For Nuclear Power in Developing Countries

CONFIDENTIAL

FOR REFERENCE ONLY NOT TO 8E REPRODUCED

MARKET SURVEY FOR NUCLEAR POWER IN DEVELOPING COUNTRIES

SINGAPORE

RESTRICTED

PRINTED BY THE INTERNATIONAL ATOMIC ENERGY AGENCY IN VIENNA SEPTEMBER 1973

FOREWORD

It is generally recognized that within the coming decades nuclear power is likely to play an important role in many developing countries because many such countries have limited indigenous energy resources and in recent years have been adversely affected by increases in world oil prices. The International Atomic Energy Agency has been fully aware of this potential need for nuclear power and has actively pursued a program of assisting such countries with the development of their nuclear power programs. So far, inter alia, the Agency has: (a) Sponsored power reactor survey and siting missions; (b) Conducted feasibility studies; (c) Organized technical meetings; (d) Published reports on small and medium power reactors; and (e) Awarded fellowships for training in nuclear power and technology. At present only eight developing countries have nuclear power plants in operation or under construction - Argentina, Brazil, Bulgaria, the Czechoslovak Socialist Republic, India, the Republic of Korea, Mexico and Pakistan. The total of their nuclear power commitments to date amounts to about 5200 MW as compared to an estimated installed electric generation capacity of about 56 000 MW. It is estimated that by 1980 only 8% of the installed electrical capacity of all developing countries of the world will be nuclear. In contrast, in the industrialized countries more than 16% of total electrical capacity will be nuclear by 1980. In view of the possible greater need for nuclear power in developing countries it was recommended at the Fourth International Conference on the Peaceful Uses of Atomic Energy, held in Geneva in 1971, and at the fifteenth regular session of the General Conference2, that efforts should be intensified to assist these countries in planning their nuclear power program. In response to these recommendations the Agency convened a Working Group on Nuclear Power Plants of Interest to Developing Countries on 11 - 15 October 1971 to review the then current status of the potential for nuclear power plants in these countries and advise on the desirability of carrying out a detailed market survey for such plants. As a result of its deliberations, the Working Group recommended that a Market Survey be carried out to determine in a more definitive way the size and timing of demand for nuclear power plants in selected developing countries where they might play an economic role in complementing conventional energy sources. The Working Group also pointed out that, although the Survey would be performed in the interests of the countries concerned, the results should be directed toward the nuclear industry, including manufacturing, engineering, construction and financial institutions, who would be looked to ultimately for meeting the requirements for equipment, facilities and financing as identified in the Survey. In response to these recommendations, the Director General decided.that the Survey should be undertaken and steps were initiated in November 1971. The objectives of the Survey as finally undertaken were as follows: (a) Examine the potential role of nuclear power in interested developing countries over the next five to fifteen years as a means of defining the size and timing of the installation of nuclear plants in this period. (b) Identify the specific market for small and medium power reactors in the countries participating in the Survey. (c) Estimate the financial requirements for the selected power system expansion programs in each of the participating countries.

Thus, this Survey will define the size and timing of the likely market for nuclear plants to be commissioned in the participating developing countries and the domestic and foreign financial requirements for that market in the 1980-1989 period3. It should be emphasized that this report provides only an indication of the need for nuclear power and associated financial considerations for the countries involved. The

1 As classified under the United Nations Development Program. 2 See General Conference Resolution GC(XV)/RES/285. 3 For convenience this will be called "study period" throughout the report.

— iii- scope of the data and information surveyed are not in such great detail as to allow the findings to be considered the equivalent of a rigorously determined feasibility study of any specific installation. The results, however, are as accurate as they could be made within the limits of data, time and manpower available. The methodology and analytical procedures used are believed to be accurate. In case the countries may need more detailed plans, an in-depth analysis will be required. It is suggested that the matter of defining the steps which would be needed to implement the suggested nuclear power programs, by all parties concerned, be the subject of further study after the participating countries have had an opportunity to thoroughly analyse the results of the Survey. In order to avoid biasing the results in favour of nuclear power, the approach and bases for analysis, including the technical and economic parameters, were subject to careful review by independent observers at the start of the study and prior to its completion. Comments by these observers were taken into consideration wherever possible. It is hoped that as a result of these reviews any bias however unintentional has been removed from the study.

SCOPE AND IMPLEMENTATION

In November 1971 letters were sent to 23 developing countries considered to be the most promising candidates for introduction of nuclear power in the time period of interest. Fourteen of these countries expressed an interest in participating and agreed to provide relevant basic data and counterpart staff to work with the visiting teams of experts. Seven Survey missions were undertaken as follows: Turkey-Greece 3-21 July 1972 Argentina-Mexico 7 August - 1 September 1972 Jamaica-Chile 4-15 September 1972 Republic of Korea-Singapore-Philippines 23 October - 17 November 1972 Pakistan-Arab Republic of Egypt 13 November - 1 December 1972 Thailand-Bangladesh 20 November - 8 December 1972 Yugoslavia 4-5 and 15-17 January 1973

The team selected for each mission was assigned the responsibility of collecting the necessary information on the characteristics of the power supply system(s) concerned, the projected power demand, current plans for expansion of the system(s), the availability of indigenous energy resources, and related economic and technical factors. This information was subsequently analysed by each mission team, reviewed by the country involved and used as a basis for the final report. Data gathered by the missions were also evaluated by the engineering staff of the Agency and by the experts assigned to the Survey. This evaluation included consideration of power flows in the basic interconnected system under normal operating conditions, the possible differences in transmission system requirements under varying generating capacity plans, an analysis of the transient stability and frequency stability of each system following an unplanned outage of one or more generating units, an analysis of alternative power system expansion plans involving nuclear and conventional plants and an estimation of the present worth of all costs for each plan. The results served as a basis for the selection of near-optimum power system expansion programs for each of the fourteen countries involved.

FINANCIAL AND MANPOWER SUPPORT OF SURVEY ACTIVITIES

Since the Market Survey was not foreseen at the time the Agency's 1972 budget was prepared, financial support was obtained from various countries and financial institutions. Furthermore, the work of the Market Survey could not have been completed within the time and manpower constraints but for the great efforts of the personnel in each country who participated in the preparation and review of data, the Agency professional and supporting staff, and the contributions of many other experts and organizations.

— iv— Support in cash funds was made available from:

Federal Republic of Germany US $ 25 000 Inter-American Development Bank 25 000 International Bank for Reconstruction and Development 50 000 United States - Export-Import Bank 75 000 Agency for International Development 25 000 Atomic Energy Commission 9 950

Total US $ 209 950

In addition, several countries provided experts on either a cost-free or partially cost- free basis: Approximate man-weeks Canada 22 Federal Republic of Germany 48 France 4 India 3 Japan 17 Sweden 9 United Kingdom 14 United States of America 19

Total 136

The fourteen participating countries contributed counterpart personnel and bore part or all of the expenses of each Survey mission during the time spent in the country in addition to the cost of preparing the responses and data required for the analyses. The Agency's contribution to the Survey included US $20 000 in cash plus approximately 260 man-weeks of professional staff, secretarial and administrative support, equivalent to about US $176 000. In addition, special consultants to the Agency provided about 170 man- weeks of support equivalent to about US $112 000. Based on the above, the total cost of the Survey is estimated to amount to US $555 000, including more than US $100 000 for cost-free services provided by its sponsors.

ACKNOWLEDGEMENTS

Mr. O. B. Falls, Jr. , USA, consultant to the Agency, was Project Manager for the Market Survey.

To list all of those who contributed in one way or another, even for one country, would be lengthy. Hence, specific recognition is limited to:

Associated Nuclear Services Ltd (ANS), London, England — who furnished, under a special contract, an electric utility system planning expert for each mission and co- ordinated the technical systems analysis work for the participating countries.

Bechtel Corporation, San Francisco, California, USA — who furnished complete heat rate data for the many sizes and types of fossil and nuclear power plants used in the Survey analyses.

Lahmeyer International GmbH, Frankfurt, FRG — who also furnished heat rate data, consulting service on costs and availability of smaller nuclear reactors, and an expert in mining of coal and lignite.

-v - Tennessee Valley Authority, Chattanooga, Tennessee, and the Atomic Energy Commission, USA — who made available TVA's basic power system planning computer program, Mr. Taber Jenkins of TVA's staff and Dr. David Joy of Oak Ridge National Laboratory (USAEC) to develop the changes required to provide the computer program capabilities especially needed for the Market Survey.

Others who contributed materially to the work of the Survey were the many organizations and the liaison officers from each country as listed in the Appendixes and the outstanding staff of consultants and Agency personnel who participated in the several missions and in the work at headquarters.

It is hoped that the information contained in this report will be of value to each country in formulating appropriate plans in regard to the potential use of nuclear energy for electric power generation in the years ahead.

- vi - INTRODUCTION

Fourteen Country Reports, one for each of the developing countries that took part in the Survey, have been prepared. These fourteen Country Reports are summarized in the General Report. Sections 1-8 of each Report contain data gathered during the visit of the team of experts and other data gathered for general accuracy. Sections 9-17 present the method of approach, the data used in the analyses, the analyses made and the results of the studies. General data and methodology common to the studies for all countries are given in the Appendixes. Section 1 concerns general economics and contains data on population, gross national product, mineral resources and energy consumption. Data on the national energy resources such as hydro potential, fossil fuel reserves, refinery capacity and production, and nuclear materials resources are given in Section 2. The electricity supply system, its development, generating and transmission facilities, costs of existing plants and plants under construction, various system operating and economic criteria, and technical data on existing generating units are given in Section 3. The historical growth of the electrical demand is described in Section 4, together with historical data on per-capita consumption, installed capacity, energy generated, load factor, and system load characteristics. Data are also given on system reliability, reliability criteria, and outage experience. The future system requirements are described in Section 5 including projections of maximum demand, generated energy, load factor and future reserve capacity. Also included are data on generating units and transmission facilities planned, under construction or projected, and on future sites. Section 6 contains data on local material and labour costs, labour practices, and the participation of local industry in the manufacture of power system components. Economic and financial aspects such as the method of evaluating the economic merit of projects, sources of funds, import duties and restrictions are described in Section 7. Section 8 contains a description of the administration and regulation practices of the Agencies responsible for nuclear power and information on nuclear legislation, licensing and safety. Section 9 describes the analytical approach used in the study; the bases of analysis, the computer programs, and the economic and technical methodology and parameters. The approach taken to determine the sensitivity of the results to certain parametric changes is also described. In Section 10 are described the bases, of the load forecasts used in the study, the future load characteristics such as seasonal peak demand, the load duration data, and the load factor. The results of the analysis of the factors limiting system development, made by Associated Nuclear Services, are given inSection 11, including data on system reliability, response of the system to loss-of-load, and recommendations on limits of generating unit sizes. The existing and committed electrical power system technical data, such as unit capacity, heat rates, fuel costs, forced and scheduled outage rates, seasonal and energy factors relating to hydro, and data on emergency hydro and pumped storage are given in Section 12. Capital cost data and the bases for their calculation are given in Section 13. The technical characteristics of the alternative generating units considered for the expansion of the power system are given in Section 14. The analyses of the alternative expansion programs are described in Section 15, including a discussion of the alternative plans considered, the method of determining the "optimum" expansion program and the consideration given to system reliability. The results of the study for the reference conditions and the sensitivity of these results to various parameters are given in Section 16. These results include the overall thermal plant additions required during the study period, the nuclear units required,and the financial requirements of the reference case expansion plan. The summary and conclusions of the study are presented in Section 17. A number of Appendixes have been included to provide additional information on the computer programs, methods of forecasting load, methodology and parameters used, fossil and nuclear fuel costs, general technical and economic data on thermal and nuclear plants, and other appropriate data.

-vii- TABLE OF CONTENTS

1. ECONOMIC BACKGROUND 1 2. NATIONAL ENERGY RESOURCES 13 3. ELECTRICITY SUPPLY SYSTEM 18 4. HISTORICAL SYSTEM DATA 34 5. PROJECTED SYSTEM DATA 48 6. NATIONAL CAPABILITIES AND LOCAL COSTS 55 7. ECONOMIC AND FINANCIAL ASPECTS 60 8. ADMINISTRATIVE AND REGULATORY 65 9. ANALYTICAL APPROACH 67 10. FORECASTS OF SYSTEM LOADS AND LOAD DURATION CURVES 70 11. LIMITING FACTORS IN SYSTEM DEVELOPMENT 73 12. CHARACTERISTICS OF THE EXISTING AND COMMITTED ELECTRIC POWER SYSTEM 77 13. CAPITAL COST DATA 81 14. CHARACTERISTICS OF ALTERNATIVE GENERATING UNITS CONSIDERED FOR EXPANSION DURING THE STUDY PERIOD 88 15. ANALYSIS OF ALTERNATIVE EXPANSION PROGRAMS 91 16. RESULTS 96 17. SUMMARY AND CONCLUSIONS 107

APPENDIXES A-N 113

- vin - 1. ECONOMIC BACKGROUND

1.1. Geographic features

The Republic of Singapore consists of the island of Singapore and 54 islets situated at the southern tip of the Malay Peninsula and separated from the latter by the Straits of Johore with a three-quarter mile causeway providing a railway link between the Singapore Island and the mainland of Malaya. It is situated between l°09' and 1°29' N latitude and 103°38' and 104°06' E longitude and is approximately 136.8 km north of the equator. To the west across the Straits of Malacca lies the elongated island of Sumatra and to the east across the South China Sea is the island of Borneo (see Fig. 1-1). The diamond shaped island measures approximately 41.8 km east to west and 22.5 km north to south with an area of 542. 6 km2 . The total area of the Republic including the smaller off-shore islands and reclaimed land amounts to 584.4 km2 (see Fig. 1 2). Of the 54 islands within the territorial waters of the Republic, the largest is the Pulau TekongBesar (17. 92 km2) lying to the east of Singapore Island. The other major offshore islands are Pulau Ubin (10. 14 km2 ) to the north-east, Sentosa(2.88 km2, formerly Pulau Blakang Mati) to the south and Pulau Bukom Besar (1.08 km2, oil-refinery and storage site of Shell) to the south-west. Other major islands situated off the south-west coast of Singapore Island are Pulau Ayer Chawan and Pulau Merlimau. The whole country is of very low elevation. It is generally undulating with low hills, the highest of which is Bukit Timah (elevation 177 m). About 64% of the country is less than 15 m above sea level. There are many small rivers, the largest of which is Sungei Seletar running 14. 5 km from its source through Seletar Reservoir to the sea in the Johore Straits.

EAST PAKISTAN n

INDIA

PACIFIC OCEAN

FIG. 1-1. SINGAPORE IN RELATION TO SOUTH-EAST ASIA. - 1 - WEST MALAYSIA (STATE OF JOHORE)

MILE 1 V2 0 1 2 3MILES

I 0 1 2 3 4 km • •••••••• nr RAILWAYS PSakijing Bendero \ C^P Seborok IS* Jotas Island) ROADS

0 PSatumu RofllK Light House

FIG. 1-2. MAP OF SINGAPORE. The lowland and mangrove forests that used to cover large parts of the island are being reduced by the advance of roads, buildings and cultivation. More specifically, the secondary jungles and swamps of Jurong, Pandan, Kranji and Seletar-Tampines areas have been cleared and reclaimed for industrial and residential purposes in recent years. Only the Bukit Timah Nature Reserve (66 ha) carries the primary forest. There are three major reservoirs on the Singapore Island, MacRitchie, Peirce and Seletar. These are situated in the central parts of the island. A total of some 2717 ha have been marked off as catchment areas for these reservoirs. The total capacity of these reservoirs, however, is inadequate to meet the water demands of the Republic which still depends heavily on Johore in Southern Malaya for the bulk of its daily needs. Geologically, the island may be divided into three distinct regions: the central hilly region of igneous rocks mainly of norite, granite and diorite, the western and southern region of sedimentary rocks mainly of shale sandstones and conglomerate forming a succes- sion of scarps and valleys, and the eastern region of Older Alluvium consisting of poorly consolidated sand and gravel with seams and patches of clay. The climate is equatorial with uniformly high daily temperatures and relative humidity. The mean shade temperature is 27.8°C with an average maximum of 30.6°C and average minimum of 23.9°C. The maximum shade temperature recorded in the last ten years was 34. 3°C and the minimum 20.1°C. The mean relative humidity is 84%. The average maximum relative humidity is 95.8% and the average minimum relative humidity 65.1%. For at least eight hours per day, the relative humidity is of the order of 90-95% and for at least a few hours per month, the relative humidity reaches saturation point. Rainfall is abundant throughout the year with December being the wettest month. The average annual general rainfall is 2438 mm. The highest annual rainfall recorded over the lasttenyears is 2921 mm, the lowest 1575 mm. The highest day's rainfall recorded was 431 mm on 9 December 1969. There are no distinct seasons except for the monsoons. The north-east monsoon lasts from mid-October to end of April and the south-west monsoon from mid-May to end of September. Lightning storms amount to 400 - 500 per annum being most prevalent in April and May.

1.2. Population

At the time of the founding of Singapore by Sir Stamford Raffles in 1819, the entire population on Singapore Island consisted of only about 150 or so fishermen and their families. The population of modern Singapore is, therefore, composed of migrants and their descen- dants who had come to make their fortune and stayed and made Singapore their home. The majority of the early migrants came from China, Malaya, Indonesia and India with smaller numbers of Arabs, Europeans and other races. The growth from about 150 indigenous fishermen to the present number of over 2 million may be divided into four phases according to the relative importance of fertility, mortality and migration. (See Table 1-1.) Migration was the single most important contribution to population increase during the first phase spanning the years 1819 - 1921. Natural increases were insignificant as the arrivals during this period were mainly men who did not intend to stay but would return to their homelands after accumulating sufficient wealth to start a better life. This unbalanced sex ratio resulted in a relatively low birth rate. On the other hand, because of poor preven- tive and curative medical measures, tropical diseases took their toll and the crude death rate was high, so much so that in the earlier years the rate of natural increase was negative because the crude death rate exceeded the crude birth rate. During the second phase from 1921 - 1941, population increase was still mainly due to migration. Natural increase, however, began to assume a more significant role in the population growth due to a combination of an increase in crude birth rate and a continuous decline in the crude death rate. The third phase covered the years 1941 - 1957. During this period natural increase replaced migration as the chief determinant of growth with, however, no slackening in the rate of increase. This was due to several factors. After the Second World War, there was stricter control on the admission of migrants which reduced the contribution to population growth from this cause. There was a rapid decline in the crude death rate with improved medical services. The crude birth rate, however, was maintained at a very high level which together with the declining death rate made up for the loss from reduced immigration to maintain the rate of population increase.

- 3 - The fourth phase which began in 1957 and continues to the present day is characterized by a fall in the population growth rate due to a sustained decrease in the rate of the natural increase and the reduction of migration to negligible proportions. The reduction in the rate of natural increase in 195 7 and the years immediately following may be attributed to several factors, one of which is the smaller proportion of women of child bearing age on account of the low crude birth rate and the high infant mortality rate duringthe Second World War. Under the Women1 s Charter of 1961, the minimum age of marriage was raised from 16 to 18 years. Women were also getting married later in life because of the raising of the standard of living, a higher literacy rate and a more sophisticated, urbanized way of life. The average age at marriage has increased steadily from 22.9 in 1961 to 24. 7 in 1965. The role played by family planning in the decline in natural increase since 1957 is dif- ficult to assess due to lack of statistics. The Family Planning Association was first formed in 1949 as a voluntary organization and until the late 1950's there was minimal interest or participation by the Government in the activities of the Association. In the early days, the primary objective of the Association was to improve the health of mothers by assisting them to avoid unplanned childbirths and to relieve the burden of those who could not afford to maintain a large family. It was not until the formation of the statutory Family Planning and Population Board in 1966 that family planning was adopted as a deliberate means to maintain and improve the standard of living of the country as a whole in the wider context of national economic planning. Under its Five-Year Plan, the FPPB aimed at persuading a total of 180 000 women, which was the estimated number of married women eligible during this period, to accept family planning and to limit the average number of offspring in each family to two or three. Abortion and sterilization have been legalized as part of the population policy to maintain population growth at a rate commensurate with economic development. As shown in Table 1-1, there has been a significant decline in the population growth rate during the past decade, by an average of more than 0.1% annually. If the growth rate sta- bilizes at the current rate of 1. 7%/yr, the population projection to the year 2000 will be as follows:

1980 2 456 000 1990 2 906 000 2000 3 440 000

If the growth rate were to decline by an average of 0.1% annually until zero growth rate were attained, the population would level off at about 2 500 000 after 1988; however, because of the present age distribution of women it is unlikely that the growth rate could continue to decline so fast for so long, in fact it could climb somewhat before starting to fall again. Much of the population is concentrated in the city area centred around the Singapore River. The population density decreases as one moves away from the city, though the eastern districts remain more densely populated than the western regions in Jurong. The population density for the country is currently about 3700 persons per km2; however, in the heart of the city the population density exceeds 100 000 per km2 . With limited land resources and the need to maximize the use of all available land, the older parts of the city which have been demolished for urban redevelopment are being re- built with multi-storey dwelling flats, offices and shops. Population density in the city areas will therefore continue to remain high if not actually be accentuated by urban redevelopment. With the development of the Jurong Industrial Estate and the policy of the Jurong Town Corporation to encourage workers to live near their place of work, the population density in the hitherto sparsely populated western part of the island has increased in recent years. This will be further accelerated by the development for public housing of the lands that formed part of the British bases and became available upon withdrawal of the British Armed Forces from Singapore. A recent areal distribution of population based on the 1970 census is shown in Fig. 1-3 (prepared by Tan Lee Wah, Department of Geography, University of Singapore, 1972). This population distribution, together with the small size of the country, severely limits the choice of sites for a nuclear power plant (see Section 6.4).

- 4 - TABLE 1-1. HISTORICAL POPULATION GROWTH

Av. annual Population , Year a growth rate ' (%)

1901 227.6a 2.3 1911 303.3a 2.9 1921 418.3a 3.3 1931 557.7a 2.9 1947 938.2a 3.3 1957 1445.9a 4.4 1960 1646.4 3.5 1961 1702.4 3.4 1962 1750.2 2.8 1963 1795.0 2.6 1964 1841.6 2.6 1965 1886.9 2.5 1966 1934.4 2.5 1967 1977.6 2.2 1968 2012.0 .1.7 1969 2042.5 1.5 1970 2074.5a 1.6 1971 2110.4 1.7

a Indicates census figure. Other numbers represent estimates of mid-year population.

1.3. National economics1

The "Key indicators" given in Table 1-2, published by the Asian Development Bank (July 1972), show Singapore to have a thriving economy. The unemployment rate has de- clined from more than 3% in 1966 to less than 2% in 1971. The Gross Domestic Product (GDP) at market prices increased at an average annual rate of 11.2% during the period 1961 to 1971 and at 1.3.9% during the period 1966 to 1971, while the retail price index was increasing at only 1.3% and 1.2% annually for the corresponding periods. Electricity production increased at average rates of 13. 6% and 15. 9% annually for the same periods, whilst for the latter of the two, the manufacturing index of production increased at an annual rate of 14.1%. In fact manufacturing growth has been greater than is indicated by this figure and the "manufacturing value added" is reported to have increased at a rate of 20.7% annually for the period 1961 to 1971 and at a rate of 25.1% annually for the period 1966 to 1971. At present, manufacturing and quarrying together provide 23% of the GDP compared with 9.8% in 1961. The traditional economic structure of the Republic was that of an entrepot port until the last decade when diversification into the manufacturing sector under an accelerated industrialization program reduced the dependence of the nation' s economy on entrepot activities. The manufacturing sector is now contributing one-quarter of the GDP compared to one- tenth a decade ago. Historically, Singapore by virtue of its strategic location in the main trade route between the West and the Far East and being endowed with a natural deep water harbour sheltered from the monsoons has been the major import/export centre for the neighbouring countries. Primary products, mainly rubber, copra, coconut oil, palm oil, coffee, pepper

1 The national currency is the Singapore dollar (S$). Its exchange rate relative to the US$ is given in Table 1-2. The rate at mid 1972 was 2. 86 S$ = 1 US$.

- 5 - 103* 55' 104" 00' 104* 05'

- 01* 25'

I ON I

EACH DOT REPRESENTS 1,000 PERSONS

- 0!* IS'

-01" 10' PREPARED BY TAN LEE WAH. DEPARTMENT OF GEOGRAPHY. UNIVERSITY OF SINGAPORE. 1972. 103" 45' 103" 50' 103" 55'

FIG. 1-3. AREAL DISTRIBUTION OF POPULATION. TABLE 1-2. KEY INDICATORS* UNIT lORQ 19 7 2 ITEM 1961 1962 1963 1964 1965 1966 1961 OU7 ( 1968 1970 1971 BASorE lull Jan Feb Mar Apr May Jun

PART A: BASIC DATA Population Mn 1.70 1.75 1.80 1.84 1.89 1.93 1.98 2.01 2.04 2.07 2.11 Labor Porcei/ 000 Unemployed 53.4 45.3 36.7 47.0 54.4 65.4 77.0 65.4 59.2 50.5 37.8 37.1 37.1 38.3 38.6 National Accounts GDP, factor cost (f.c.) S$ Mn 2240 2371 2684 2700 3043 3365 3692 4257 4833 5675 6471" GDP, market prices (m. p.) 2364 2501 2819 2846 3213 3570 3901 4496 5118 6020 6838 * Gross Domestic Savings 132 192 240 266 454 584 632 994 1312 1636 1838" Index of Production Manufacturing 1968=100 76.3 84.8 100.0 118.9 139.2 147.5 Electricity Production Mn kWh 720 776 823 914 1048 1236 1424 1639 1876 2205 2585 230 218 248 246 External Trade S$ Mn Trade Balance -655 -619 -805 -707 -803 -692 -915 -1193 -1503 -2778 -3284 -262 -273 Exports (f.o. b.) 3308 3417 3474 2772 3004 3374 3491 3891 4741 4756 5371 456 404 Rubber Ribbed Smoked Sheet 656 559 506 365 422 475 446 471 756 583 442 30 26 Motor Spirits Refined 96 84 88 76 67 89 85 108 112 116 159 14 13 High Speed Diesel Fuel 59 70 78 67 60 95 105 148 153 119 148 II 10 Other Fuel Oils 32 45 42 70 97 105 119 153 173 202 350 27 23 Palm Oil 20 20 23 25 36 36 39 40 49 87 134 10 6 Imports (c. i.f.) 3963 4036 4279 3479 3807 4066 4406 5084 6244 7534 8655 718 677 Rice Milled Whole 94 111 139 84 88 94 121 126 93 81 68 7 9 Rubber Ribbed Smoked Sheet 592 509 443 296 352 373 281 274 436 374 299 27 31 Saw Log, Veneer Log 17 19 24 27 31 42 53 66 76 87 95 9 8 High Speed Diesel Fuel 71 60 72 53 54 65 71 92 101 96 93 4 6 Other Fuel Oils 123 106 113 136 138 116 119 124 143 187 145 9 13 Balance of Payments S$ Mn Goods and Services -249 -176 -300 -93 -101 48 -188 -435 -660 -1881 -2635 Exports (f.o.b.J 2750 2860 3291 2601 2810 3168 3239 3589 4471 4428 5200 Imports (f.o.b,) 3358 3431 3995 3252 3570 3825 4149 4759 5863 7048 8539 Travel 19 27 44 50 53 72 108 128 196 248 299 Investment Income 340 368 360 508 006 633 614 607 536 491 405 Unrequited Transfers (net) -30 -23 -32 -73 -49 -45 -39 -41 -39 -24 -27 Capital Flows 14 27 41 22 81 51 113 271 173 444 462 Private 39 24 16 33 58 94 104 122 145 365 389 Official -25 3 25 -II 29 -43 9 149 28 79 73

Soutce: Asian Development Bank,Economic Office, Vol.Ill, No. 2 (July 1972). The abbreviations used in this table do not conform to those laid down for the Market Survey. In particular: Mn = million = 106, 000 = thousand = 103, MnkWh = GWh. TABLE 1-2. KEY INDICATORS (cont. )

UNIT i g 7 2 ITEM or 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 BASE Jan Feb Mar Apr May Jun

Net Errors and Omissions 279 363 372 54 31 133 473 753 994 1922 3045 Overall Surplus/Deficit 15 191 81 -90 -32 187 359 548 468 461 845 Change in Foreign Exchange Reserves -15 -191 -81 90 32 -187 -359 -548 -468 -461 -845

Public FinanceL/ S$ Mn Revenue 397 419 470 433 508 585 663 803 1261 1266 1423 141 125 187 99 Expenditure 385 352 601 326 392 530 592 702 1102 1206 1119 125 79 248 62 Current 279 231 462 159 209 316 403 476 769 830 621 75 45 176 38 Development 106 121 139 167 183 214 189 226 333 376 498 50 34 72 24

External Public Debt $ Mn Outstanding 8.8 6.5 21.0 20.4 26.7 50.9 74.8 169.6 206.9 269.2 Service Payments 0.7 0.7 i.O 1.4 2. i 4.7 7.1 II. 0

International Reserves S$ Mn ...... 1325.8 1318.7 1316.1 1509.1 1723.7 2279.0 2927.3 3584.7 4412.1 4370.6 4405.8 4484.9 4441.4 i CO Exchange Rate (Selling) s$/$ 3.06 3.06 3.07 3.06 3.08 3.07 3.08 3.09 3.08 2.90 2.86 2.85 2.84 2.83 2.82 2.86 i Price Index 1963=100 Retail (or Consumers')U 97.1 97.4 100.0 101.6 101.8 103.9 107.3 108.0 107.8 108.2 110.2 III.7 III.6 III.4 110.1 Money & Banking (Outstanding) S$ Mn Money Supply 734 795 841 854 895 1006 985 1 f 97 1421 1651 1915 1824 1919 1916 Currency in Circulation 375 389 404 439 469 508 423 £02 617 726 851 854 895 898 Demand Deposit 359 406 437 415 426 498 562 695 804 925 1064 970 1024 1018 Commercial Banks Time Deposit 31*7 415 438 484 549 634 919 1240 1563 1818 2168 2211 2251 2293 Savings Deposit 168 194 209 217 241 302 321 355 413 480 478 493 506 Loans 630 731 792 877 914 1026 1106 1351 1727 2168 2615 2690 2796 2787

PART B: GROWTH RATE %

Population 3.0 2.9 2.9 2.2 2.7 2.1 2.6 1.5 1.5 1.5 1.9 Labor Force Unemployed -1.8 -15.2 -19.0 28.1 15.7 20.2 17.7 -15.1 -9.5 -14.7 -25.1 -1.9 3.2 0.8 Cross Domestic Savings -6.4 45.5 25.0 10.8 70.7 28.6 8.2 57.3 32.0 24.7 12.3 Index of Production Manufacturing II.1 17.9 18.9 17.1 6.0 Electricity Production 9.3 7.8 6.1 II.1 14.7 17.9 15.2 15.1 14.5 17.5 17.2 3.6 -5.2 13.8 -0.8 Exports -1.9 3.3 1.7 -20.2 8.4 12.3 3.5 11.5 21.8 0.3 12.9 -3.0 -11.4 TABLE 1-2. KEY INDICATORS (cont. )

UNIT 19 7 2 1QR1 1QR? i qcq IQfM 1 QKR 19RR 1OR7 1QRQ 1QRQ ITEM 13D 1 1 tJUi. 1 OUU 1 9UJ 1 OUU 1 uU / I OUU 1 OUv 197I3(U0 1971 v <1 1 o r i out Jan Feb Mar Apr May Jun BASE

Imports -2.8 1.8 6.0 -18.7 9.4 6.8 8.4 15.4 22.8 20.7 14.9 0.6 -5.7 External Public Debt Outstanding ... -26.1 223.1 -2.9 30.9 90.6 46.9 126.7 22.0 30.1 Retail Price Index O.I 0.3 2.7 1.6 0.2 2.1 3.3 0.7 -0.2 0.4 1.8 1.4 -0.1 -0.2 -1.2 >1oney Supply -2.3 8.3 5.8 1.5 4.8 12.4 -2.1 21.5 18.7 16.2 ICO -4.8 5.2 -0.2

PART C: PER CAPITA

GDP, m.p. s$ 1391 1429 1566 1547 1700 1850 1970 2237 2509 2908 3241 Gross Domestic Savings 78 no 133 145 240 303 319 495 643 790 871 Electricity Production kWsih 42H 443 457 497 554 640 719 815 920 1065 1225 Exports 1946 1953 1930 1507 1589 1748 1763 1936 2324 2298 2545 Imports esi 2331 2306 2377 1891 2014 2107 2225 2529 3061 3640 4102 Capital Flows S$ 8 15 23 12 46 26 57 135 85 214 219 Government Revenue S$ 234 239 261 235 269 303 335 400 618 612 674 Government Expenditure S$ 226 201 334 177 207 275 299 349 540 583 530 Current 164 132 257 86 no 164 204 237 377 401 294 Development 62 69 77 91 97 III 95 112 163 182 236 External Public Debt $ Outstanding 5.21 3.76 11.80 11.21 14.35 26.65 38.16 85.23 102.43 130.05 Service Payments ... 0.4 0.4 0.5 0.7 I.I 2.4 3.5 5.3 International Reserves s$ ... 737 717 696 782 871 1134 1435 1732 2091 lloney Supply s$ 432 454 467 464 474 521 497 596 697 798 908

PART D: RATIO Development Expenditure/Government Expenditure 0.28 0.34 0.23 0.51 0.47 0.40 0.32 0.32 0.30 0.31 0.45 External Public Debt Out- standing/GDP m.p. 0.01 0.01 0.02 0.02 0.03 0.05 0.06 0.12 0.12 0.14 External Public Debt Out- standing'Intemational Reserves • • # ,,, 0.05 0.05 0.06 0.10 0.13 0.22 0.22 0.23 International Reserves/Imports (Merchandise) ...... 0.31 0.38 0.35 0.37 0.39 0.45 0.47 0.48 0.51 Service Payments/Exports (SIdse.) . . . 0.001 0.001 0.001 0.001 0.002 0.004 0.005 0.007 Money Supply/GDP m. p. 0.31 0.32 0.30 0.30 0.28 0.28 0.25 0.27 0.28 0.27 0.28 Currency in Circulation/Demand Deposit I.0H 0.96 0.92 1.06 1.10 1.02 0.75 0.72 0.77 0.78 0.80 0.88 0.87 0.88 TABLE 1-2. KEY INDICATORS (cont.)

