every destination. Instead, and the Institute of Comprehensive we exogenously define which origins can Transportation (ICT). serve which destinations, and also 5. The total number of O-D pairs exogenously define how many different defined for this study was around 550. The routes between the origin and destination total number of routes (paths) used for this will be generated. study was about 2,500. This number is 2. The O-D pairs data are based limited by the hardware and software on several principles. The first principle is requirements. Of course, not all viable that paths should be generated for all routing options can be included in this existing O-D flows, by coal type. The limited number. current and future O-D pattem is to ship 6. The number of transportation coal from west to east and from north to routes is uniquely specified for each O-D south. However, sometimes reverse flows pair. Generally, for short distance O-D) are necessary because of the difference in pairs, a small number of routes are defined. the coal types between regions. Sometimes Sometimes only one route is defined, for coal must be shipped to the coal-rich areas instance from Beijing to Tianjin. As the if the coal-rich areas do not have any distance between the origin and destination resources of a particular kind of coal. lengthens, more routes are defined. If 3. The second principle is to coastal shipping is an option for an O-D allow some new scenarios or policies. One pair, an even larger number of routes is such new policy is to supply nearly all defined. The largest number of routes places in China from the major new coal (paths) between any O-D pair is 20 (e.g., fields in the Coal Base region, particularly from Datong to Guangzhou or Shanghai), Shenmu, Dongsheng, Zhungeer, and so on. but the CTS software can accommodate up Another new coal distribution policy is to to 35 paths for any O-D pair. accommodate new plans for industrial 7. The paths used in this study distribution. Industrial development in were generated by the CTS model's path western areas is expected to increase, and generator. These paths were then edited one therefore more O-D pairs must be defined to at a time to check them, delete unwanted supply coal to these western areas. A third paths, and add missing paths. This task took new policy to be evaluated is to increase about three person-weeks. coal imports to coastal economic zones. Although there have been sporadic imports Transportation Network Primary Data if coal in the late 1980s during the severe Bases (existing and new) shortages, there has been no steady stream of imports to China, and no import policy. Spatial and Technological Assumptions The policy tested in the base case is to allow only minor (less than 10(,000 tons) coal 8. The transportation network is imports to three of the nodes most highly divided into two separate primary data 15-2 Annex 15 bases. The existing arcs data base example, the rail arcs from Beijing to (PDBTRANA) is designed for the transport Shijiazhuang are bi-directional because coal network links which have no expansion from Taiyuan (west of Shijiazhuang) must be projects, that is, they have no investment able to flow northeastward to the steel plants costs. Both brand new arcs and existing arcs in Northeastern China. Other example of bi- with proposed expansion projects are input directional corridors are from Hailar to in the new and expansion arc primary data Dalian; Tangshan to Baotou; Beijing to base (PDBTRANB). The existing part of an Zhengzhou; Changsha to Hengyang; and expansion project is not input in Tianjin to Fuliji. PDBTRANA. 12. Generally, most ports are 9. The modes included in the CTS defined as a single arc, which means that model are rail, port, and all shipping, both various berths have been aggregated into a coastal and inland waterway. Canals are also single variable. These various kinds of described in the CTS network, linking berths include state-owned and user-owned Xuzhou and Peixian to the Yangzi River. berths, both at the port and in port power Figure J shows a schematic map of the CTS plants. However, some ports have been transport network which shows all of the defined with several arcs in order to existing and new rail, port and pipeline arcs, accurately describe the real shipping but does not show the shipping arcs because situation. Because some ports have different there are too many to be displayed. kinds of facilities that can handle different 10. The network of existing arcs is sizes of ships or have drastically different based on the current pattem of major coal costs, they are represented by two or more flows, and also considers coal distribution arcs. For instance, Qinhuangdao port is policy. Each arc represents a rail line defined as three arcs: the first is the old 9m between two major points, but small existing port; the second is the moder.u 9- substations are ignored. Rail nodes. that is 12m existing port; and the third is a existing origins and destinations of single arcs, cover 9-14m port with proposed expansion all supply and demand nodes in the CTS measures. data base. The network also covers all major 13. Two kinds of transshipment are junction points where arcs meet, but which allowed at ports in the CTS network: rail-to- do not have supply or demand activities, water, and water-to-water. Some dummy such as Yuanping in Shanxi province or arcs have been defined for technical reasons Baxian in Hebei province. Marshalling yards for water-to-water transshipment, such as at are not treated as arcs in their own right. Shanghai, Shenzhen, and Nanjing. Some Marshalling yard costs and capacity are ports can have incoming and outgoing flows. factored into the cost and capacity of several For instance, Nanjing both receives and arcs that begin or end at the marshalling ships coal by river. Also, Shenzhen is yard. defined as a port where large ships can 11. Because the network design come in small ships can go out. follows the current flow patterns and current 14. There are no existing coal distribution policy, some arcs are considered slurry pipelines in China. Three new as bi-directional arcs, but most are uni- proposed pipelin_ projects are evaluated in directional. By prohibiting backhauls on the CTS. The network had to be designed in most arcs, the CTS model will not achieve a special way to prevent ridiculous results a perfect global optimum, but it adheres using the slurry pipelines. The slurry more closely to the current policies and pipelines are not designed to handle helps t* calibrate the model. For the most transshipments at either end of the pipeline; part, arcs flow only in a single direction, they are dedicated links between a single generally from north to south and from west origin and a single destination. In order to to east, based on the geographical prevent coal being transshipped from rail to distribution of supply and demand. For pipeline, a separate supply node is defined Annex 15 15-3 as the pipeline's origin, with no rail arcs to Is. Operating Costs. The financial it, as in the node Changzhil (node ID = operating cost for all modes includes labor, 8003). A dummy rail arc is defined from the fuel, maintenance, management, and other special supply node to the main supply node variable costs. The financial operating cost (e.g., from Changzhil to Changzhi) so that never indudes depreciation of any kind. if the pipeline is not built in the model run, Depreciationoftransportation equipment and the mining capacity can still be built and infrastructume on existing arcs is not sent by rail. Furthermore, the coal cannot be included anywhere, because these things transshipped from the end of die pipeline to have already been built. For new arcs, the railroad network because no O-D pairs investment in new lines, rolling stock, ships, are defined for the pipeline origin node etc. is considered as an up-front investment except to the pipeline destination node. cost instead of as a depreciation cost per 15. Project Packages. The CaS ton. These assumptions put the operating model allows for as many as two investment cost of existing and new lines on an equal packages in order to choose between the footing when finding least-cost paths. different project alternatives. However, in 19. Financial operating costs for the Yellow Cover model runs, only one expansion projects are considered to be the package was considered for each railway and same as for the existing line. No adjustment port arc because because of the model's is made in this study for adjusting the running time. operating cost to lower or higher post- 16. Rail projects are of two types: expansion cost levels, even though the brand new lines and expansion projects. model has the capability to do so. Railway expansion measures mainly indude 20. For rail, financial operating double-tracking and electrification. Port costs range from a low of .012 yuan per tkm expansion projects are for increasing the to a high of .05 (e.g., for the Da-Qing line) number of berths and improving the loading, depending on geographic conditions, unloading or storage yard facilities. Brand technology used, and modernity. For new rail projects include single track lines; shipping, financial operating costs vary double track lines; double track electrified depending on the ship's size, type, and lines; and unit train lines on the DaQing line whether existing or new (see Table 15.1). and the proposed line from Shuoxian to Operating costs for slurry pipelines average Huanghua port. Some packages on crucial around .75 per ton-kin, based on the lines were made larger than they are in the estimate for the Yuexian to Weifang line. Plans. They were given a larger new 21. The so-called fake operating capacity, and also a proportionally larger costs in the primary data bases are only used investment cost. to fool the path generator into generating paths which may not actually be the cheapest Cost and Price Data Assumptions ones. Generally, the fake costs are set equal to the financial operating costs, with the 17. The cost data for the exception of a few brand new rail lines, for transportation network comes from several which lower fake costs are used. This is sources. Rail operating and investnent cost because the financial operating costs of the data are based on internal data bases from new lines is very high, and the path the Railway Investment Study (RIS) and generator-like many shippers!-would try from ICT. Shipping, inland waterway, port, to avoid those arcs. But since the CTS and pipeline data are also based on ICr data model would have a lot of urused rail bases. All cost data are in 1989 financial capacity and a lot of shortages if these arcs costs, with no adjustment made to 1990 were avoided, they are forced in by using a levels. No published data are available yet lower -fake" cost. But for the objective for 1990. function coefficient in the optimization 15-4 Annex 15 model, the rcal financial operating cost is LOTUS spreadsheet model was developed to used. estimate rolling stock requirements, using 22. 7>pe of 0-1 Variable Used rin the same method used by the China the New Arcs Data Base). Ideally, all International Engineering Consulting transportation investnent decisions would be Company (CIECC) in appraising new made on a 0-1 basis, but we have had to projects. In general, the assumption is that if limit the number of 0-1 variables to maintain the new line provides, say, 50 million tons good computational performance. In the of capacity, then enough rolling stock to results reported here, about 35 of the most haul 50 million tons annually on this important brand new projects are defined as particular line must be included in the 0-1 or "B"-type variables. To save on the project's cost. We do not assume that rolling number of 0-1 variables, about 90 transport stock can be shifted from other lines to a projects-mostly expansion arcs, or less new line. Generally speaking, rolling stock important brand new projects-are defined accounts for approximately one-third of the as continuous "C"-type variables. The total investment cost. remaining 60 projects are already under 27. A special situation is involved construction or the investment funds have in the case of new or expansion bi- definitely been allocated. These are directional arcs. In the CTS model, the two represented as "U"-type variables, that is, directions of an arc are considered as the 0-1 variable is set to a value of 1. completely separate arcs. But in reality, 23. Investnent Costs. The figures when investment is added to a line, both for investment costs are based on the total directions get higher capacities. To deal financial investment costs which have been with this situation, the data had to be estimated for each project by MOR, MOC, manipulated and a new constraint was added ICT, or, in the case of slurry pipelines, the by hand. The total investnent cost is given local coal companies. The total investment to the direction of the arc that has heavier for a project does not consider inflation-all traffic volume. The reverse, low-volume investment costs are in 1989 Y. direction is given no investnent cost. Thus, 24. The actual investmentcost used to use the main direction's new capacity, the in the CTS primary data bases is coal's investment cost must be paid, but to use the share of the total investment cost. The reverse direction's new capacity is free. Bu. percentage of a project's capacity that will the added constraints act to prevent usage of be used for coal has been exogenously the reverse direction's new capacity unless estimated by ICT (see capacity data, below). the 0-1 variable for the main direction is This same percentage is applied to the equal to 1. project's total investment cost. 25. Each railway and port project Capacity Data Assumptions has been appraised uniquely to determine its investment cost. Examples of some of the 28. There are two kinds of more costly projects are Y 500 million for capacities for arcs in the CTS model: each half of the proposed new port at existing and new. The existing arc capacities Ruanghua; Y 1-2 billion for new slurry are based on 1989 data. New capacities are pipelines; and Y 16 billion for the proposed estimated for the future by MOR and ICT. new rail line from Shuoxian to Huanghua 29. Generally speaking, existing port. arc capacities are considered constant over 26. For railways, the total the model's 15 year time horizon. Arc investment cost for both expansion projects capacities are the capacity for coal only. But and brand new lines includes line investment since capacities are based on the estimated and rolling stock investment- The amount of ratio of coal traffic to all traffic on the arc, rolling stockl needed for each project has there are instances where this ratio is been estimated based on ICT data. A expected to be adjusted over time. For arcs Annex 15 15-5 Table 15.1: Shipping Costs in the CTS Model Deadweight Financial Financial Tons Operating Cost Investment Cost Type of Ship (thousand tons) (yuan per tun) (million yuan per ship) General Bulk 9m 20-25 0.0067 120 12m 60 0.0067 260 14m 100 0.0067 300 Self-Unloader 9m 27 0.0120 150 Pusher Barge 9m 0.0200 Shallow Draft 9m 35 0.0120 150 with new projects, the ratio of coal traffic to Fleet Primary Data Base total traffic is forecasted exogenously by MOR and by the Plans. This ratio ranges Spatial and Technological Assumptions from 20 percent to 90 percent. 30. Expansion arcs have both an 33. Rail and shipping vehicle fleet existing capacity and a new capacity, where are handled differently in the CIS. The the new capacity is the additional or investment cost for new rail vehicles is incremental capacity. Brand new arcs have included in the rail line investment cost. But no existing capacity. The largest of the new the investment for new ships could not be arcs have around 100 million tons of annual included in the new port investnent cost capacity. because the necessary ship capacity would 31. Lead times and lifatimes of depend on the shipping distance. Since ports projects are specified based on MOR and can serve different destinations, the shipping ICT estimates. Different phases of a single distances for a port are indinae. project sometimes have different lead times. Therefore, the cost of new ships is handled Lead time for rail projects usually depends in a different way an the cost of rail on the length of the line and the scale of the vehicles. project. Lead times range from 2 to 10 years 34. Vehicle fleet data are organized for rail; from 2 to 8 years for ports; and by transport mode. The modes included in from 3 to 5 years for slurry pipelines. the CrS are the following (although the fleet Lifetimes of transportation arc projects are data assumptions apply only to the shipping generally 30 years. modes, not to RAIL, PORT, and 32. In Case 15, extra transport PIPELINE): capacity was given to selected arcs, RAIL railroads including imports, some railway lines from PORT ports the energy base (DaQing line +60 million PIPELINE slurry pipelines tons), most ports (+ 10 percent in cost and INLANDWW inland waterway capacity), and the three proposed slurry 9MGENRL 9-meter draft pipelines, general class vessels 15-6 Annex 15 9MPUSHB 9-meter draft 300,000 dwt of inland waterway barges (and pusher barges a proportional number of tug boats); 2.29 9MSELFU 9-meter draft self- million dwt of 9m general class ships; and unloaders 100,000 dwt of 12m general class ships. 