UNIT 19 7 2 1 gc 1 IQR1 1Q.KR 1QRK 1QR7 iqco IQRQ 1970 1171 ITEM 130 1 1962 1 uU*t 1 3UJ I 30/ 1 jQO 1 \JUO lull or Jan Feb Mar Apr May Jun BASE

PART E: PERCENTAGE % GDP, f. c. by Industrial Origin 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Agriculture, Fishing & Forestry 6.0 5.8 5.5 5.3 4.6 4.5 4.0 3.5 3.2 3.0 2.9 Manufacturing & Quarrying 9.8 10.4 II.0 12.2 13.6 14.5 16.7 16.8 18.6 20.2 23.0 Construction 3.0 3.0 3.5 4.2 4.3 3.8 4.1 4.2 4.3 5.3 5.5 Electricity, Water Supply, etc. 2.1 2.2 2.0 2.2 1.8 2.2 2.5 2.6 2.5 2.4 2.4 Wholesale & Retail Trade 31.4 30.3 32.0 26.2 25.4 26.1 27.0 30.7 32.0 30.7 29.5 Ownership of Dwellings it. 5 4.4 4. i 4.4 4.2 4.2 4.1 3.9 3.8 3.C 3.7 Government Services 6.4 6.9 7.0 7.1 7.0 7.3 7.2 7.1 6.6 6.9 7.0 Services 36.8 37.0 34.9 38.4 39.1 37.4 34.4 31.2 29.0 27.9 26.0 CUP, m.p. by Expenditure 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Private Consumption 81.8 80.4 79.2 76.3 75.6 76.3 76.1 72.1 68.6 64.3 60.4 Public Consumption 8.7 9.4 9.5 9.7 9.7 10.0 10.2 10.5 11.4 11.8 13.6 Gross Domestic Capital Formation 9.5 10.2 11.3 14.0 14.7 13.7 13.7 17.4 20.0 23.9 26.0 Principal Exports as % of Total

Exports 19.8 16.4 14.6 13.2 14.0 14.1 12.8 12.1 15.9 12.3 8.2 6.6 6.4 MotoRubber r SpiriRibbet d RefineSmoked Sheet 2.9 2.5 2.5 2.7 2.2 2.6 2.4 2.8 2.4 2.4 3.0 3.1 3.2 High Speed Diesel Fuel 1.8 2.0 2.2 2.4 2.0 2.8 3.0 3.8 3.2 2.5 2.8 2.4 2.5 Other Fuel Oils 1.0 1.3 1.2 2.5 3.2 3.1 3.4 3.9 3.6 4.2 6.5 5.9 5.7 Palm Oil 0.6 0.6 0.7 0.9 1.2 I.I i.l i.O 1.0 i.8 2.5 2.2 i.5

tJ Figures refer to persons on the "live register" of the Employment Exchange (Ministry of Labour).

^J Figures for 1969 refer to Financial Year 1st January, 1969 to 31st March, 1970. Figures for 1970 refer to Financial Year 1st April, 1970 to 31st March, 1971.

SJ original base: April-May 1960 = 100. and timber, were imported from the neighbouring countries for sorting, grading and processing for re-export to the user countries. In turn, manufactured goods, foodstuffs, and machinery from Europe, the USA, Japan and other manufacturing countries were imported for distribution to the neighbouring countries. Besides the primary produce of the neighbouring countries and the manufactured goods of the industrialized countries, a third group of commodities which feature prominently in the entrepot trade are the petroleum products of the local oil refineries. Until the Second World War, economic development of Singapore was concentrated in the entrepot trade which provided wide scope for good profits. Development inthe manufacturing sector was neglected and the few factories that were set up were mainly associated with the sorting, grading and processing of primary produce from the neighbouring countries. Since the Second World War, however, there has been an increasing tendency for goods from these countries to bypass Singapore. In recognition of the need to diversify the economic structure of the country and to reduce the dependence of the country1 s economy on entrepot trade alone, the Government embarked upon an export oriented industrialization program a decade ago to develop the manufacturing and service industries. The major fields of industrial activities are in petroleum refining and distribution, ship building and ship repairing, manufacture of electronic/electrical goods and components and timber (sawmilling and plywood manufacture). While other entrepot goods, especially those associated with the neighbouring countries, have decreased in significance in relation to the nation1 s economy, entrepot trade in petroleum products refined in the Republic has been increasing over the last few years. Petroleum products are now the single largest earner of revenue for the customs. There are four petroleum refineries operating in Singapore (Shell, BP, Esso and Mobil) with a fifth (Singapore Petroleum) under construction. Another major oil company, Caltex, maintains a bunkering depot in the Republic. The ship building and ship repair industries experienced rapid expansion in the last decade. It was given added impetus by the taking over for civilian use of the British Naval Base facilities at Sembawang upon the withdrawal of the British Armed Forces. The Sembawang Shipyard has dry dock facilities for vessels of up to 100 000 dwt. Other major shipbuilders/repairs are the Keppel Shipyard, the Jurong Shipyard Ltd and Vosper Thornycroft. In anticipation of the arrival of the supertankers, one drydock is being extended to accommodate vessels up to 350 000 dwt. Shipbuilding in the past was mainly associated with small craft, but the industry now plans to build ocean- going vessels beginning with the Freedom-type freighters. The electronics industry which began with the manufacture of electronic components for export is now entering into the production of consumer goods such as radios and television sets. Besides the export oriented industries, manufacturing plants were also established over the last decade catering for local demands. These include those involved in food and drinks (sugar refinery, dairy products, canned foodstuffs etc.),textile, leather and rubber products (cotton goods, footwear, tyres etc.), chemicals and oils including plastics, and construction materials. The last category is fairly well developed and capable of supplying most of the materials required for the construction trade such as steel bars (from the National Iron and Steel Mill), ceramics and aluminium ware, cement, wood products (plywood, veneer), and other materials. Some of these manufacturers are now looking beyond the home market and are making efforts to increase their overseas trade especially in textiles and garments. The Republic has a well developed and efficient service sector constituting, in the main, banking, financial, transportation and communication services. There were 37 banks with a total of 181 banking offices operating in Singapore in June 1971. Eleven of the banks are locally incorporated. Singapore has excellent port and bunkering facilities. It replaced London as the fourth largest port (by shipping tonnage) in the world in 1968. The port facilities are being enhanced by the construction of a container complex in anticipation of increasing container traffic through Singapore. Economic growth is projected to continue at about 15%/yr during the 1970's and at about 10%/yr during the 1980's.

1.4. Total energy consumption

Singapore has no known energy resources of any form and depends on imported fuels for its energy requirements. The consumption of petroleum products which therefore indicates the total energy consumption and the details are given in Table 1-3. The fuel oil and

- 11 - TABLE 1-3. CONSUMPTION OF PETROLEUM PRODUCTS (103 tons (UK))

PUBa consumption for Total consumption electricity generation Year Major spirit Fuel oil Gas/diesel oils Fuel oil Burning oil Diesel oil including naphtha Bunker 'C

1963 108 333 113 41 1.5 249.5 1964 170 379 130 49 6.7 274.1 1965 186 406 119 48 18.2 304.7 1966 263 452 132 60 10.9 362.0 1967 273 500 145 62 5.4 409.3 1968 228 593 154 51 1.5 465.3 1969 293 680 173 55 2.3 521.7 1970 353 819 205 66 0.8 614.4 1971 391 921 209 54 0.9 706.7

a PUB = Public Utilities Board

diesel oil consumption for electricity generation are also entered in the same table for comparison purposes. The total energy consumption for 1970 is equivalent to 1.68X106 tons of coal equivalent (TEC2), or about 0.818 TEC per capita. The percentage share of electricity production in the total increased from about 42% in 1963 to about 45% in 1971.

1.5. Interest in nuclear power

Consideration of nuclear power for Singapore is based mainly on economic factors. The cost of nuclear energy as compared with other sources of energy (e.g. fossil fuel) for the generation of electricity must be economically justified before a decision could be made on the erection of nuclear power stations. It must, however, be borne in mind that oil prices, particularly from Middle East countries, have recently increased markedly and that the Republic's power supply is totally dependent upon imported oil and hence on the stability of oil supply. Power from nuclear energy may, therefore, prove to be attractive both in economic terms and also as a means of diversifying fuel supplies. No full scale feasibility study on the introduction of nuclear power has yet been carried out due to the comparatively small grid size of the Singapore system. However, officers of the Electricity Department of the Public Utilities Board (PUB) have constantly kept them- selves abreast of nuclear energy developments with a view to incorporating nuclear electricity generation as part of its long-term development strategy. With the rapid increase in system load demand, the PUB is planning to conduct a full scale study in the near future to determine the feasibility of integrating nuclear power into its system in the early/mid 1980s. Several preliminary studies have been undertaken and the reports are included in the list of references at the end. If Singapore were to build a nuclear power station, it is most likely that the unit size would be in the 400/500 MW range. The selection is largely governed by the system generating capacity and the peak load demand at the time of commercial operation. In the past, PUB has sent some officers to training courses/seminars in nuclear power organized by the IAEA. The Nuclear Committee of the Ministry of Science and Technology, is currently looking into training schemes for the manpower which would be required for future nuclear projects. Additionally, the Ministry has invited two officials from the Australian Atomic Energy Commission to visit Singapore to advise on Singapore' s requirement for the training of personnel in reactor assessment and reactor economics.

2 TEC = metric tons of coal equivalent = 28.5 x 106 Btu = 7.25 x 106 kcal.

- 12 - 2. NATIONAL ENERGY RESOURCES

2.1. Hydroelectric potential

In Singapore there is no hydroelectric potential of significance.

2.2. Coal and lignite

There are no indigenous fossil fuel resources of significance.

2.3. Oil and gas

As mentioned in Section 1.4, Singapore1 s energy consumption is based essentially on fuel products which, prior to 1961, were imported. Most of the local market requirement since 1961 has, however, been manufactured locally from imported crude. It is, therefore, fair to say that Singapore' s energy consumption is essentially based on fuel imports. Singapore has become one of the biggest centres of oil refining, blending and manufacturing in the Far East. There are four petroleum refineries in operation with a fifth under construction. Most of the products of the refineries are for bunkering and re-export. The Shell Refinery at Pulau Bukom off the south-west coast of Singapore has facilities for the bunkering of vessels up to 210 000 dwt. The internal market for petroleum is relatively small, the biggest item being fuel oil, most of which is consumed by the PUB, as already shown in Table 1-3. Crude oil for the refineries is imported mainly from Kuwait, Iran and other Middle East sources. In 1971, the cost amounted toS$720 million. The present total capacity of all the oil refineries in the Republic is about 494 000 barrels (bbl)/day which will be further expanded with the coming into service of the fifth refinery due for commissioning in mid 1973. Table II-1 shows the historical and projected refinery capacity in Singapore. By 1975 it is expected that the total refining capacity will increase to approximately one million barrels per day. The bulk of this future capacity will be designed to meet the growth in demand for residual fuel oil and naphtha in Japan.

TABLE II-1. HISTORICAL AND FORECAST REFINERY CAPACITY

Capacit3 y At end of year (10 bbl/d)

1961 48

1962 68

1963 68

1964 68

1965 68

1966 93

1967 185

1968 185

1969 185

1970 379

1971 384

1972 494

1973 937

1974 937

1975 999

1980 1267

1985 1447

- 13 - The future development of the oil refining industries could be influenced by three important factors. The first to be considered is the attitude of the Oil Producing and Exporting Countries (OPEC) towards the oil companies. The second is the running down of the war in Viet-Nam and the third is the desire of neighbouring countries now dependent upon imports for their oil and other petroleum products to be self-sufficient in refining capacity. The success of the off-shore oil exploration activities in this region will also have an important bearing on the future of the petroleum industries in the Republic. In the last few years, major oil companies have been stepping up their exploration in this part of the world. By virtue of its excellent transport and communication services and the availability of organized skilled personnel, Singapore has emerged as the base of operations for these activities. PUB has at present two fuel oil contracts, one with Caltex for the supply of Bunker 'C fuel oil to the Pasir Panjang Power Stations and the other with Mobil for the Jurong Power Station, both contracts running from 1969 to 1976. In each contract, the nominal contract price is subject to adjustment according to changes in the "reference price" which in the case of Caltex is the average of the lowest quotation per barrel for light fuel oil at Pulau Bukom and Ras Tanura and in the case of Mobil, the lowest price per barrel quoted for light fuel oil at Bandar Mah-Shahr as reported in Platt' s Oilgram Price Service. Table II-2 shows historical actual average prices paid by PUB through 1971. PUB has been exempt from fuel oil taxes since 1965. Table II-3 shows historical posted prices for fuel oil at the three reference f.o.b. points. The difference between Pulau Bukom (Singapore)

TABLE II-2. ANNUAL AVERAGE COST OF FUEL OIL

Year S $/t US $/t a US $/bbl b

1949 27.75

1950 35.53

1951 36.56

1952 42.67

1953 42.77

1954 42.99

1955 42.68

1956 42.67

1957 56.49

1958 53.67

1959 48.77

1960 52.26

1961 53.61

1962 41.75 13.64 1.98

1963 41.59 13.59 1.97

1964 42.37 13.80 2.00

1965 53.50 17.48 2.53

1966 46.40 15.06 2.18

1967 35.01 11.40 1.65

1968 36.50 11.85 1.72

1969 36.25 11.73 1.70

1970 30.18 9.80 1.42 1971 32.67 11.27 1.63

US $/t based on exchange rates shown in Table 1-2. b US $/bbl based on 6. 9 bbl/t.

- 14 - TABLE II-3. HISTORICAL FUEL OIL POSTED PRICES (US $/bbl)

Persian Gulf Pulau Bukom a Bandar-Mah-Shahr a Ras Tanura b

1967

July 11 1.95 1.60 1.55 July 21 2.00 1.60 1.55 August 28 2.10 1.60 1.55 December 11 2.05 1.60 1.55

1968

April 4 2.00 1.50 1.55 June 19 1.95 1.60 1.55 September 5 1.95 1.50 1.55 September 11 1.90 1.50 1.55

1969

May 20 1.90 1.42 1.45 May 26 1.80 1.42 1.45

1970

July 1 1.90 1.42 1.45 October 27 1.90 1.52 1.45 October 29 1.95 1.52 1.45 December 1 1.95 1.62 1.55 December 7 2.05 1.62 1.55

1971

February 15 2.30 1.92 1.85 June 1 2.55 2.05 2.05 December 31 2.60 2.05 2.05

1972

January 20 2.72 2.17 2.07

1973

January 1 Withdrawn 2.22 2.23

Shell. Esso.

and Persian Gulf postings reflects mainly the transportation costs from the Persian Gulf to Singapore. In August 1972, PUB was paying about S $41.5/ton (UK) (US $14.5/ton (UK)). Following the OPEC Agreement of Teheran in 1970/71, crude prices have increased in steps in November 1970, February 15 and June 1, 1972. There was also an unscheduled crude price increase of 8.49% in January 1972 due to international currency adjustment. However, this increase generally had no financial impact on local crude oil importers because the Singapore dollar was effectively revalued by 8.49%. For the future, negotiations with the Oil Producing and Exporting Countries (OPEC) are continuing. The Teheran Agreement in 1970/71 allows for future crude oil cost increases of approximately 7 US j£/bbl in January 1973, 1974 and 1975. The total cost increase which came into effect on January 1st, 1973, as a result of the various OPEC agreements (including the Teheran Agreement) was in excess of the scheduled 7 US ^/bbl. The future trend of crude oil prices will be dictated to a large extent by the attitude of the OPEC countries. One

- 15 - possible forecast is given in Table II-4. However, with the use of larger tankers, freight costs are expected to decline marginally in the near future. Looking further into the future, freight costs could be influenced by the issue of whether the larger super tankers (exceeding 200 000 dwt) would be allowed through the Malacca Straits. It is clear from the historical trend shown in Table II-3 that fuel oil prices move closely with crude oil prices. For example, the crude price increase in February 1971 resulted in an increase of fuel oil postings in both areas by approximately an equivalent amount, US $0.30/bbl. Based on this association, and the scheduled crude cost increases, fuel oil prices can be expected to increase in the future. Table II-5 shows the forecast, made in September 1972 for financial planning purposes, by PUB of fuel costs for their oil-fired stations. At present the PUB is burning fuel oil with the usual 3 to 3^% sulphur content. Environ- mental protection regulations could conceivably require changing to low-sulphur fuel in the future. It costs today about 0.70 US $/bbl to reduce the sulphur content from 3|% to l|% and an additional 0.50 US $/bbl to reduce it to 0.3% from l|%. A more than proportionate increase in cost would be involved if the sulphur content were to be decreased below 0.3%. Some consideration has been given to using naphtha as an alternative fuel, since it is essentially sulphur-free and currently has a posted price equivalent of only about US$0.24/bbl more than fuel oil (based on 20 January 1972 postings at Ras Tanura). The naphtha price is at present probably abnormally low due to temporary over-supply (the price differential was US $0.44/bbl in 1969). Also because of its greater volatility some technical problems would be involved in converting the power plants to burn it.

TABLE II-4. PAST AND PROJECTED CRUDE OIL COSTS (US $/bbl)

1971 1972 1973 1974 1975 1980 1985

Crude cost f. o. b. a 1. 78 1. 90 1. 97 2. 04 2. 11 2.50 3. 00 Freight a 0. 35 0. 29 0. 28 0. 28 0. 30 0.32 0. 32 Crude cost c. i. f. 2. 13 2. 19 2. 25 2. 32 2. 41 2.82 3. 32

Mainly from Persian Gulf.

TABLE II-5. FORECAST AVERAGE FUEL OIL COSTS FOR SINGAPORE PUBa

b c Year S $/t US $/t US $/bbl

1972 41.41 14. .48 2.10 1973 43.62 15.25 2.21 1974 45.95 16.07 2.33 1975 49.24 17.22 2.50 1976 51.62 18.05 2.62 1977 54.20 18.95 2.75 1978 56.91 19.90 2.88 1979 59.76 30.89 3.03 1980 62.75 21.94 3.18 1981 65.89 23.04 3.34 1982 69.18 24.19 3.51

Based on September 1972 forecast by the PUB Electricity Dept. b Based on 2. 86 S $/l US $. c Based on 6. 9 bbl/t.

- 16 - The distribution of refined products reflects primarily the distribution of the demand for them. Historically, the industry has been producing approximately 55-60% residual fuel oil (see Table II-6). However, it is expected that the production of residual fuel would increase to over 60% in the future to cater for increased local demand (including Bunkers) (see Table II-7) and as a result of Japan relying more on the Far East for residual fuel oil and placing less reliance on Persian Gulf sources. Consumption of crude oil would increase in the future (see Table II-1) along with growth in refining capacity in response to increased demand. Singapore' s declining reliance on petroleum product imports would continue as the diversification of refining capacity and the wider range of products enables the industry to meet its local and export markets. It should be emphasized that besides having the unique role of supplying virtually 100% of Singapore1 s vital energy requirement for economic growth, the oil industry in Singapore has also an important role in supplying the other markets in the Far East.

TABLE II-6. HISTORICAL REFINERY PRODUCTION {%]

1962 1963 1964 1965 1966 1967 1968 1969 1970 1971

LPG 0.1 0.2 0.3 0.2 0.3 0.2 0.2 0.3 0.4 0.6 Gasolines/naphtha 13.1 13.3 13.8 12.8 13.5 14.4 14.3 13.8 14.0 12.1 Distillates (kerosene, diesel) 31.7 31.2 29.4 24.9 26.5 28.6 31.0 30.9 31.4 33.7 Fuel oil 55.1 55,3 56.5 62.1 59.7 56.8 54.5 55.0 54.2 53.6

TABLE II-7. ACTUAL AND FORECAST FUEL OIL DEMAND

1971 1972 1973 1974 1975 1980 1985

103 bbl

Inland 5 992 6 370 7 274 8 312 9 090 13 378 16 804 Bunkers 22 925 28 580 30 085 32117 34158 47 404 60 315 Total 28 917 34 950 37 359 40 429 43 248 60 782 77119

103 ton (UK)a

Inland 868 923 1054 1205 1319 1939 2 435 Bunkers 3 323 4142 4360 4655 4 950 6 870 8 741 Total 4191 5 065 5 414 5 860 6 267 8 809 11176

Based on conversion rate of 6. 9 bbl to 1 ton (UK) of fuel oil.

- 17 - 3. ELECTRICITY SUPPLY SYSTEM

3.1. Past history and development of industry to present

The Electricity Department, which is responsible for the overall electricity supply in the Republic, is one of the three trading departments of the Public Utilities Board, the other two being the Water and Gas Departments. The Electricity Department was formerly part of the City Council but was incorporated into the Public Utilities Board when the latter was formed by statute in 1963. Tables III-1 and III-2 list financial and technical indicators of the state of development at the end of 1971.

(a) Generation

There are at present four generating stations in Singapore, the St. James power station, Pasir Panjang power stations 'A' and 'B', and Jurong power station. Work on a new power station at the Senoko Basin has started and the first set is scheduled for commissioning in 1975. Work on a gas turbo-alternator station comprising two units of 22. 5 MW each located at the same site is in progress and the first unit is scheduled for commercial service at the end of 1972. St. James, commissioned in 1926, was the first municipally owned generating station in Singapore. Prior to this and dating back to 1906, electricity was purchased in bulk from the Singapore Tramway Co. for distribution to consumers. The initial capacity of St. James power station was 2 MW, This was increased over the years by the installation of additional units. The total capacity of St. James reached 37 MW in 1948. By 1958, the generating plant at St. James had reached the end of its useful life and was dismantled and replaced by six free-piston gas turbine sets of 4. 5 MW.. each. The free-piston gas turbines were commissioned in 1960. Two package type gas turbine units of 11 MW each were added to the St. James power station in 1963 to meet the rapid increase in electricity demand. The total capacity of Pasir Panjang 'A' power station is 175 MW, consisting of seven 2 5 MW sets. The first set was commissioned in 1953 to relieve the acute shortage of generating plant in Singapore. Until then about one third of the city was subject to scheduled load shedding every night. Pasir Panjang 'A' power station reached its original design capacity of 150 MW in 1958 with the commissioning of the sixth 25 MW set. The station was extended to include the seventh set in 1962. Pasir Panjang 'B' power station, adjacent to Pasir Panjang 'A', was built to meet the increasing demand for electricity brought about by industrial development and improved living conditions. The first phase of two 60 MW sets was commissioned in 1965 and the second phase, also of two 60 MW sets, in 1966, bringing the station to its design capacity of 240 MW. Even while Pasir Panjang 'B' power station was being built, it was foreseen that new generating plants would be required by the late 1960's due to the high rate of load growth experienced and the forecast based upon the trend in the progress of industrialization. It was decided to construct a new power station in the Jurong Industrial Estate where factories and manufacturing plants were being built under the industrialization program. The original design capacity of Jurong Power Station was 240 MW consisting of four 60 MW sets. The first and second 60 MW sets were commissioned in the last quarter of 1969, the third set in 1970 and the fourth set in March 1971. With urban redevelopment and industrialization maintaining a high growth rate in electricity demand, it was found necessary to extend Jurong power station from its original design capacity of 240 MW to 600 MW by the addition of three 120 MW sets. These units are due for commissioning in 1973. The commissioning of the first 120 MW set at the Jurong power station will increase the largest unit size in the system from 60 MW to 120 MW. To meet the load requirements of the mid and late 1970's a new power station of three 120 MW sets is being built at the Senoko Basin in the north of Singapore Island. The first set is scheduled for commissioning in mid-1975 to be followed by the remaining two sets at six-monthly intervals. Future additional generating plant requirements immediately after completion of Senoko Stage I development would be met by further development of the Senoko site up to a total maximum of 1610 MW by the tentative addition of five 2 50 MW units, although preliminary investigations indicate that the Senoko Basin site may be capable of a maximum capacity of 2000 MW.

- 18 - TABLE III-l. DATA ON 1971 OPERATIONS OF THE PUB ELECTRICITY DEPARTMENT

Increase (decrease) 1971 over 1970 (%)

Installed capacity (MW) 704 9.3 Maximum demand (MW) 422 11.9 Energy generation (GWh) 2 585 17.2 Energy sales (GWh) 2 257 16.2 Number of consumers 291074 8.8 Number of employees 4143 4.1 Average revenue/kWh sold (S?i) 6.6 (3.2) Total operating revenue (106 S $) 148.4 13.0 Operation and maintenance expense (106 S ! 56.8 17.6 Depreciation expense (106 S $) 23.0 6.0 Interest expense (106 S $) 22.4 12.0 Net income (106 S $) 46.2 11.9

a Year-end PUB estimates.

TABLE III-2. SOME IMPORTANT INDICATORS OF THE 1971 OPERATIONS OF THE PUB ELECTRICITY DEPARTMENT

REFLECTING CHARACTERISTICS OF THE ELECTRICITY DEPARTMENT

Thermal generation as % of total energy generation 100 Average growth rate of energy sales in the last 5 years (%) 16.0 Annual load factor (%) 70.0 Industrial and commercial sales as "jo of total sales 78.8 Average fuel cost per 106 Btu (USj£) 25.2 Average consumption per consumer (kWh) 7 755 Total investment per kw installed capacity (US$) 374.4 Debt as "jo of total capitalization

REFLECTING EFFICIENCY OF THE OPERATIONS

System loss as "jo of total energy sent out 7.2 Average thermal efficiency of plants (<#>) 29.9 Average revenue per kWh sold to all consumers (USfi) 2.3 Average revenue per kWh sold to residential consumers (USji) 3.2 Average revenue per kWh sold to industrial and commercial consumers 1.6 Average O & M expense per kWh sold (US^) 0.9 Operating ratio {"jo) 53.8 Return on average net fixed assets in operation (%) 14.7 Net internal cash generation as % of capital expenditure 55.8 Number of times debt service covered by internal cash generation 2.0 Number of consumers per employee 70

- 19 - Meanwhile, the extremely high rate of growth in demand for electricity since the beginning of this year has necessitated advancing the installation program of the Senoko gas turbine station originally scheduled for commissioning in 1974. Construction of this station which consists of two packaged units of 22. 25 MW each is in an advanced stage. The first unit is scheduled for service by the end of 1972 followed by the second unit one month later. These two units will be required to make up for the anticipated shortfall in generating capacity in 1973 due to the abnormally high rate of growth in 1972.

(b) Transmission and distribution

The distribution voltage at the time of commissioning of St. James power station was 6. 6 kV. This remained the system's highest voltage for many years as the system remained small and distribution was confined mainly to the city areas. By 1952, when the first 25 MW set at Pasir Panjang 'A' power station was commissioned, the system had grown sufficiently to warrant the introduction of the 22 kV for transfer of bulk power from the generating station to major load distribution centres. The seven 25 MW sets at Pasir Panjang 'A' power station generate at 22 kV and feed directly into the 22 kV transmission system without the need for step-up transformers. With residential, commercial and industrial developments spreading outwards from the central city areas, the physical size of the transmission and distribution systems expanded accordingly. At the same time, the electrical loads in the built-up areas continued to rise so that the 22 kV transmission system became increasingly inadequate to meet the rising system demands economically and.technically. In 1965, the 66 kV system was introduced to supersede the 22 kV system as the primary transmission system for the transfer of bulk power, relegating the 22 kV system to serve as the sub-transmission system and in selected areas as the distribution system where the load density was very high. Except for some rural areas, high voltage transmission and distribution are carried out entirely by underground and submarine cables due to physical limitations imposed by the urban nature of development in the country. In the more remote rural areas, to save cost, distribution is by self-supporting aerial cables and pole-mounted transformers. Submarine cables supply power to the oil refineries on Pulau Bukom, Pulau Ayer Chawan and Pulau Merlimau at 66 kV while the Satellite Tracking Station on Sentosa is served by 22 kV submarine cables. By the end of 1971, there were 1403 transmission and distribution substations in service.

PUBLIC UTILITY BOARD

General Manager Deputy General Manager

General Manager Secretariat Personnel Finance Architectural & Office (3 Divisions) Department Department Maintenance (3 Divisions) (3 Divisions) (3 Divisions) Department (1 Division)

Gas Electricity Water Department Department Department (3 Divisions) (4 Divisions)

Chief Electrical Engineer Dep. Ch. Electrical Engineer

Generation Transmission Planning Commercial Inspectorate Administrative Division & & Division Division Division Distribution Design Division Division

FIG. 3-1. ORGANIZATION OF THE PUBLIC UTILITIES BOARD. - 20 - 3.2. Present organizational structure

PUB is a statutory body responsible for the organization of three utilities, namely, electricity, water and gas, in the Republic of Singapore. In accordance with the PUB Ordinance, it is headed by a Board comprising twelve members including the Chairman and the Deputy Chair- man. The overall administration of the Board is under the General Manager. The organizational structure of the Board is shown in Fig, 3-1 down to the department level except for the Electricity Department which is carried down to division level.

3.3. Geographical areas of responsibility

The PUB Electricity Department is responsible for the overall supply of electricity in the whole of Singapore.

3.4. Generation and transmission facilities

(a) Location and rating

With the exception of six free-piston engines and two gas turbines totalling 49 MW, all generation is by oil-fired steam plant. Table III-3 lists the generating units in service at the end of 1972 with their ratings and commissioning dates. The 2 5 and 60 MW units are normally operated at 80% of rated capacity, and usually at not less than 60% of rated capacity, though operation at as low as 40% could be permitted before shutting down a unit. Similar operating practice is contemplated for the 120 MW units now under construction. The gross energy production by the four stations for the past five years is shown in Table III-4. Figure 3-2 shows the locations of existing and future plants.

TABLE III-3. INSTALLED GENERATING PLANT 1972

Year in Installed Power station a Unit No. ^.^ gross capacity (MW)

St. James Free piston 1-6 1960 27 Gas turbines 1,2 1964 22 Pasir Panjang 'A1 1-7 1952/62 175 Pasir Panjang 'B' 1-4 1965/67 240 Jurong Stage I 1,2 1969 120 3,4 1971 120

Total installed capacity 1972 704

a Oil-fuelled.

TABLE III-4. GROSS ENERGY GENERATION (GWh)

Pasir Pasir Jurona Year St. James ... ,„, „ Panjang A Panjang B Stage I

1967 16.4 609 799 1968 4.2 618 1018 1969 6.3 651 1144 76 1970 2.2 575 1165 464 1971 2.3 504 1119 960

- 21 - STATE OF JOHORE N

POSSIBLE FUTURE NUCLEAR PLANT SITE SINGAPORE ISLAND

t-o ho

LEGEND • EXISTING THERMAL POWER STATIONS • PROPOSED THERMAL POWER STATION

? 1 2 3 U 5, 6 7 8 9 1.° I i i i i I i i i i I SCALE-km

FIG. 3-2. LOCATION OF GENERATING PLANT. STATE OF JOHORE

SINGAPORE

to EXISTING 66kV CIRCUITS PROPOSED 230kV CIRCUITS 66kV SUBSTATIONS FUTURE 230 kV SUBSTATIONS EXISTING THERMAL POWER STATION Pasir FUTURE THERMAL POWER STATION Panjang PS.

FIG. 3-3. ROUTES OF TRANSMISSION CIRCUITS. -Woodlands

LEGEND

INDICATES SYSTEM IN y- Mandai — Thomson -Changi LATE 1972

INDICATES ADDITIONS BY F MID 1973

Juron< Jurong IndL "Harbour * I Toa Payoh • I • Aljunied

.Jurong Kallang .ndan • Basin Sigtap - Crawford ~i '• Kallang Park

Telok Blancas Armenian

Merlimau - Queenstown "11•Out ram • ShentonWay

• PA.Chawan _i. Pasir .; PanjangB • SI. James Trafalgar

•Pulau Bukom . Pasir , . -J-!_Pasir Pan Jang Panjang A 'BExt.

FIG. 3-4. SINGLE LINE DIAGRAM OF THE 66 kV NETWORK.

(b) Transmission system

Since 1965 the main transmission voltage has been 66 kV and all 66 kV circuits are cables, either underground or submarine. Their routes are indicated in Fig. 3-3. A superimposed 230 kV underground system of primary transmission is planned, the first 230 kV substation being scheduled for mid-1974. The proposed configuration of this system at 1980 is also shown in Fig. 3-3. Figure 3-4 is a single line diagram of the existing 66 kV system and Fig. 3-5 shows the planned 2 30 kV development.

The estimated construction costs for Jurong power station Stage II (units 5 and 6, 2 x 120 MW) and Jurong power station Stage II Extension (unit 7, 1 x 120 MW) are shown in Table III-5 and III-6, respectively. They are broken down by major accounts and are also separated into their local currency and foreign currency components. The average cost per kilowatt for the three 120 MW units is about S $390, or about US $136 at an exchange rate of 2. 86. The local currency component of total cost for the three units is about 29%. The September 1972 financial planning forecast for Senoko power station costs is summarized as follows:

106 S 1972 Land and improvements 7 1972 Gas turbines (2 x 22 MW) 11 1971/77 Stage I (3 x 120 MW) 185

(d) Transmission line costs

Table III-7 indicates typical costs at 1972 prices for supply and installation of copper- cored oil-filled pressure cables.

- 24 - 2 . 1x50MVAR 11975) PER 500mm CIRCUIT ON 100 MVA BASE '2x50MVAR (1978) Z = IO.9*j3.8)1O"'p.u./km B = 3.8x10'2p.u./km SENOKO (ESTIMATED VALUES) POWER STATION 19JJ0 11760 METRES (1974) 32 550 METRES (1974) 11760 METRES (1978) 32550 METRES (1980)

CHUA CHU KANG SUBSTATION 1 x 150 MVA (1974) 2x150 MVA (1977) 4x150 MVA (1980)

KALLANG BASIN OTH SUBSTATION -IH 2x50 MVAR 1x150 MVA 11974) (1974) 2x150 MVA (1977) 3x150 MVA (1980)

2x150 MVA (1975)

13700 METRES AYER RAJAH 13 700 METRES (1975) (1975) SUBSTATION

: 1x50 MVAR (1977)

FIG. 3-5. PROPOSED 230 kV NETWORK.

3. 5. System reserve capacities

(a) General operating practices

It is the policy of PUB to provide a total plant reserve equivalent to the sum of the maximum scheduled capacity outage and the rating of the largest unit on the system. In operation, the spinning reserve is not less than the rating of the largest unit generating, and the load is distributed according to plant economic ratings. These are generally 80% of maximum continuous ratings. Generating units are loaded strictly in accordance with their economic merit order. There is no automatic frequency control system and all machines operate under speed governor control. No provision is made for automatic load shedding in emergency.

(b) Quantitative measures

In accordance with PUB policy the level of spinning reserve has fallen from 21% of peak load in 1968 to about 11% at the end of 1972 since the largest unit has been 60 MW throughout this period (see Table IV-3). With the commissioning of the first 120 MW unit in 1973 the spinning reserve will rise to 18% of peak load. System frequency is normally maintained between 49. 8 and 50.1 Hz. In emergency 66/22 kV transformers are isolated in successive blocks, each representing about 5% of the system peak load.