9MSHALL 9-meter draft These capacities were assumed to decline by shallow draft about one-tenth every five years due to vessels vessel retirement. 12MGENRL 12-meter draft 36. In the existing transportation general class arcs data base, there is a column of numbers vessels called DELTAAV. This is the number of 12MPUSHB 12-meter draft trips per year made by a ship serving that pusher barges particular O-D route. This figure is used by 12MSELFU 12-meter draft the model to determine the amount of self-unloaders deadweight tonnage capacity needed to carry 12MSHALL 12-meter draft a given number of annual tons. For instance, shallow draft if a ship makes 10 trips per year on a certain vessels route, then the annual tonnage divided by 10 14MSUPER 14-meter draft is the amount of dwt capacity needed. The supercolliers average number of trips per year on each shipping route is determined according to distance. Capacity Data Assumptioms Cost Data Assumptions 35. 'he shipping fleet capacities are measured in deadweight tons (dwt), not 37. It is assumed that new ships are in annual capacity. Unfortnately, the only produced domestically, rather than data available to us were the annual tonnages purchased second-hand from abroad. The by vessel class in 1989. Data on the number lifetime of new ships is considered to be 30 of ships used for carrying coal were years; while the lead time ranges from one unavailable. From this we worked year for barges to five years for 14m super- backwards to infer the deadweight ton colliers. The financial investment capacity of the ships. Since we know that costranges from Y 2,147 per dwt for barges the tonnage shipped was limited by ship or tugs to Y 7,500 per dwt for 14m availability, we assumed that the number of supercolliers. The unit investment cost tons carried by each vessel class in a year, increases with size because Cbinese ship- divided by the average number of voyages building companies charge intemational per year, is a reasonable estimate of the prices for all except the standard 9m general deadweight tonnage of the ships used for bulk vessels. The average tow size for carrying coal. The final estimates were inland waterways is assumed to be 6 barges. Annex 16 16-1 ELECaRICITY SECTOR DATA AssumrnoNs Electricity Demand Nodes Primary Data Base Electricity Demand Data Assumptions Spatial and Technological Assumptions 4. Determiningelectricitydemand for China for 1995, 2000, and especially for 1. Electricity demand for the 2005 is a very difficult research task. If we entire country was aggregated into nodes try to forecast electricity demand for these according to where the consumption centers three key years by methods such as from are, present electricity consumption patterns, input-output tables, we could not finish the and characteristics of transportation and CTS in a timely fashion. Fortunately, China geography. There is at least one node for has several institutions which have been every province, and several provinces are studying energy demand nationally. With split into two or three or four nodes. For practical experience of the past few years, instance, Shanxi province is defined into some demand forecasts are very accurate. four nodes (northern, middle, southern, and The CTS team studied these reports and then southeastem) according to current patterns used some useful study results from them, of coal distribution. Likewise, Liaoning including electricity demand projections. province is defined into two nodes (Sheny- 5. Total electricity demand for the ang and Dalian) according to present indus- medium demand case in the CTS for the trial allocation and consumption. Presently, year 1995 is about 10,441 (10' kwh); de- the CTS model has 38 electricity demand mand for the year 2000 is 14,436 (10' kwh); nodes. and demand for the year 2005 is 19,891 (10' 2. All electricity demand nodes kwh). The demand elasticity with respect to are also electricity generating nodes, but the GNP is assumed to be 1.0. converse is not true. The set of generating 6. The electricity demand projec- nodes also includes some remote generating tions from this earlier study were checked stations with no significant independent by a team of State Planning Commission electricity demand. (SPC) experts in a data conference in Janu- 3. Electricitv consumers at each ary 1991. Several changes were identified node are defined into five sectors. Agri- and made to the primary data base. In addi- cultural electricity demand is defined as the tion, the overall level of electricity demand electricity demand for agricultural produc- was scaled during the calibration process to tion. Light industry electricity demand is make it consistent with other CTS primary defined as the electricity demand for light data bases. However, the relative demand industry production. Heavy industry electric- levels from node to node used in the CTS ity demand is defined as the electricity are based on this earlier study. demand for heavy industry production. 7. The "Comprehensive Equiib- Urban electricity demand is defined as the rium and Systematic Optimal Study," upon nonproduction electricity demand in cities, which CIS electricity demand projections consisting primarily of residential, commer- were based, was one of the Chinese pre- cial, and transportation demand. Rural cursor models that led up to the CTS. To electricity demand is defined as the nonpro- briefly explain the basis for these demand duction electricity demand in rural areas, projections, electricity and coal demand consisting of the same components as urban were endogenously produced by an optimi- demand. zation model not only considering produc- 16-2 Annex 16 tion, transportafion, and transmission operat- 9 percent shortage at a cost of 0.15 yuan per ing cost and investment, but also considering kwh. Following that, if there were still a productive allocation and investment cost shortage of electricity supplied at under 0.20 and local utilization efficiency of energy. yuan per kwh at a node, then a shortage The study tried to model the tradeoffs be- would be incurred for light industry at 0.20 tween production and transportation costs on yuan per kwh up to 9 percent of the light the one hand and utilization efficiency on the industry demand, and so on. For the node as other hand. For instance, according to coal a whole, the total electricity shortage cannot production and transportation costs, the exceed 9 percent of the total demand. This model would tend to supply more coal to be approach effectively approximates the condi- consumed near the coal base, but consider- tion that marginal shortage costs are increas- ing the utilization efficiency and economic ing, and that the electricity users who can eFficiency, it may tend to ship coal to Shang- afford to pay the least, in monetary terms, hai. So the model used for this study traded would be the first ones to be outbid for off these facts to produce reasonable elec- electricity in a shortage situation. tricity and coal demand or consumption. 8. Electricity demand scenarios in Hydropower Primary Data lase the CTS are generated in two ways. One way is to amplify the base case demand by 11. All data including hydropower a given percentage across the board (about 6 projects, investment, leading time, and to 7 percent) to keep the same growth rate technical data were supplied by the State with the economy but increasing speed for Planning Commission, the Ministry of Ener- the different zones and time periods. Anoth- gy, and the Planning Institute of Hydropow- er way is to collect demand data forecasted er Resources, and were checked at a CTS- by authoritative agencies like the State Plan- sponsored conference held in January 1991. ning Commission or the MOEP. Hydropower Plant Spatial Assumptions Electricity Shortage Data Assumptions 12. As we know, China has very 9. Shortage costs for electricity abundant hydropower. How best to exploit vary by sector. When looked at as a group, hydropower has very important implications they form a step function. Shortage costs for changing the energy structure, transpor- start at 0.21 yuan per kwh for heavy indus- tation system, and environmental situation of try and increase in steps of 0.03 yuan. The China, not to mention its implications for shortage cost is 0.24 yuan per kwh for light loss of agricultural land, flood control, industry, 0.22 yuan for agriculture, 0.30 water supply, etc. yuan for the rural residential sector, and 13. Most hydropower sites are 0.33 yuan for the urban sector. These short- located far from consumer centers, meaning age costs were calibrated to create a reason- that in several cases, separate nodes inde- able balance between coal shortages and pendent of the set of electricity demand electricity shortages. They are slighdy less nodes had to be defined. Because sites and than the cost of diesel electricity generation locations are different, the economic and (0.50 per kwh), which represents a higher technological characters can vary significant- cost substitute for thermal, hydro, or nuclear ly, so in the CTS we had to define each new power (see Annex 13 for an explaiiation of hydro power plant uniquely. However, to shortage costs). have a separate node for each hydropower 10. The upper bound on the size of site would have created an unmanageably any electricity shortage, for any sector or large set of nodes. The approach taken in time period, is 9 percent of demand. There- the CTS is to aggregate hydro power plants fore, the way the step function would work into a smaller set of nodes, but have up to is that heavy industry would incur up to a twelve separate projects in each node. In Annex 16 16-3 order not to make the transmission line hydropower plants can be built by the year system in the CTS model too complicated, 2005 given projected technological appreci- we consider the investment for building ation. Total new hydropower plant capacity transmission lines from hydropower stations available for construction in the CTS is to consumer center as part of the investment about 10,000 x IO4kw. for the plant itself, thus taking into account the distance from the remote sites to the Other Technical Assumptions consumption centers. I8. Lead and life times of hydro- Cost Data Assumptions power investments are taken from planning data. Lead times are uniquely defined for 14. Hydropover plant investment each project, and range from 4 to 12 years. costs average around Y 3,000 to Y 4,500 All projects except the big hydro projects per kw of capacity, depending on different that are modeled with 0-1 variables can be locations of hydropower plants. They range built in any of the three time periods. The from an extreme low of Y 1,200 per kw of large darn projects can either be built for the capacity to a maximum high of Y 6.000 per years 2000 or 2005, depending on their xeal kw. Operating costs include water fees, lead times. Lifetimes are assumed to be 50 materials, salaries, worker welfare, mainte- years. nance fees, and other fees, but do not in- 19. The annual hours of operation, dlude depreciation. Operating costs range which is a technical data for a hydropower from Y 5.10 to Y 37.51 per kwh for exist- plant, is determined by natural conditions ing plants. and varies by hydropower site. 15. Transmission line investment costs range from 210 yuan to 1,098 yuan Thermal Power Primary Data Base per kw depending on different distance and different voltages. Operating costs include Thermal Power Plant Spatial Assumptions materials, salaries, worker welfare, mainte- nanice, and other fees, but do not include 20. Thermal power plant nodes depreciation. They range from Y 0.43 to were created according to coal allocation Y 0.68 per kwh for existing lines. patterns, existing thermal distribution, and 16. For some very large projects, the transportation network. At least one the investment cost of the dam is separated node is defined for every province, while out from the investment cost of the generat- several provinces are defined as two or three ing station and considered as a 0-1 invest- or four nodes in order that major intra- ment decision. In these cases, the size of the provincial coal flows can be accounted for. dam and the dam's investment cost is exoge- Presently, the CTS model has 47 thermal nously determined from the Five-Year plant nodes. Plans. 21. Each generating node has an existing thermal power plant variable (EX_- Capacity Data Assumption COAL) and a new base load thermal power plant variable (NEWC 1 in the primary data 17. In the CTS's hydropower base, operation hours defined as 5,000 to primary data base, existing capacity for 5,500 hours per year). A few nodes where hydropower plants is defined as the total electricity is in severe shortage have a new capacity of plants both already built and peak load thermal power plant variable currently under construction and due to be (NEWC2, operation hours defined as 3000 completed by 1995. All new hydropower hours per year) plant. Also, nodes that are plant options have posted feasibility studies in both the sets of thermal power nodes and with the Grid Planning Institute of Hydro- coking coal production nodes have a mid- power Resource, meaning that these new 16-4 Annex 16 dling coal power plant variable (MID) relat- Datong. With the exception of mine mouth ed with coking coal washing. thermal power plants, and Qinhuangdao, where the rail line capacity is mainly re- Thermal Power Plant Technological As- served for the port, there is no upper limit sumptions on new thermal capacity. 22. In the current Chinese situation Nudear Power with widespread coal shortages, one of the Primary Data Base Assumptions key parameters is the coal consumption rates of thermal power plants. For existing ther- Nuclear Power Plant Spatial Assumptions mal power plants, the CTS uses present (1990) average conversion factors of elec- 26. At present, there are two nodes tricity supply, which ranges from 525 gram with existing (that is, under construction) per kwh (standard raw coal) in Shanghai nuclear power plants. Because nuclear power regions to 826 gramn per kwh in Qinhai plant equipments heavily rely on imports, regions. For new thermal power plants, the and because its investment is very much CTS assumes higher efficiency conversion higher than that of thermal and hydropower factors for electricity generating, based on plants, the proposals for increased scale and 300-60 mW generator units that are pro- new nodes for nuclear power are very limit- posed to be the main type used for the next ed- Further limitations are imposed by 15 years. We also assume that the eneru, nuclear resources. In the CTS model, the conversion factors for new plants range fromr two existing nodes are made available for 460 gram per kwh to 500 gram per kwh. nuclear capacity expansion, and three new nodes are defined at Shenyang, Changsha, Cost and Investment Data Assumptions and Jinan-zones with severe electricity shortages. 23. For all thermal power plants, the operating cost is the 1990 average oper- Cost Data Assumptions ating cost of electricity supply (financial cost minus fuel cost and depreciation charges). 27. Because there is not a single 24. For new power plants, the nuclear power plant built yet in China, investment cost used in the CTS is defined operating and investment cost data from as the investment for a 300 mW generator. other countries were revised and adopted. The investment cost for new thermal capaci- The assumed cost of new capacity is ty is Y 2,200 per kw of capacity. This Y 6,000 per kw, compared with 2,200 per figure is assumed to be the same everywhere kw for thermal base load plants. The oper- because the investment in equipment ac- ating cost is assumed to be four times higher counts for the major part of total investment. than thermal power plant. All data were checked by experts. Capacity Data Assumptions Capadty Data Assumptions 28. No upper bound was used for 25. Because water resources are the nuclear capacity construction. Because the main constraining factor for building thermal cost is so high, we have assumed it would power plants at mine mouth locations in the be the choice of last resort. coal base of northern China, new thermal power plant capacity was determined mainly Other Technical Assumptions according to water resource availability. For instance, new thermal capacity proposed at 29. The lead time is assumed to be mine mouths from 1991 to 2005 has been five years and the lifetime is assumed to be constrained to be less than 10,000 mW in 30 years. It is assumed that it is feasible to Annex 16 16-5 construct new nuclear capacity by 1995 in are usually adopted. all three nodes. Cost and Investment Data Assumptions Transmission Line Primary Data Base 35. Operating costs and investment Transmission Line Spatial and Technolog- data for transmission lines used in the CTS ical Assumptions were based on practical data for 1990, from the Ministry of Energy and statistical re- 30. As we know, transmission lines ports. Investment for building new proposed are one of the options for energy shipment transmission lines includes line investment which is especially significant ia the Chinese and transformer investment, the cost of case where the transportation network is which is related to which level voltage is very congested. But because the physical adopted and the length of the line in miles. transmission network is very complex, it is The four linear regression equations for necessary to simplify it. The criteria of estimating line investnent costs are shown simplification are as follows. below. 