- 25 - TABLE III-5. ESTIMATED COST FOR UNITS 5 AND 6 OF JURONG POWER STATION (103 S $)

Unit No. 5 Unit No. 6 Total for Stage II It em Foreign Local Total Foreign Local Total Foreign Local Total

Civil Main foundations - 4 900 4 900 - - - 4 900 4 900 Steel frame buildings 1900 1900 3 800 - - 1900 1900 3 800 Circulating water works - 1417 1417 - - - - 1417 1417 Superstructures - 1753 1753 - 104 104 - 1857 1857 Bulk oil storage works - 1500 1500 - 500 500 - 2 000 2 000 Painting 150 448 598 92 276 368 242 724 966 Chimney 482 200 682 - - - 482 200 682 Subtotal 2 532 12118 14 650 92 880 972 2 624 12 998 15 662

Mechanical Boiler plant 8 800 2 800 11600 7 400 2 500 9 900 16 200 5 300 21500 Turbo-alternator plant 11250 1000 12 250 11250 1000 12 250 22 500 2 000 24 500 High pressure piping equipment 1460 410 1870 1460 410 1870 2 920 820 3 740 Low pressure piping equipment 540 150 690 340 110 450 880 260 1140 Boiler feed pumps 580 65 645 580 65 645 1160 130 1290 Circulating water pumps 1200 188 1388 400 62 462 1600 250 1850 Circulating water screening plant 268 19 287 268 19 287 536 38 574 Water treatment dosing plant 18 3 21 18 3 21 36 6 42 Chlorination plant 260 26 286 - - 260 26 286 Vacuum cleaning plant 100 20 120 15 2 17 115 22 137 Fire fighting equipment 415 75 490 181 56 237 596 131 727 Fuel oil tanks 640 320 960 320 160 480 960 480 1440 Air conditioning - 80 80 - - - - 80 80 Subtotal 25 531 5156 30 687 22 232 4387 26 619 47 763 9 543 57 306

Electrical 66 kV switchgear 300 24 324 300 24 324 600 48 648 Auxiliary switchgear 640 69 709 322 31 353 962 100 1062 66 kV transformers 1015 100 1115 835 80 915 1850 180 2 030 Auxiliary transformer 183 11 194 125 8 133 308 19 327 Cabling 958 425 1383 858 375 1233 1816 800 2 616 Lighting - 385 385 - 65 65 - 450 450 Communications equipment - 12 12 - 12 12 - 24 24 Subtotal 3 096 1026 4122 2440 595 3 035 5 536 1621 7157

Total, net 31159 18 300 49459 24 764 5 862 30 626 55 923 24162 80 085 TABLE III-5 (cont.

Unit No. 5 Unit No. 6 Total for Stage II Item Foreign Local Total Foreign Local Total Foreign Local Total

Contingencies Civil, 10% 253 1212 1465 9 88 97 262 1300 1562 Mechanical and electrical, 5<7o 1433 303 1736 1237 250 1487 2 670 553 3223

Total, including contingencies 32 845 19 815 52 660 26 010 6 200 32 210 58 855 26 015 84 870 Engineering 2 800 2 800 1720 1720 4 520 4 520 Site supervision 1360 1360 840 840 2 200 2 200

Total, including engineering and site supervision 35 645 21175 56 820 27 730 7 040 34 770 63 375 28215 91590 Interest during construction 1600 1230 2 830 1010 130 1140 2610 1360 3 970

Total gross 37 245 22405 59 650 28 740 7170 35 910 65 985 29 575 95 560 TABLE III-6. ESTIMATED COSTS FOR UNIT 7 (EXTENSION) OF JURONG POWER STATION (103 S $)

Item Foreign Local Total

Civil Main foundations 2 606 2 606 Steel frame building 1212 197 1409 Superstructures 864 864 Painting 100 300 400

Subtotal 1312 3 967 5 279

Mechanical Boiler plant 8 500 1500 10 000 Turbo-generator plant 13 086 1414 14 500 Piping equipment 1745 955 2 700 Boiler feed pumps 750 30 780 Circulating water pumps 627 73 700 Circulating water screening plant 120 10 130 LP chemical dosing plant 53 7 60 Vacuum cleaning pipework 33 9 42 Fire fighting equipment 115 35 150 Air conditioning 42 42 Subtotal 25 029 4 075 29104

Electrical 66 kV switchgear 1150 305 1455 Auxiliary switchgear 897 153 1050 66 kV transformers 228 222 250 Cabling 535 310 845 Lighting 175 175 Communications equipment 25 25

Subtotal 3 580 1070 4 650

Total, net 29 921 9112 39 033

Contingencies Civil, 10<7o 131 397 528 Mechanical and electrical, 5<7° 1430 257 1687

Total, incl. contingencies 31482 9766 41248

Engineering 2 212 2 212

Site supervision 1000 1000

Total, gross 33 694 10766 44460

TABLE III-7. UNDERGROUND CABLE COSTS'

S $/m US $/m

66 kV, 3-core, 290 mm 240 85 230 kV, 3x1 core, 630 mm2 521 185

a Includes supply and installation of copper cored oil-filled pressure cables including accessories and auxiliary cables.

- 28 - TABLE III-8. MONTHLY RATED STAFF ESTABLISHMENT - JURONG POWER STATION (STAGE I)

No. of posts Consolidated Designation of post Class or Grade salary scale 1971 1972 (S$/month)

Superintendent 1 1 Superscale Bl 1910-2210 Deputy Superintendent 1 1 Superscale B2 1750-2110

Senior Engineer (Operation & Efficiency) M Senior Engineer (Turbine Room) 1 1 Senior Engineer (Boiler House) 1 y Class I 1600-1950 Senior Engineer (Electrical) 1

Senior Chemist 1 r- i Senior Engineer (Instrument) 1 1 Engineer (Instrument) 2 2 Shift Charge Engineer 6 6 Engineer (Combustion) 1 1 Engineer (Turbine Room) 2 - Class II 1050-1850

Engineer (Boiler House) 2 C O Engineer (Electrical) 2 2 Chemist 1 2

Engineer (Operation & Efficiency) 1 i- i Assistant Shift Charge Engineer 6 6 Assistant Engineer (Combustion) 1 1 Assistant Engineer (Shift Control) 6 6 Assistant Engineer (Turbine Room) 2 3 Assistant Engineer (Boiler House) 2 3 Assistant Engineer (Electrical) 2 3 Assistant Engineer (Instrument) 2 3 • Class III 880-1650 Workshop and Service Engineer 1 1 Assistant Chemist 2 2 Assistant Engineer (Mechanical) 4 8 Assistant Engineer (Electrical) 3 3 Assistant Engineer (Instrument) 3 3 Assistant Charge Engineer 6 6 Administrative Officer 1 1 Administrative Assistant 1 1 Administrative Assistant Scale 650-1040 General Engineering Assistant (Shift) - 5 Mechanical Engineering Assistant (Turbine Room) - 1 Mechanical Engineering Assistant (Boiler House) 1 - Technical Special Grade 700-965 Mechanical Engineering Assistant (Combustion) _ 1 Electrical Engineering Assistant - 1 Instrument Engineering Assistant - 1 Engineering Assistant 3 3 Technical Timescale 190-800 Chemistry Technician 8 9 Technical Timescale 190-800 Unit Control Operator 40 40 Miscellaneous Grade 2 450-520

- 29 - TABLE III-8 (cont.)

No. of posts Consolidated Designation of post Class or Grade salary scale 1971 1972 (S$/month)

Control Room Operator 5 5 Miscellaneous Grade 2 450-520 Shift Technician 15 20 Maintenance Technician (Boilei House) 3 5 Maintenance Technician (Turbine Room) 3 5 Maintenance Technician (Electrical) 3 5 • Technical Timescale 190-800 Maintenance Technician (Instrument) 3 5 Combustion Technician 2 5 Electrical Draughtsman 1 1 Foreman (Instrument) 1 1 Foreman (Workshop) 1 1 Boiler House Maintenance Foreman 1 1 - Supervisor Grade II 680-855 Turbine Room Maintenance Foreman 1 1 Foreman Electrician 1 1 Assistant Foreman (Instrument) - 1 Assistant Foreman (Workshop) 1 . Overseer Grade A/Senior Assistant Foreman (Boiler House) - 1 450-645 Overseer Grade Assistant Foreman (Turbine Room) 1 Assistant Foreman (Electrical) - 1 Laboratory Assistant 3 3 Laboratory Assistant Timescale 190-670 Stenographer 1 1 Stenographer Grade 365-625 Clerical Assistant 5 6 Clerical Assistant Grade 165-435 Typist 2 3 Typist Grade 285-505 Comptometer Operator. 1 1 Comptometer Operators' Grade 270-520 Telephone Operator/Receptionist 2 2 Telephone Operator/ 205-385 Receptionist Grade Plan Keeper 1 1 Subordinate Division 13 175-190 Office Attendant 2 2 Subordinate Division 1 140-165

Total 175 318

3.6. Operating and maintenance costs

Tables III-8 and III-9 show the types, number and salary scales of monthly-rated and daily-rated personnel of the Jurong power station, as it is at present, with four 60 MW oil- fired units. (The three Stage II 120 MW units are scheduled to begin operation in 1973.) In 1972 there were 218 monthly rated staff, such as engineers, technicians, administrative and supervisory staff, and 258 daily-rated staff, such as electricians, fitters, welders, and labourers. Table 111-10 shows a breakdown of estimated 1972 generation costs for Pasir Panjang 'A', Pasir Panjang 'B' and Jurong (Stage I) power stations. The cost of monthly-rated staff, employee expenses, daily-rated labour as shown plus labour included in the "plant and

- 30 - TABLE III-9. DAILY RATED STAFF ESTABLISHMENT - JURONG POWER STATION (1972)

No. of Salary scale Designation of post Grade posts (S $/day)

Artisan I (Chargehand) 11 VI $13.80-16.20 Artisan I (Fitter I) 18 VI 13.80-16.20 Artisan I (Electrician I) 6 VI 13.80-16.20 Artisan II (Mechanic II) 6 V 10.30-13.20 Artisan I (Mason I) 3 VI 13.80-16.20 Artisan II (Machinist II) 3 V 10.80-13.20 Artisan II (Welder II) 3 V 10.80-13.20 Artisan II (Blacksmith II) 1 V 10.80-13.20 Artisan II (Carpenter II) 2 V 10.80-13.20 Artisan III (Leading Hand, Sp.) 6 IV 8.80-11.20 Artisan III (Motor Transport Drive) 4 IV 8.80-11.20 Artisan II (Fitter II) 22 V 10.80-13.20 Artisan II (Electrician II) 7 V 10.80-13.20 Artisan III (Machinist III) 7 IV 8.80-11.20 Artisan III (Mason II) 3 V 10.80-13.20 Artisan III (Machinist III) 3 IV 8.80-11.20 Artisan III (Welder HI) 4 IV 8.80-11.20 Artisan III (Carpenter III) 4 IV 8.80-11.20 Artisan III (Painter II) 3 V 10.80-13.20 WMSS I (Plant Attendant II) 35 III 7.60- 9.20 Artisan III (Leading Hand, Sp.) IV 8.80-11.20 Workman (semi-skilled) I III 7.60- 9.20 WMSS I (Engine Operator II) 1 II 6.40- 7.90 WMSS II 4 II 6.40- 7.90 WMSS I (Handyman) 36 III 7.60- 9.20 WMSS I 18 II 6.40- 7.90 Workman (unskilled) 30 I 4.60- 6.00 Workman (cleaning) 4 SG(A) 5.40- 6.40 WMSS I (Leading Hand) HI 7.60- 9.20 WMSS I IV 7.60- 9.20 Artisan I (Chargehand) I 13.80-16.20

Total 258

machinery" category, accounts for about 70% of non-fuel operation and maintenance costs. The unit costs shown in the table are based on gross energy generated. For the system, station losses are about 5. 7% of energy generated. Transmission and distribution losses are about 6. 6% of energy generated.

3. 7. Energy costs

The average system cost per unit sold is given in Table III-ll. A breakdown of depreciation, amortization and interest charges by station was not available, but a station energy cost for Jurong Stage I in 1972 can be approximated as shown in Table 111-12.

- 31 - TABLE 111-10. ESTIMATED 1972 GENERATION COSTS (103 S$)

Pasir Pasir Jurong Panjang Panjang (Stage I) 'A'

Operation & maintenance

Monthly rated staff 1277 1231 1214 A dministrative expenses 32 84 90 Employee expenses 210 202 199 Upkeep of land and buildings 206 128 100 Transport and haulage 43 37 84 Daily rated labour 570 305 100 Materials-sundries (excluding fuel) 215 166 200 Plant and machinery 1350 710 600

188 280 Other expenses 174

Subtotal, O & M 4 077 3 051 2 867

Fuel 10425 12 093 13 014

Total 14 502 15144 15 881

Number and size of units 7x25 MW 4 X 60 MW 4X60 MW

Gross generation (GWh) 794 1124 1165 Unit O & M cost (S mill/kWh) 5.12 2.71 2.46 Unit fuel cost (S mill/kWh) 13.13 10.76 11.17

TABLE III-11. TOTAL COSTS PER UNIT SOLDa (S £ /kWh)

1968 1969 1970 1971

Administration 0.112 0.158 0.154 0.107 Generation Fuel 1.170 1.181 0.962 1.015 Others 0.510 0.464 0.454 0.423 Distribution 0.363 0.303 0.292 0.252 Installation and inspectorate 0.039 0.040 0.036 0.035 Property tax 0.482 0.471 0.459 0.440 Service Departments' costs 0.167 0.158 0.128 0.097 Depreciation 1.431 1.297 1.117 1.021 Interest and amortization .1.095 1.083 1.031 0.979

Total 5.378 5.155 4.633 4.369

a The relationship between the gross and net generations and the sales can be obtained from Table IV-1.

- 32 - TABLE 111-12. APPROXIMATE TOTAL AND UNIT COSTS FOR JURONG STAGE I, 1972 s$xios

Station cost 112.7

Annual fixed charges at lO^o 11.27 Fuel costs 13.01 Other costs (O & M) 2.87

Total annual charges 27.15

Units generated gross (GWh) 1165 Unit cost on gross generation (Smill/kWh) 23.30 Units sent out (GWh) 1099 Unit cost on units sent out (Smill/kWh) 26.19 Unit cost on units sent out (USmill/kWh) 9.16

- 33 - 4. HISTORICAL SYSTEM DATA

4.1. Historical load growth

Table IV-1 gives, for 1965-1971, installed total capacity and capacity for each station (at year end and mean for year), firm capacity, maximum demand, annual increase in maximum demand, load factor (system, commercial, and plant), units (kWh) generated (total, and by station), station losses (by station, and system average), distribution and transmission losses, heat rates (by station, and system average), fuel prices (by station), and fuel consumption (by station, and total), and unit fuel costs.

(a) Demand and energy

Table IV-2 gives the historical load growth in Singapore from 1953 to 1971. From 1965 (the year in which Singapore became fully independent) to 1971, the compound annual growth rates of electric energy consumption and peak demand have been 16.3% and 14. 0%, respectively. Figure 4-1 shows the gross annual generation, energy sold and annual growth rate of generation. Figure 4-2 shows installed capacity, peak load, and annual growth rate of peak load. Figure 4-3 shows the annual system load factor. After hitting a high of 70% in 1971, the load factor is expected to average about 68% in future years (see Section 5). Figure 4-4 shows the annual load duration curves for 1970 and 1971. The system load factor was 66. 8% in 1970 and 69. 9% in 1971. Figure 4-5 shows the same information plotted on a relative basis (fraction of time against fraction of peak load). Figure 4-6 shows typical weekday load curves for 1968, 1969, 1970 and 1971. The corresponding daily load factors are 78.7, 80.4, 79.4 and 82.7%. Figure 4-7 shows the range of weekday (Monday through Friday) load curves, and typical Saturday and Sunday load curves, for October 1972. These curves correspond to a load factor of 78. 7% for Monday through Friday and of 76.1% for Monday through Sunday.

(b) Generating capacity

Figure 4-2 shows the growth of installed generating capacity from 1961 to 1971 in relation to the annual peak loads. Reserve margins are shown on Table IV-3.

Until 1965 the transmission voltage was 22 kV and the system was supplied directly from Pasir Panjang 'A1 power station at generator voltage. In 1965, with the commissioning of Pasir Panjang 'B1, 66 kV transmission was introduced, but this in turn will become inadequate by 1975 and will be supplemented by a new 230 kV system accepting the output from Seneko power station.

(d) Per-capita consumption

Table IV-2 gives per-capita consumption (based on gross generation) and per-capita maximum demand since 1960, based on the population figures of Table 1-1.

4.2. System reliability

(a) Reliability criteria

The system should be capable of accommodating the forced outage of the largest unit without load shedding. The transmission system should be capable of accommodating a single contingency, such as loss of a circuit or a transformer, without interruption of supply. Where this is not feasible supply is to be restored within a short time by transfer of the load to adjacent stations.

- 34 - TABLE IV-1. ANNUAL GENERATION STATISTICS - PUB ELECTRICITY DEPARTMENT (1965-1971)

1965 1966 1967 1968 1969 1970 1971

1 Installed capacity (MW) - Total - December 344 404 464 464 584 644 704 2 — Mean 279 364 464 464 484 587 694 3 St. James — December 49 <• 49 A AQ r <± ~ Mean Pasir Panjang 'A' 175 - \ 5 - December 7 C i nc f D 7 Pasir Panjang 'B' - December 120 1°0 f p T An * o *"* IV163n 00 Jurong Stage I - 120 180 240 9 - December 10 - Mean - - - - 20 123 230 21 Maintenance and spinning reserve (MW) — December 145 < x 22 Firm capacity (MW) - December 199 259 319 319 439 499 559 23 Maximum demand (MW) 192.5 223 248 283 320 377 422 24 Annual increase in maximum demand (<7o) 14.6 15.8 11.2 14.1 13.1 17.8 11.9 25 System load factor 62.1 63.3 65.6 65.9 66.9 66.8 69.9 26 Units generated (GWh) - Total 1 047. 6 1236.5 1424. .5 1639.4 1 876.1 2 205.2 2 585. 3 27 St. James 57.9 33.6 16.4 4.2 6.3 2.2 2.3 28 Pasir Panjang 'A1 853.6 713.1 609.4 617.6 650.6 574.7 504.2 29 Pasir Panjang 'B' 136.1 489.8 798.7 1017.6 1143. 5 1164. 8 1118.5 30 Jurong Stage I - - - - 75.7 463.5 960.3 36 Work units (°/o) - St. James 4.49 6.55 10.37 30.95 19.05 51.36 48.77 37 Pasir Panjang 'A' 4.96 5.23 5.17 4.78 5.39 6.01 5.99 38 Pasir Panjang 'B' 7.13 6.33 5.58 5.29 5.24 5.44 5.52 39 Jurong Stage I - - - - 7.46 6.26 5.97 45 System 5.26 5.69 5.54 5.30 5.43 5.81 5.82 46 Units sent out (GWh) 992.5 1166.1 1345.5 1 552.6 1774.3 2 077.1 2434.8 47 Free supplies ("Jo) generated - - - 0.01 0.01 0.01 - 48 Distribution and transmission losses °lo of units generated 7.64 7.38 7.51 6.44 6,46 6.13 6.42 49 "Jo of units sent out 8.07 7.82 7.95 6.80 6.83 6.51 6.82 50 Units sold 912.5 1 074. 9 1238. 5 1446.8 1652.9 1 941. 6 2 268.9 51 Commercial load factor 54.1 55.0 57.0 58.2 59.0 58.8 61.4 52 Plant load factor - Total 47.1 41.-6 37.0 42.5 46.7 44.8 44.1 St. James 13.5 7.7 3.7 0.9 1.2 0.5 0.5 Pasir Panjang 'A' 65.0 54.3 46.4 47.0 49.5 43.7 38.4 Pasir Panjang 'B' 28.2 40.0 38.6 48.4 54.4 55.4 53.2 Jurong Stage I 43.2 43.0 47.7 TABLE IV-1. (cont.)

1965 1966 1967 1968 1969 1970 1971

62 Heat rate (per kWh) St. James — Generated 13 655 14054 14276 15 807 16187 16 965 17 044 63 - Sent out 14305 15 017 15 907 22 614 19 908 34876 33270 64 Pasir Panjang 'A' — Generated 12 975 13226 13425 13 398 13215 13 311 13291 65 — Sent out 13 651 13956 14157 14 071 13 968 14163 14138 66 Pasir Panjang 'B' - Generated 11737 11223 10 985 10 837 10 988 11162 11209 67 — Sent out 12 673 11982 11653 11442 11597 11804 11864 68 Jurong Stage I - Generated - - - - 11217 10 716 10 654 69 - Sent out - - - - 12120 11432 11331 80 Total system (weighted) — Generated 12 851 12 454 12 065 11797 11794 11634 11414 81 Fuel oil prices (S $/t) St. James on fi 83 Pasir Panjang 53.30 46.60 35.01 36.50 36.25 30.00 29.95 84 Jurong - - - - 40.16 30.86 37.25 86 Fuel Consumption St. James - t 18 272 10 915 5 431 1542 2 370 883 915 87 - 103 S $ 1583 957 478 136 208 77 80 90 Total industrial diesel - t 18272 10 915 5 431 1542 2 370 883 915 91 - 10s S $ 1583 957 478 136 208 77 80 92 Pasir Panjang 'A' - t 266 505 227 431 198233 199 840 208 340 184770 160 558 93 - 103 S $ 14332 12103 6 919 7 274 7 545 5 566 4 872 94 Pasir Panjang 'B' - t 38251 134560 211143 265461 302 389 311 358 300110 95 - 103 S $ 2 057 4922 7 362 9 652 10 952 9 376 8 989 96 Total Pasir Panjang - t 304756 361991 409 375 465 301 501265 496 058 460 668 97 - 103 S $ 16 389 17 025 14281 16 926 18497 14942 13 861 98 Jurong Stage I - t - - - - 20 437 118 419 243 978 99 - 103 S $ - - - 821 3 655 9 089 112 Total Bunker C - t 304756 361991 409 375 465 301 521702 614475 704 646 113 - 103 S $ 16 389 17 025 14281 16 926 19 317 18 596 22 951 114 Total fuel - 103 S $ 17 973 17 981 14759 17 061 19 525 18 674 23 031 115 Fuel cost (S mill/kWh) On units generated 17.15 14.55 10.36 10.41 10.41 8.47 8.91 116 On units sent out 18.10 15.52 10.97 10.99 11.01 8.99 9.46 117 On units sold 19.71 16.73 11.92 11.79 11.81 9.62 10.15 TABLE IV-2. HISTORICAL LOAD GROWTH Energy consumption Maximum demand Energy consumption Maximum demand per capita per capita Year Annual growth Annual growth Annual growth Annual growth MW GWh kWh W •(.% (%) 0»)

1953 34.2 212.5 1954 65.5 91.5 306.7 44. 3 1955 74.0 13.0 366.9 19.6 1956 87.5 18.2 434.0 18.3 1957 95.5 9.1 496.8 14.5 1958 105.7 10.7 571.3 15.0 1959 112.9 6.8 616.2 7.9 1960 118.5 5.0 658.5 6.9 400 72 1961 128.5 8.4 719.6 9.3 423 5.8 75 4.2 1962 139.0 8.2 775.7 7.8 443 4.7 79 5.3 1963 151. 8 9.2 822.9 6.1 458 3.4 85 7.6 1964 168.0 10.7 914.2 11.1 496 8.3 91 7.1 1965 192.5 14.6 1047.6 14.6 555 11.9 102 12.1 1966 223.0 15.8 1236.5 18.0 639 15.1 115 12.7 1967 248.0 11.2 1424. 5 15.2 720 12.7 125 8.7 1968 283.0 14.1 1639. 4 15.1 815 13.2 141 12.8 1969 320.0 13.1 1 876.1 14.4 919 12.8 157 11.3 1970 377.0 17.8 2 205.3 17.5 1063 15.7 182 15.9 1971 422.0 11.9 2 585.3 17.2 1225 15.2 300 9.9

4000

Q LU 2000 LU ^^. LU 13

ERG Y i 1000 LU

500 1961 62 63 64 65 66 67 68 69 70 71 72

FIG. 4- 1. GROSS GENERATION AND SALES OF ENERGY 1961 - 1972.

- 37 - in o o d z z z — m m o § z o z 1-1 z *t (B z

PA N UJ 2 Q:

JA M U) .SI R ^SI R •»* 0- 35 a. 1000

t J 500 r CAF'ACITY 1 s J r o 2 PEAK LO 1^ J 200 20 a J ANN JAL GROWTH OF / z v / PEAK •/' 15 \* \_ < —• \ \ o \ 100 s 10

1961 62 63 64 65 66 67 66 69 70 71 72

FIG. 4-2. INSTALLED CAPACITY AND PEAK LOAD 1961 - 1971.

- /

- on / o t— o 66 / < ^— u. Q < - o /

1965 66 67 68 69 70 71 72 FIG. 4-3. ANNUAL LOAD FACTOR 1961 - 1971.

- 38 - LOAD MW ro o o o o p o o o o -"I FRACTION OF PEAK DEMAND •j / ro I G / ro / o p p o o b ro in CD d / ad

0. 2 if / CD £|

IO N / \. * J CO a> co 00 o j / / / co o o o / o / / Tl o / f to o -C P m en v> d i—

/, CD c it / O 197 0 w / I a> 197 1 o 1 1 (D a> 1 1 o / / Ij 1 1 o 1 1 d 1 1 1 1 o t 1 CD -» / a o /

CO 400

380 s 1971 360 \ 340 r 320 1 1970 r s 300 \ Z. 280 A Q < 260 -—— N Y\ —1969 Jr Q 240 220 v^^.— / 1968 7 200 fr 180 ^^ 160 140 :\^ 120 0 1 2 3 4 5 6 7 8 9 10 11 12 13 H 15 16 17 18 19 20 21 22 23 24 TIME, HOURS

FIG. 4-6. TYPICAL WEEK DAY LOAD CURVES FOR 1968 - 1971.

WEEKDAYS: MONDAY 2ND - FRIDAY 6TH0CT. 521.0 MW AT 1641 HRS. SATURDAY: 28TH OCT. ^ THURSDAY 10.12.72 - SUNDAY: 29™ OCT. A RANGE OF WEEKDAY *'\ / LOADS

10 12 1( 16 20 22 24 HOURS

FIG. 4-7. MAXIMUM AND MINIMUM LOAD DEMANDS FOR OCTOBER 1972.

- 40 - TABLE IV-3. GENERATION RESERVES

1968 1969 1970 1971 1972

Peak load (MW) 283 320 377 422 540 a

Installed capacity (MW) 464 524 584 704 704

Maintenance reserve (MW) 85 85 85 85 85 (as °lo of peak) 30.0 26.6 22.6 20.1 15. 8 a

Spinning reserve (MW) 60 60 60 60 60 (as °]o of peak) 21.2 18.8 16.0 14.3 11. la

Minimum required total reserve (MW) 145 145 145 145 145 (as °lo of peak) 51.2 45.3 38.5 34.4 26. 9 a

Actual year end margin (MW) 181 204 207 282 164 a (as % of peak) 63.9 63.7 55.0 66.9 30.4 a

a Calculated on forecast peak load for 1972. (b) Outage records

Table IV-4 gives a summary of the forced outages of the Jurong power station. Table IV-5 gives a summary of generator outages on other units for the period 19 March 1969 to 30 September 1972. For all the stations there were fourteen outages in 1969, fifteen in 1970, ten in 1971 and three in the first nine months of 1972. System load shedding occurred twice in 1970 and twice in 1971 with 10 to 128 MW being shed. For the Jurong power station alone the load shed varied from 6 to 60 MW. Many of the outages are attributable to start-up troubles and cannot be considered typical. This can be seen from the following summary of these outages:

Unit Hours forced out

No.l - first 5 months 278 next 30 months 7 No.2 - first 7 months 1 363 next 25 months 18 No.3 - first 5 months 286 next 20 months 0 No .4 - first 4 months 716 next 13 months 1

Table IV-6 gives transmission and distribution outage records for January 1971 through August 1972. Table IV-7 describes the faults of the 66 kV network from 1 January 1970 through 6 September 1972. The 66 kV cable trips can be summarized as follows:

Cause T97() 1971 1972a

Electrical failure 10 0 Shipping 0 10 Relay malfunction 0 0 2 Excavation 0 0 1

1972 data to November only.

- 41 - TABLE IV-4. FORCED OUTAGES RECORD OF JURONG POWER STATION

Load rejected Date Cause Duration (MW) Unit No. 1 - Commissioned on 7th September 1969

11. 9.69 30 DC control failure

15. 9.69 30 ID fan tripped 24i h

6.10.69 60 Station transformer tripped; flame detector lih source failed; total flame out resulted

13.10. 69 30 Automatic voltage regulator 11 kV PT fuse failure 28 h

7.11.69 48 Indiscriminate operation of fuel oil shut-off 12 h protection 11.11.69 60 Indiscriminate operation of fuel oil shut-off 78 h protection

20.11. 69 25 On-load airheater washing; leakage of water into 8h wind box and steam blanketing of flame eyes

20.11.69 30 Loss of atomizing steam etc. 18 h

9.12.69 30 Generator earth fault 43 h 24.1. 70 Cleaning of cooling water intake culvert 48 h 29. 1.70 48 Flame detector AC source failure 11 h 14.10.71 50 DC supply to shut-off valve failure 3 h 27.10.71 48 Loss of unit auxiliary transformer 4 h

Unit No. 2 - Commissioned on 23rd December 1969

23. 1.70 40 Economiser tube leak 43 d 15. 5.70 Commutator arcing and feed pump pipework 9 d adjustments

14. 6.70 45 Low bearing oil pressure 8 h 26. 6.70 48 Indiscriminate operation of fuel oil shut-off valve 11 h 2. 7.70 Deterioration of 415 V switchgear insulation 4 d resistance 23. 9.70 Condenser tube leak 12 h

24. 10.70 50 Operational error 4 h 15. 11.70 18 Chain flame failures 2 h

Unit No. 3 - Commissioned on 11th December 1970

12.12.70 15 Malfunction of automatic voltage regulator 20 h 15.12.70 30 Economiser tube leak 6 d 22. 2.71 50 Failure of governor's control switch 2 h 21. 4. 71a 60 Faulty thrust bearing wear transmitter 5 d

Unit No. 4 - Commissioned on 3rd March 1971

6. 3.71 10 Saturated steam tube leak 8 d Economiser tube leak during hydraulic test

25. 4.71 Controlled Economiser tube leak 21 d 21. 6.71 50 Loss of control air on feedwater control 20 h element 12. 11.71 6 Chain flame failures 1 h

Load shedding carried out.

- 42 - TABLE IV-5. RECORD OF GENERATOR OUTAGES EXCLUDING JURONG POWER STATION

Time Date Generator units Duration Remarks (h)

19.3.69 02.04 'B' unit No. 3 2 h 34 min Force shutdown due to generator. Transformer fans' closing-coil burnt.

23.4.69 09.36 'A' TG4 Force shutdown. One gasket of TG4. main exutor loosened.

10.6.69 11.57 *B* unitNo.4 Force shutdown. Feed pump E bearing temp.high.

13.7.69 14.26 *A' TG4 49 min 22 kV bus-section B monobias protection operated.

23.12.69 05.56 'B' unit No.3 41 min Complete burner failure (load 30 MW).

18.5.70 06.23 'B' unit No.4 Tripped off due to faulty air heater damper (shut by itself).

14.8.70a 08.25 'B' unit No. 3 Unit No. 3 tripped by vacuum un- 1 h 50 min 'B' unit No.4 loading device. Unit No.4 tripped reverse power relay. System load shed 90 MW.

23.11.70 05.39 •B' unit No. 2 66 kV OCB oil leak. Unit tripped manually.

5.10.71a 19.28 •B1 unit No. 3 Tripped on total flame failure. •B1 unit No.4 3h 25 min System load shed 128 MW.

25.10.71 00.07 'B' unit No. 2 37 min Tripped off on complete flame failure due to dirty fuel oil strainer.

22.4.72 12.52h 'B' unit No.4 Force shutdown due to OP 2C control mechanism faulty.

29.9.72 15.27 SJ O/D No. 2 Tripped at 3 MW. Cause unknown.

30.9.72 19.45 'B' unit No. 1 48 min Tripped on total flame failure. Air vane of F/D fan A jammed.

a Resulted in system load shedding.

Figure 4-8 shows the major overhaul program for generating units for October 1971 through March 1974.

- 43 - TABLE IV-6. TRANSMISSION AND DISTRIBUTION OUTAGE RECORDS

Average Others Maximum duration Cable System cable fault System equipment fault Faults duration of (e.g. trees falling Total number of interruption Year and month damaged by in consumers' interruption Auxiliary on O/H lines of faults of supply LV HV Switch gear T/former installation of supply construction equipment etc.) (hr) work (hr)

1971

Jan. 4 2 _ 1 3 3 13 3.1 29.4 Feb. 2 4 2 - 3 1 - - 12 2.5 4.8 Mar. 1 1 4 1 1 2 - 3 13 2.1 4.0 Apr. 5 3 5 3 3 - 1 4 24 2.1 4.5 May 5 2 7 4 2 - 1 - 21 2.2 9.5 Jun. 3 1 2 - 1 - - - 7 5.4 19.0 Jul. 1 4 3 1 1 1 1 4 16 4.4 9.5 Aug. - 3 5 - - 1 1 1 11 4.7 2.2 Sep. 2 4 3 3 1 1 2 4 20 1.5 8.5 Oct. 4 3 3 1 - 1 - la 13 3.5 9.0 Nov. 1 2 2 - - 1 - 1 7 2.2 4.5 Dec. 1 5 1 - 1 1 - 3 12 3.1 5.5

Total 29 32 39 13 13 10 9 24 169 3.1 29.4

1972

Jan. 4 5 2 2 6 3 2 24 3.1 13.5 Feb. 3 3 5 1 - - 4 1 17 2.1 6.0 Mar. - 6 4 - - 1 1 1 13 3.8 9.6 Apr. 2 3 1 - - 3 1 1 11 2.3 4.0 May 3 1 1 1 - 1 2 4 13 2.3 4.2 Jun. 6 2 3 1 1 2 2 3 20 3.1 8.6 Jul. 3 1 3 - - 1 1 6 15 2.4 8.5 Aug. 1 1 2 2 1 3 2 1 13 3.5 7.2 Sep. Oct. Nov. Dec.