31. The small transmission lines with voltage under 110 kV are ignored. For 500 kV DC lines: Because low levei voltage is mainly suitable I = 700 + .31*L (800 < L < 1500) for very short distance intra-provincial electricity transmission, ignoring these For -500 kV lines: transmission lines just ignores the problem I = 100 + .70*L (600 < L < 800) of coordinating short distance intra-provin- cial allocation, which is not part of the scope For -330 kV lines: of the CTS. I = 125 + 1.458*L (400 < L < 700) 32. Existing transmission lines were aggregated and defined to be consistent For -220 kV lines: with the spatial aggregation of the electricity I = 150 + 1.25*L (L < 600) generating and demand nodes, and the exist- ing network of transmission lines. where: 33. Proposed new transmission lines for the CTS were based ort options I = financial investment cost per kw of proposed by the Planning Institute of Electri- capacity; and cal Power, the State Planning Commission, L = distance, in km. and the Ministry of Energy. 34. The voltages of new proposed For reference purposes, a 500 kV DC linr transmission lines are consistent with exist- has a capacity of 2.4 million kw. ing levels of each grid. For instance, the northwestern grid has two voltage levels: Capacity Data Assumptions 220 kV and 330 kV; and other grids have two or three voltage levels: 220 kV, 500 kV 36. Capacity for existing transmis- (AC) and 500 kV (DC). Which voltage is sion lines is defined as the natural trans- assumed for each line depends on the length mission capacity. Because both generating of the transmission line proposed. Generally, plants and demand are aggregated into artifi- for distances under 300 km, 220 kV technol- cial nodes, the existing line capacity is also ogy is adopted. For distances between 300 aggregated. Capacity of transmission lines is and 800 km, 500 kV (AC) lines are adopted handled in two different ways in the CTS. except for the northwestern grid where 330 New transmission lines with voltages under kV lines are adopted. For distances over 800 500 kV (AC) are defined as contimuous km, which for the most part are lines con- variables, while new transmission lines with necting separate grids, 500 kV (DC) lines voltage of 500 kV (DC) are defined as 0-1 16-6 Annex 16 variables. For the continuous variables, no proposed origin-destination pair, with a a priori upper bound on new capacity is capacity of 2,400 mW_ imposed. The solution results of the vari- 37. The actual volume of electricity ables gives a total kw of capacity built, and in kwh that can be transmitted on a line is from these it is possible to calculate how equal or less to the capacity built in kw many new lines are needed in each time times the annual hours of operation. The period. For the 0-1 variables, the a priori number of hours for existing and new lines assumption is that a maximum of one new is exogenously set at 7000 hours per year, large transmission line can be built on any based on technical standard. Annex 17 17-1 ENVIRONMENT SECTOR DATA AssumpIoNs 1. As there are almost no sulfur model, as up to now there is no plan to scrubbers equipped for coal boilers in China equip sulfur scrubbers to small and medium at present, the data assumptions for sulfur size boilers. The boilers of power plants are scrubbers are roughly estimated. It is esti- all of big size, and are most likely to be the mated that sulfur scrubber will cost about 20 first to be equipped with scrubbers in the percent of the total investrnent for power future. plant, the investment for scrubber is 500 3. No specific constraints on yuan per kw power generator. Assuming sulfur and ash contents are given exogenous- that the average sulfur content of coal is 1.5 ly to the nodes for the base case. What we percent, that 3 tons of coal will be consumed suppose is that once we get the emission for I kw generator capacity with 6,000 volume of ash and sulfur in each node from hours of operation time, and that the dispos- the base case, we will use 90 percent or 95 al rate of sulfur is 90 percent, then the unit percent of the original emission volume as investment cost for scrubbers is calculated as the constraint for scenario analysis, to find 12,500 yuan. The operation cost for removal out the percentage cost increase for lowering of I ton of sulfur is estimated as 300 yuan. ash or sulfur emissions for the energy sup- 2. Sulfur scrubbers are only ply system, and what is tie most economical designed for coal power plants in the CTS strategy for doing so. Annex 18 18-I METHODOLOGY FOR ESTIMATING BENEFITS BY COMPARING SCENAMUO COSTS: STEAM COAL WASHING, ELECTRICrrY TRANSMISSION, SFENMU-HUANGHIUA RAILWAY COMPLETION Overview system to the absence of this option, substi- tuting other measures and incurring greater 1. The total national benefit from shortages. The differential between the some kind of investment measure is estimnat- optimal cost with and without these options ed in dhe following way. The method relies is the basis for the benefit estimation. The on comparing systemwide costs in "with" difference between objective functions of the and "without" scenarios, and then adjusting base case and the no-new-intergrid-transmis- that cost differential. The two scenarios must s;on case is Y 1.6 billion, or slightly less differ by only one assumption for it to work than I percent of the overall costs. properly. 5. Note that only the intergrid lines are eliminated from this scenario, Cost Differentials Between Scenarios because intergrid lines represent the real policy change taking shape in China. They 2. To estimate die benefit of have long used intragrid lines to create a steam coal washing, a separate scenario was reliable power grid. It should be further run assuming that no new steam coal wash- noted that intergrid lines number only 16, ing is allowed to be built in any of the three compared to about 100 intragrid lines. In time periods. This is done by imposing an fact, Shanxi and Nei Monggol are in the upper bound of 0 on all ZW (new washing same grid as Beijing and Tianjin, and like- capacity) variables. Of course, with no wise, Nanning and Guangdong are in the steam coal washing investment allowed, the same grid. Examples of intergrid lines elimi- model must adjust the entire optimal system nated from the scenario are Tangshan-Shen- to compensate and replace it. Consequently, yang, Chongqing-Wuhan, Guiyang-Guangd- there are more shortages, more transmis- ong, Changzhi-Xuzhou, and Changzhi-Shan- sion, more hydropower, more shipping and ghai. railway flows (where possible). Readers may 6. Scenarios 93-1 and 93-2 can be refer to the summary tables in Annex 9 to compared to estimate the benefit of building see exactly which substitution measures are the Shenmu-Huanghua railway. Despite the adopted when steam coal washing is elimi- added cost of building the line, systemwide nated. Overall, the systemwide cost is in- costs are reduced by 0.2 percent. creased from Y 197 billion to Y 201 billion. 3. This cost differential of about Adjustments to Cost Differentials Y 4 billion is considered as an estimate of the cost savings made possible by new steam 7. In percentage terms, the benefit coal washing. Were we to optimize the (cost savings) calculations are z:curate, but system without considering steam coal wash- adjustments must be made to get an accurate ing, and if the Chinese failed to build any measure in terms of yuan. This is because new plants, the cost of satisfying coal and the model's objective function includes only electricity demands would be about 2 perc- the discounted costs for the three "key ent higher. years" (1995, 2000, and 2005) of each FYP. 4. Likewise, for electricity trans- However, the cost differential cannot simply mission, a scenario was run in which no be multiplied by five, because that would new 500 kV intergrid power lines were overstate the benefits. The following adjust- allowed to be built. The scenario adjusts the ments must be made to get a more acctuiate 18-2 Annex 18 measure of the savings in economic terms. separatelyfor coal and electricity because the 8. First, estimate the ratio of the growth rate assumptions are different (1.8 p-. average demand for the entire FYP to the ercent for coal, 9 percent for electricity). key year demand. This must be done For electricity: (1L09) + (1.09)2 + fl.09)? + (1.09')' + (109? = (1.099 = 6.5211.538 - 4.23 For coal: (1.018) + (1,018)2 + (1.0189 + (1.018)4 + (1.018? = (1.0189 5.2765/1.0933 = 4.826 9. Second, factor in the different ity demand in 2000 converts to 793.98 mil- size of the coal and electricity demands. lion tons of coal. Thus, coal accounts for Nonelectricity demand for coal in 2000 in 53.3 percent of demand, and electricity the 9 percent demand cases is 905.78 mil- accounts for 46.7 percent. We apply these lion tons. Electricity demand in 2000 is percentages to the two ratios calculated 14,436 x1I0 kwh. Using a conversion factor above to arrive at an average ratio of key of 55,000 kwh per ton of coal, the electric- year demand to total demand: (.533 * 4.826) + (.477 * 4.23) = 4.54775 10. Finally, we multiply the cost cost savings over the model's 15 year time differentials between scenarios by 4.54775 horizon. to arrive at an estimate of the total For Steam Coal Washing: Y (2.0104 - 1.9718) x1O01 * 4.54775 = Y 17.55 billion For Intergrid Power Transmission: Y (1.9873 - 1.9718) xlO" * 4,54775 = Y 7.05 biltion For Accelerating Shenmu-Huanghua Railway: Y (1.9718 - 1.9684) x101' 4.54775 = Y 1.55 billion Annex 18 18-3 11. This same method is used in billion x 4.54775 = Y 897 billion for all 15 estimating the total systemwide discounted years. cost in Table 9.1 in Annex 9, e.g., Y 197 Annex 19 19-1 CHINESE EVALUATION OF THE CTS (DECEMBER 1991) 1. The China Coal Transport 3. From the point of view of the Study (CTS) aims at finding out the rela- function, scale and structure of the model, it tionships amnong energy production, trans- is the first of its kind that has ever been portation, consumption and environment built and is also quite advanced to compare protection as well as issues concerning their it with those similar models in the world. long-term developments. The project is 4. Tbis study has produced large coordinated by the Institute of Economics. number of data and infornation concerning Under the supervision of the World Bank coal production, selection and washing, experts and with the participation of the power production and transmission, con- Institute of Energy, the Institute of General struction of transportation network, alloca- Transportation and the Institute of Technical tion of traffic flows, environment control Economy of the State Planning Commission and the demand for coal and electric power and other related institutions of China, the etc. These data and information are very project has completed studies on the system- valuable for decision makers of related atic optimization for the production and departments. In addition, policy options and transportation (transmission) of coal and suggestions about relationships among differ- electric power, consumption and environ- ent industries made on the basis of the ment protection for 8th, 9th and 10th (1991- calculation and analysis are new and applica- 2005) five-year plans separately in three ble. years. This study is important both in theory 5. The latest software (Mathpro- and reality. It is a great achievement in the express) for linear planning and advanced field of soft science studies which has pro- GIS graphic display system software have vided methodology and tools of high level been used in the study. In addition, good for the improvement of existing management systematic users interface software also have work for economic planning. Great number been established for a more convenient data of valuable information and data for deci- inputs and output. Therefore, it is highly sion-making provided by CsS has enabled atomized and is easy for making corrections long-term development plans and policies for of the data and even the structure of the energy production and transportation, de- model. Mix integer programming model mand and environment protection to be (MEP model) has also been applied in the made scientifically. Further development, study to get solution by using PC 486 desk- application and popularization of CTS model top computer. This has provided better base is of great importance for acceleration of for the maintenance and expansion in usage scientific and modernized decisioa-making of the model. for China's economic planning. 6. Further development and im- 2. CTS not only has made good proveinent could be done to the study so as use of modem systematic engineering theory to provide a more efficient service to deci- but also advanced theories and methods for sion-makers and planners. New subjects for economic management. CTS model is an further study could be determined on the outcome of a systematic study of China's basis of this study. Priorities should be present situation of energy production, given tO those subjects like study on the transportation and consumption and their transportation price of coal and electric future trends. The model is an optimized power; comparative study on the energy re- one which is suitable for China's reality of serve and opening up; study on the coordi- coal and electric power production, transpor- nated development of economy, energy and tation, demand and environment control. enviromment and so on. 19-2 Annex 19 This evaluation was issued by the CTS Evaluation Committee and is quoted here in its entirety. Members of Chinese Evaluation Committee for the CTS No Namc Department Title 1 Guo Hongtao China Communica- Chairman, tion and Transport Central Association, SPC Advisory Commission 2 Sun Shangqing Development Re- Vice search Center State President Council 3 Gui Shiyong State Planning Com- Vice mission (SPC) Chairman 4 Wang Chaowong Planning Depart- Deputy ment, Director China Energy In- vestment Co. 5 Shi Dinghuan Industry Department Director State Science and Technology Com- mission 6 Wu Jiapei National Information Deputy Center Director 7 Fu Jiaji Management Re- Director search Institute, Professor Qinghua University NI N1ame Department Title 8 Li Xuesheng Planning Depart- Deputy ment, China National Director Coal Corp. 9 Li WVeimin Long-term Planning Deputy Department, SPC Director Annex 19 19-3 10 Li Duanshen Communication, Deputy Transport and Tele- Director communication De- partment, SPC 11 Yang Zhengguo Management Insti- Professor tute, Harbin Industry University 12 Zhou Dawen Planning Depart- Director ment, MOR 13 Lin Pinya Planning Depart- Director ment, MOC 14 Zhang Jianqiu Institute of Com- Assistant prehensive Trans- Researcher port, SPC 15 Qin Shengtao Science and Techno- Director logy Department, SPC 16 Xu Zhen Economic Research Deputy Center, SPC Director 17 Tu Zhuming Energy Department, Director SPC 18 Huang Zhijie Energy Research Institute, SPC Researcher No Name Department Title 19 Huang Yuqing China Communica- Deputy tion and Transport Secretary-in- Association, SPC General 20 Huang Fanzhang Economic Research Deputy Center, SPC Director 19-4 Annex 19 21 Lei Tang China Communica- vice tion and Transport Chairman Association, SPC 22 Wei Quanling Information Depart- Professor ment, People's Uni- versity Annex 20 20-1 LIST OF PARTICIPANTS AT CTS REPORT MEETING (OCrOBER 1993) Mr. Wei Liqun General Secretary of SPC Mr. Xu Zhen Deputy Director of Economic Research Center (ERC) Mr. Huang Fanzhang Deputy Director of ERC Mr. Ling Zoomu Deputy Director of ERC Mr. Meng Guangbing Deputy Director of ERC Mr. Tu Zuming Director of Energy Department of SPC Mr. Lan Shiliang Director of (Long-term) Planning Department of SPC Mr. Jiang Junru Deputy Director of Science Department of SPC Mr. MIa Deqing Director of Policy, Regulation, and Law Department, Ministry of State-owned Coal Mines Mr. Lu Qi Huaneng Energy Group Co. Mr. Shi Shanxin Ministry of Railways Ms. Ren Hong Ministry of Communications Mr. Zhang Jianxian State Energy Investment Co. Mr. Zhou Caiyu Director of Institute of Economy of ERC Mr. Xu Shoubo Director of Technical Economic Research Institute of ERC Ms. Liu Liru Director of Institute of Comprehensive Transportation of ERC Ms. Li Lianpu China Offshore Oil Development Co. Mr. Zhang Jianping Transport Department of SPC Annex 21 21-1 BACKGROUND oN EDELM4N AwARD 1. The Coal Transport Study has been (b) The Columbus-America Discovery awarded one of the six Finalist prizes for the Group, which made innovative use of prestigious 23rd Annual International Com- classical search theory and a collection petition for the Franz Edelman Award for of supporting models to successfully Management Science Practitioners. All find the sunken SS Central America, finalists can be considered prize winners, after many other groups had failed to since each finalist has a paper published it locate it. They recovered an estimated the January 1995 issue of the journal Inter- $1 billion in gold coins. (Runner-up, faces, has their presentation videotaped for 1991 competition) sale to universities, and receives a cash (c) The AIDS Division of the New Haven prize. Tnc Grand Prize Winner receives a Health Department, for evaluating the $10,000 award. New Haven needle exchange program. 