This interruption of supply was occasioned by the forced outage of two sets at Pasir Panjang 'B' due to water in the fuel oil. TABLE IV-7. FAULTS ON 66 kV NETWORK DURING THE YEARS 1970, 1971 AND 1972

Date Description Remarks

1970

1 Jan. Siglap-Changi feeder cable. Suspected vandalism. Oil leak in punctured copper pipe within No tripping. asbestos conduit at joint hole 20.

2 Jan. Siglap-Toa Payoh feeder. Feeder tripped on main and back up Cable exploded at the edge of the cable bridge. protection.

3 Aug. Newton to Toa Payoh feeder. No tripping. Cable oil leak along Braddell Road. Leakage from lead sheath to PVC cover. Caused by sinking of earth after extensive excavation work by Sewerage Department.

4 Aug. /Sept. PPPS to Toa Payoh feeder. As in item (3). Oil leakage along Braddell Road.

5 Dec. JPS-Pulau Ayer Chawan via Pulau Merlimau. No tripping. Blue phase and pilot cables severed by ship Cable out of service. along shipping channel between Danar Laut and Pulau Merlimau.

1971

1 Jan. Toa Payoh-Aljunies Road along Braddell Road. No tripping. Oil leakage due to tiny cracks on lead sheath. Oil leakage caused by sinking of earth as a result of extensive excavation work by Sewerage Department and movement of heavy traffic along the road.

2 Feb. Trafalgar-Kallang feeder. No tripping. Cable oil leakage. Oil leakage caused by considerable damage to the oil pressure tank housing along Shenton Way. The OG Box was bulldozed and oil feed system ripped apart.

3 Mar. Aljunied Substation 31. 25 MVA transformer. Malfunction of the temperature gauge on the Tripped. transformer.

4 May Earth fault on 6c pilot cable. No tripping of main power cable. Damage by excavation work. Unable to trace department responsible.

5 July JPS-Jurong Harbour No. 1 feeder. No tripping. Damaged. 20c pilot cable damaged. Damage was caused by the Jurong Power Cable oil leak. Station Contractors.

6 Aug. Jurong Harbour 31.25 MVA transformer. Tripping caused by an explosion of the 22 kV Tripped, on instantaneous element O/C. dividing box of the Chemical Industries to Bridgestone feeder.

7 Aug. Changi 31. 25 MVA transformer. The operation of the Buchholz relay was Tripped on Buchholz relay. caused by the loss of transformer oil. One of the transformer drain plugs was removed by person or persons unknown. Suspected vandalism.

- 45 - TABLE IV-7. (cont.)

Date Description Remarks

i Sept. Crawford 30 MVA No. 1 Transformer. Tripping was caused by a defective tempera- Tripped on oil temperature. ture gauge which had failed to operate the cooling fans when the ON rating of the transformer was exceeded.

9 Sept. Outram Substation 31. 25 MVA No. 1 transformer. Tripping was caused by the operation of the Tripped on Buchholz gas. Buchholz relay as a result of excessive loss of transformer oil. Deliberate removal of drain plug by vandals.

10 Sept. JPS-Esso No. 2 feeder. Tripping caused by damage to red phase cable, Tripped on main protection. by ship or ships unknown.

11 Dec. Jurong Industrial T/F No. 2. Tripping caused by malfunction of the Tripped. temperature gauge.

1972

1 Jan. JPS-Esso No. 2. Operation of the relay caused by vibrations Tripped on main protectioi or maloperation of the tripping relay.

2 Mar. JPS-Jurong Industrial. Faulty tripping relay at JPS end. Tripped on DPDL.

3 Mar. JPS-PPPS I/C 4. No tripping. Cable oil leak. Caused by the cracking of lead sheath along 9 m. s. and 9

4 Apr. Toa Payoh to Newton feeder. No tripping. Oil leak at Stop jt. 4. Leakage caused by a crack in the wiping of the stop joint.

5 Apr. Pulau Ayer Chawan-Pulau Bukom No. 2. Feeder tripped. Oil leakage on blue phase. Damage caused by Shell contractors who Tripping on yellow phase. were doing piling/dredging work off Bukom Kechil.

6 Sept. JPS-Esso No. 2. No tripping. Cable oil leak. Oil leakage was traced to be at the oil feed system in the JPS basement.

- 46 - MAJOR OVERHAUL PROGRAMME - 1972 & 1973

1971 19 72 19 73 1974 OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEPT OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEPT OCT NOV DEC JAN FEB MAR 25 MW TGI ai aa-a IO N 25 MW TG2 Bi BB-i B BBI • i BBTJI-a E 25 MW TG3 25 MW TG4 Bl B-fB I BfB Bl -*-a-ji IB i 25 MW TG5 25 MW TG6 BB* BB-J Bl ASI R 25 MW TG7 -*a Bl Baa a

BB B BB-JBi a Bam s 60 MW TG1 22/12, 60 MW TG2 B BB 15/8 -*-*B B-l 18/11 if 60 MW TG3 1Bi BB-J Bi 1 2? 60 MW TG4 -•a -a 11/5 Baa B-J a*ar-j 6/6 ra-B -a NOli V 60MW TGI 60 MW TG2 Bl BBI B BB-J V) 11/11 BB -J-J-J-JI-m-j 50 MW TG3 — • I 60 MW TG4 BBI I 4/2 1BB-i Bl B-a l BB-J Bi o 120 MW TG5 I C 120 MW TG6 OC URO t ~* 120 MW TG7 >c

1. INSTALLED CAPACITY* 655 655 655 655 655 655 655 655 655 655 655 655 655 655 655 7 75 775 775 775 695 895 895 695 895 1015 1015 1015 1015 1015 1015

2.OVERHAUL {MAX] 85 110 85 85 85 85 85 60 60 60 60 60 25 25 25 60 60 60 25 85 145 120 120 85 145 145 145 145 120 120 3.AVAILABLE CAPACITY* 570 570 570 570 570 570 570 595 595 595 595 595 630 630 630 715 715 715 750 810 750 775 775 810 670 870 570 870 89S 395 4.PREDICTED LOAD 444 451 458 466 4 75 483 492 500 509 517 526 534 543 551 560 571 582 593 604 615 626 637 648 659 670 681 692 700 709 717

5.SPINNING RESERVE 60 60 60 60 60 60 60 60 60 60 60 60 60 60 60 120 120 120 120 120 120 120 120 120 120 120 120 120 120 120 6. TOTAL 141 .[5 504 511 518 526 535 543 552 560 569 577 586 594 603 611 620 691 702 713 724 735 746 757 768 779 790 801 812 620 829 837 7.RESERVE13) -(6) 66 34 52 44 35 27 18 35 26 18 9 1 27 19 10 24 13 2 26 75 4 18 7 31 80 69 58 50 66 58

NOTES: HI'NOT INCLUDING ST. JAMES POWER STATION (2) DATES ARE EXPIRY DATES OF BOILER REGISTRATION I3I ©C COMMISSIONING OF JURONG 120 MW UNITS

FIG. 4-8. MAJOR OVERHAUL PROGRAM FOR 1972 AND 1973.

- 47 - 5. PROJECTED SYSTEM DATA

Table V-l gives the January 1973 forecast of the PUB Electricity Department, for 1972-1981, of installed capacity (mean for year, and year end) in total, and for each station, firm capacity, maximum demand, annual increase in maximum demand, load factors (system, commercial, and plant), units generated (total and by station), station losses (by station and by station average), distribution and transmission losses, heat rates (by station and system average), fuel prices (by station), fuel consumption (by station and total), and unit fuel costs. The fuel price forecast in Table V-l has been superseded by the one of September 1972 which is presented in Table II-5.

5.1. Projected energy requirements

Figure 5-1 shows gross energy generation, energy sold and annual growth of generation to 1981, based on Table V-l. Figures 5-2 and 5-3 show the projected breakdown of sales by class of consumer, for the existing and the proposed new tariff groups. Table V-2 shows the projected per-capita electrical energy consumption based on the forecast of Table V-l and on the assumption that population will continue to increase at the current rate of 1.7%/yr (see Section 1.2).

5.2. Peak demand and load factor

Figure 5-4 shows peak demand and peak demand growth rate forecasts consistent with Table V-l. The corresponding load factors vary between 66.7% and 68.6%.

5.3. Plant installation program

The installed capacity at the end of 1972 is 704 MW, made up as shown previously in Table III-3. Figure 5-4 shows the plant installation program consistent with the forecast of Table V-l. Table V-3 gives the details of the committed and tentatively planned plant installation and retirement program consistent with Fig.5-4.

5.4. Committed and planned transmission line program

A study of requirements of the transmission and distribution system up to 19.80 has been carried out by the Board' s consultants, Montreal Engineering Co., Ltd. It is intended that expansion of the transmission and distribution networks will follow the recommended programs outlined in the consultants' report suitably modified where necessary to accommodate deviations from load conditions predicted in the report. At present, the primary transmission voltages are 66 kV with 22 kV as the sub- transmission voltage. The primary distribution voltage is 6.6 kV. The consultants' report concludes that by the mid-1970s the concluded system will have grown to such an extent that it will no longer be economical or technically desirable to carry out transfer of bulk power to the major load distribution centres at 66 kV. It recommends the introduction of 230 kV to replace 66 kV as the primary transmission voltage with a network of 230 kV substations to serve as the major load distribution centres. The 230 kV network will be designed to provide strong ties between the generating stations and to facilitate the tie-in of new generating stations to the primary transmission network. The existing 66 kV system will be developed into several sub-transmission systems to serve the eastern, central city, western and northern areas. A contract for the first phase development of the new 230 kV transmission system has already been awarded. The first 230 kV substation is scheduled for commissioning in mid-1974. This planned development of the 230 kV system is shown in Figs 3-3 and 3-5 (Section 3).

- 48 - TABLE V-l. ANNUAL GENERATION STATISTICS FORECAST FOR 1972 - 1981 (made in January 1973)

1 Installed capacity (MW) Total - December 727 1109 1202 1442 1442 1942 1942 2192 2692 2 - Mean 704 1109 1162 1382 1442 1692 1942 2 067 2567 3 St James - December 49 — 22 4.5 MW F. P. sei retire 4 - Mean 49 — 22 in October 1975 5 Pasir Panjang ' A" - December 175 2 MW House set excluded 6 - Mean 175 7 Pasir Panjang * B" - December 240 • 8 - Mean 240 - 9 Jurong Stage I - December 240 10 - Mean 240 11 Jurong Stage II - December 360 360 12 - Mean 150 360 13 Senoko/Gas turbine - December 22.5 45 45 14 - Mean 0 43 • 45 15 Senoko Stage I - December 360 360 16 - Mean 300 • 360 • 17 Senoko Stage II - December 500 500 500 500 18 - Mean 250 500 500 500 19 Senoko Stage in - December 250 750 20 - Mean 125 625 21 Maintenance & spinning reserve (MW) - December •» 265 265 300 325 455 500 500 585 22 Firm capacity (MW) - December 582 964 844 937 1142 1117 1487 1442 1692 2107 23 Maximum demand (MW) 521 623 777 834 940 1059 1194 1345 1516 1708 24 Annual increase in MD (<&) 24.9 18.2 15.1 12.7 12.7 12.7 12.7 12.6 12.7 12.7 25 System load factor 67.9 66.7 68.6 68.3 68.6 67.8 68.1 68.3 68.3 68.5 26 Units generated (GWh) - Total 3143.6 3642.3 4308.2 4989.8 5664.6 6288.8 7123.2 8 046.6 9090.2 10249.0 1 kwh =3412Btu 27 St James 3.2 3.0 10.0 20.0 1.0 1.0 1.0 1.0 1.0 1.0 28 Pasir Panjang ' A' 617.4 500.0 316.2 249.5 197.6 197.1 184.0 184.0 158.1 131.4 29 Pasir Panjang 'B' 1234.3 1210.5 1082.2 1136.8 1021.5 967.1 900.0 645.4 816.8 262.0 872.0 326.1 30 Jurong Stage I 1288.7 1263.7 1135.3 1198.4 1069.4 1022.0 951.0 707.6 2055.5 1892.2 31 Jurong Stage II 591.3 1734.5 2 049.8 2 055.5 2 049.8 2 049.8 2 049.8 2.0 2.0 32 Senoko/Gas turbine 73.8 30.0 98.6 2.0 2.0 2.0 2.0 33 Senoko Stage I 236.5 1317.6 2 049.8 2 049.8 2049.8 2 055.5 2 049.8 34 Senoko Stage II 985.5 2409.0 2635.2 2847.0 35 Senoko Stage in 494.1 2737.5 36 Work units (<&) St James 35.82 40.00 16.00 6.00 50.00 - 50.00 37 Pasir Panjang *A' 5.64 6.50 7.80 8.30 8.70 8.90 9.20 9.40 38 Pasir Panjang *B' 5.38 5.40 5.90 5.70 6.10 6.30 7.70 6.90 9.80 39 Jurong Stage I 5.56 5.50 5.90 5.70 6.20 6.40 6.60 7.70 7.00 9.80 40 Jurong Stage II 7.00 6.30 5.55 5.50 5.55 5.60 5.65 5.65 5.80 41 Senoko/Gas turbine 5.50 6.50 4.50 50.00 - 50.00 42 Senoko Stage I 7.30 6.70 5.55 5.50 5.55 5.60 5. 65 43 Senoko Stage II 7.70 6.80 6.40 6.10 44 Senoko Stage III 7.70 7.20 45 System 5.53 5.88 6.23 5.82 6.15 5.93 6.23 6.29 6.52 46 Units sent out (GWh) 2 969.7 3428.1 4 039.8 4 699.2 5316.2 5915.9 6679.4 8518.4 9580.8 47 Free supplies (%) generated 48 Distribution & transmission losses °fo of units generated 5.71 7.52 7.42 7.33 . 6.38 6.38 6.33 6.3S 49 <£of units sent out 6.07 8.02 7.88 7.81 6.79 6.80 6.82 6.75 6.77 50 Units sold 2762.2 3220.2 3 715.8 4329 4 901 5514 6225 7 020 7 943 8932 51 Commercial load factor 59.7 59.0 59.2 59.3 59.4 59.4 59.5 59.6 59.6 59.7 52 Plant load factor - Total 52.7 47.7 45.4 50.1 47.5 50.7 48.8 47.9 50.7 46.0 53 St James 0.7 0.7 2.3 5.4 0.5 0.5 11 MW & 4.5 MW units. 54 Pasir Panjang *A' 46.9 38.1 24.1 19.0 15.0 15.0 14.0 14.0 12.0 10.0 25 MW units 55 Pasir Panjang "B' 58.5 57.6 51.5 54.1 48.5 46.0 30.7 38.7 12.5 60 MW units 56 Jurong Stage I 61.1 60.1 54.0 57.0 50.7 48.3 45.2 33.7 41.4 15.5 60 MW units; 57 Jurong Stage II 45.0 55.0 65.0 65.0 65.0 65.0 65.0 65.0 60.0 120 MW units, 58 Senoko/Gas turbine 19.6 7.6 25.0 0.5 - 0.5 22.5 MW units 59 Senoko Stage I 45.0 50.0 65.0 65.0 65.0 65.0 120 MW units. 60 Senoko Stage II 45.0 55.0 60.0 250 MW units 61 Senoko Stage III 45.0 250 MW units. TABLE V-l (cont.)

1981

62 Heat rate (per kWh) St James - Generated 16852 16980 16250 15160 19300 CV= 19300 Btu/lb 63 - Sent out 26250 28300 19350 16130 38600 64 Pasir Pan Jang 'A' - Generated 12953 13280 13280 = 25.7 % CV = 18550 " 65 - Sent out 13 727 14200 14 550 14 560 14580 14580 14 630 14 660 66 Pasir Panjang 'B' - Generated 11251 10830 10830 = 31.5* CV = 18550 " 67 - Sent out 11890 11450 11530 11560 11600 11730 11630 12 010 68 Jurong Stage I - Generated 10714 10 830 10830 = 31.5 * CV = 18 550 " 69 - Sent out 11344 11480 11510 11480 11550 11570 11600 11730 11650 12 010 70 Jurong Stage II - Generated 10010 -i 10 010 = 34.1 % CV = 18550 " 71 - Sent out 10760 10680 10 600 10590 10 600 10600 10610 10610 10630 72 Senoko/Gas turbine - Generated 13650 13930 13120 19300 CV »19300 " 73 - Sent out 14440 14900 13 740 38600 74 Senoko Stage I - Generated 10010 = 34.1 Is CV = 18550 " 75 - Sent out 10 800 10730 10 600 10590 10600 10600 10610 76 Senoko Stage II - Generated 9610 = 35.5 It CV = 18550 - 77 - Sent out 10310 10270 10230 78 Senoko Stage III - Generated 1610 = 35.5 % CV = 18550 " 79 - Sent out 10410 10360 80 Total system (weighted) -Generated 10 085 9384 81 Fuel oil prices (S $/t) St James • 87.60 Industrial diesel 82 Gas turbine • 87.60 Industrial diesel 83 Pasir Panjang 40.13 41.64 44.36 45.77 47.18 48.59 50.00 50.00 50.00 Bunker C 84 Jurong 39.17 40.58 43.40 44.81 46.22 47.63 49.04 50.00 50.00 Bunker C 85 Senoko 44.40 45.81 47.22 48.63 50.00 50.00 50.00 Bunker C 86 Fuel consumption St James - t 1230 1179 3759 7 013 446 1. 000 lb/kwh Ln 87 - 103 S $ 108 103 329 614 39 • 39 88 Senoko/Gas turbine - t 0 23303 9667 29 923 893 • 893 1.000 " 89 - 10s S $ 0 2041 847 2621 78 - 78 90 Total industrial diesel - t 1230 24482 13426 36 936 1339 -1339 91 - 103 S $ 108 2144 1176 3 235 117 - 117 92 Pasir Panjang 'A* - t 191837 159799 101057 79740 63153 62 993 58 806 50528 41995 0.7159 93 - 103 S $ 7 698 6638 4340 3 537 2891 2 972 2 857 2940 2 526 2100 94 Pasir Panjang 'B' - t 333 097 315487 282048 296 279 252050 234563 168207 212879 68284 0.5838 95 - 103 S $ 13 367 13105 12114 13143 12185 11892 11397 8410 10 644 3414 96 Total Pasir Panjang - t 524934 475286 383105 376 019 329381 315043 293369 227013 263407 110279 97 - 10s S $ 21065 19743 16454 16 680 15 076 14864 14 254 11350 13170 5514 98 Jurong Stage I - t 329547 329352 295 888 312 333 278712 266359 247880 184418 227265 84990 0.5838 99 - 103 S $ 12 908 13365 12424 13 555 12489 12311 11807 9044 11363 4250 100 Jurong Stage n - t _ 142440 417289 493 782 495155 493782 493782 495155 455817 0.5396 101 - 10s s $ - 5780 17 522 21430 22188 22823 23519 24215 24758 22 791 102 Total Jurong - t 329547 471792 713177 806 115 773867 760141 741662 678200 722420 540 807 103 - 103 S $ 12 908 19145 29946 34 985 34677 35134 35326 33259 36121 27 041 104 Senoko Stage I - t 56 971 304754 493 782 493782 493782 495155 493782 0.5396 105 - 103 S $ 2530 13 961 23316 24013 24689 24758 24689 106 Senoko Stage II - t 227 941 557 189 609508 658496 0.5181 107 - 10s S $ 11085 27 859 30475 32 925 108 Senoko Stage III - t 114283 633169 0.5181 109 - 10s S $ 5714 31658 110 Total Senoko - t 56 971 304 754 493 782 721723 1050 971 1218946 1785447 111 -103S$ 2 530 13 961 23 316 35098 52548 60 947 89272 112 Total Bunker C - t 854481 1239105 1408002 1568 966 1756754 1 956184 2204773 2436533 113 - 103 S $ 33 973 46400 54195 63 714 73314 84678 97157 110238 121827 114 Total fuel - 103 S $ 34081 47 576 57430 63 831 73431 84795 97274 110355 121944 115 Fuel cost(S mill/kWh) On units generated 10. 84 11.27 11.04 11.51 11.27 11.68 11.90. 12.09 12.14 11.90 116 On units sent out 11. 48 11.97 11.78 12.22 12.01 12.41 12.70 12.92 12.95 12.73 117 On units sold 12. 34 12.74 12.80 13.27 13.02 13.32 13.62 13.86 13.89 13.65 20 000

10000 GROSS GENERATION ^ ^-—' J

^ *— o 5000 SALES z o UJ o < 30 g o a: z i— 2000 20 ANNUAL GROWTH RATE

1000 10 1971 72 73 74 75 76 77 78 79 80 81 82 FIG. 5-1. PROJECTED GROSS GENERATION AND SALES OF ENERGY.

100

LIGHTING AND FANS 90 _. — — . " .—- ' ^- 80

DOMESTIC POWER 70

O 50 SMALL COMMERCIAL* INDUSTRIAL POWER ^. u. o —.

/ / 30

LAf?GE COMME:RCIAL 20 / /

1966 1966 1970 1972 1974 1976 1976 I960 1982

FIG. 5-2. HISTORICAL AND FORECAST BREAKDOWN OF ENERGY SALES BY EXISTING TARIFF GROUPS.

- 51 - STREET LIGHTING AND ENTERTAINMENT

90 HIGH TENSION INDUSTRIAL

HIGH TENSION NON-INDUSTRIAL

— . —•—,

• • . 40 •

LOW TENSION NON-DOMESTIC

20 •i . 1 . 1—-^-

LOW TECJSION DOME.STIC 10

1973 1975 1977 1977 19B1

FIG. 5-3. FORECAST BREAKDOWN OF ENERGY SALES BY NEW TARIFF GROUPS.

TABLE V-2. PROJECTED ELECTRICAL ENERGY CONSUMPTION PER CAPITA

Year kWh/capita

1972 1468 1973 1670 1974 1942 1975 1976 2467 1977 2 692 1978 3000 1979 3 332 1980 3701 1981 4103

a Based on estimates of gross generation given in Table V-l and on a 1.7% annual growth rate of population starting at 2.110 x 106 in 1972.

- 52 - It z z O 55 o in oin o z o o m CM UJft a X X <— r- (M CO f) ^ In z< i 1 i g I I I CO 5NNN m CO V- LU ID ID ID LU LU ID UJ o © < totn to t/> ST , «r o o gooo o o o o o fflUJ Q CC O O Q o o •z.o o 2: 55 £ UJ ID UJ LU ID IzD UJ 0)> coo: to u> 5000 ^- 4000 —- INSTALLED CAPACITY 30 = 3000 o \ I 20 S 2000 "***•—^ H o * 1 12.7 12% g

1000 10 ^-" ANNUAL GROWTH OF PEAK LOAD 700 ^^ K PEAK LOAD a ^^ < 500 o 400

1972 73 75 76 77 78 79 80 81 82 83 85 86 87 88 89 90

FIG. 5-4. PLANNED INSTALLED CAPACITY AND FORECAST PEAK LOAD 1972 - 1981.

5.5. Planned future reserve capacity

The criteria that govern the provision of reserve capacity up to date apply equally for the planned future reserve capacity, i.e. minimum generating plant margin is to be provided to cover for outage of the largest unit in the system, as well as scheduled outage for maintenance purposes. Table V-3 indicates total future year-end reserves. The planned reserve capacity for the transmission system as mentioned previously is outlined in a report of Montreal Engineering Co., Ltd.

5.6. Siting data

It is planned to develop the Senoko site to 1360 MW by about 1981 and it may prove capable of further development up to 2000 MW. The site for the next station is not yet clear. If nuclear, it is most likely to be on Pulau Tekong island, but a detailed site survey has yet to be made. (See Sections 1.5 and 6.4 and also Section 8.2 in respect of pollution problems.)

- 53 - TABLE V-3. PLANT INSTALLATION PROGRAM

Unit System Peak Total reserve Year Plant capacity capacity demand (MW) (MW) (MW) MW of peak)

Committed 1972 704 527 177 33.6 1973 Senoko GT Nos 1, 2 45 Jurong Stage II No.l 120 Jurong Stage n No.2 120 Jurong Stage II No.3 120

1974 1109 717 392 54.6 1975 Senoko Stage I No.l 120 Retire free piston units (27) 1202 834 368 44,1 1976 Senoko Stage I No. 2 120 120 1442 940 502 53.4

Tentative 1977 1442 1059 383 36.2 1978 Senoko Stage II No. 1 250 Senoko Stage II No.2 250 1942 1194 748 62.6 1979 1942 1345 597 44.4 1980 Senoko Stage III No. 3 250 2192 1516 676 44.6

1981 Senoko Stage III No.4 250 Senoko Stage III No. 5 250 2 692 1706 986 57.8 1982 2 692 1911 781 40.9 1983 Possible nuclear 500 Retire Pasit Panjang ' A' (175) 3 017 2140 877 41.0

- 54 - 6. NATIONAL CAPABILITIES AND LOCAL COSTS

6.1. Contribution of local industry to past or present power projects

Table III-5 and III-6 show the degree of local participation in power projects currently being constructed. The contribution has been high in the supply of material and labour particularly for civil work. Little contribution is made to the supply of sophisticated mechanical and electrical equipment.

6.2. Targets for future local industry participation

There is no fixed target for the participation of local industry in future power projects as such, but it is expected that the experience acquired in the erection of projects in the past will enable it to play an increasing role in the future. In all tender documents issued by the Public Utilities Board for the supply of plant and equipment both for power stations and transmission and distribution projects, the following two clauses aimed at encouraging local participation are included:

Use of local resources

Whenever possible, the tenderer shall include the use of local resources and sub- contractors where these are available and adequate.

Goods produced in Singapore

In the case of goods produced in Singapore, the Public Utilities Board may, after taking into account such factors as quality, delivery date, performance, guarantee and service facilities, award a contract to the lowest qualified Singapore tenderer provided that his offer at the project site does not exceed the offer at the project site, excluding customs and other duties on imports, of the lowest qualified foreign tender by more than 15% of the c.i.f. value of the foreign tenderer.

It is of interest to note that recent tenders received for major plant and equipment for the new Senoko power station include a number of proposals by suppliers to enter into joint venture with local engineering companies, principally the shipyards, to fabricate portions of the supplies in Singapore. These portions relate principally to such items as lowpressure pipework, structural steels, boiler cladding and ducting etc. In addition to civil engineering, piling and building contractors, Singapore has a wealth of civil, structural, mechanical and electrical engineering firms which are potential participants in future power projects. Some of the contractors and engineering firms are the Singapore or South-East Asia branches of international firms.

6. 3. Local construction costs and practices

Civil construction works measurement is usually based on the standard measurement for building works agreed between the Institution of Surveyors (Singapore) and the Contrac- tors Association (Singapore). Use is also made of the Institution of Civil Engineers1', "Standard Method of Measurement of Civil Engineering Quantities". The wages for construction workers are fairly standardized. Examples are given in Table VI-1. Fringe benefits, including social welfare costs, amount to about 20% of construction labour wages. (The corresponding figure for industrial workers is about 50%.) Table VI-2 shows the cost of a variety of construction materials inSingapore. Table VI-3 shows the rental cost of construction equipment. Other construction job costs are given in Table VI-4.

- 55 - TABLE VI-1. CONSTRUCTION LABOUR WAGE RATES (late 1972)

Item Rate (s$/h) Labourer or mate (any trade) 1.20 Unskilled labourer (male) 1.00 Unskilled labourer (female) 0.80 Concretor 1.40 Steel bender and fixer 1.50 Carpenter 1.60 Bricklayer 1.50 Joiner 1.50 Pavior 1.50 Plasterer 1.50 Filer 1.80 Steelworker 1.50 Plumber 1.50 Glazier 1.40 Painter 1.50 Welder 2.50 Fitter 2.00 Winchman or pile driving rig operator 2.00 General plant operator 2.50 Lorry driver 2.00 Crane driver 2.50 Watchman 1.00

Jurong Town Corporation building construction costs were estimated to be approximately as follows: Low cost flats, reinforced concrete structures up to 20 storeys, 8'6" ceilings S $12/ft2

High cost flats S $20/ft2

Standard flatted factories 7 storeys, 12.5 ft/storey S $15/ft2

Type C-3 factory, steel frame, concrete floors, brick walls S $11/ft2

6.4. Problems and costs associated with possible power plant sites

The problems of power plant siting in Singapore are primarily related to the small total land area and the high population density. The Senoko site is suitable for up to 2000 MW(e) of oil-fired power plant. Depth limitations in the Johore Strait do, however, require that 100 000-ton tankers be only partially laden when delivering fuel oil to the site. Although there is adequate cooling water for 2000 MW(e), the surface temperature of the water would reach 41 °C during periods of slack tide. The island of Pulau Tekong has received some preliminary consideration as a possible site for a future nuclear power station. The IAEA sent a three-man siting mission to Singapore in March 1972 and this was followed with a visit by a meteorological expert in September 1972. The locality is relatively free from earthquakes, floods, hurricanes, faulting etc. The dominant factor concerning the selection of a suitable site, however, is the population distribution. In comparison with US and European plants, the population density a few miles from the site is higher by an order of magnitude. Further studies must be made to ascertain the effects of additional engineered safeguards, of reducing power rating, and of additional quality assurance programs on both the probability and severity of the accident. In addition a field measurement program on micro-meteorological diffusion conditions and ocean currents must be undertaken to facilitate the determination of possible dose distribution in both space and time as a result of accidents and of normal operations.

- 56 - TABLE VI-2. CONSTRUCTION MATERIAL COSTS (late 1972)

Description of materials Cost (S$)

Sand For concrete works 7.00/yd3 For roadwork and filling 6.50/yd3 Coarse aggregate 1{" down to I" 16.00/yd3 Under I" down to J" 17. 50/yd3 Under i" down to 1/16" 19.00/yd3 Cement Ordinary Portland 3.00/cwt Rapid Hardening ' 5.00/cwt Mild steel rod reinforcement Below i" diameter 410.00/t 5/8" to 14" diameter 400.00/t Above li" diameter 390.00/t High tensile deformed bar reinforcement Below i" diameter 400.00/t 5/8" to li" diameter 410.00/t Above li" diameter 420.00/t Granite blocks Over 6" and up to 9" size 14.00/yd3 Over 9" and up to 12" size 13.00/yd3 Laterite 3.50/yd3 Crusher run granite metal 14.00/yd3 Formwork for concrete works Sawn (rough) 4.50/yd2 Wrot (smooth) 6.30/yd2 Hardwood of any cross-sectional area 7.50/ft3 Bricks Common bricks (4|" X 9" x 3" high) 0.12 each Facing bricks (4|" x 9" x 3" high) 0.15 each Hollow concrete block 3" thick of various sizes 2.50/yd2 4i" thick of various sizes 3.75/yd2 9" thick of various sizes 7.50/yd2 Structural steel of various sections 650.00/t Ready-mix concrete 44.00/yd3

6.5. Plans for staffing of future power plants

Staffing of future conventional power stations will evolve from experience at the Jurong power station. No firm plans are now required for staffing of the first nuclear power station. This would be considered at the feasibility study stage. The construction period for a nuclear plant provides adequate time for a formal training program for operating staff; however, it would be desirable to have certain key professional staff trained far enough in advance to participate fully in the necessary decision making. PUB, the Ministry of Science and Technology, and the University of Singapore have already taken some steps in this direction. (See Section 1.5.)

- 57 - TABLE VI-3. RENTAL COST OF CONSTRUCTION EQUIPMENT (late 1972)

Hourly rate Description of plant (S$)

Mobile cranes 2-ton 8.00 3-ton 10.00 5-ton 12.00 Portable air compressors complete with air lines Over 200 and up to 300 ft3/min 7.00 Over 300 and up to 400 ftVmin 9.00 Over 400 and up to 500 ft3/min 11.00 Pneumatic tools Sump pump 3" 3.00 Rock drill or concrete breaker 2.00 Concrete vibrator li" diameter 1.00 Concrete vibrator i\" diameter 1.00 Petrol driven concrete vibrators 2|" diameter 1.20 Under 2|" diameter 1.20 Concrete mixers 14/10 ft3 3.50 7/10 ft3 3.00 Concrete weight-batcher 1.00 Concrete dumper 1.50 Road rollers Over 4 and up to 6 tons 5.00 Over 6 and up to 8 tons 6.00 Over 8 and up to 10 tons 7.00 Over 10 and up to 12 tons 8.00 Dumpers (petrol or diesel)

Up to 4.00 Over 2 and up to 4| yd3 5.00 Generating sets 3 kW 2.00 15 kW 4.00 30 kW 6.00 Welding and cutting sets Oxy-acetylene 1.50 Electric - 300 A 1.50 Mechanical rammer 1.00 Lorries Ordinary 3-ton 5.00 Tipper 3 -ton 5.00 Motor grader 12.00 Excavators (face shovels, skimmers, back-actors, draglines etc.) Up to and including I yd3 20.00 Over i and up to 1 yd3 22.00

- 58 - TABLE VI-4. CONSTRUCTION JOB COSTS (late 1972)

Excavation

First 5 ft Next 5 ft Next 5 ft Next 5 ft

Painting

(including labour and material, for plastered wall not exceeding 25 ft height)

No. of coats

Enamel paint 1 10 2 15 3 18

Emulsion paint 1 5

2 8 3 10

Scaffolding

(including labour and material) (onesquare = 10 ftxlO ft = 100 ft*)

Up to 10 squares S $6.20/sq. More than 10 squares S $5.50/sq.

- 59 - 7. ECONOMIC AND FINANCIAL ASPECTS

7.1. Economic ground rules for evaluation of power projects

The determining factor in the evaluation of a power project, whether conventional or nuclear, is the economic justification both in terms of capital expenditure and operating costs. The prime rate of interest in Singapore is currently 7.5%/yr. In the Senoko station feasibility study, PUB's consultants used the following assumptions:

Interest rate 8.0%/yr Capital charge rate, including interest, sinking fund depreciation, insurance, interim replacements 10.0%/yr Interest during construction 7.5%/yr Escalation rate -- labour 1.5%/yr materials 3.0%/yr Fuel oil prices -- 1972 S $41.50/t 1975 S $48.50/t 1980 S $58.50/t

7.2. Current methods and sources of financing

The past and current methods of financing power projects are:

(i) Loans from international lending organizations, e.g. IBRD, Asian Development Bank (ADB). (ii) Loans from Government (Singapore), (iii) Deferred payments from suppliers. PUB's experience has been that attractive suppliers' credits are readily available for large contracts, particularly those for power stations with a high percentage of foreign exchange component. The Board is, therefore, increasingly availing itself of this mode of financing for its power station projects. For instance, it is expected that the new Senoko power station will largely be financed through suppliers' credits. (iv) Own resources. (v) Loans from the public. This is done through issues of debenture stock to the public, but this practice has not been carried out by the Public Utilities Board since it took over from the former City Council.

Singapore is no longer eligible for "soft" loans from international financial institutions such as the International Development Association, since such loans are restricted to countries with per-capita incomes below US $300 and Singapore now has an income of about US$1200. PUB has loans from IBRD and ADB at interest rates ranging from 5|% to 7|%/yr. The interest rate on Singapore Government loans to PUB is about 7^%/yr. The Government recently floated a 10-year bond issue in the Federal Republic of Germany for 400 million DM face value, discounted to 98. 5% with a nominal interest rate of 7%/yr; which, considering the discount, is effectively less than 7|-%/yr. A 15-year "Asian Dollar" bond issue of US$20 million was floated at par with a 7f %/yr interest rate. Loans from IBRD and ADB to the PUB are repayable in from 12 to 19| years after a grace period for construction which ranges from 1 to 5 years. Suppliers' credits typically have interest rates ranging from 5|% to 7i%/yr; one recent example involved the repayment of the foreign portion in 12|- years at 5i%/yr and the local portion in 5 years at 5^-%/yr. The PUB Electricity Department generates a healthy internal cash flow which is available for expansion. IBRD depreciation guidelines (straight line) are followed and according to the IBRD definition the return on investment is in the range of 8 to 10%/yr. Tables VII-1 and VII-2 summarize the sources and amounts of capital expenditure financing and debt service requirements of the PUB Electricity Department, giving actual figures for 1967 - 1971 and forecasts for 1972 - 1976. Table VII-3 states the assets and

- 60 - TABLE VII-1. SOURCES OF CAPITAL EXPENDITURE FINANCING (103 S $)

Actual Forecast

1967 1968 1969 1970 1971 1972 1973 1974 1975 1976

Capital expenditure financing Capital expenditure 40 878 80 563 56 397 84 875 74104 136 665 222488 211921 193 691 279 690

Financed from: (1) Internal sources 27 154 30 386 37 685 26791 35496 64 308 79 888 93 024 102 757 115 994 (2) IBRD Loan I (44 178) 4203 ------IBRD Loan II (29 269) 4 647 2472 ------IBRD Loan III (47 594) 42 17 694 17 890 10 072 1896 - - - - - IBRD Loan IV (60 912) - - 455 34 539 15918 10 000 - - - - (3) ADB loan (43 621) - - - - - 18 667 24 926 28 - - Proposed ADB loan (55 272) ------26 925 28 347 - (4) UK Aid/Govt loan - - - - - 9 700 6100 - - - (5) Government loans 10 000 7 000 ------(6) Other sources of finance ------40000 a 70 000 51000 - (7) Suppliers' credit 9258 28715 6 352 22 561 8157 47 637 43444 23 460 11731 160000b (8) Other internal sources -14426 -5 704 -5 985 -9 088 12 637 - 13 647 28130 - 1516 -144 3 696

40 878 80 563 56 397 84 875 74104 136 665 222488 211921 193 691 279 690

a To be covered by loan or suppliers' credits for Senoko power station — Stage I. b For Senoko power station - Stage II. TABLE VII-2. DEBT SERVICE REQUIREMENTS PUB ELECTRICITY DEPARTMENT (103 s$) A ctual Forecast

1967 1968 1969 1970 1971 1972 1973 1974 1975 1976

Interest payable Debenture stock 6 982 6 982 6 982 6 982 6827 6 099 5214 4126 4126 3 166 IBRD Loan I 2211 2 130 2051 1997 1874 1782 1654 1460 1335 1204 IBRD Loan II 1428 1473 1466 1412 1352 1275 1229 1166 1099 1028 IBRD Loan III 24 271 1697 2 681 2 848 2 746 2 577 2 152 2 041 1924 IBRD Loan IV _ - 245 1458 3211 3 935 3 976 3 530 3 389 3 238 UK Aid - - - 87 1016 1127 1047 968 Suppliers' credit- secured 461 1088 1713 2 030 2 951 2 550 3196 5 881 5335 4 566 Proposed suppliers' credit (power station) ------5 200 ADB loan - - - - 30 567 2 370 3 272 3 228 3 049 Proposed ADB loan ------1178 3 259 4145 Government loans 1963-1967 3 266 3 511 3 247 2 982 2 717 2 474 2 296 2118 1940 1761 Government loans 1968 120 449 427 404 381 358 336 313 290 Other sources of finance — 1913 - - - - - 1500 2 964 2 812 2 664 1974 _ - - - _ - - 2 625 5187 4 921 1975 ------1913 3 779 14372 15 575 17 850 19 969 22214 21896 25386 31935 37 024 41903 to I Principal repayment IBRD Loan I 1813 1760 1708 2 049 1837 2 032 2230 2 242 2354 2495 IBRD Loan II - 812 827 964 1026 1001 1001 1069 1136 1201 IBRD Loan III - - 1546 1579 2 722 2484 1813 1922 2 048 IBRD Loan IV - - - - 959 2 002 2143 2284 2 459 UK Aid ------1054 1054 1054 Suppliers' credit - secured 1205 2 247 3 937 6 664 17 616 7 273 9 566 13 413 13145 13 277 Proposed suppliers' credit (power station! ------ADB loan ------2348 2 527 Proposed ADB loan ------Government loans 1963-1967 3 980 4480 4480 4480 4480 3 100 3100 3 100 3100 3 100 Government loans 1968 _ - 350 350 350 350 350 350 350 350 Other sources of finance — 1973 ------2 000 2 000 2 000 1974 ------3 500 3 500 1975 ------2 550

6 998 9 299 11302 16 053 26 888 17 437 20 733 27 184 33193 36 541

Principal repayment 6 998 9 299 11302 16 053 26 888 17 437 20 733 27184 33193 36 541 Sinking fund contribution 3 442 4 526 4304 4 303 4227 2 811 2406 1754 1754 1351 Interest 14372 15 575 17 850 19 969 22214 21896 25386 31935 37 024 41903

24 812 29 400 33456 40 325 53329 42144 48 525 60 873 71971 79 795 TABLE VII-3. STATEMENT OF ASSETS AND LIABILITIES OF THE PUB ELECTRICITY DEPARTMENT (103 S $)

1967 1968 1969 1970 1971

ASSETS 1. Cash/Departmental current account 8103 13 809 19 794 28 882 16245 2. Inventories — stores, capital stock etc. 14 632 18384 31580 50 633 59 659 3. Deferred charges 375 111 55 _ _

4. TOTAL CURRENT ASSETS AND DEFERRED CHARGES 23110 32 304 51429 79515 75 904

5. Gross fixed assets in operation 458 337 494447 599 893 631920 720 551 6. Less: Accumulated depreciation 142 184 162 258 182 807 203 974 227 012

7. Net fixed assets in operation 316153 332189 417 086 427 946 493 539

8. Works in progress 37 988 81459 17 194 68 802 53 626

9. TOTAL FIXED ASSETS 354141 413 648 434280 496 748 547 165

10. TOTAL ASSETS 377251 445 952 485 709 576 263 623 069

LIABILITIES 1. Refundable deposits 2 994 2 630 2144 1207 745

2. Debenture stocks 158 059 158 059 158 059 158059 141629 3. Less: Sinking Fund Investment and Cash 85 722 94323 103118 113 372 106 885

72 337 63 736 54941 44 687 34 744 4. IBRD loans 64 352 81945 98 939 138991 153 531 5. UK aid/Government loans - - - - - 6. Government loans 62102 64623 59793 54963 50133 7. Other sources of finance - - - - - 8. Suppliers' credit — Secured 8120 34 588 36787 52 684 45 587 — Power stations - - - - - 9. ADB loan - - - - - 10. Proposed ADB loan - -

11. NET LOANS OUTSTANDING 206 911 244 892 250460 291325 283 995

12. General/Development reserves 64789 84737 108 896 145 779 204 929 13. Sinking fund reserve 85722 94 323 103118 113 372 106 885 14. Capital contributions 16 835 19 370 21091 24 580 26 515

15. TOTAL EQUITY 167 346 198430 233 105 283 731 338 329

16. TOTAL LIABILITIES 377 251 445952 485 709 576 263 623 069

Average net fixed assets in operation 315 984 324171 374637 422 516 460 659 Return from operations 29 302 40 312 47 359 61350 "2 255 Rate of return on average net fixed assets in operation Oft) 9.30 12.44 12.64 14.52 15.69 Debt/Equity ratio 55/45 55/45 52/48 51/49 46/54

- 63 - liabilities of the Electricity Department, together with the return from operations and the debt/equity ratio for 1967-1971.

7.3. Foreign exchange and consideration in evaluating capital and fuel costs

There are no restrictions on foreign exchange in the purchase of capital equipment or fuel. The Singapore dollar is a strong convertible currency. As was mentioned in Section 6. 2, there is a local manufacturers preference of up to 15% of the c.i. f. value of the foreign tenderer, but this is not an important consideration in Electricity Department projects.

7.4. Import duties and taxes

All generation and transmission equipment and fuels which have no local manufacturers are exempt from import duties. The only item which may be subject to customs duties are low voltage cables of which there are several local manufacturers.

- 64 - 8. ADMINISTRATIVE AND REGULATORY

8.1. Atomic energy matters

There is no formal organization, such as an Atomic Energy Commission, set up specifically to deal with the administrative and regulatory aspects of nuclear energy. The Ministry of Science and Technology is the official body through which all dealings with the International Atomic Energy Agency are channelled. This Ministry has initiated the formation of two committees, the Nuclear Energy Committee and the Radioisotopes and Radiation Protection Committee, to advise the Minister on nuclear matters. Members of these committees are drawn from appropriate Government departments, institutions and statutory boards which are or are expected to be involved in the development of nuclear energy in Singapore. The Nuclear Energy Committee advises the Minister on all aspects concerning nuclear energy. It maintains close liaison with other governmental authorities, e.g. with PUB, on the generation of electricity by nuclear energy. The Radioisotopes and Radiation Protection Committee deals with all aspects of safety and radiation protection in the industrial use of radioisotopes. It is drafting legislation to cover the protection and safety of personnel in the handling of radioisotopes for industrial and medical purposes.

8.2. Interrelationship between Government organizations

The Public Utilities Board and the Anti-Pollution Unit are under the Prime Minister's office. The PUB has been described in Section 3. The Anti-Pollution Unit is concerned with all aspects of pollution control, including sulphur emission, smoke emission, and thermal pollution. Currently, as far as oil-fired power plants are concerned, the sulphur dioxide emission criterion is based on ground-level concentration and is equivalent to USA Federal standards. The ground-level concentration is controlled by making the power plant stacks high enough to give the required dilution and dispersion. The stacks at Jurong are 380 ft high and those at Senoko will be 600 ft. (The Anti-Pollution Unit wanted 800 ft, but the civil aviation authorities objected.) The prevailing winds are 3-10 miles/h and the isothermal layer extends up to about 5000 ft, which conditions are much more favourable than in Los Angeles where the isothermal layer extends up to only about 600 ft. The three main sources of sulphur pollution are the PUB power plants, the refineries, and automobiles. The refineries are required to recover sulphur whenever possible. Ground-level sulphur dioxide concentration is felt to be at acceptably low levels at present; however, it is estimated that it could increase to objectional levels sometime between 1975 and 1980, and restrictions on the sulphur content of PUB fuel could be imposed as one means of control. In addition to its atomic energy interests, mentioned in Section 8-1, the Ministry of Science and Technology is responsible for the National Museum, the Chemistry Department, the Metrication Board, the Science Council, the Science Centre and the Regional Marine Biological Centre. The Ministry has two units — the Economics and Statistics Unit and the Industrial Liaison Unit — to deal with matters relating to the training and utilization of scientific and technical manpower at professional level and to promote research and development on the scientific and technological aspects of industrial problems that are of special significance to the economic development of Singapore. The Ministry of Health deals with public medical and health services. The Ministry of Finance is made up of the Treasury and Economic Development Divisions. The Treasury Division is responsible for the formulation of the national budget, taxation, and the control and administration of Government finances. The Economic Develop- ment Division is responsible for the formulation and administration of general economic, industrial and trade policies. It deals among other things with matters related to development programming, external financial and technical assistance, regional economic cooperation, the compilation of statistics, the Economic Development Board, the Jurong Town Corporation, the Asian Development Bank and the International Bank for Reconstruction and Development.

- 65 - 8.3. Regulatory bodies and procedures for licensing and safety assessment

These will have to be developed before a nuclear power project can be implemented.

8.4. Nuclear legislation

In addition to the legislation now being drafted on protection and safety of personnel handling radioisotopes, mentioned in Section 8.1, it will be necessary to develop legislation establishing regulatory and licensing bodies and procedures and also legislation covering indemnification in case of a nuclear accident.

- 66 - 9. ANALYTICAL APPROACH

9.1. Approach and bases of analysis

The major objective of this study is to determine the size and timing of nuclear power plants that could, on economic grounds, justifiably be built in Singapore during the period 1980 to 1989, and to determine the sensitivity of the results to certain of the key parameters. The economic criterion, which is explained fully in Appendix D, is that the total operating and capital costs of the expansion plan for the generating system should be near to the minimum, when calculated in terms of present worth at 1. January 1973 and in terms of constant US dollars at that date. That is, normal price escalation is not treated explicitly. The implicit treatment of escalation is discussed in Appendix D. Any expansion plan must clearly be consistent with the forecast of load growth during the period of the study, and with other technical constraints of the system. Two forecasts of the growth in system demand have been used, one high and one low, and the method of deriving them is given in Section 10. A number of alternative expansion plans were then taken consistent with the appropriate forecast and the near-optimum plan determined by the use of a series of computer programs, the principal one being the Wien Automatic System Planning Package (WASP). This program evaluated the capital and operating costs of each alternative expansion plan over the period 1980 to 2000. The reason for extending the evaluation for a decade beyond the study period proper is to take account of at least ten years of operation of all plants introduced during the study period. The low- forecast case was evaluated using the dynamic programming optimization feature of WASP. Time limitations prevented the use of the dynamic program for the high forecast case. The analysis was based partly upon data obtained during the visit of the Market Survey mission to Singapore in November 1972 and partly on data developed to permit a consistent approach to the fourteen-country survey. A summary of the computer programs used in the analysis is given below together with a summary of the data required for the evaluation. These data and the results obtained in the analysis are discussed in more detail in the sections that follow.

9. 2. Description of computer programs

The basic tool used in the analysis of the alternative system expansion plans was the WASP program. Two subsidiary programs were used to provide specific data for the WASP program — the ORCOST program for calculating the capital costs of various fossil and nuclear units and the polynomial regression analysis program used to fit a polynomial equation to the load duration data.

(a) Wien Automatic System Planning Package (WASP)

The WASP program utilizes six blocks of input data as the basis for simulating the operation of the power stations on a seasonal (quarter-by-quarter) basis, evaluating the operating costs of each plant, present-worth discounting these operating costs and the capital costs associated with all additions beyond the start of the study and determining the total system costs to the year 2000. The data required for this analysis are as follows: (i) System load description — consisting of the year-by-year peak demands for the power system during the study period, quarterly load duration data expressed as the coefficients of a polynomial equation, and factors relating the quarterly peak loads to the annual peak loads. (ii) Fixed system description — consisting of a list of the generating units that will be in operation at the start of the study year (1977 for both the high and the low forecast cases), their maximum and minimum operating levels, their minimum load and incremental heat rates, 1 January 1973 fuel costs, expected scheduled and forced outage rates, and expected operating and maintenance costs. The description also includes data on the retirement of existing plants, and on specific firmly planned additions.

- 67 - (iii) Alternative generating units — consisting of technical data on the various sizes and types of generating units that may be considered for an alternative expansion plan during the study period. The data required are the same as those required for the fixed system.

(iv) A series of alternative expansion plans — each consisting of a year-by-year definition of the generating units to be added during the study period. For dynamic programming optimization the minimum and maximum number permitted is specified for each type and size of unit for each year and the code selects the most economical plan within these limits as defined by the objective function (see below).

(v) Loading order for both the plants in the fixed system and those considered as expansion alternatives.

(vi) Capital costs of the alternative generating units — broken down into foreign and domestic costs, and the expected economic life of the units.

The output from the WASP program consists of a quarter-by-quarter, plant-by-plant tabulation of the energy generation and associated costs for the study period. The total of these costs, plus the capital costs of the additions minus their salvage value at the study horizon, all present-worthed to 1973, is the "objective function" used to measure the economic merit of the system being analysed. That is, the expansion plan with the smallest value for the objective function was considered to be the "best" or "near-optimum". A detailed description of the data input to the WASP program is included in the following sections and the results of the analysis are described in Section 16. For further information on the WASP program, see Appendix A.

(b) Capital cost program

The capital cost data required by the WASP program were determined by utilizing the ORCOST computer program. This program, which was obtained from Oak Ridge National Laboratory of the US Atomic Energy Commission, had been prepared by them to provide estimates of power plant capital costs in the USA for PWR, BWR, HTGR, coal, oil and gas-fired plants. Provision had been made in the program to adjust equipment, material and labour costs from region to region. This made it possible to adjust the costs to conditions prevailing in Singapore by utilizing local labour, materials and equipment cost information. Section 13 describes how these cost data were developed. For a more detailed description of the ORCOST program, see Appendix B.

Load duration curves were obtained from the Singapore Public Utilities Board. The WASP program requires quarterly load duration curves expressed as the coefficients of a fifth order polynomial. The coefficients were calculated by a least-squares curve-fitting program that is described in more detail in Appendix C. The coefficients and the actual shapes of the quarterly load duration curves defined by the polynomial expressions are shown and discussed in Section 10.

9.3. Economic methodology and parameters

The economic merit of the various alternative expansion plans was determined and used as a basis for selecting the near-optimum case. External or social costs were disregarded, as were taxes and restraints on foreign capital. Definitions of the costs and other economic parameters are given in Appendix D.

- 68 - The paramters for the reference case were assumed to be as follows: Study values Equivalent real values (at 4% inflation) Discount rate 8% 12% Capital cost escalation rate 0% 4% Oil and gas price escalation rate 2% 6% Other fuels escalation rate 0% 4% Loss-of-load probability (fraction Maximum - 0. 0082 Average — less than 0. 005

The fuel oil costs are those prevailing in the Persian Gulf at 1 January 1973, plus ocean and inland transport costs.

9. 4. Technical methodology and parameters

In order to facilitate preparation of data for the WASP program, the characteristics of the alternative generating units which might be installed on the system were standardized as described in Appendix E. The range of plant types and sizes considered are shown in Table IX-1. Characteristics of the alternative generating units are described in more detail in Section 14, and the supporting data on operating and maintenance costs, expected outage rates and plant life are described in Appendix E.

TABLE IX-1. PLANT SIZES AND TYPES CONSIDERED AS POSSIBLE SYSTEM ADDITIONS

Rated capacities Type of plant (MW)

Oil-fiied 150, 200, 250, 300, 400, 500, 600, 800, 1000 Nuclear 200, 250, 300, 400, 500, 600, 800, 1000 Gas turbine 50

9.5. Sensitivity studies

In order to evaluate the sensitivity of the results obtained for the reference case to the various economic parameters used, studies were carried out for other values of these parameters. These are summarized as follows: Study values Equivalent real values (at 4% inflation) Discount rate 6%, 10% 10%, 14% Oil price escalation rate 0%, 1%, 4% 4%, 5%, 8% Nuclear fuel price escalation rate 2% 6% (one forecast only) Shadow exchange ratio 1. 1

In addition to these sensitivity studies, evaluations were also carried out for two different load forecasts; namely, a low forecast developed for the Market Survey as described in Appendix F and a projection to 1990 by the Singapore PUB, extrapolated to 2000 by assuming some regression in the growth rate. Two sets of capital cost data were used. These were ORCOST-1 (lower differential capital costs between nuclear and conventional plants) and ORCOST-3 (the reference capital costs as of 1 January 1973). For details of these costs see Appendix B. In the sensitivity studies, all parameters listed above were kept constant except for the parameter being studied. The results of these studies are discussed in Section 16.

- 69 - 10. FORECASTS OF SYSTEM LOADS AND LOAD DURATION CURVES

10. 1. Review of load forecasts used in the study

(a) Survey forecast

Appendix F describes in detail the method of forecasting gross electricity production developed by the forecasting expert, H. Aoki, assigned to the Market Survey. The method is based on a correlation between electricity production per capita and Gross National Product (GNP) per capita, and hence a GNP forecast and a population forecast are required in order to make a gross electricity generation forecast. Table X-l shows the Aoki forecast for Singapore. The peak demand forecast is based on the energy forecast and an assumed annual load factor of 65%.

TABLE X-l. FORECAST OF ELECTRICITY GENERATION IN SINGAPORE8

1971 1976 1981 1986 1991 1996 2001

Population (106) 2.11 2.27 2.45 2. 64 2. 84 3. 03 3.23 Population growth rate C/o/yr) 1.50 1.50 1. 50 1. 50 1. 30 1.30 GNP/capita (1964 US $) 858 1290 1809 2309 2742 3180 3690 GNP/capita growth rate (°SVyr) 8.5 7.0 5. 0 3. 0 3. 0 3.0 Gross electricity generation kwh/capita 1225 2335 3796 5189 6444 7700 9150 Total (GWh) 2585 5300 9300 13700 18300 23300 29600 Load factor (°lo) 69.9 65 65 65 65 65 65 Peak demand (MW) 422 930 1630 2400 3215 4090 5200

a Forecast prepared by H. Aoki, January 1973.

(b) PUB forecast

The forecast of peak demand to 1990 by the Singapore Public Utilities Board is given in Fig. 5-5. A more detailed forecast through 1981 is given in Table V-l, which projects system annual load factors of about 68%.

Since the Aoki (low) and PUB (high) forecasts are significantly different by 1990, it was decided to analyse expansion plans for a low forecast and for a high forecast. The computer study period was selected to be 1977-1999, inclusive. The first year was selected on the basis that PUB has already made expansion commitments through 1976. It so happens that no expansion is needed in 1977, so that the study, in effect, begins with 1978; however, starting with 1977 presented the opportunity of simulating operation of the system as PUB has planned it, before considering additions to it. The last year of the study was selected on the basis that it permitted consideration of at least ten years of operation of all plants commissioned during the 1980s, which are the plants of interest to the Market Survey. Since the PUB forecast ended with 1990, it was necessary to extend it through 1999. This was done by assuming a 10% annual growth rate during the 1990s, i. e. allowing for some regression in the growth rate from the 12% value used by PUB in the 1980s. The details of the two forecasts studied are discussed in Section 10.2.

- 70 - 10. 2. Load description data required for WASP program (Module 1)

(a) Study increment

In carrying out the computations associated with the load duration curve, it is necessary to select a "study increment", as discussed in Appendix C. Values of 25 MW for the low forecast case and 30 MW for the high forecast case were found to be suitable.

(b) Annual peak loads

Table X-2 gives the annual peak loads for each of the two forecasts, as used in the computer studies. The low forecast numbers do not correspond exactly to the Aoki peak demands in Table X-l since the year-by-year values in Table X-2 were calculated by doing a polynomial regression analysis on the five-year Aoki values to get a best-fit smooth curve. The high forecast peak loads in Table X-2 are those of the PUB through 1989, assuming a 10% annual growth rate for 1990-99.

In addition to the annual peak loads the WASP input data include quarterly peak loads expressed as a fraction of annual peak and quarterly load duration curves expressed as the coefficients of a fifth order polynomial.

TABLE X-2. ANNUAL PEAK LOADS USED IN COMPUTER STUDIES

Low forecast High forecast Year (MW) (MW)

1977 1112 1059 1978 1236 1194 1979 1365 1345 1980 1497 1516 1981 1633 1708 1982 1744 1911 1983 1918 2140 1984 2066 2397 1985 2218 2684 1986 2374 3007 1987 2534 3367 1988 2698 3771 1989 2865 4224 1990 3037 4646 1991 3212 5111 1992 3391 5622 1993 3574 6184 1994 3761 6803 1995 3952 7483 1996 4146 8231 1997 4345 9055 1998 4547 9960 1999 4753 10956

- 71 - Quarterly peak loads were taken to be 94, 96, 98 and 100% of the annual peak for the first, second, third and fourth quarters, respectively, in the low forecast. For the high forecast the corresponding numbers were taken to be 91, 94, 97 and 100%. The difference is due primarily to differences in growth rate, since Singapore has little seasonal dependence of demand. The annual load duration curves of Fig. 4-5, for 1970 and 1971, were subjected to a polynomial regression analysis as described in Appendix C and the coefficients adjusted to give the desired quarterly load factors. The shape of the dimensionless quarterly load curve was taken to be the same for each quarter. The desired quarterly load factors were 67. 0% for the low forecast, corresponding to an annual load factor of 65. 0%, and 71. 4% for the high forecast, corresponding to an annual load factor of 68. 2%. The corresponding polynomial coefficients are given in Table X-3, and the load duration curves are plotted in Fig. 10-1.

TABLE X-3. POLYNOMIAL COEFFICIENTS DESCRIBING QUARTERLY LOAD DURATION CURVES3

Forecast bo b2 b3 b4 bS

Lowb 1. 000 000 -2. 210 580 11,553 940 -29. 684 555 31. 896 286 -12. 204 209

High0 1. 000 000 -1. 915 520 10.011769 -25. 722 412 27. 638 916 -10. 575 250

a The coefficients are those of a fifth order polynomial relating load and duration:

x = b0 + bx + b2 + b3 + b4 + b5 where x = load, expressed as a fraction of the quarterly peak load, t = fraction of time during which the load is equal to or greater than x. The number of decimal digits used reflects only the precision of the computer in making the regression analysis, not the accuracy of the fit, which is much poorer, b See curve labelled "SINGA 1" on Fig. 10-1, c See curve labelled "SINGA 2" on Fig. 10-1.

100- \

80-

60- "••-... '•'•••.... SINGA 2 .OA D

SINGA 1 '•••.. '•-. 40- PERCEN T O F PEA K

20-

0 20 40 60 80 100

PERCENT OF TIME

FIG. 10-1. QUARTERLY LOAD DURATION CURVES.

- 72 - 11. LIMITING FACTORS IN SYSTEM DEVELOPMENT

11.1. Transmission -network

Singapore Island measures approximately 35 kmX 20 km and transmission distances are necessarily short. All existing 66 kV circuits (except for submarine cables) are underground, and all proposed 230 kV circuits will be underground. The 230 kV cables planned are 1000 MCM or 630 mm2 with site ratings in the range of 140 to 210 MVA per circuit depending on circuit proximity. It is unlikely that a higher voltage could be justified in the foreseeable future.

11.2. Load flow/transient stability

If the Senoko site proves capable of development up to 2000 MW, the total system installed capacity will be brought to a little over 3000 MW which, according to various growth projections, would be appropriate to the load at some time between 1986 and 1990. At this stage some 10 to 15 underground 230 kV circuits can be expected to feed out from Senoko. As- suming the next power stations also to be capable of development to 2000 MW, it will eventu- ally require a similar number of circuits and there should be little difficulty in developing the system to meet the required load flows. Special problems will arise if an island site such as Pulau Tekong is developed and the station cost estimates should include a submarine cable link. In a system of this kind transient instability is unlikely to influence system design.

11.3. Frequency stability

There is no system of automatic frequency-sensitive load shedding in use at present and in situations requiring rapid load reduction reliance is placed on operator action according to pre-arranged procedures. The load duration curve for 1971 shows that system load was equal to or greater than 50% of the annual peak for 90% of the time and in the following comments the minimum load is assumed to be 50% of the maximum load. Table XI-1 shows the relationship between the first unit of each rating and the system peak load during the year of its commissioning. Except for case (e), these plans employ large units relative to the system and their adoption implies acceptance of load shedding on loss of the largest unit if frequency is to be maintained within the limits normally acceptable. For a thermal system these may be taken as 1 Hz deviation at peak load and 1. 5 Hz at light load.

TABLE XI-1. RELATIONSHIP BETWEEN THE SIZE OF THE FIRST UNIT OF EACH RATING AND THE SYSTEM PEAK LOAD IN THE YEAR OF ITS COMMISSIONING

Unit rating System peak Ratio, rating/peak Casea Year (MW) (MW) (°lo)

(a) 1965 60 192.5 31 (b) 1973 120 527-623 19-23 (c) 1978 250 1059-1194 21-24 (d) 1983 500 1911-2140 23-26 (e) 1987 300 2534 12 (f) 1986 600 3007 20 a Case (a) based on information in Sections 3 and 4. Cases (b), (c) and (d) based on PUB projections in Fig. 5-5. Case (e) based on Tables X-2 and XV-1, column 6 (low forecast expansion with "small" units). Case (f) based on Tables X-2 and XV-2, column 2 (high forecast expansion with "large" units).

- 73 - An approximate analysis indicates that if these frequency limits are to be met, the situations (a) to (d) will necessitate increasingly severe load shedding on loss of the largest unit as an increasing proportion of spinning reserve capacity is carried by reheat plant. In case (a) it is probable that without load shedding the frequency would have stabilized at around 47. 5 to 48 Hz, but this will not occur in later situations. Typical situations from plans (e) and (f) have therefore been investigated to assess the stability of system frequency following loss of the largest generating unit. Figures ll-l(a) and (b) show the computed behaviour of system frequency at peak load and minimum load respectively on loss of a 600 MW unit when maximum demand is about 3000 MW. Figures 11-2 (a) and (b) show the computed behaviour of system frequency at peak load and minimum load respectively on loss of a 300 MW unit when maximum demand is about 3000 MW.

11.4. Limits to the introduction of large units

In assessing the acceptability of large units it is necessary to consider actual machine loading to provide an adequate spinning reserve in conditions of minimum system load. In the situation of Fig. ll-l(b) there is a large spinning reserve capacity implying either a turn down of all running plant to about half load or maintenance of full load on the largest (base load) units with turn down of the remaining plant to about a quarter load. This apparently generous provision does not avoid severe load shedding because of the time required by the reserve capacity to pick up load.

TIME, SECONDS TIME, SECONDS (3 12 3 4 5 6 7 8 9 10 3 4 5 6 0-

-0.6- I / z 1- LOAD SHED /

-1,2-

-PER C /

-1.8-

\« LOAD SHED / JENC Y DEVIATIO N

FRE Q \ / -2.4- -1.6 • \ /

r-l—-LOAD SHED \ -3.0- -2.0 • y

MINIMUM LOAD 507. 1 SPINNING RESERVE = 487. • SPINNING RESERVE = •3QO/. F PEAK L0A GENERATION LOSS = 17.8V, (• OF PEAK LOAD ( GENERATION LOSS = 1787. f ° ° LOAD SHED = 5% .OAD SHED 157. J

FIG. ll-l(a). FIG. ll-l(b). SYSTEM FREQUENCY VARIATION SYSTEM FREQUENCY VARIATION FOLLOWING 18% GENERATION LOSS AT FOLLOWING 18% GENERATION LOSS AT PEAK LOAD. LIGHT LOAD.

- 74 - SHOULD OCCURONLYINLIGHTLOACONDITION S transmission ofthenetwork. generation affectingallconsumers(unlesoruntirelievedbyloashedding),afailure of TABLE XI-2.LIMITSOFGENERATORSIZ IFLOADSHEDDINGSTOBEMINIMALAN of powerattheiterminals.Thismaybeunderminedeitheninstantaneoushortag f should beminimalanoccuronlyinlightloaconditions. acceptable onlossfaunit.TablXI-2ibasedthassumptiothatloashedding FIG. ll-2(a) FOLLOWING 10%GENERATIONLOSSAT 11. 5Systemreliability—loss-of-loadprobability PEAK LOAD. SYSTEM FREQUENCYVARIATION

FREQUENCY DEVIATION

The abilityofasystemtmeethdemandsitconsumerdependnavailabilit In therangofunitratingsconsideredalimiisebyseveritloashedding PER CENT -0.9- -0.6 - Below 2000 -1.2- -0.3- -1.5- Peak load (MW) LOAD SHE GENERATION LOSS= SPINNING RESERVE= I V 2 000 4 000 3 000 TIME, - ^ SECONDS NONE J 21V. | 9.8V. \OFPEAKLOAD 5 678910 Low forecast 1996 1990 1984 Year — 75_ FOLLOWING 10%GENERATIONLOSSAT FIG. ll-2(b) LIGHT LOAD. SYSTEM FREQUENCYVARIATION High forecast o- -1.2 -2.4 -0.6 -3.0 - 1989 1986 1983 SPINNING RESERVE=20V. LOAD SHE=1% GENERATION LOSS=9.87. MINIMUM LOAD=507. 3 4567 TIME, SECONDS Maximum rating 0F PEAKL0AD (MW) 400 250 330 120 Transmission failure can result in (i) isolation of a complete power station producing a power shortage more severe than the loss of a single machine; (ii) loss of a radial feed with the associated load; or (iii) loss of part of a network. Precautions against loss of a power station include provision of an adequate number of transmission circuits, standby transformer capacity and high security busbar arrangements. Similar measures apply to radial feeds to the extent justified by their importance; networks are designed to provide alternative paths. In Singapore there will be a number of cables from each power station which, to the extent that they are separately switched, will assist security. The proposed 230 kV system, being in the form of a ring, should be inherently secure against loss of any one section. Loss-of-load probability and energy not served, as calculated in the WASP probabilistic simulation code, are discussed in Appendix A and, as applied in Market Survey studies of Singapore, in Section 15. In the reference expansion plans adopted, the annual loss-of-load probability averaged less than 0. 5% and the energy not served averaged less than 0. 05%.

76 - 12. CHARACTERISTICS OF THE EXISTING AND COMMITTED ELECTRIC POWER SYSTEM

12.1. Existing system and commitments

Table III-3 lists installed generating units as of the Market Survey team visit in November 1972. The first 22. 5 MW Senoko gas turbine unit was to come on stream in December 1972 and the second in January 1973. • By the end of 1973 all three 120 MW units of Jurong Stage II are to be in operation, and, as shown in Tables V-l and V-4 and Fig. 5-4, the three 120 MW Senoko Stage I units will be commissioned in 1975-76. The six St. James free piston units, totalling 27 MW, are scheduled for retirement in 1975. Figure 3-2 shows the location of existing and proposed generating plant. Thus the system at the beginning of the first study year, 1977, will have 1442 MW of installed capacity, all oil-fired steam units except for 67 MW of gas turbines. No expansion is planned in 1977. There are tentative plans for a series of 250 MW units at the Senoko station beginning in 1978, but these were not taken as committed.