2. The Edelman Award is given by The They designed data collection mecha- Institute of Management Sciences (TIMS), nisms for tracking needles and built a the main professional association for the mathematical model for the transmis- operations research/management science sion of HIV, which showed convinc- (OR/MS) community (along with their sister ingly that the program had reduced the organization, the Operations Research Soci- HIV incidence rate by 33 percent. ety of America (ORSA)). The award is (Grand Prize Winner, 1992 compe- given for completed, practical applications tition) 'that had significant, verifiable, and prefera- (d) BelIcore. the R&D consortium for the bly quantifiable impact on the performance seven regional telephone operating of the client organization." companies in the United States, for de- 3. This year's finalists include, in addi- veloping a mathematical progran for tion to the CTS: Digital Equipment Corp., maximizing the total project utility re- The U.S. Military Downsizing, Bellcore, the ceived by the operating companies Hanshin Expressway (Japan), and Tata Iron givent heir budget constraints. This re- and Steel Co. (India). Each finalist must sulted in an average of 23 additional give an oral presentation of the project at the R&D projects for each operating Boston Meeting of TIMS on April 24, 1994. company at no additional cost in 1991. A high-ranking official of the client organi- (Runner-up, 1992 competition) zation (in our case, the SPC) must attend the 5. Other Finalists during 1991 and 1992 conference and be available to answer ques- induded the Gas Research Institute; Mexi- tions about the impact of the modeling work co's Vilpac Truck Company; The City of on the client organization. New York arrest to arrmignent system; 4. Recent Edelman Award winners have Merit Brass Co.; the New Haven Fire De- included: partment; GE Capital Corporation; GTE (a) American Airlines Decision Technolo- Telephone Operatons; The U.S. Military gies, for their Sabre Reservation Sys- Airlift Command; Prudential Securities Inc.; tem, a series of st- tistical and mathe- The U.S. Postal Service; and Yellow Freight matical models to 'sell the right seats System, Inc. to the right customers at the right prices." This yield management work has generated more than $1.4 billion Speech by Mr. Gui Sbiyong in revenue for American Airlines at Edelman Competition during the last three years. (Grand Prize Winner, 1991 competition) 6. The results and conclusions of the CTS are guidance for acbieving the long 21-2 Annex 21 term goals of the coal, electricity, and trans- 12. Sixth, CTS has shown that portation system. The CTS, using scientific such an energy structure with coal as the methods, systematically described the deeper main cnergy resource will bring heavy and problems and contradictions in the system. long term burden on the environment protec- It is of great value for the adjustment of tion of China. Relevant departments are planning and decision-making, as demon- thinking about how to deal with these prob- strated by the following seven policy im- lems by the measures of energy conserva- pacts. t.on, increasing coal washing, expanding 7. First, one conclusion from hydropower, and installing more de-sulfur CTS, is that, for the capacity of coal, elec- and de-ash equipment, etc. tricity, and transportation system in China, 13. Seventh, opening of the Chi- it is possible to meet GNP growth of 8-9%. nese domestic market to the outside world but it is difficult to meet that of 10%. This will bring an even greater impact upon the conclusion was reflected in the Eighth Five energy equilibrium of China. Considering Year National Economic Plan. these reforms, the CTS used shadow prices 8. Second, relevant departments and international market prices to adjust are preparing to adjust their plans for coal, plans according to market requirements. For electricity, and transportation system, taking example, CTS proposed for the first time CTS results as reference. For example, we that it is reasonable to import coal in the are considering to increase the coal output coastal areas to alleviate the coal shortage. from 1.4 billion tons in the former Eighth Presently, some provinces like Guangdong Five Year Plan to more than 1.5 billion have followed this proposal. tons. 14. In December 1991,1 served as 9. Third, the CTS showed that co-chairman of the CTS Inspection Commit- coal transportation is still the main factor tee. In that meeting, the officials from dif- limiting the future coal supply, and strongly ferent agencies, and directors of several SPC supports the urgency of constructing a sec- plawning departments reviewed the CTS ond major railway from the energy base to optimization model and found it to be suit- the ports. Recently, the Vice Premier, Mr. able for China's reality. Zhu Ronugi, held a special meeting to dis- 15. In 1993, another conference cuss how to finance it and finish its con- was held to report updated results of the struction before 2000. model. In this meeting, the officials and 10. Fourth, the CTS gives quantita- experts said that the CTS has had an impor- tive analysis for the first time showing that tant influence on China's economy and coai washing is an effective way to transport should be used more widely. The CTS more energy over the same railway capacity. represents the first time that model results of In 1993, China passed a rule that all new this kind have been used at this level in steam coal mines that export coal to other China. provinces should include coal washing 16. The uses of CTS in China are plants. just beginning. As the Vice Chairman of the 11. Fifth, some provinces and State Planning Commission, I will support enterprises are considering to carry out some further applications of the CTS, including electricity transmission projects suggested by making economic forecasts for the year 2010 the CTS. For example, transmission from and planning in the departments of Trans- the coal base to the Shanghai region has portation, Energy, and Science and Techmol- entered the stage of preparation, and hydro- ogy. Also, some large enterprises such as electricity transmission to Guangdong is China Energy Group, China Ocean Oil under consideration by the Central Govern- Corp, and China Oil and Natural Gas Cor- ment. Furthermore, China has announced poration, are preparing to make analysis and plans to integrate its separate grids into a forecasts using CTS results. single power grid. Anne21 21-3 17. In mmary, in a country dha of Cbin. nd the environmeta protecion of uses - much coal - Cbhn. soving the th endre world. problems of coal production, tban_p--o, and consunpton scientiflcally wiM benefit gi': W d t bh __ the ownomic and socal developnet of all ., |d4 rn d o ClihEC S v_hi Annex 22 22-1 CTS IMPLEmEN T,iON CONFERENCE AND FRANz EDELMAN AwARD CEREMONY (OCTOBER 1994) 1. On October 28, 1994, a conference risk that no electricity will be delivered to was jointly sponsored by the World Bank them, and the difficulty of managing that and the Economic Research Center (ERC) of risk, is a barrier. the SPC in Beijing to discuss implcmentation 4. Distortions can be removed and issues relating to the recommendations of the barriers can be overcome with market CTS Phase 2 policy analysis. Attendees mechanisms such as prices that are were also introduced for the first time to representative of costs, reliable contracts, Ms. Xie Zhijun's findings on energy open markets, and corrective taxes. The conservation using the enhanced CTS model Central Government must try to provide an she developed during her McNamnara enabling legal and market environment for Fellowship. After-wards, a ceremony was them to choose the best decisions. held honoring the CTS team for being 5. Coal washing was used to Finalists in the 1994 Franz Edelman illustrate this kind of policy analysis. The Competition for Management Science recent guidelines issued by the Government Achievement. Publication and continuation that call for new coal mines to build of the CTS were also discussed. A Chinese washeries do not address the underlying translation of the Green Cover report was reasons why the mining