12. 2. Derivation of computer input data for WASP Module 2

(a) Grouping of plants by size, type of fuel, and fuel cost

For the purpose of WASP computations it is desirable to group the units comprising the "fixed system" (the system existing at the beginning of the first study year) into plants with hypothetical identical unit capacities, fuel types, fuel costs, heat rates, forced outage rates, maintenance requirements, and operating and maintenance costs. Table XII-1 shows how this grouping was done. The computer input data format requires integer values for capacities; hence, the Senoko gas turbines were entered as 22 MW each. The computer studies of the low forecast were started before receiving the country's comments on the draft Sections 1 through 8 and, as a result of misinterpretation of information brought back by the Market Survey team from Singapore, the 2X11 MW St. James gas turbine plant was not included because it was thought they were to be retired before 1977. They were included, however, in the studies of the high forecast.

TABLE XII-1. GROUPING OF UNITS IN THE FIXED SYSTEM

"PPAA". 7 x 25 MW. oil-fired Pasir Panjang A, units 1-7, 25 MW each

„ „ . ., , . Pasir Panjang B, units 1-4, 60 MW each "PBJ1", 8 x 60 MW, oil-fired ' B .. , . ' „,,, =—s • JuronT g Stage IT, units 1-4, 60 MW each

„ „ „ „ , ., , , Jurong Stage II, units 5-7, 120 MW each "J2S1", 6 x 120 MW, oil-fired , „ , „„*„,, ' Senoko Stage I, units 1-3, 120 MW each

"SNGT, 2 x 22 MW, gas turbines Senoko gas turbines, 2 units, 22.5 MW each

"SJGT". 2 x 11 MW. gas turbines St. James gas turbines, 2 units, 11 MW each

(b) Minimum and maximum operating levels

The WASP input requires specification of the minimum operating level, i.e. the lowest level a unit will be operated at before shutting down, and maximum operating capacity. The maximum capacities were taken to be those given in Table XII-1. For the gas turbines the minimum operating level was taken to be the same as maximum capacity, i.e. they are assumed to operate at full load or not at all. For the oil-fired units, the minimum operating level was taken to be 50% of maximum capacity, which is within the 40-60% range mentioned in Section 3.4(a), and which is the same minimum level selected for standard Market Survey expansion alternatives.

- 77 - (c) Heat rates

Gross heat rates at minimum operating level and average incremental gross heat rates between minimum and full load were estimated from the average values in Tables IV-1 and V-l using the standard values in Appendix G as a guide.

(d) Fuel costs

The cost of high sulphur heavy fuel oil in Singapore as calculated under Market Survey standard assumptions is given in Appendix I as 140 US ^/106 kcal, as of 1 January 1973. This figure, which is a little lower than the estimated price to the PUB as of that date, was used to be consistent with Market Survey analyses of the other 13 countries. Under reference conditions the fuel oil cost was taken to escalate at 2%/yr in terms of 1 January 1973 "constant US dollars", i.e. at 2% faster than the general rate of inflation (assumed 4%/yr). Alternative oil cost escalation rates of 0% and 4% were considered. The 4% figure might also be thought of as equivalent to a 2% basic rate for high-sulphur fuel oil plus another 2% to represent the cost of gradual reduction in sulphur content of fuel oil burned by PUB. The cost of gas turbine fuel was taken to be 229 US ^/106 kcal, based on the most recently available information from PUB. This cost was also assumed to escalate from 1 January 1973 in constant dollar terms on the same basis as heavy fuel oil. Actually, fuel cost escalation is handled in WASP Module 6, not in Module 2.

(e) Forced outage rates

Average forced outage rates were estimated considering the information in Section 4. 2 and the standard Market Survey values of Appendix E.

(f) Scheduled maintenance requirements .

The average number of days per year required for scheduled maintenance for each unit were estimated considering the information in Section 4.2 and the standard values of Appendix E.

(g) Operating and maintenance costs

PUB operating and maintenance cost (O&M) information, based on their past experience and financial plan for the next several years, was used to estimate the average O&M costs for each group of units in the fixed system at end-1976, expressed as fixed costs in US $/kW- month at 1 January 1973 price levels. These were assumed to remain level in constant US dollars, i. e. to escalate at the same rate as inflation in general. The WASP code can treat O&M costs as either fixed or variable costs, or both; however, it was decided to treat them as fixed costs since this is simpler and is not believed to introduce any singificant error in the evaluation of the nuclear power market of the 1980s.

(h) Retirement schedule

Any retirements of units in the fixed system before the study horizon must be specified in the WASP input. In preparing the input data for the low forecast it was assumed that all units would be retired 30 years after commissioning. Thus, it was decided to assume retirement of all seven Pasir Panjang units in 1983 for the high forecast.

12.3. Computer printout showing the characteristics of the fixed system assumed to be in existence at start of study year

Table XII-2 shows the actual printout from the WASP Module 2 computer program summarizing the characteristics of the system at the start of the first study year.

- 78 - TABLE XII-2(a). COMPUTER PRINTOUT FROM WASP MODULE 1 (FIXED SYSTEM)1

BASE' AVGE FUEL COSTS L FRCD FULL NO. MIN. CAP- LOAD I NCR CENTS/MILL ION C OUT- DAYS LOAD OF LOAD CITY HEAT HEAT T AGE SCHL MAIN ENRGY O&M O&M HEAT NAME SETS MW MW RATE RATE DMSTC FORGN TYPE N RATE MAIN CLAS GWH (FIX) (VAR) RATE

1 PPAA 7 12 25 3347. 3199. 0.0 140.00 1 1 7.60 34 40 0. 0 .700 0.0 3270 VD 1 2 PBJ1 8 30 60 2958. 2500, 0.0 140.00 1 1 3.80 47 40 0. 0.370 0.0 2729

3 J2SI 6 60 120 2574. 2406. 0.0 140.00 1 1 6.20 28 120 0. 0 .150 0.0 2490

4 SNGT 2 22 22 3528. 3528. 0.0 229.00 2 1 2.00 4 40 0. .0.230 0.0 3528

5 SJGT 2 11 11 4864. 4864. 0.0 229.00 2 1 2.00 4 40 0. 0.780 0.0 4864

For legend see Table XII-2(b). TABLE XII-2(b). LEGEND FOR TABLE XII-2(a)

NAME WASP code for existing plants (see Table XII-1 for fossil plants).

No. OF SETS Number of units of a given size located at a given plant.

MIN. LOAD, Minimum load at which units will be operated (see 12.2(b)). MW

CAP-CITY, Maximum load at which units will be operated (see 12.2(b)). MW

BASE LOAD Unit heat rate at base load, in kcal/kWh (see 12.2(c)). HEAT RATE

AVGEINCR Unit heat rate for each kW above base load, in kcal/kWh HEAT RATE (see 12.2(c)).

FUEL COSTS, Fuel costs in US(i/kcal x 106 (see 12.2(d)). DOMESTIC

FUEL COSTS, Same as above. FOREIGN

TYPE A code where; -1 = emergency hydro 0 - nuclear 1 = oil-fired 2 - 4 = optional 5 = hydro

LCTN Not used. Defaulted to 1 in all cases.

FRCD OUTAGE Days lost due to forced outage (see 12.2(e)). RATE

DAYS SCHL Days lost due to scheduled outage (see 12.2(f)). MAIN

MAIN CLAS An arbitrary assignment of unit size, for maintenance calculations.

ENRGY, Used only for hydro. GWh

O &M Average O & M costs, in US$/kW -month (see 12.2(g)). (FIX)

O &M Not used. (VAR)

FULL LOAD Full load heat rate, as calculated by WASP based on the HEAT RATE base load heat rate and average incremental heat rate data above.

- 80 - 13. CAPITAL COST DATA

13.1. Basis for thermal plant cost estimates

Appendix B describes in detail how capital costs for thermal power plants were developed using the ORCOST computer program. The required input data are shown in Table 5 of Appendix B. Except for the cost indices for equipment, materials and labour, which varied for each of the countries covered by the Survey, the input data were kept the same for all countries to provide consistency among results.

(a) Interest during construction

ORCOST-1 and ORCOST-3 capital cost estimates were based on an assumed 8% interest rate during construction, i.e. the same as the reference present-worth discount rate used in the economic evaluation. However, the interest during construction was not changed when the discount rate was varied from 6% to 10%. The possible effect of changing the interest rate during construction to keep it equal to the discount rate is discussed in Section 16.3.

(b) Construction schedules

The construction schedule for each size and type of plant was based on current experience in the USA. Extrapolations were made for the smaller plant sizes. The values used are given in Table XIII-1.

TABLE XIII-1. LENGTH OP CONSTRUCTION PERIOD ASSUMED FOR CALCULATION OF INTEREST DURING CONSTRUCTION AND FOR CASH-FLOW ANALYSIS (yr)a

Plant size Nuclear plants Oil-fired plants

150 - 2.60 200 4.75 2.75 300 5.00 3.00 400 5.20 3.20 600 5.50 3.50 800 5.75 3.75 1000 6.00 4.00

a The cash-flow analysis is reported in Section 16.

As mentioned in Appendix B, contingency factors were taken to be 5% on equipment and 10% on construction labour. The spare parts factor was assumed to be 1% of equipment and materials costs.

(d) Other considerations

The ORCOST program allows for the inclusion of extra costs for such items as special materials, cooling towers (instead of using standard ocean or river-water cooling), sulphur dioxide removal equipment, overtime pay, etc.; however, none of these costs were included in the capital cost estimates. Capital costs of all fossil-fuelled plants include costs for electrostatic precipitators for stack-gas clean-up.

- 81 - 13. 2. Cost indices

(a) Equipment cost index

A review of recent world market prices of conventional thermal power plant equipment indicated that on an international competitive bidding basis these should be about 85% of those used in ORCOST. Allowing 5% additional for extra transportation costs gave an equipment index for conventional thermal plants of 0. 9. In the case of nuclear (PWR) plant equipment, however, it was decided that world market prices should be about 95% of those used in ORCOST. After allowing 5% for extra transportation costs this gave an equipment index of 1.0. This evaluation was made as of end-1972, and costs are expressed in 1 January 1973 US dollars. Thus any possible changes as a result of the recent dollar devaluation are not considered.

(b) Materials cost index

A comparison of costs of construction materials in Singapore (see Table VI-2) with reference costs and amounts used in ORCOST (see Appendix B), and also taking into account recent available information on materials and equipment cost breakdown for oil-fired and nuclear plants in another country in the same area, where weighted average construction materials prices are practically the same as in Singapore, it was decided that a materials cost index of 0. 85 would be appropriate for use in ORCOST for adjusting US costs to Singapore conditions.

Construction labour wage rates in Singapore are given in Table VI-1, and Section 6. 3 indicates an adder of about 20% for fringe benefits. The weighted construction labour wage, including fringe benefits, as used in ORCOST, in Singapore is less than 8% of the corresponding US value. Low-cost labour is used differently and more freely in construction work than is high-cost labour; for example, more labour and less equipment rental are used. Considering all of the power plant and building construction cost information in Section 3 and 6, and taking account of the recent labour cost estimates referred to in (b) above, for conventional and nuclear power plants in Luzon in comparison with those in Singapore and in ORCOST, it was decided to use an overall labour cost index of 0. 27 in estimating power plant construction costs in Singapore for Market Survey purposes.

(d) Indirect cost indices

These were taken to be the same as in ORCOST. See Appendix B for details.

13.3. Capital costs of oil-fired and nuclear steam plants

(a) ORCQST-3 and ORCOST-1 cost summaries

ORCOST-3 printouts of capital costs for 600 MW oil-fired and nuclear plants in Singapore are shown in Tables XIII-2 and XIII-3. Summaries of costs calculated by ORCOST-3 for other plant sizes considered as possible additions are given in Tables XIII-4 and XIII-5. A summary of the capital costs calculated by ORCOST-1 and used in a sensitivity study is given in Table XIII-6.

(b) Local and foreign components of capital cost

Based on past cost experience with conventional steam-plants built in Singapore it was estimated, for purposes of calculating local and foreign financing requirements, that 30% of the capital cost of oil-fired plants to be built in Singapore would represent local currency costs. This is slightly higher than the average for Jurong units 5-7 (Tables III-5 and III-6),

- 82 - TABLE XIII-2. ORCOST-3, CAPITAL COST ESTIMATES FOR A 600 MW OIL-FIRED PLANT (106 US $)a

DIRECT COSTS

20 LAND AND LAND RIGHTS 0.1

PHYSICAL PLANT EQU. MAT. LABOUR TOTAL 21 STRUCTURES AND SITE FACILITIES 0.5 5.9 2.9 9.3 22 REACTOR/BOILER PLANT EQUIPMENT 16.2 3.9 3.4- 23.6 23 TURBINE PLANT EQUIPMENT 17.1 5.1 2.8 25.0 24 ELECTRIC PLANT EQUIPMENT 3.9 1.4 1.4 6.8 25 MISCELLANEOUS PLANT EQUIPMENT 0.9 0.6 0.5 2.0 26 SPECIAL MATERIALS 0.0 0.0 0.0 0.0 INCREMENTAL ALLOWANCE FOR ENVIRONMENTAL ASSURANCE SO-2 REMOVAL SYSTEM 0.0 0.0 0.0 0.0 COOLING TOWERS 0.0 0.0 0.0 0.0 SUBTOTAL (PHYSICAL PLANT) 38.7 16.9 11.0 66.6 CONTINGENCY ALLOWANCE 3.9 SPARE PARTS ALLOWANCE 0.6 SUBTOTAL (PHYSICAL PLANT) 71.0 OVERTIME ALLOWANCE < 40.0 HR WORKWEEK) 0.0 SUBTOTAL (TOTAL PHYSICAL PLANT) 71.0

INDIRECT COSTS

91 CONSTRUCTION FACILITIES, EQUIPMENT, AND SERVICES - 6.5 92 ENGINEERING AND CONSTRUCTION MANAGEMENT SERVICES - 10.7 93 OTHER COSTS - 3.1 94 INTEREST DURING CONSTRUCTION ( 8.0 PCT- 3.50 YRS) 11.5 SUBTOTAL (TOTAL INDIRECT COSTS) <- 31.9

SUBTOTAL (TOTAL PLANT COST) • 103.0 CAPABILITY PENALTY ( 0.0 PCT- 0.0 MW(E)) 0.0

TOTAL PLANT CAPITAL COST (AT START OF PROJECT) 103.0 / f\ n \ Ci --—•-—»•----—•-—-—————-^•—.—.—.———————.———————— -»——— ^ \ c.+

ESCALATION DURING CONSTRUCTION ( 0.0 PCT ) 0.0

TOTAL PLANT CAPITAL COST (AT COMMERCIAL OPERATION) 103.0 $ / KW(E) 172.

Costs are in 1972 constant dollars.

but seemed reasonable since, as unit size increases, it was assumed that the ratio of local to foreign costs will stay about the same, though it might be expected to increase somewhat if the type and size of plants were repeated over an extended period. For nuclear plants the local participation would be relatively lower than for oil-fired plants, the estimated local cost component being about 20%. Because nuclear plant capital costs are much higher, local participation, in absolute terms, will actually be higher. For example, for 600 MW units local costs would be US $65/kW for a nuclear plant and US $52/kW for an oil-fired plant.

13.4. Treatment of transmission costs

Costs of transmission lines added to the system during the study period were assumed to be the same regardless of the type of unit being considered as an expansion alternative. Since nuclear plants are more likely to be built on Pulau Tekong Island than on the mainland,

- 83 - TABLE XHI-3. ORCOST-3, CAPITAL COST ESTIMATES FOR A 600 MW NUCLEAR PLANT (106 US $)a

DIRECT COSTS 20 LAND AND LAND RIGHTS 0.1 PHYSICAL PLANT EQU. MAT. LABOUR TOTAL 21 STRUCTURES AND SITE FACILITIES 1.2 11.6 6.5 19.3 22 REACTOR/BOILER PLANT EQUIPMENT 29.5 10.1 4.1 43.8 23 TURBINE PLANT EQUIPMENT 26.9 7.2 3.9 38.0 24 ELECTRIC PLANT EQUIPMENT 4.7 5.9 2.4 13.0 25 MISCELLANEOUS PLANT EQUIPMENT 1.8 0.2 0.7 2.7 26 SPECIAL MATERIALS 0.0 0.0 0.0 0.0 INCREMENTAL ALLOWANCE FOR ENVIRONMENTAL ASSURANCE UPGRADED RADWASTE SYSTEM 0.0 0.0 0.0 0.0 COOLING TOWERS 0.0 0.0 0.0 0.0 SUBTOTAL (PHYSICAL PLANT) 64.1 35.0 17.6 116.8 CONTINGENCY ALLOWANCE ^ 6.7 SPARE PARTS ALLOWANCE 1.0 SUBTOTAL (PHYSICAL PLANT) 124.5 OVERTIME ALLOWANCE ( 40.0 HR WORKWEEK) 0.0 SUBTOTAL (TOTAL PHYSICAL PLANT) 124.5 INDIRECT COSTS 9-1 CONSTRUCTION FACILITIES, EQUIPMENT, AND SERVICES - 8.9 ?2 ENGINEERING AND CONSTRUCTION MANAGEMENT SERVICES - 22.9 93 OTHER COSTS 5.0 94 INTEREST DURING CONSTRUCTION { 8.0 PCT- 5.50 YRS) 33.4 SUBTOTAL (TOTAL INDIRECT COSTS) 70.2 SUBTOTAL (TOTAL PLANT COST) 194.8 CAPABILITY PENALTY ( 0.0 PCT- 0.0 MW(E)) 0.0 TOTAL PLANT CAPITAL COST (AT START OF PROJECT) 194.8 $ / KW(E) 32 5. ESCALATION DURING CONSTRUCTION ( 0.0 PCT ) 0.0 TOTAL PLANT CAPITAL COST (AT COMMERCIAL OPERATION) 194.8 $ / KW(E) 325.

Costs are in 1972 constant dollars.

it might be considered that they should be penalized for the extra cost of a submarine cable link with the mainland; however, it could be that oil-fired plant expansion after the Senoko site capacity is used up would also be on Pulau Tekong. Though they were not added to the unit capital costs in the computer studies, the costs (in 1972 US $) of a 5 km submarine cable link at 230 kV were estimated to be as follows:

Firm transfer capacity 1000 MW 6. 5 US $/kW Firm transfer capacity 2000 MW 5.0US$/kW

13.5. Gas turbine costs

As mentioned in Appendix B, the 1 January 1973 installed cost of a 50 MW gas turbine was taken to be US $125/kW.

- 84 - TABLE XIII-4. SUMMARY OF CAPITAL COSTS CALCULATED BY ORCOST-3 FOR OIL- FIRED STEAM ELECTRIC PLANTS (106 US $)a

Plant size

Account No.b 200 MW 300 MW 400 MW 800 MW 1 000 MW

Direct costs

20 0.1 0.1 0.1 0.1 0,1 21 4.1 5.5 6.8 11.5 13.6 22 8,8 12.6 16.4 30.5 37.3 23 10.4 14.4 18.1 31.4 37.6 24 4.1 5.0 5.6 7.7 8,5 25 1.4 1.6 1.8 2.2 2.3

Subtotal0 28.9 39.2 48.8 83.4 99.4

Contingency 1.7 2.3 2.8 4.8 5.8 Spare parts 0.2 0.3 0.4 0.7 0.8

Subtotal 30.8 41.8 52.0 89.0 106.0

Indirect costs

91 3.7 4.7 5.5 7.0 7.5 92 6,1 7.7 9.0 11.1 12.0 93 1.9 2.4 2.7 3.1 3.3 94 4.2 6.1 8.0 15.1 18.8

Subtotal 15.9 20.9 25,2 36.2 41.6

Totald 46.7 62.7 77.2 125.2 147.6

Unit cost (US$/kW) (233) (209) (193) (156) (148)

a 600 MW size covered in Table XIII-2. b See Table XIII-2 for account names corresponding to the account number. c Physical plant cost including arbitrary US$100 000 for account 20. Total plant capital cost at 1 January 1973 dollar price levels (no escalation during construction).

- 85 - TABLE XIII-5. SUMMARY OF CAPITAL COSTS CALCULATED BY ORCOST-3 FOR NUCLEAR (PWR) STEAM ELECTRIC PLANTS (106 US $)a

Plant size

Account No. 200 MW 300 MW 400 MW 800 MW 1 000 MW

Direct costs

20 0.1 0.1 0.1 0.1 0.1 21 12.4 14.6 16.4 21.6 25.7 22 22.7 28.9 34.3 52.1 59.6 23 15.8 21.8 27.5 47.8 57.2 24 6.7 8.6 10.2 15.5 17.7 25 1.9 2.2 2.4 2.9 3.1

Subtotal0 59.6 76.2 90.9 140.0 163.4

Contigency 3.5 4.4 5.2 8.0 9.3 Spare parts 0.5 0.6 0.8 1.2 1.4

Subtotal 63.6 81.2 96.9 149.2 174.1

Indirect costs

91 6.9 7.6 8.0 9.7 10.6 92 16.7 19.5 20.7 25.0 27.3 93 4.1 4.3 4.5 5.5 5.9 94 16.1 20.9 25.3 41.2 49.7

Subtotal 43.7 52.3 58.5 81.4 93.5

Totald 107.4 133.5 155.4 230.6 267.6

Unit cost (US$/kW) (537) (445) (388) (288) (268)

a 600 MW size covered in Table XIII-3. b See Table XIII-3 for account names corresponding to account numbers. c Physical plant cost including arbitrary US $100 000 for account 20. d Total plant capital cost at 1 January 1973 dollar price levels (no escalation during construction).

- 86 - TABLE XIII-6. SUMMARY OF CAPITAL COSTS CALCULATED BY ORCOST-1 FOR STEAM ELECTRIC PLANTS IN SINGAPORE

Totala Plant type and size Direct costs Indirect costs 6 6 (MW) (10 US$) (10 US$) 106 US$ US$/kW

Oil-fired

200 30.1 17.8 47.9 240 300 40.6 23.3 63.9 213 400 50.3 28.0 78.3 196 600 68.5 35.3 103.8 173 800 85.5 39.7 125.2 157 1 000 101.7 45.2 146.9 147

Nuclear

200 52.6 39.2 91.8 459 300 67.2 47.7 114.9 383 400 80.2 54.1 134.3 336 600 103.4 63.7 167.1 278 800 124.0 73.1 197.1 246 1 000 3 44.7 83.1 227.8 228

Total plant capital cost at 1 January 1973 dollar price levels (no escalation during construction).

- 87 - 14. CHARACTERISTICS OF ALTERNATIVE GENERATING UNITS CONSIDERED FOR EXPANSION DURING THE STUDY PERIOD

14. 1. Review of local plans for system expansion

The PUB plans for system expansion through 1976 are covered in Section 5. There are no commitments after 1976; however, there is a ten-year expansion forecast for planning purposes in Table V-3 and Fig. 5-4. It calls for five 250 MW oil-fired units at Senoko during 1978-81, and a possible 500 MW nuclear plant in 1983. The list of standard unit sizes considered as alternatives for expansion in the Market Survey, in Appendix E, does not include the 250 MW and 500 MW sizes, and so these were not considered in the search for a near-optimum expansion plan for the 1980s; however, the PUB plan can be considered to be equivalent to a like number of megawatts of standard size units during the same time period. Thus, in the reference expansion plan for the high forecast, described in Section 15, there are 200 MW units in 1978-7 9, 300 MW units in 1980-82, 400 MW units in 1983-85 and 600 MW units in 1986-89. Actually, descriptions of 250 MW and 500 MW plants were included in the WASP computer input data mentioned below, but computer time and calendar time limitations prevented their being evaluated on an equal basis with the standard units.

14. 2. Derivation of computer input data for WASP Module 3

The WASP module for processing the descriptive information concerning units to be considered as expansion alternatives requires essentially the same information already described in Section 12 for units in the fixed system, and in the same format. The only significant differences are that the number of sets of a given size and type is input as zero, to be determined in a subsequent program module, and there is no provision for retirements.

(a) Unit sizes and types

Table XIV-1 lists the 18 different units included in the Module 3 input data for the high forecast. For the low forecast the 250, 500, 800 and 1000 MW units were omitted. As mentioned above, the 250 and 500 MW units received only minimal attention since it was desired that the reference expansion plans consist only of standard size units, to permit uniform treatment of the results of the fourteen-country studies making up the Market Survey.

(b) Minimum and maximum operating levels

As with the fixed system, maximum operating levels were taken to be equal to nominal capacity, and minimum operating levels taken to be 50% of maximum except for gas turbines, for which minimum was taken to be the same as maximum. See Appendix E.

Half-load and incremental heat rates were taken to be equal to those given in Appendix E for standard oil-fired and nuclear plants. The values for 250 and 500 MW sizes were obtained by interpolation from standard unit data. The heat rate for the 50 MW gas turbine was taken to be equal to that for the 22.5 MW Senoko gas turbines in the fixed system.

(d) Fuel costs

Standard fuel costs for oil-fired plants and nuclear from Appendixes I and J, were used. The gas turbine fuel cost used was the latest reported by PUB. See Section 12. 2(d) for treatment of fuel cost escalation, which is handled in WASP Module 6.

OO TABLE XIV-1. LIST OF UNITS CONSIDERED AS POSSIBLE EXPANSION ALTERNATIVES

Unit type and size Code name (MW)

Oil fired 150 O150 200 O200 250 O250 300 O300 400 O400 500 O500 600 O600 800 O800 1 000 O1T0

Nuclear 200 N200 250 N250 300 N300 400 N400 500 N500 600 N600 800 N800 1 000 N1T0

Gas turbine 50 GT50

(e) Other data

Standard forced outage rates and scheduled maintenance requirements were used, from Appendix E. The standard operating and maintenance costs of Appendix E, which are based on US conditions, were reduced, taking into account actual PUB experience, the big difference being wages and salaries even after allowing for differences in staffing practices. The staffing requirement for a 600 MW nuclear plant was assumed to be the same as that estimated in a 1972-73 feasibility study for a nuclear power plant in the same area, and the operation and maintenance costs for other unit sizes were taken in proportion to the standard US values.

14. 3. Computer printout summary of characteristics of expansion alternative generating units

Table XIV-2 shows the actual Module 3 code output summarizing the expansion alternative input data, as described in Section 12. 3 for the corresponding fixed system information.

- 89 - TABLE XIV-2. COMPUTER PRINTOUT FROM WASP MODULE 3£

BASE AVGE FUEL COSTS L FRCD FULL NO. MIN. CAP- LOAD INCR CENTS/MILLION C OUT- DAYS LOAD OF LOAD CITY HEAT HEAT T AGE SCHL MAIN ENRGY O&M 0£M HEAT NAME SETS MW MW RATE RATE DMSTC FORGN TYPE N RATE MAIN CLAS GWH (FIX) (VAR) RATE

1 GT50 0 50 50 3528. 3528. 0.0 229.00 2 1 2.00 4 38 0. 0.200 0.0 35 28.

2 0150 0 75 150 2347. 2193. 0.0 140.00 1 1 5.30 28 120 0. 0.240 0.0 2270.

3 0200 0 100 200 2280. 2146. 0.0 140.00 1 1 5.40 28 200 0. 0.190 0.0 2213.

4 0250 0 125 250 2308. 2164. 0.0 140.00 1 1 6.00 28 300 0. 0.170 0.0 2236.

5 0300 0 150 300 2335. 2183. 0.0 140.00 1 1 6.50 28 300 0. 0.150 0.0 2259.

6 0400 0 200 400 2324. 2098. 0.0 140.00 1 1 9.80 28 400 0. 0.130 0.0 2211.

7 0500 0 250 500 2326. 2137. 0.0 140.00 1 1 10.90 28 600 0. 0.120 0.0 2232. 1 8 0600 0 300 600 2328. 2172. 0.0 140.00 1 1 12.00 28 600 0. 0.100 0.0 2250. o 1 9 0800 0 400 800 2334. 2170. 0.0 140.00 1 1 12.20 35 800 0. 0.090 0.0 2252.

10 01T0 0 500 1000 2344. 2152. 0.0 140.00 1 1 12.20 35 1000 0. 0.090 0.0 2248.

11 N200 0 100 200 2648. 2359. 0.0 59.00 0 1 5.40 28 200 0. 0.430 0.0 2504.

12 N250 0 125 250 2647. 2360. 0.0 58.50 0 1 6.00 28 300 0. 0.380 0.0 2504.

13 N300 0 150 300 2645. 2360. 0.0 58.00 0 1 6.5 0 28 300 0. 0.320 0.0 2503.

14 N400 0 200 400 2643. 2362. 0.0 57.00 0 1 9.80 28 400 0. 0.260 0.0 2503.

15 N500 0 250 500 2640. 2364. 0.0 56.00 0 1 10.90 28 600 0. 0.230 0.0 2502.

16 N600 0 300 600 2637. 2365. 0.0 55.00 0 1 12.00 28 600 0. 0.200 0.0 2501.

17 N800 0 400 800 2632. 2368. 0.0 53.00 0 1 12.20 35 800 0. 0.170 0.0 2500.

18 N1T0 0 500 1000 2627. 2372. 0.0 51.00 0 1 12.20 35 1000 0. 0.140 0.0 2500.

For legend see Table XII-2(b). 15. ANALYSIS OF ALTERNATIVE EXPANSION PROGRAMS

15. 1. Method of analysis

(a) Selection of unit sizes

Starting with the low load forecast, trial computer runs were made to establish the relationship between unit sizes, timing, loss-of-load probability (LOLP) and overall cost. Table V-l shows that 1976 is the last year for which system additions are committed and that no additions in 1977 are required but that further additions in 1978 are anticipated. Based on this, 1977 was selected as the initial year, with no additions during this year. Simulation by WASP of the fixed system during 1977 indicated an average annual LOLP of 0. 0082 for the low load forecast and 0. 0033 for the high forecast. It was decided to select expansion plans which give annual LOLP values not higher than 0. 0082 and which give an average LOLP during the study period of less than 0. 0050. Table XV-1 shows two expansion

TABLE XV-1. ALTERNATIVE EXPANSION PLANS TO MEET GENERATION REQUIREMENTS - LOW LOAD FORECAST CASE

Expansion with large units Expansion with small units Year Installed Reserve Installed Reserve Additions LOLP Additions LOLP capacity15 margin capacity" margin Wo) (MW) (MW) TO (W (MW) (MW)

1977 - 1419 28 0.82 - 1419 28 0.82 1978 200 1619 31 0.57 150 + 50 1619 31 0.47 1979 200 1819 33 0.35 150 1769 30 0.45 1980 200 2019 35 0.25 150 1919 28 0.48 1981 50 2 060 27 0.57 200 2119 30 0.37 1982 300 2 369 36 0.29 200 2 319 33 0.30 1983 300 2 644 38 0.17 200 2494 30 0.25

1984 50 2 669 29 0.43 200 2 669 29 0.29 1985 400 3 044 37 0.24 200 2 844 28 0.32 1986 - 3 019 27 0.77 200 3 019 26 0.34 1987 400 3 394 34 0.42 300 3 294 30 0.23 1988 400 3 769 40 0.23 - 3 269 21 0.80 1989 - 3 769 32 0.56 300 3 569 25 0.47 1990 600 4 369 44 0.23 300 3 869 27 0.31 1991 - 4 369 36 0.49 - 3 869 20 0.81 1992 600 4 944 46 0.22 300 4144 22 0.66 1993 - 4 944 38 0.46 300 4444 24 0.47 1994 50 4 994 33 0.73 400 4844 29 0.28 1995 600 5474 39 0.48 400 5124 30 0.27 1996 600 5 954 44 0.35 50 5 054 22 0.79 1997 - 5 954 37 0.63 400 5454 26 0.48 1998 600 6 554 44 0.33 400 5 854 29 0.32

1999 - 6434 35 0.80 2 X 50 5 834 23 0.72

23-yr average - 36 0.45 27 0.47

a This plan was selected as the reference schedule for the dynamic programming optimization study. See text, b Taking retirements into account. .

- 91 - plans meeting these criteria. The expansion plan with "large" units, in this case with unit size up to 20% of annual peak load, shows a slightly better 23-year average reliability than the expansion plan with "small" units, in this case unit sizes meeting the no-load-shedding criteria of Section 11, but at the expense of a 36% average reserve margin compared to 27%. On the other hand, the large-unit expansion plan was also at slightly lower cost when the two plans were compared on the basis of all oil-fired (and gas turbine) additions, since the economics of scale outweighed the extra reserve margin, and the large-unit plan should benefit more from the introduction of nuclear power which has even greater economies of scale. Thus, the large-unit expansion plan was selected as the basis for a dynamic programming optimization study for the low-forecast case. As shown in Tables IV-1 and XI-1, it has been the practice in the past in Singapore to expand with relatively large units and to maintain relatively high reserve margins. Table V-l and Fig. 5-4 show that the PUB presently plans to continue this practice. Thus for the high load forecast also an expansion plan based on unit sizes up to 20% of peak load, with a 1000 MW limit on unit size, was adopted as the reference plan. In the high forecast an alternative expansion plan was also evaluated, based on the 20% size limit up to 600 MW units and then continuing to use 600 MW units until 800 MW units could be introduced without violating the no-load-shedding criterion of Section 11. These two expansion plans are shown in Table XV-2. The alternative plan is the same as the reference plan through 1988, so that the difference is mainly in the 1990s and the comparison of the two cases amounts to a sensitivity study of the effect on the near-optimum solution for the 1980s of assumptions made about the 1990s. The system stability study referred to in Section 11 indicates that expansion with the relatively large units implies acceptance of load shedding on loss of the largest unit if frequency is to be maintained within limits normally acceptable. The WASP probabilistic simulation code calculates the loss-of-load probability and the probable amount of energy not served, both of which were indicated to be acceptably low for the expansion plans involving the relatively large units (the energy not served averaged less than 0. 05%). The energy not served measures the product of the frequency and the severity of load shedding, but does not indicate how severe an individual occurrence might be. The balancing of cost against risk must be left to the local authorities. A change to a smaller relative unit size limit would lead to a smaller total market, because less reserve margin would be required for a given loss-of-load probability, and to a smaller nuclear fraction of the market, because nuclear units are less competitive in smaller sizes.

(b) Selection of schedules of capacity additions for optimization studies

For the low forecast the WASP dynamic programming optimization code was used to select the lowest cost expansion plan by choosing between fossil and nuclear for all of the steam plants added during 1980-1999, i.e. it "optimized" additions for both decades. The optimum solutions obtained under the various assumed combinations of economic conditions are, however, actually only "near-optimum", since calendar time and computer time limitations did not permit full relaxation of the constraints imposed on the choices available to the dynamic program. For example, it was not allowed to hunt for the lowest-cost size sequence within the imposed limit on loss-of-load probability, being restricted to the sequence given in Table XV-1 for the large unit case. For the high forecast, time did not permit using the dynamic programming optimization technique. Instead, a limited number of alternative expansion plans were evaluated, involving a fixed schedule of capacity additions. In this case the capacity addition mix during the 1990s was held fixed at approximately half nuclear and half oil-fired, as a reasonable though arbitrary basis for optimizing the nuclear/oil addition mix during the 1980s. Starting with an all-oil expansion plan during the 1980s, nuclear plants were substituted for oil plants beginning with 1989 and working back toward 1980 a year at a time until a minimum-cost solution was obtained. This solution was considered to be "near-optimum"; it cannot be said to be the optimum since all possibilities for the 1980s could not be considered and since only two possible schedules of additions during the 1990s were considered. The two capacity addition schedules evaluated are those given in Table XV-2. The footnotes indicate which additions during the 1990s were assumed to be nuclear and which oil-fired.

- 92 - TABLE XV-2. ALTERNATIVE EXPANSION PLANS TO MEET GENERATION REQUIREMENTS - HIGH LOAD FORECAST CASE Reference expansion plan Alternative expansion plan Year Installed Reserve Installed Reserve Additionsa LOLP Additions21 LOLP capacity margin capacity margin (MW) (MW) (%) (%) (MW) (MW) (.1°) m 1977 - 1441 36 0.33 - 1441 36 0.33 b 1978 200b 1641 37 0.31 200 1641 37 0.31 b 1979 200b 1841 37 0.23 200 1841 37 0.23 1980 300 2141 41 0.18 300 2141 41 0.18 1981 300 2441 43 0.16 300 2441 43 0.16 1982 300 2 741 43 0.13 300 2741 43 0.13 1983 400 2 966 39 0.35 400 2 966 39 0.35 1984 400 3 366 40 0.28 400 3 366 40 0.28

1985 400 3 766 40 0.28 400 3 766 40 0.28 1986 600 4 366 45 0.25 600 4 366 45 0.25 1987 600 4 966 47 0.22 600 4 966 47 0.22 1988 600 5 566 48 0.23 600 5 566 48 0.23 1989 800 6 366 51 0.21 600 6166 46 0.28 1990 50 6416 38 0.62 600b 6 766 46 0.29 1991 1000C 7 416 45 0.42 600° 7 366 44 0.31 1992 1000b 8416 50 0.29 600b 7 966 42 0.37 1993 1000° 9416 52 0.26 600° 8 566 39 0.46 1994 1000b 10 394 53 0.27 2 X 600d 9 744 43 0.23 1995 1000c 11274 51 0.33 600b 10 224 37 0.46 d 1996 1000b 12154 48 0.44 2 X 600 11304 37 0.38 1997 1000° 13154 45 0.49 600° 11904 31 0.70 1998 1000b 14154 42 0.60 2 X 800d 13 504 36 0.40 1999 1000c 15 324 39 0.69 800b 14 234 30 0.81 4 X 50 50

23-yr average - 44 0.33 40 0.33

a The 50 MW units are gas turbines. The others are either oil-fired or nuclear except where footnotes indicate the choice was limited to one or the other. b Oil-fired. 0 Nuclear One oil-fired, one nuclear.

15. 2. Input data for WASP Modules 4, 5 and 6

(a) Module 4 input

WASP Module 4 generates alternative generation expansion plans for each year in the study period, within the input restraints imposed. These restraints are: (i) the minimum number of units of each size and type in the expansion alternative list (see Table XIV-2) required to be in the system, for each year; (ii) the maximum number of units of each size and type permitted to be considered, for each year; (iii) the minimum and maximum reserve margins, during the critical quarter, permitted to be considered, for each year.

- 93 - The restraints must not conflict with each other. For a predetermined type and size capacity addition schedule the minimum number required and the maximum number permitted are equal, and the minimum and maximum reserve margins are set such that the predetermined schedule is not excluded. For an all-thermal system as in Singapore, the critical quarter is the one with the highest peak demand. For the low forecast, having found by trial and error a capacity addition schedule considered acceptable from the reliability point of view (Table XV-1, large unit size case), the minimum and maximum reserve margins were set close together to force selection of a combination of units giving the desired total capacity. No choice was permitted for the oil- fired units added in 1978-79 and the gas-turbine units. For the other units any combination of oil or nuclear was permitted. For the high forecast eleven predetermined type-and-size addition schedules were prepared for each of the reference and the alternative expansion plans of Table XV-2. The eleven schedules correspond to introducing nuclear in each year from 1980 to 1990 inclusive. The 1990 introduction date corresponds to no nuclear additions during the 1980s. Each other case corresponded to all steam capacity additions being nuclear from the introduction date through 1989.

(b) Module 5 input

Module 5 simulates the system operation corresponding to each alternative generation expansion plan permitted by the restraints imposed on Module 4 for each quarter of each year in the study period. It uses the information from Modules 1 through 4 and requires, in addition, the loading order for all units in the fixed system and those under consideration as expansion alternatives, in order to calculate the fuel and operating and maintenance costs for each type and size of plant for each quarter. For units having a minimum operating level different from the maximum, which in the Singapore case includes all units except gas turbines, the loading order must be specified separately for the minimum-load capacity and the load-following capacity (the capacity between minimum and maximum operating levels). The loading order used was as follows, in order of perference:

(1) The minimum-load portions of nuclear capacity, in order of size, with largest units first. (2) The load-following portions of nuclear capacity, in the same order as (1). (3) The minimum load portions of the oil-fired capacity, in order of size, with largest units first. (4) The load-following portions of the oil-fired capacity, in the same order as (3). (5) Gas turbines, in order of size, with largest units first.

WASP Module 6 is the economic evaluation code in the package. For a pre-determined size- and-type schedule of capacity additions Module 6 merely calculates the value of the "objective function", which is the present worth of all capital, fuel, and operating and maintenance costs during the study period, less a credit for salvage value at the study horizon. If Module 4 has generated more than one permissible addition schedule in one or more years during the study period, Module 6 uses the forward dynamic programming method of selecting the lowest cost solution (i. e. the capacity addition schedule with the smallest value of the objective function) from among the alternatives available. Module 6 requires the following input information:

(1) Capital costs, domestic and foreign, and plant life for each of the generating units considered as expansion alternatives. See Section 13 for cost values and lives used. (2) Penalty factor, if any, to be assessed against foreign currency costs. In the Singapore case no penalty was assessed under reference conditions, and a 10% penalty was used in a sensitivity study. (3) Domestic and foreign escalation rates on capital costs. As explained in Section 9, these were assumed to escalate at the same rate as general price inflation, which corresponds to using 0%/yr in Module 6 under the Market Survey methodology adopted.

- 94 - (4) Escalation rates on domestic and foreign generation costs (fuel plus operating and maintenance). Operating and maintenance costs were taken to be all domestic currency costs, and were assumed to escalate at the same rate as general price inflation, i. e. 0%/yr under reference conditions. Fuel costs were taken to be all foreign currency costs. Under reference conditions fuel oil was assumed to escalate at 2%/yr faster than the rate of general price inflation. Sensitivity studies were made at 0% and 4% relative escalation rates. Nuclear fuel was taken to have a 0%/yr relative escalation rate under reference conditions. A sensitivity study was made for 2%/yr, for the low load forecast. (5) Present-worth discount rate. As mentioned in Section 9, the reference value chosen was 8%/yr (equivalent to about 12%/yr if the general inflation rate is 4%/yr). Values of 6% and 10% were used for sensitivity studies. (6) Critical loss-of-load probability. Module 6 will reject any alternative expansion plan having an annual loss-of-load probability higher than the critical value. For the Singapore studies the critical value selected was 0. 0082. (See Section 15.1.) (7) Present-worth base year and horizon year. For these studies all costs were discounted to 1 January 1973. The reference horizon was end of 1999. (8) Method of calculating salvage value at horizon. For the Singapore studies the reference salvage value was based on linear depreciation. Sensitivity studies were made using the sinking-fund depreciation method.

Other Module 6 input information (e. g. that applicable to hydro and pumped storage) is not pertinent to the Singapore studies.

- 95 - 16. RESULTS

16.1. Low load forecast

(a) Basis for analysis

As described in Section 10, the Aoki forecast of gross electricity generation at five-year intervals for Singapore from 1976-2001 was converted into a peak demand forecast by assuming an annual load factor of 65% (see Table X-l). A polynomial regression analysis was used to fit a smooth curve through the five-year forecast to give the annual peak demand forecast of Table X-2 to be used as input to WASP (see Appendix A). The derivation of quarterly peak demands and quarterly load duration curves is also given in Section 10. The study period selected was 1977-1999 inclusive. Assumptions regarding characteristics of the system existing at end 1976 and retirements during 1977-99 are given in Section 12, and characteristics of units considered as expansion alternatives are given in Section 14. As discussed in Section 15, the capacity addition schedule described in the first five columns of Table XV-1 was selected as the basis for an optimization study using the dynamic programming feature of WASP. The additions during 1977-79 were assumed to be oil-fired, and the few 50 MW gas-turbine additions were taken to be pre-determined. The remaining additions during 1980-1999 were permitted to be either oil-fired or nuclear, the selection being made by the dynamic programming method using the criterion of the lowest present- worth total of all costs during 1977-99 allowing a credit for salvage value at the end of 1999. The capacity addition schedule selected involves relatively large unit sizes, up to 20% of annual peak load during the 1980s decreasing to 13% during the 1990s. This large-unit policy is consistent with past practice in Singapore and tentative expansion plans through 1983, and is indicated by WASP studies to be lower cost than an alternative policy based on relatively smaller unit sizes and to have acceptably low values of loss-of-load probability and energy not served. It does imply, however, acceptance of load shedding on loss of the largest unit, as discussed in Sections 11 and 15.1 (a), if system frequency is to be maintained within limits normally acceptable. Other analytical assumptions are discussed in Sections 9, 14 and 15.

(b) Lowest-cost solution under reference conditions

The "optimum" expansion plan under reference conditions and within the restraints imposed on the choice of alternative expansion plans, as discussed in Section 15. 2, is given in Table XVT-1. Because of the restraints imposed to meet computer time limitations and calendar deadlines, the solution is better described as a "near-optimum", considered to give a reasonable approximation to the optimal mix of nuclear and oil-fired capacity additions duringthe 1980s. It is seen that no market for nuclear power plants in Singapore is indicated during the 1980s for the low forecast under the defined reference conditions, consisting of the reference economic parameters of Section 9. 3, ORCOST-3 capital costs (see Section 13), study horizon at end of 1999, and salvage values based on linear depreciation. As indicated in the sensitivity studies reported below, nuclear power is very close to being competitive in the late 1980s under these conditions, a change in the method of calculating salvage values being sufficient to cause two 400 MW nuclear units to be substituted for oil-fired units in 1987-88. The WASP dynamic program optimizes capacity additions during the 1990s as well as during the 1980s. Though the 1990s are beyond the scope of the Market Survey, and were included in the study period only to minimize possible end effects of using too short a time period, it is interesting to note in the bottom part of Table XVI-1 that nuclear units captured essentially the entire market for new capacity (5 x 600 MW) during the last study decade. Thus while 400 MW units were indicated to be barely competitive, if at all, in the late 1980s, 600 MW units are indicated to be strongly competitive in the 1990s.

- 96 - TABLE XVI-1. LOWEST COST SYSTEM CAPACITY EXPANSION SCHEDULE Reference conditions, modified Aoki load forecast, "large" unit sizesa

Installed capacity (MW) Annual loss- Conven- % Reserve b Year Retire- Gas -of-load Nuclear tional Total ments turbines probability steam

Total system 1977 1375 44 1419 28 0.0082

Additions Average Average 1977-1979 400 400 31 0.0058

Total system 1979 1775 44 1819 33 0.0035

1980c 200 200 35 0.0025

1981 50 50 27 0.0057

1982 300 300 36 0.0029

1983 25c 300 275 38 0.0017

1984 25 50 25 29 0.0043

1985 25 400 375 37 0.0024

1986 25 -25 27 0.0077

1987 25 400 375 34 0.0042

1988 25 400 375 40 0.0023

1989 32 0.0056

Total additions Average Average 1980-89 -150 2 000 100 1950 34 0.0039

Total system 3 775 1989 -150 144 3769 32 0.0056 3 625

Additions Average Average 1990-99 -385 e 3 000 50 2 665 40 0.0047

Total system 3 625 1999 3 000 -385 194 6434 35 0.0080 3 240

a "Large" units are defined to indicate size limits up to 20% of annual peak demand during the 1980s, gradually decreasing to by 1999. ° Critical quarter. c Additions and retirements each year. d Pasir Panjang A/1-6. e Pasir Panjang A/7, Pasir Panjang B, Jurong I/I.

Table XVI-2 indicates the range of results for the variety of sensitivity studies carried out. Six nuclear capacity addition schedules, labelled A through F, include the results of the reference case (A) and 12 sensitivity studies. The potential nuclear market is most sensitive to the oil cost escalation rate, amounting to 1400 MW duringthe 1980s, starting with 300 MW in 1983, at a 4% escalation

- 97 - TABLE XVI-2. POTENTIAL MARKET FOR NUCLEAR POWER PLANTS IN SINGAPORE - HIGH LOAD FORECAST (MW)

Sensitivity studies8 Year Reference conditions Discount rate Oil cost escalation rate ORCOST-1 6<7<>/yr 10%/yr •0%/yr 4<}'o/yr capital costs

1980-82 • i 1983 300 300 1984 300 1985 1986 1987. 400 400 400 1988 400 400 400 1989

Total 1980-89 - 800 1400 1100

Totalb 1990-99 5x 600 5x600 4x600 5 x 600 5 x 600

Schedule8 A B C D ' E F

The same nuclear plant addition schedules were obtained for additional sensitivity studies: Parameter varied Resulting schedule 10% foreign exchange cost penalty A Sinking fund salvage value (SFSV) instead of linear B 2% nuclear fuel cost escalation (NFCE) rate C 2% NFCE with &to discount rate A 2% NFCE with lCft discount rate D SFSV with lOfo discount rate A SFSV with e°/o discount rate F SFSV with 0<7o oil escalation rate D

The dynamic program optimized both the 1980s and the 1990s.

rate (schedule E). Using the lower capital cost differential between nuclear and oil-fired plants given by ORCOST-1 gave almost the same market (1100 MW, Schedule F). Using a 6% present-worth discount rate, or using sinking fund salvage values instead of linear, gave a market of 800 MW (Schedule B), while using them both at the same time increased the market back to 1100 MW (Schedule F). Using a 2% nuclear fuel cost escalation rate in combination with a 6% discount rate reduced the nuclear market back to zero. The 4% oil cost escalation rate, which resulted in the largest nuclear market among the sensitivity studies, is not necessarily to be thought of as an extreme, however, since the base oil cost is for fuel containing 3. 5% or more sulphur. Fuel with only 1. 5% sulphur costs about 40% more in Singapore. Thus the difference between 2% oil cost escalation rate (the reference condition) and 4% corresponds to a very gradual reduction in sulphur content from 3. 5 to 1. 5% over a 17-year period. It is well possible that environmental protection regulations will call for a more rapid rate of reduction than this and an ultimate sulphur content lower than this. The WASP dynamic program optimized the nuclear/oil capacity addition mix in the 1990s as well as the 1980s. Table XVI-2 shows the potential nuclear market during the 1990s to be 3000 MW (5 x 600 MW) for most cases. Increasing the discount rate to 10%, or increasing the nuclear fuel cost escalation rate to 2%, decreased the market to 2400 MW. Reducing the oil cost escalation rate to 0%, or using 2% nuclear fuel cost escalation rate in combination with 10% discount rate, reduced the nuclear market to zero.

- 98 - 16. 2. High load forecast

(a) Basis for analysis

As described in Section 10, the PUB forecast to 1990 was extended to the end of 1999 by assuming a 10%/yr growth rate during the last decade. The resulting annual peak load forecast is given in Table X-2. An annual load factor of 68. 2%, based on the average forecast by PUB for 1977-1981 (see Table V~l), was assumed for the whole study period, which was taken to be 1977-99 inclusive. The derivation of quarterly peak demands and quarterly load duration curves is also given in Section 10. Assumptions regarding characteristics of the system existing at end 1976 and retirements during 1977-99 are given in Section 12, and characteristics of units considered as expansion alternatives are given in Section 14. As discussed in Section 15, the capacity addition schedule described in the first five columns of Table XV-2 was selected as the reference expansion plan for the high forecast. As with the corresponding plan for the low forecast,it assumes addition of relatively large units, up to 20% of annual peak load during the 1980s, but decreasing to 9% by 1999. An alternative expansion plan, based on the 20% limit for units up to 600 MW and then continuing to use 600 MW units until 800 MW units can be introduced without violating the no-load- shedding criterion of Section 11, was also evaluated. Both expansion plans gave acceptably low values for loss-of-load probability and energy not served but, as mentioned also for the low forecast, the use of large units implies acceptance of load shedding on loss of largest unit if system frequency is to be maintained within normal limits. Other analytical assumptions are discussed in Sections 9, 13 and 15. As discussed in Section 15.1 (b), time did not permit using the WASP dynamic programming code to optimize the nuclear/oil capacity addition mix during the 1980s and 1990s, as was done for the low forecast. Instead, the size-and-type addition schedules for 1977-79 and 1990-99 were held fixed as shown in Table XV-2, and a limited number of alternative schedules for 1980-89 were evaluated to determine the minimum present-worth cost solutions over the range of economic parameter values studied. Table XVI-3 shows the schedules evaluated for the reference expansion plan, and Table XVI-4 shows the similar schedules for the alternative expansion plan. These two plans differ only in 1989-99, so their comparison amounts to a sensitivity study of effect in the 1980s of assumptions concerning the 1990s.

TABLE XVI-3. ALTERNATIVE NUCLEAR/OIL CAPACITY ADDITION SCHEDULES EVALUATED - HIGH LOAD FORECAST - REFERENCE EXPANSION PLANa

Schedule b Year 1988R 1987R 1986R 1985R 1984R 1983R 1982R

1980 O300 1981 O300 1982 O300 N300 1983 O400 N400 N400 1984 moo N400 N400 N400 1985 O400 N400 N400 N400 N400 1986 O600 N600 N600 N600 N400 N400 1987 O600 N600 N600 N600 N600 N600 N600 1988 N600 N600 N600 N600 N600 N600 N600 1989 N800 N800 N800 N800 N800 N800 N800

Total nuclear market (MW) 1400 2000 2600 3000 3400 3800 4100

a See Table XV-2 for 1977-79 and 1990-99. b The date indicates the year of introduction of nuclear power and the R refers to reference expansion plan. See Table XVI-4 for the similar A schedules for the alternative expansion plan. Before the introduction date the additions are all oil-fired of the same size as in the 1988R schedule. O = oil-fired; N = nuclear.

- 99 - TABLE XVI-4. ALTERNATIVE NUCLEAR/OIL CAPACITY ADDITION SCHEDULES EVALUATED - HIGH LOAD FORECAST - ALTERNATIVE EXPANSION PLANa

Scheduleb

1988A 1987A 1986A 1985A 1984A 1983A 1982A

1980 O300 1981 O300 1982 O300 N300 1983 0400 N400 N400 1984 O400 N400 N400 N400 1985 O400 N400 N400 N400 N400 1986 O600 N600 N600 N600 N600 N600 1987 O600 N600 N600 N600 N600 N600 N600 1988 N600 N600 N600 N600 N600 N600 N600 1989 N600 N600 N600 N600 N600 N600 N600

Total nuclear market (MW) 1200 1800 2400 2800 3200 3600 3900

See Table XV-2 fox 1977-79 and 1990-99. The date indicates the year of introduction of nuclear power and the A refers to alternative expansion plan. See Table XVI-3 for the similar R schedules for the reference expansion plan. Before the introduction date the additions are all oil-fired, of the same size as in the 1988A schedule. O = oil-fired; N = nuclear.

(b) Lowest cost solution under reference conditions

The near-optimum solution for the reference expansion plan under reference conditions is given in Table XVI-5. Because of the limited number of possible expansion plans considered it cannot be considered to be "the optimum", but it is felt that it gives a reasonable approximation to the potential market for nuclear power in the 1980s in Singapore if the high forecast is correct. It indicates a nuclear market of 2600 MW, beginning with a 600 MW unit in 1986, which amounts to 57. 5% of new capacity added during the 1980s. The reference conditions are the same as those defined in Section 16. 1 (b) for the low forecast. For the alternative expansion plan, the near-optimum solution under reference conditions is the same as for the reference plan except that a 600 MW nuclear plant is substituted for the 800 MW nuclear plant in 1989, reducing the nuclear market to 2400 MW.

Tables XVI-6 and XVI-7 show the values of the objective function calculated by WASP Module 6 for the schedules evaluated, under reference conditions and for various sensitivity studies involving changing the values of certain key economic parameters one at a time. The objective function is defined as the present worth, as of 1 January 1973, of all capital, fuel and operation and maintenance costs during 1977-99, less a salvage value at end 1999 calculated on the basis of linear depreciation. In each column the minimum value of the objective function indicated the near-optimum solutionfor the set of economic parameters assumed. The nuclear market corresponding to this near-optimum is indicated in the bottom line of the table. For the reference expansion plan, Table XVI-6, the nuclear market during the 1980s varies from zero (estimated value) to 3800 MW as the oil cost escalation rate varies from 0% to 4%. The 4% oil cost escalation rate is not necessarily an extreme case, as discussed for the low forecast. The market is relatively insensitive to the other parameters being varied. For the alternative expansion plan, Table XVI-7, the results are similar, but vary by

- 100 - TABLE XVI-5. LOWEST COST SYSTEM CAPACITY EXPANSION SCHEDULE Reference conditions, PUB load forecast, "large" unit sizesa

Installed capacity (MW) Annual loss- Conven- °/o Reserve -of-load Year Retire- Gas Nuclear tional Total probability ments turbines steam

Total system 1977 1375 1441 36 0.0033

Additions Average Average 1977-1979 400 400 37 0.0029

Total system 1979 1775 66 1841 37 0.0023

1980 ' 300 300 41 0.0018

1981 300 300 43 0.0016

1982 300 300 43 0.0013

1983 -175' 400 225 39 0.0035

1984 400 400 40 0.0028

1985 400 400 40 0.0028

1986 600 600 45 0.0025

1987 600 600 47 0.0022

1988 600 600 48 0.0023

1989 800 800 51 0.0021

Total additions Average Average 1980-89 -175 2 600 2100 4 525 44 0.0023

Total system 3 875 1989 2 600 -175 6 366 51 0.0021 3 700

Additions Average Average 1990-99 -382 d 5 000 4 000 250 8 868 46 0.0044

Total system 1999 7 600 7 340 294 15 234 39 0.0069

a "Large" units are defined to indicate size limits of 20<7o of annual peak demand during the 1980s, gradually decreasing to %"!o by 1999. Additions and retirements each year. c Pasir Panjang A. d St. James gas turbines, Pasir Panjang B, Jurong I/I. e Critical (fourth) quarter.

-200 MW to +600 MW compared to the equivalent conditions in Table XVI-6. This indicates that the optimum solution for the 1980s depends on the assumption made concerning additions during the 1990s, but is not extremely sensitive in this regard. It is interesting to note that every value of the objective function in Table XVI-6 is smaller than the corresponding value in Table XVI-7, implying that expansion with larger units is more economical, since the average loss-of-load probability is the same for the two plans (see Table XV-2). This is not strictly proven by the cases evaluated since the only valid comparison would be between the two truly optimum solutions for the two plans under

- 101 - TABLE XVI-6. OBJECTIVE FUNCTION VALUES FOR NUCLEAR/OIL CAPACITY ADDITION SCHEDULES OF TABLE XVI-3 - REFERENCE EXPANSION PLAN (106 US $)

Sensitivity studies

Schedule a Discount rate Oil escalation rate ORCOST-1 10<7o foreign Reference capital exchange conditions 6% in 0% 4% costs penalty

1982R 1201.2 1487.5 971.9 1165.6 1247.2 1126.5 1296.0 1983R 1183.6 1471.4 954.3 1140.1 1240.1 1117.7 1277.2 1984R 1171.2 1462.7 940.8 1117.3 1242.0 1114.3 1264.2 1985R 1162.7 1458.5 930.5 1097.9 1248.6 1113.8 1255.4 1986R 1157.3 1458.0 923.0 1081.3 1258.8 1115.7 1249.8 1987R 1161.7 1472.1 922.2 1069.3 1286.7 1128.7 1255.2 1988R 1168.7 1489.4 923.6 1059.8 1317.5 1143.3 1263.3

Nuclear b market (MW) 2600 2600 2000 _ c 3800 3000 2600

a See Table XVI-3. b Nuclear market during the 1980s corresponds to near-optimum solution. c No minimum was found for O°/o oil cost escalation rate, since the objective function was still decreasing at the latest nuclear introduction date studied. Nuclear market during the 1980s is no more than 1400 MW and probably zero for this condition.

TABLE XVI-7. OBJECTIVE FUNCTION VALUES FOR NUCLEAR/OIL CAPACITY ADDITION SCHEDULES OF TABLE XVI-4 - ALTERNATIVE EXPANSION PLAN (106 US $)

Sensitivity studies

Schedule a Discount rate Oil escalation rate ORCOST-1 10% foreign Reference capital exchange conditions 6<7o 10% 0% 4<7o costs penalty

1982A 1214.6 1509.7 980.0 1171.8 1271.6 1143.8 1310.5 1983A 1199.6 1497.5 964.1 1147.7 1268.8 1137.6 1294.6 1984A 1190.6 1493.7 952.8 1126.8 1276.3 1137.5 1285.3 1985A 1185.2 1494.3 944.7 1109.1 1288.3 1140.2 1280.0 1986A 1182.6 1498.0 939.1 1093.9 1303.3 1144.9 1277.5 1987A 1191.3 1518.5 941.2 1084.3 1338.3 1162.1 1287.5 1988A 1200.8 1539.7 944.3 1076.1 1373.4 1179.3 1298.4

Nuclear marketb (MW) 2400 3200 2400 c 3600 3200 2400

a See Table XVI-4. Nuclear market during the 1980s corresponds to near-optimum solution. c No minimum was found for Qffo oil cost escalation rate, since the objective function was still decreasing at the latest nuclear introduction date studied. Nuclear market during the 1980s is no more than 1200 MW and probably zero for this condition.

a given set of economic conditions, and Tables XVI-6 and XVI-7 indicate only near-optimum solutions in each case. For any given column in Table XVI-6 or XVI-7 the variation of the objective function with nuclear introduction date in the vicinity of the minimum appears to be quite "flat"; however, this, is somewhat misleading. In the first place it is the absolute value of the difference between the numbers in a given column which is important, not the relative difference, since all of the numbers in a column have the following costs in common: (i) Capital, fuel and operation and maintenance costs for 1977-82. (ii) Capital costs for 1988-99.

- 102 - These costs-in-common make up a large part of the total. In the second place, the costs have been discounted to 1 January 1973 at an 8% (or 6% or 10%) discount rate, which makes them appear smaller than they are by a factor of 2 to 4, depending on the year and discount rate.

16. 3. Treatment of interest during construction

As mentioned in Section 13. 1 (a), the amount of interest during construction was calculated at 8%/yr in ORCOST-1 and ORCOST-3, and was not varied when the discount rate was varied from 6% to 10%. While the interest rate paid on construction loans is not necessarily the same as the discount rate to be used in economic evaluation, it might be argued that they should be consistent. If the interest rate had been set equal to the discount rate in the sensitivity studies,the effect would have been in the direction of increasing somewhat the indicated sensitivity to discount rate.

16. 4. Financial considerations associated with the reference case expansion program

Capital costs of alternative generating units considered in the expansion plans were calculated by the ORCOST program described in Section 13 and in Appendix B. These capital costs were used by the WASP program in determining the objective functions of each expansion alternative. As a supplement to the basic analyses described above, it was decided to determine the year-by-year domestic and foreign cash requirements of the reference case expansion plan, as a guide to planners and financial institutions. In order to accomplish this, a computer program was written (cash-flow program). The input data required for the cash-flow program for each year of the study period and for each plant that became operational during that year are as follows. Plants were assumed to become operational on 1 January and capital costs were assumed to have been fully expended by the end of the preceding year. These assumptions are consistent with the WASP program.

(a) Plant construction schedule (same schedule, in years, that was used in the ORCOST calculations). The ORCOST-3 total plant capital costs (including interest during construction) are distributed over the construction period according to the expenditure-time schedules (S-curves) assumed in ORCOST. (b) Per cent of expenditure that was domestic (the foreign being 100 minus this value). (c) Capital cost, in US $/kW (same value as used in the WASP program; this value includes interest during construction). (d) Unit capacity, in MW.

The cash-flow program, using a 4th order polynomial approximation of the time- expenditure curve used in the ORCOST program, developed the year-by-year domestic and foreign expenditures associated with each plant. These values were printed in tabular form, together with the annual totals. It should be noted that nuclear plants were entered in two parts,(i) the cash requirements of the plant excluding the first (fuel) core, and (ii) the cash requirements of the first core. These first core requirements were calculated on the basis of 90% cash required during the year preceding operation, and 10% being required one year earlier. Table XVI-8 displays the domestic and foreign cash flows associated with capital investments for the near-optimum solutiontothe low forecast case under reference conditions, i.e. the all oil-fired expansion plan of Table XVI-1. The cash flows, representing the ORCOST-3 total plant capital cost including interest during construction, were distributed over the construction period according to the expenditure-time schedules assumed in ORCOST. The cash flows are given for each plant and for the total program. Only plants commissioned duringthe 1980s are included, so the total cash flows begin in 1977 and peak in 1986, though in fact they will continue to increase after that because of expenditures on plants to be commissioned in the 1990s. The total financing requirements for plants commissioned duringthe 1980samount to US $ 416 million (at 1 January 1973 price levels), about 70% being foreign currency financing.

- 103 - TABLE XVI-8. DOMESTIC AND FOREIGN CASH FLOWS ASSOCIATED WITH THE NEAR-OPTIMUM SOLUTION TO THE LOW FORECAST CASE

DOMESTIC CASH FLOW YEAR PLANT 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 TOTAL

1980 0200 .0 .0 .0 .7 7 .0 6.1 .0 0 .0 .0 .0 .0 .0 .0 .0 13.9 1981 GT50 .0 0 .0 .0 .0 .0 .0 0 .0 .0 .0 .0 0 .0 .0 .0 1982 0300 .0 .0 .0 .0 .0 1.9 9.7 7 1 .0 .0 .0 .0 0 .0 .0 18.8 1983 0300 .0 .0 .0 .0 .0 .0 1.9 9 .7 7.1 .0 .0 .0 0 .0 .0 18.8 1984 GT50 .0 .0 .0 .0 .0 .0 .0 0 .0 .0 . 0 .0 0 .0 .0 .0 1985 0400 .0 .0 .0 .0 .0 .0 .0 1 3.1 11.9 7.9 .0 0 .0 .0 23.1 1986 No Plants Added 1987 0400 .0 0 .0 .0 .0 .0 .0 0 .0 .1 3.1 11.9 7 9 .0 .0 23.1 1988 0400 .0 0 .0 .0 .0 .0 .0 0 .0 .0 .1 3.1 11 ,9 7.9 .0 23.1 1989 No Plants Added - 10 4 Domestic total .0 .0 .0 . 7 7 .0 8.0 11.6 17 .0 10.3 12.0 11.2 15.1 19.8 7.9 .0 121.0 i FOREIGN CASH FLOW YEAR PLANT 1974 19751976 1977 19 78 1979 1930 1931 1982 1983 1984 1985 1986 1987 1988 TOTAL

1980 0200 .0 .0 .0 1.8 16 .4 14.3 .0 .0 .0 .0 .0 .0 .0 .0 .0 32.6 1981 GT50 .0 .0 .0 .0 .0 .6 5.6 .0 .0 .0 .0 .0 .0 .0 .0 6.2 1982 0300 .0 .0 .0 .0 .0 4.4 22.6 16 .7 .0 .0 .0 .0 .0 .0 .0 43.8 1983 0300 .0 .0 .0 .0 .0 .0 4.4 22 .6 16.7 .0 .0 .0 .0 .0 .0 43.8 1984 GrT^O .0 .0 .0 .0 .0 .0 .0 .0 .6 5.6 .0 .0 .0 .0 .0 6.2 1985 0460 .0 .0 .0 .0 .0 .0 .0 .2 7.3 27.8 18.4 .0 .0 .0 .0 54.0 1986 No Plants Added 1987 04OO .0 .0 .0 .0 .0 .0 .0 .0 .0 .2 7.3 27.8 18.4 .0 .0 54.0 1988 0400 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .2 7.3 27.8 18.4 .0 54.0 1989 No Plants Added

Foreigr1 total .0 .0 .0 1.8 16 .4 19.4 32.7 39 .7 24.8 33.7 26.1 35.2 46.3 18.4 .0 295.0 TABLE XVI-9. DOMESTIC AND FOREIGN CASH FLOWS ASSOCIATED WITH THE NEAR-OPTIMUM SOLUTION TO THE HIGH FORECAST CASE

DOMESTIC CASH FLOW YEAR PLANT 1974 1975 1976 19 77 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 TOTAL

1980 0300 .0 .0 .0 1 .9 9.7 7.1 .0 .0 .0 .0 .0 .0 .0 .0 .0 18.8 1981 0300 .0 .0 .0 .0 1.9 9.7 7.1 .0 .0 .0 .0 .0 .0 .0 .0 18.8 1982 0300 .0 .0 .0 .0 .0 1.9 9.7 7.1 .0 .0 .0 .0 .0 .0 .0 18.8 1983 0400 .0 .0 .0 .0 .0 .1 3.1 11.9 7.9 .0 .0 .0 .0 .0 .0 23.1 1984 O400 .0 .0 .0 .0 .0 .0 .1 3.1 11.9 7.9 .0 .0 .0 .0 .0 23.1 1985 0400 .0 .0 .0 .0 .0 .0 .0 .1 3.1 11.9 7.9 .0 .0 .0 .0 23.1 1986 N600 .0 .0 .0 .0 .0 .0 .2 1.9 7.2 12.3 12.7 4.4 .0 .0 .0 39.0 Fuel .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 1987 N600 .0 .0 .0 .0 .0 .0 .0 .2 1.9 7.2 12.3 12.7 4.4 .0 .0 39.0 Fuel .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 1988 N600 .0 .0 .0 .0 .0 .0 .0 .0 .2 1.9 7.2 12.3 12.7 4.4 .0 39.0 Fuel .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 1989 N800 .0 .0 .0 .0 .0 .0 .0 .0 .0 .4 3.1 9.2 14.3 14.2 4.6 46.0 Fuel .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0

1 Domestic total .0 .0 .0 1..9 11.6 18.9 20.4 24.6 32.4 41.8 43.4 38.7 31.5 18.6 4.6 288.9 I-J o 1 FOREIGN CASH1 FLOW YEAR PLANT 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 TOTAL

1980 0300 .0 .0 .0 4,.4 22.6 16.7 .0 .0 .0 .0 .0 .0 .0 .0 .0 43.8 1981 0300 .0 .0 .0 .0 4.4 22.6 16.7 .0 .0 .0 .0 .0 .0 .0 .0 43.8 1982 0300 .0 .0 .0 *.0 .0 4.4 22.6 16.7 .0 .0 .0 .0 .0 .0 .0 43.8 .0 1983 0400 .0 .0 .0 i,0 .0 .2 7.3 27.8 18.4 .0 .0 .0 .0 .0 54.0 1984 0400 .0 .0 .0 .0 .0 .0 .2 7.3 27.8 18.4 .0 .0 .0 .0 .0 54.0 1985 0400 .0 .0 .0 .0 .0 .0 .0 .2 7.3 27.8 18.4 .0 .0 .0 .0 54.0 1986 N600 .0 .0 .0 i,0 .0 .0 .9 7.8 29.0 49.3 51.1 17.6 .0 .0 .0 156.0 Fuel .0 .0 .0 i,0 .0 .0 .0 .0 .0 .0 1.-7 15.0 .0 .0 .0 16.8 1987 N600 .0 .0 .0 4.0 .0 .0 .0 .9 7.8 29.0 49.3 51.1 17.6 .0 .0 156.0 Fuel .0 .0 .0 1.0 .0 .0 .0 .0 .0 .0 .0 1.7 15.0 .0 .0 16.8 1988 N600 .0 .0 .0 4.0 .0 .0 .0 .0 .9 7.8 29.0 49.3 51.1 17.6 .0 156.0 Fuel .0 .0 .0 1.0 .0 .0 .0 .0 .0 .0 .0 .0 1.7 15.0 .0 16.8 1989 N800 .0 .0 .0 ,0 .0 .0 .0 .0 .0 1.7 12.6 36.8 57.4 56.8 18.7 184.3 Fuel .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 1.9 17.2 19.1

Foreign total .0 .0 .0 4 .4 27.1 44.1 48.0 61.1 91.5 134.3 162.3 171.8 143.0 91.5 36.0 1015.7 Total -- nuclear fuel - .0 .0 ,0 .0 .0 .0 .0 .0 .0 1.7 16.7 16.7 16.9 17.2 69.2 only Table XVI-9 shows the corresponding domestic and foreign cash flows associated with the capital investments for the near-optimum solution to the high forecast case under reference conditions, i.e. the oil-fired and nuclear expansion plan of Table XVI-5. In this case the fuel cycle working capital requirements for each nuclear plant are also shown. Though individual fuel purchases are normally financed over short terms, e.g. three to five years, there is in fact a substantial investment outstanding in fuel over the life of the plant. Though the fuel capital investments used in the WASP economic evaluation are the present-worth levelized average investment over plant life, they may be used as an approximation to the cost of the first core and in Table XVI-9 they have been distributed over the two years preceding commissioning more or less according to the payment schedule for the first core (see Appendix J). The cash flows begin in 1977 and peak in 1985, one year sooner than in the low forecast because of the longer construction period for nuclear plants. The total requirements are US $ 1305 million, about 78% being foreign financing. This total is about three times that for the low forecast, partly because 4700 MW are added instead of 2100 MW and partly because 2600 MW of the 4700 MW are capital-intensive nuclear plants.