companies do not distributed to all attendees. want do it. While scarcity of capital is 2. The main purpose of this meeting perhaps the greatest problem, the fact is that was to review the CTS recommendations there is a lot of capital being invested in and to discuss issues relating to their other part the coal-electricity system. implementation. The CTS model suggests Enterprises are choosing not to spend scarce economically efficient strategies for capita on coal washing because it does not producing and delivering enough coal and look as profitable as some other uses of electricity to satisfy China's projected capital in the coal-electricity system. economic growth. The central question of 6. The biggest distortion in the coal the meeting, then,was how should China washing area is that coal buyers are not implement these recommendations, given penalized for polluting the environment. tat it is in transition to a arket economy Pollution is a real cost to the economy as a and the central government can no longer whole, but polluters are not accountable for dictate all investments? Because of barriers it. Market mechanisms used in other and distortions, the market often does not countries to internalize such externalities send the right economic signals to were reviewed, including pollutiontaxes and enterprises. tradeable pollution allowances. Higher 13 The Bank's presentations defined rail rates were mentioned as another price a distortion as something that makes an signal that would give an incentive to wash enterprise want to do the wrong thing, and coal, so that coal with 30 percent of ash a barrier as something that prevents an would not be shipped over thousands of km. enterprise from doing the right thing even 7. Barriers were discussed next. when they want to do it. An example of a Even if distortions were removed and buyers distortion is a subsidized price, such as low began to demand washed coal, there would energy costs, which encourages an enterprise still be barriers making it risky for the to be wastefil. An example of a barrier is buyers and sellers. The proper legal when a coastal province cannot make a famework is necessary for willing buyers long-term contract with a minemouth and sellers to enter into long-term power plant that cam be re-ied upon. Tne contracts that guarantee a reliable market for 22-2 Annex 22 the washery's products and a reliable supply sold at the load center end of a grid at a for die special boilers designed to burn high price, this high price is not passed washed coal. Otherwise, buyers won't invest back to power generators at the input end of in the special boilers, and sellers won't the grid, i.e., the grid benefit is not shared. invest in the washeries. (Of course washing coal for minemouth 8. The responses of the Chinese plants would not create any transport savings discussants focused on the coal washing anyway.) example. Most experts agreed that the CTS 9. In sunmnary, while the CTS is recommendation to wash at least 30 percent very useful for identifying least cost of coal by 2000 is reasonable, and perhaps strategies, the meeting demonstrated how low, from an efficiency standpoint. The essential it is that Government officials try Chinese participants mentioned a nunber of to remove the distortions and barriers that different barriers and distortions: make least-cost strategies look Like high (a) the inability to pass higher costs on to cost strategies, and then set up market electricity consumers; mechanisms to send the right price signa!s. (b) how to finance coal washing investnent 10. After the discussion meeting, a (c) how to dispose of waste from coal ceremony was held to present the Franz washing; Edelman Finalist Award to the CTS team. (d) inefficient methods for matching needs The ceremony was chaired by Mr. Pieter of coal suppliers and coal buyers; Bottelier, head of the Bank's Resident (e) coal washing investnent costs can be Mission in China. A plaque was presented to Y 100 per ton, compared with Y400 per ton Mr. She Jiarnming, Vice Chairman of the for mining investment costs; SPC, and certificates were presented to the (1) resource taxes are higher for good ERC and to CTS team members. ne quality coal, lower for poor quality coal; Government leaders expressed strong (g) local governments will levy taxes for support for continuing the CTS. The expansion of existing facilities, such as ceremony was filmed and reported on both adding washeries to mines; and Chinese and English language TV news (h) the electricity price is not high enough broadcasts over the next few days. A to support coal washing at minemouth power complete list of attendees is included in the plants, because although electricity can be next page. Annex 22 22-3 List of Attendees: Gui Shiyong Former Vice Chairman of SPC and Deputy President of the Administrative Institute of China; She Jianming Vice Chairman of SPC; Wu Jingru Director, Electricity Bureau, State Development Bank Li Qun Director, Policy Bureau, State Bank Development Bank Zhao Kunin,g Division Chief, Transportation Bureau, State Development Bank Zhang Guocheng Division Chief, Industry Department, State Development Bank, Science and Technology Commission Jia Qinxiu Ministry of Coal, Senior engineer Sun Jianqi Division Chief, Information Division, SPC Zhang Xiaojian Information Division, SPC Xu Zhen Deputy Director of ERC Huang Fanzhang Deputy Director of ERC Wang Xingjia Deputy Director of ERC Lin Zbaomu Deputy Director of ERC Bian Bingyin Deputy Director of ERC Li Ping Division Chief of ERC Jiang Chunze Division Chief of ERC Tian Jun Division Chief, Long-term Planning Department of SPC Wei Yuanpeng Division Chief, Energy and Transportation Department of SPC Ren Long Division Chief, Policy Department of SPC Li Peng General Secretary, Academic Committe, of SPC Guo Yibing Huaneng Energy Group Co. Zhou Fengqi Director of Energy Research Institute of ERC 22-4 Annex 22 Zhou Xiaping Director of Institute of Comprehensive Transportation of ERC People's Daily CCTV Central Broadcasting Co. Economic Daily Guangming Daily China News Agency 4 RUSSIAN FEDERATION KAZAKHSTAti -7- J P MONGOLIA KYRGHYZ N REPUBLIC UZOI ISTAN N O.. 'M . . ........ hum .... d 73, d. -p W., a P." f f Id 0'0.p' u,,y fdwpelll he wpo y Wml iA b--i ....... .. .... sm Lo 02 Q1,ed L7:1 Sma e Z. PJEPA4 CHINA PROVINCIAL COAL PRODUCTION AND CONSUMP'l ION IN 1992 "C'CN 'ONCCAL IW." , ", MILLION 7cr4s %A- CNA. BY FROVINCE D;C,l By RCVNFE .4c! aou%04;1.s JV 4 FOR '997 90Q '992 3- 4 es F' "km e"N C, :E-A ;. C ? -,, Js sm 4_ RUSSIAN FEDERATION 4<-'. A KAZAKHSTAN -k'' ( -If ; T~~~~~~~~~~~~~~~~J /v _x, /-.--- -7 - PCF AND CONSUMPTION IN 1993 w - * -. ELECTRI C T ; ELECTRIClwy Ril9ER5 _s' S > PRODUCTION I CONSUMPTION o NATIONAL CAPTAL _ ' CTERAWATT HOURSI f TERAWATT HOURSi _V E I- BY PROVINC BY PROViNCE ceOVINCE BOLNOARLES /'N EL FOR 1993 L FOR 1993 - ,*NTERNAT'ONAIL OUNDARIES Z 4?<..VIET1i9 /9 -J *1 TEiAWATT HOURS 2AMILON MEGAWATT HOURS j-LS w vlo0yriO o 200 AoO o00 eoo JL _ - D C)wr , PrJUrPlPES - > , ' ' r - ~~~~~~~~~~~~~~~~THAILAND, EiE£ .......... - 3 91EE5 0 00 o 7x 34¶ 5o 2- 00 I00~~~~~~~~~ lIT ~~~~~~~~~~20 130 11Ihefbadwn codL. C H I N A dennn hcscd RAILWAY NETWORK 7h- WdnG r /P t _ DEM PEOPLE'S (-Ca. DEM. REPI ~ ~ ~ ~ ~~Hhh i-NANRP F OE usleo- \ lil 15 PHILIPPINES z91 DW X_ Yinchzw; k '\S>_ > t ~~~ Sh ngB / 0 KOREA nSB 1, \ Ynn'en L r aw $hWANDONG/ OINGHA I j i z4 5iiw-- L p _,,,;5 HAA t ; ; t J!4^j - 6 < x) 9 - ( Chzmgshnjit ~~~~~JIA' Xi hnX xs {/ v A - . . < ',_ f l, e ao c u jS aLAO PEOPLE'S Hio DEM. REP- 0; > i, \ ~~~~~~~~HAINAN ZTHAILAND; s c leo } g-> Z X <, lo- IXe ~~~~~~~~~~ ~~~~~PHILIPPINES CHINA 3r 120' -- - POWER PLANT AND / * N TRANSMISSION NETWORK RUSS I A UNDERt PANNED CONSI1RJCUGN EJIDSNG / N O O * THERMAL POWER PRANTS o ii * HYDROOWER PLANS * 0 4 NIEARPOWERP PNS A o \ 5D0 kV UNES - flseee __- 333-50DDEVUNES . a 0 LESS THAN 5C0 kV UNES / H L *N POWER GRID BOUNDARIES i *05 PROVINCE BOUNDARIES INIERNAIGMAI BOUNDARIES NCG - Nor1 C6in Gid MEG - NoE.Eal Gid ECG - to..I diM GM CCG - Ce.Eal Cie. G.io WCG - We,l Cdie GMd SCG - Seen ChiLe Gido 5DG Shod.nn Pince GMdd FiG - Fujien Provinc Gda YNDG Y-ece PlCvi;K GCid MONGOL*A w r DEM, pE*SevEM. P s -i NEI MONGOL - - - 0 l w4iN s ma REP. OF < V fi. A 2 P SH Yellnw S~~~~~~~~~~~KOREA W 0.t W /LTHAI[ANGHA I >X1~~~~~~ *.~'~ ~ 6 MP-IPNS e~- a::; | Yb < / A~~~~~~~~~~~~~~~~~~~~~J av ;uGXI mL _ r v t - '" *7 --__ _ _ _ _ _ _ _ _ _ _ __u ._ _ _ _ YUN~~~~~~~~~~~~~~~~~ .......... TAIWANl '7,, ~~~Y N-l G°XG MYANiMAR OL&W ITN f * LAO > ;< (IET NA Souih China Sea zo0r PEOPLE'S * f |THAILAND 0 E E..S o 93PI!PIE wo 1zB0> - fr r \ \< V 10w bU~~~~~~~~~~~~~~~~~~AILES 110- '0