- 106 - 17. SUMMARY AND CONCLUSIONS

17.1. Basic conditions

Table XVII-1 summarizes the conditions used in analyses of various generation expansion plans for the Singapore electricity system during the decade 1980-89. As shown in this table, two load forecasts were considered. The low load forecast was developed by a Market Survey expert using the method described in Appendix F. The high load forecast is that of the Singapore Public Utilities Board to 1990, extrapolated to 2000. From the figures given in Table XVII-1, it is seen that there is a total market of about 2100 MW of thermal capacity additions during the study decade for the low load forecast and of about 4700 MW for the high load forecast. The low load forecast was studied using the dynamic programming optimization feature of WASP, though it was necessary to impose some restraints on the range of possible solutions in order to keep within computer time and calendar time limits. The high load forecast was studied by considering a limited number of alternative expansion plans involving a fixed schedule of capacity additions, because of time limitations. In this case during the 1990s the capacity addition mix was held fixed at approximately half nuclear and half oil-fired as a reasonable though arbitrary basis for optimizing the mix of the nuclear/ oil additions during the 1980s. Starting with an all-oil expansion plan for the 1980s, nuclear plants were substituted for oil-fired plants beginning with 1989 and working back toward 1980 a year at a time until a minimum cost solution was obtained. This was considered to be a near-optimum solution; obviously it cannot be said to be the optimum since all possibilities for the 1980s could not be considered and since only two possible schedules of additions during the 1990s were considered, the second one as a sensitivity study.

TABLE XVII-1. SUMMARY OF CONDITIONS USED IN THE ANALYSES

Low forecast High forecast

1979 1989 1979 1989

e Population (106f 2.38 2.76 n. a. n.a. GNP/oapita (1964 US $) a 1580 2560 n. a. n. a. b Electric energy production kWh/yr/capita 3266 5910 n.a. n.a. GWh/yr total 7772 16313 8035 25235 Peak demand (MW) 1365 2865 1345 4224 c Total installed capacity (MW) 1819 3769 1841 6366 New capacity added, 1980-89, inclusive (MW) 2100 4700 Average reserve margin, 1980-89 (fy) 34 44 Average annual loss-of-load probability, 1980-89 fify 0.39 0.23

a These population and GNP/capita forecasts were the basis of the low energy production forecast (by Aoki: see Appendix F). The corresponding information for the high forecast is not applicable, b Annual load factor 65<7o for low forecast, 68,2% for high forecast. c All thermal (oil-fired, nuclear, or gas turbine), d Retirements during the period account for the difference between new capacity added and increase in total installed capacity. e n. a. - not available.

- 107 - 17.2. Economic basis

The economic merit of the various alternatives was measured by an objective function representing the present worth (to 1 January 1973) of all costs (expressed in US $ at 1 January 1973 price levels) associated with the operation of all units in the system during 1977-1999 and with the construction of units added to the system during this period (taking a credit for their salvage value at the end of 1999). External or social costs were disregarded, as were taxes and restraints on foreign currency expenditure. The study period was extended through 1999 in order that at least 10 years of operating history of all plants added during the 1980s would be considered. In the low load forecast the dynamic programming optimization method gave optimum solutions (withinthe constraints imposed) for both the 1980s and the 1990s. In the high load forecast the capacity additions during the 1990s were held constant and therefore their capital costs contributed a constant amount to the objective function, so that the changes in the objective function were caused by changes in additions during the 1980s. Similarly, operation and additions during 1977-79 were the same for all cases and hence did not affect the optimization during the 1980s. The data used as a basis for the economic evaluations are summarized in Table XVII-2. The capital costs were derived for construction conditions in Singapore, as described in Appendix B. The availabilities, heat rates, and fuel costs are based on Appendixes E, H, I and J. The operation and maintenance costs are based on Appendix E with adjustment for local conditions.

TABLE XVII-2. TECHNICAL AND ECONOMIC DATA USED AS BASIS FOR ANALYSIS

d Plant type Capital Availability Heat rates (kcal/kWh) e 3 Fuel cost O & M cost^ and size* costc (W (USj

Oil-fired

150 253 87 2347 2270 140 0.24 200 233 87 2280 2213 140 0.19 300 209 86 2335 2259 140 0.15 400 193 83 2324 2211 140 0.13 600 172 81 2328 2250 140 0.10 800 156 79 2334 2252 140 0.09 1000 148 79 2344 2248 140 0.09

Nuclear

200 573 87 2648 2504 59 0.43 300 479 86 2645 2503 58 0.32 400 420 83 2643 2503 57 0.26 600 353 81 2637 2501 55 0.20 800 312 79 2632 2500 53 0.17 1000 288 79 2627 2500 51 0.14

Gas turbine

50 125 97 n. a. 3528 229 0.20

Costs given at constant 1 January 1973 price levels. Plant life 30 years for oil-fired and nuclear plants, 20 years for gas turbine. c Reference values (ORCOST-3). Sensitivity study made for one alternative set of values (ORCOST-1). Nuclear plant capital costs shown include levelized fuel cycle working capital. Indicates combined effect of scheduled and forced outages. The WASP simulation code schedules maintenance by quarters, and treats forced outages on a probabilistic basis. e Only variable nuclear fuel cost (fixed cost added to capital costs). Treated as fixed cost, for simplicity, though the computer code can handle them as a combination of fixed and variable costs.

- 108 - 17.3. Summary of cases considered

For both load forecast cases the reference schedules assumed a continuation of past and projected local practice of adding relatively large-size units, i.e. up to about 20% of annual peak load. Preliminary studies with the WASP computer codes indicated that expansion with these relatively large units gave a lower cost solution than for smaller units, for the same loss-of-load probability. The system stability analysis in Section 11 indicates, however, that the use of larger units implies acceptance of load shedding on loss of the largest unit if frequency is to be maintained within the limits normally acceptable. The WASP probabilistic simulation code calculates loss-of-load probability and the probable amount of energy not served, both of which were indicated to be acceptably low. The energy not served measures the product of the frequency and severity of load shedding, but does not indicate how severe an individual occurrence might be. The balancing of cost against risk must be left to the local authorities. A change to a relatively smaller unit size limit would lead to a smaller total market, because less reserve margin would be required for a given loss-of-load probability, and to a smaller nuclear fraction of that market, because nuclear units are less competitive in smaller sizes. The reference expansion plans for the two forecasts are summarized in Table XVII-3.

TABLE XVII-3. REFERENCE EXPANSION PLANS Low forecast High forecast

Year Capacity Type of Loss-of-load Capacity Type of Loss-of-load added capacity probability added capacity8 probability (MW) <*) (MW) (W 1977 - - 0.82 - - 0.33 1978 200 Oil 0.57 200 Oil 0.31 1979 200 Oil 0.35 200 Oil 0 23 1980 200 Oil/nucl. 0.25 300 Oil/nucl. 0.18 1981 50 Gas turbine 0.57 300 Oil/nucl. 0.16 1982 300 Oil/nucl. 0.29 300 Oil/nucl. 0.13 1983 300 Oil/nucl. 0.17 400 Oil/nucl. 0.35 1984 50 Gas turbine 0.43 400 Oil/nucl, 0.28 1985 400 Oil/nucl. 0.24 400 Oil/nucl. 0.28 1986 - 0.77 600 Oil/nucl, 0.25 1987 400 Oil/nucl. 0.42 600 Oil/nucl. 0.22 1988 400 Oil/nucl. 0.23 600 Oil/nucl. 0.23 1989 - 0.56 800 Oil/nucl. 0.21 1990 600 Oil/nucl. 0.23 50 Gas turbine 0.62 1991 - 0.49 1000 Nuclear 0.42 1992 600 Oil/nucl. 0.22 1000 Oil 0.29 1993 - 0.46 1000 Nuclear 0.26 1994 50 Gas turbine 0.73 1000 Oil 0.27 1995 600 Oil/nucl. 0.48 1000 Nuclear 0.33 1996 600 Oil/nucl. 0.35 1000 Oil 0.44 1997 - 0.63 1000 Nuclear 0.49 1998 600 Oil/nucl. 0.33 1000 Oil 0.60 1999 - 0.80 1250 Nuclear & 0.69 Gas turbine

Totals 5550 14400

Oil/nucl. indicates oil or nuclear.

- 109 - 17.4. Summary of sensitivity studies

Table XVII-4 gives the reference values, assumed for the major economic parameters, together with the alternative values used in studies to determine how sensitive the near- optimum solution was to the values chosen for these parameters. For the sensitivity studies the parameters were varied only one at a time. The results of the sensitivity studies are summarized in Section 17.5.

TABLE XVII-4. VALUES ASSUMED FOR ECONOMIC PARAMETERS8 (%/yr)

Reference case Sensitivity studies Item Study Equivalent Study Equivalent values real values values real values

Present-worth discount rate 8 12 6 & 10 10 & 14 Capital cost escalation rate 0 4 - - Escalation rate for oil and gas turbine fuel 2 6 0 &4 4 &8 0 Escalation rate for nuclear fuel 0 4 2 6

a Reference conditions are based on study horizon at end of 1999, and credit for salvage values at horizon is based on linear depreciation. For the low forecast, sensitivity studies were also made using the sinking fund depreciation salvage value. Assuming that the general price inflation rate is about 4%/yr, c Sensitivity study for low forecast only.

17.5. Potential 1980-1989 nuclear power market

The potential market for nuclear power plant additions in Singapore during the 1980s is summarized in Tables XVI-2 and XVII-5 for the low and high load forecasts, respectively. It is seen that for the low load forecast there is no nuclear market indicated for the 1980s under reference conditions. The potential market is most sensitive to the oil cost escalation rate, amounting to 1400 MW during the 1980s, starting with 300 MW in 1983, at a 4% escalation rate. Using the lower capital cost differential between nuclear and oil- fired plants given by ORCOST-1 gave almost the same result as a 4% oil cost escalation rate. The 4% oil cost escalation rate is not necessarily to be considered an extreme, however, since the base oil cost is for fuel containing 3.5% or more sulphur. Fuel with only 1.5% sulphur costs about 40% more in Singapore. Thus, the difference between 2% oil cost escalation rate (the reference condition) and 4% corresponds to a very gradual reduction in sulphur content from 3.5% to 1.5% over a 17-year period. It is well possible that environmental protection regulations will call for a more rapid rate of reduction than this and an ultimate sulphur content lower than this. The dynamic programming method used in the low forecast optimized the nuclear/oil addition mix in the 1990s as well as in the 1980s. Table XVI-2 shows that under reference conditions the potential nuclear market in the 1990s is five 600 MW units, which is essentially all of the new capacity additions. This same result was obtained for 6% discount rate, for 4% oil escalation rate and for ORCOST-1 capital costs. Increasing the discount rate to 10% reduced the market to four 600 MW units, and decreasing the oil cost escalation rate to 0% reduced the nuclear market to zero. For the high load forecast, the indicated nuclear power market during the 1980s is 2600 MW for reference conditions, and varies from zero MW at 0% oil cost escalation rate to 3800 MW at a 4% rate. The market during the 1990s is not shown as the mix of the nuclear/oil additions during that decade was held constant as shown in Table XVII-3. As mentioned in Section 16, a similar sensitivity study assuming expansion in the 1990s with 600 MW and 800 MW units instead of 1000 MW units gave somewhat higher cost solutions, but the near-optimum nuclear addition schedules in the 1980s were not very different from those of Table XVII-5.

- 110 - TABLE XVII-5. POTENTIAL MARKET FOR NUCLEAR POWER PLANTS IN SINGAPORE HIGH LOAD FORECAST (MW)

Year Sensitivity studies Reference conditions* Discount Oil cost escalation rate ORCOST-1 rate Ofo/yr 4%/yr capital costs 10<7o/yr

1980-82 - - - 1983 - - 400 - 1984 - - 400 - 1985 - - 400 400 1986 600 - 600 600 1987 600 600 . 600 600 1988 600 600 600 600 1989 800 800 800 800

Total 1980-89 2600 2000 3800 3000

The results in this column were obtained also for a 6% discount rate and for a lCf/o penalty on foreign exchange costs.

BIBLIOGRAPHY

Public Utilities Board Singapore (PUB), Annual Reports 1970 and 1971. Singapore Facts and Pictures, Ministry of Culture (1972). Yearbook of Statistics, 1971/72 (1972). Singapore: Economic Pattern in the Seventies, Ministry of Culture (1972). Desalination — A Study of the Methods of Desalination with Particular Reference to a Dual-Purpose Plant for Electric Power Generation and Water Desalination, Electricity Department, Dec. 1967. Report on the possible application of nuclear energy in Singapore, Science Council, Singapore (1968). Major Development Scheme, Report on Future Power Generation Requirements, Rep. No. 01/70, Planning and Design Section, Electricity Department, PUB (1970). Review of System Planning, 1970-1980, Montreal Engineering Co. Ltd.

- Ill -

APPENDIXES

A. WIEN AUTOMATIC SYSTEM PLANNING PACKAGE (WASP) B. GENERATING PLANT CAPITAL COSTS (ORCOST) C. LOAD DESCRIPTION DATA FOR WASP PROGRAM D. ECONOMIC METHODOLOGY AND PARAMETERS E. STANDARDIZED DATA FOR GENERATING UNITS CONSIDERED AS EXPANSION ALTERNATIVES F. LONG-RANGE FORECASTING OF THE DEMAND FOR ELECTRICAL ENERGY G. BASIS FOR HEAT RATE DETERMINATION H. GENERALIZED POWER SYSTEM ANALYSIS APPROACH TO DETERMINE SYSTEM LIMITATIONS I. FUTURE FOSSIL FUEL PRICES J. NUCLEAR FUEL CYCLE COST TREATMENT K. SENSITIVITY STUDIES L. IAEA SERVICES AND ASSISTANCE IN CONNECTION WITH NUCLEAR POWER M. ABBREVIATIONS USED IN THE MARKET SURVEY REPORTS N. COOPERATING ORGANIZATIONS AND PARTICIPANTS IN THE MARKET SURVEY MISSION

- 113 -

APPENDIX A

WIEN AUTOMATIC SYSTEM PLANNING PACKAGE (WASP)

R. Taber Jenkins*

INTRODUCTION The WASP package is a series of six computer codes which include capabilities especially developed for the needs of the IAEA Market Survey. At the same time, it is a second generation of an earlier power system planning program, developed by and for the Tennessee Valley Authority in the United States of America. The package is designed to find the "optimum" power system expansion plan within established constraints. By optimum is meant that the discounted cash flow (capital and operating expense) is minimized over the study period with provision made to reduce effects of uncertainties beyond that period. Until recent years the choice of generating equipment available to an electric utility was fairly limited. In many cases only one fuel could be considered and it was only necessary to determine the appropriate unit size. The major questions to be resolved were, firstly, the extent to which it was sensible to increase the unit size in order to benefit from the economy of scale at the expense of early investment and of possible system operating problems and, secondly, how much should be spent to reduce heat rates. The traditional method of solution was for the system planner to assume two or three possible expansion plans and to determine their present-worth values either by hand calculations, or, more recently, with computer assistance, but with the planner intervening at various stages of the calculation. Such solutions required many hours of engineer's time in spite of the fact that the range of cases studied was extremely limited. The choice of generating equipment is now much wider and includes nuclear units, gas turbines, combined cycle, quick start intermediate fossil fuel units and pumped storage stations. Dynamic programming, in its most general sense, is an ideal method for solving the system planning problem. However, even with a limited range of possible expansion plans this method of solution was impractical without the aid of a computer. With the additional range of units now available the number of possible expansion plans is so large that even with the aid of computers general linear programming is impractical. The WASP package attempts to tread the ground between the two extremes. The system planner is given the facility to direct the area of study to configurations which he believes most economic, but the program will tell him if his restrictions were a constraint on the solution. The WASP program then permits him to modify his constraints and, without re- peating all the previous computational effort, to determine the effect of the modification. This process can be repeated until an optimum path conforming with the user-imposed constraints is determined. The WASP package consists of six modular programs which may be operated sequentially in a single run, or may be operated individually. The six modules are:

(1) a program to describe the forecast peak loads and load duration curves for the system; (2) a program to describe the existing power system and all future additions which are firmly scheduled; (3) a program to describe the alternative plants which could be used to expand the power system; (4) a program to generate alternative expansion configurations; (5) a program to determine if a particular configuration has been simulated and, if not, to simulate operation with that configuration; and (6) a program to determine the optimum schedule for adding new units to the system over the time period of interest.

Tennessee Valley Authority, United States of America,

A-l Module 1 Module 2 Module 4 LOAD PROGRAM FIXED SYSTEM PROGRAM EXPANSION CONFIGURATION PROGRAM

Read Data Describing Load Read rules defining expansion Read data on fixed generation system, Duration Curves Seasonally for configuration generation including including arty firmly scheduled Duration of Study 1. Reserve Range. additions. 2. Minimum No. of units of each Requirements Include: expansion alternative to be 1. Thermal plant: name, capacity, considered each year. number of units, outage rate, heat rate, and fuel price. 3. Additional No. of units of each expansion alternative 2. Hydro: basic capacity, to be considered each yeai. energy and seasonal variations. 3. Pumped storage, capacity, LOAD efficiency and limiting energy. FILE

From Fixed System Data, Load Data and Minimum Installations, Calculate total operational costs determine season of critical capacity for each plant (S/MWh) Module 3 EXPANSION ALTERNATIVE PROGRAM

Read Data on Candidate Plants From expansion alternatives for system expansion similar to Define calendar time of plant find all expansion configurations data on existing system additions and / or plant retirements which meet rules

FIXED SYSTEM FILE

Module 5 Module 6 MERGE AND SIMULATION PROGRAM OPTIMIZATION PROGRAM

Read Capital Costs of Plants, Plant Life, Interest Rates, Critical Loss of Load Probability, Escalation Rates, Base Year for Financial Discounting, No. of years in study

PREVIOUS OPERATING COST & RE- LIABILITY FILE

For each previous year's reliable Compute Discounted Capital Cost of configuration from which this con- all construction for this configuration figuration is accessible, compute this year. Calculate discounted end of total discounted cash flow by adding study value of all construction and to all discounted cash flow associated apply as credit. Compute Discounted FROM DATA ON LOAD, FIXED with previous slate, capital costs Cost of System Operation. Combine SYSTEM and EXPANSION FILE (corrected for end of study value) to form first year's contribution to simulate and calculate cost and associated with transfer from that stale objective function. reliability of expansion configuratioi to this state augmented by discounted operating costs of this state this year. Preserve as the suboptimum objective function, the lowest value of total discounted cash flow and the previous WRITE DATA on current system stale as the path associated with the operating cost & reliability file suboptimum. 1 - , ~

Increment Configuration

CURRENT OPERATING COST & RE- Print Optimum & Near Optimum Paths. LIABILITY Indicate any configurations limited FILE by minimum and maximum rules.

FIG. A-l. WASP PROGRAM FLOW SHEET.

A-2 Each of the first three programs creates data files which are used in the remaining programs. Additional files are created by the fourth and fifth program and are used in the sixth. Each program produces a printed summary. Figure A-l shows a flow chart of this program. An immediate advantage of the modular program approach is that the first three programs (loads, existing system, expansion alternatives) can be run separately and in parallel to eliminate the bulk of the data errors. These programs are very fast to run, thus avoiding extensive long runs with incorrect data. The separation of the expansion configuration generator from the simulation produces further savings in computer time by permitting elimination of a large number of expansion configurations from being simulated when data errors are made in defining configurations to be considered. The ability to save simulation results on a data file is the major time-saving feature of the program. While searching through successive re-runs of the last three programs for the unconstrained optimum, only those simulations which have not been performed are executed. Since simulation is the most time-consuming part of examining an expansion configuration, the computation time saved can be very large. The program permits consideration of up to 20 alternative generating units (size, fuel, heat rate etc.). In addition to thermal units, hydro and pumped-storage units can be included in the list of alternatives. If a series of hydro or pumped-storage projects are to be considered by the program, projects of each type must be identified in the chronological order in which they would be installed in the system. Up to 20 such projects may be included in the list. When hydro or pumped-storage units are added to the system, they are merged with existing hydro or pumped-storage units. Therefore, all of the hydro projects count as only one alternative and all of the pumped-storage projects count as an additional alternative. The expansion configurations to be chosen for simulation in any year are controlled by three factors:

(i) The configuration must satisfy the specified minimum and maximum reserve margin, (ii) The choices must lie within minimum and maximum constraints (tunnels) specified by the user, (iii) They must be accessible from at least one of the previous years' alternatives.

The logic of modules 5 and 6 is broken into three general areas: firstly, the simulation of the power system operation which makes use of a probabilistic simulation method which has generated much interest in recent years; secondly, the handling of financial cash flows and their effects on the function to be minimized; thirdly, the actual optimization procedure utilizing a dynamic programming algorithm. These three aspects and their handling in the program are described briefly below. More complete information is available from the references and textbooks.

Simulation

The purpose of the simulation is to provide an estimate of production costs associated with a given system configuration. This is the most time-consuming part of the program. The program permits the years to be broken into as many as 12 periods each of which may have its own peak load, load shape, hydro operating characteristics and maintenance schedule. The running time of the simulation is directly proportional to the number of periods chosen. Consequently, for the purposes of the Survey, the year was divided into four periods or seasons. On the basis of seasonal peak loads and seasonal capacity variations caused by hydro conditions, a heuristic method is used to develop a "reasonable" distribution of maintenance among the seasons. By 'reasonable' is meant that maintenance on the largest units will be in that season which has the greatest difference between installed capacity and peak load, while maintenance on smaller units is distributed in those seasons having less excess capacity. Having decided in which season maintenance on a particular unit will occur, the actual maintenance within the season is randomly distributed. The heart of the simulation is the algorithm which distributes the energy among the units on the system. It is an extension of the old load duration curve method which rigorously accounts for random outages of thermal units and has the effect of causing units higher on the loading order to supply more energy at a higher unit price than would otherwise be

A-3 HYDRO PEAKING RATING PUMPED HYDRO GENERATING PUMPED HYDRO GENERATOR RATING

HYDRO PEAKING GAS TURBINES AND ANTIQUE FOSSIL

(NOTE-IGNORES SPINNING RESERVE AND FORCED OUTAGES)

LOAD-FOLLOWING FOSSIL

Q < o VARIABLE FOSSIL

PUMP RATING (MW)'

PUMPED HYDRO (PUMPING)'

BASE FOSSIL

NUCLEAR

BASE HYDRO-RUN-OF-RIVER

%OF TIME FIG. A-2. IDEALIZED PLACING OF VARIOUS TYPES OF STATION UNDER THE LOAD DURATION CURVE. experienced. Figure A-2 illustrates the idealized placement of various capacity types under a typical load duration curve. The above procedure is illustrated by the simple diagrams shown in Fig. A-3. Figure A-3(a) shows a load duration curve with ten thermal units "stacked" under the load curve. As long as all units are running, units 1-4 run 100% of the time; units 5-9 run part of the time; and unit 10 does not run at all. However, if a unit fails, for example unit 1, unit 2 assumes the position of unit 1; 3 the position of 2; and so on. The same effect can be achieved by raising the load curve by the capacity of unit 1, as shown in Fig.A-3(b), in which case units 5 to 9 inclusive have their energy requirements increased and unit 10, which formerly did not generate at all, is carrying significant load. If it is assumed that outages of unit 1 are random, and occur x% of the time, then (100 - x)% of the time the system operates like Fig.A-3(a) and x% of the time like Fig.A-3(b). Therefore, a resultant "expected" load curve (called the equivalent load) which is shown as the solid line in Fig. A-3(c) can be computed. An algorithm computes the resultant equivalent load curve recursively as one considers all of the units in the merit order of their loading. Figure A-4 shows the resultant equivalent load curve after all the plants have been considered. If the total system generating capacity is plotted on the ordinate, the corresponding value on the abscissa, p*, represents thepercent of time the equivalent load exceeds the system generating capacity. In other words, the value p* represents thepercent of time that the system cannot meet the expected load. The probability of not meeting the load is simply p* 100.

A-4 (a) UNIT NO. (b) UNIT NO. (c)

10 \ 10

\ 9 \ V V \^ 8 \^^ 7 ^\ 7

Q ^\ 6 o \ • \ » 5 4 U

3 3

2 2

1 1

TIME TIME %TIME FIG. A-3. ILLUSTRATION OF THE METHOD OF STACKING THERMAL UNITS UNDER THE LOAD CURVE.

TOTAL INSTALLED CAPACITY

'L 7o OF TIME 100

FIG.A-4. EQUIVALENT LOAD CURVE FOR AN ENTIRE SYSTEM.

The loss-of-load probability calculated in this model only considers the generating system. To get a true measure of system reliability, the transmission and distribution systems must also be considered, but consideration of the system aspects is beyond the scope of the model. The true system loss-of-load probability can never be less than the loss-of-load probability calculated by the model since the model assumes a perfect transmission system. The area

A-5 FIG. A-5. SCHEMATIC DIAGRAM OF LOAD GROWTH AND CORRESPONDING CASH FLOWS. between the equivalent load curve and the ordinate above the total installed capacity is a measure of the probable value of energy demand not served. The simulation code calculates loss-of-load probability and the amount of energy not served for each time period of the study (usually quarterly). The more complicated aspects of the probabilistic simulation are beyond the scope of this simplified description. These aspects include the simulation of pumped storage and hydro units and the use of multiple capacity blocks for thermal units to better represent actual unit loading.

Treatment of economics

Consider the situation illustrated in Fig. A-5(a). This shows, in diagrammatic form, three years in the history of a power system experiencing load growth. It is seen that at the beginning of year 2 and year 3 an increase in system capacity is required by the growth in load. The capital expenditure which is equivalent to all of the construction costs of these plants is considered to be concentrated at a single point in time when the plant becomes operative. The operating expense to serve the given load duration curves is assumed for simplicity to be concentrated at the middle of each year. The corresponding cash flow diagram is shown in Fig. A-5(b). The present worth, to some reference year, of such a cash flow (ignoring the effects of the study horizon) is a measure of the cost of that particular expansion scheme. The method chosen to deal with the end effects caused by a finite study horizon is to assume that the salvage value of any piece of equipment installed during the study is proportional to the unused portion of its plant life. Therefore, the present worth of the cash flow calculated in the previous paragraph should be reduced by the present worth, measured from the horizon, of a credit for each plant's salvage value. The function (present worth) to be minimized then may be stated symbolically as

NYRS-1 NINSTW NFUELS u NYRS+k F = y [ y k /C|) - p (cr - )i +y F "(k+1) k=0 4=1 m=l

A-6 where F — objective function NYRS — number of years in the study NINSTk — number of installations in the kth year — present-worth factor for the kth year and ithplant — capital cost of the ith plant — present-worth factor for the horizon and the ith plant — present-worth factor for the mth fuel in the kth year ). m — plant life of the ith plant — operating cost of the mth fuel system for the (k+ 1) year NFUELS — number of different fuel types considered

Dynamic programming

In optimization terminology, the above function is known as an objective function or performance criterion. The value of the objective function denotes the relative benefit of a particular expansion schedule. The purpose of the optimization package is to determine which one of the selected alternative expansion schedules minimizes the value of the objective function. Dynamic programming is a powerful optimization tool and requires the definition of three types of variables: the stage variable, the state variable, and the control variable. The stage variable defines the sequence of events and, in the WASP program, is defined as the year being considered. The state variable describes the state of the system under study and is defined as the configuration of installed units in any given year. Once the values of the state variable are defined for all stages, any question concerning the system can be answered. The change between the states that might occur from stage to stage is determined by the value of the control variable between stages. Hence the control variable determines the capital investment and operating costs from year to year. In simple terminology, the control variable is the independent variable and the state variable is the dependent variable. In operation a number of configurations are generated for each stage (year) of the study. These configurations must satisfy the constraints of reserve margin and capacity-mix specified by the user. The production cost and reliability of each of these configurations is determined in the simulations for the appropriate year (stage). All of these calculations are performed before going to the dynamic program. In Fig. A-6 a number of states are represented, by dots, for two successive stages, k and (k+1). It should be kept in mind that the value of the objective function associated with each state in the kth stage is the minimum cost path from the beginning of the study to that state. In calculating the cost of the paths from state B to state A, the capital cost corresponding to the transfer from state B to A and the operating costs for state A are added to the value of the objective function of state B. This represents the present-worth cost of expanding the system to state A and passing through state B. The costs for the other paths from states C, D, E and F converging at state A are calculated in a similar manner, The path which yields the lowest value of the objective function at state A is retained by storing the objective

STAGE STAGE K K • 1

(B) • • -_ IC\ » "—-*M IM

(0)

(E) •

(F) •

FIG. A-6. ILLUSTRATION OF A DYNAMIC PROGRAM STEP.

A-7 function and sufficient information for determining the state in the previous stage. The other paths are discarded as they cannot possibly be part of the optimal trajectory. This procedure is repeated for all of the states in stage (k + 1). Then the next stage (k+2) is considered, with the calculations proceeding until the study horizon is reached. Then the lowest value of the accumulated objective function in the final stage is traced back from that state through the various stages to determine the optimal expansion strategy. In order to provide flexibility in representing real system situations, many features have been included in the WASP package. All cash flow is separated into domestic and foreign exchange in computing total expenditure. Total operating costs and cost of the fuel used in the plant are separately stated. Thus discounting and escalation may be applied separately to the domestic and foreign costs of operating plants consuming different fuels. In the same manner, the capital cost of each expansion alternative is separated into foreign and domestic components. Different discount rates and escalation rates on capital costs (foreign and domestic) are permitted on each alternative. Consequently, many sensitivity studies can be carried out with a minimum of computational effort after a basic optimum has been reached. Studies of the effects of plant capital cost, capital cost interest rates and escalation, exchange ratio (foreign/domestic), plant life, interest rate on operating cost, and critical loss-of-load probability require only reruns of the sixth (dynamic programming) step. If the operating policy does not change and if there are no pumped-storage installations, the escalation of operating costs may also be included in sensitivity studies.

LIMITATIONS OF THE PACKAGE

The program suffers mainly from approximations in the simulation. When the year is divided into large time blocks, the maintenance schedule is only approximate. Since the simulation uses a load duration curve technique, the chronological sequence of events during the individual periods is lost. The hydro representation includes two approximations. All hydro is lumped into a single pseudo-plant with an "always-run" and a "peak-shaving"component. The peak-shaving component is removed from the load duration curve prior to thermal plant simulation. This is not rigorous since hydro is also normally used to cover forced outages of thermal units. All pumped-storage units are also lumped into a single pseudo unit and will not exactly simulate multiple plants with widely varying weekly capacity factors.

ACKNOWLEDGEMENTS

Dr. R.R. Booth of the State Electricity Commission of Victoria (Australia) introduced the probabilistic simulation method and provided the Tennessee Valley Authority with early versions of his simulation and dynamic programming system expansion model. Dr.David Joy of Oak Ridge National Laboratories (USA) provided extensive consultation and development work on the simulation and generous advice and instruction on the dynamic programming method. At TVA Mr. Roger Babb (formerly with TV A) did extensive work to speed the simulation, Miss Katherine McQueen of the Power Research Staff coded and debugged and extensively modified the first generation TVA program and T.R. Jackson of TVA's Power Supply Planning Branch struggled with the difficulties of TVA's first generation program to produce useful results. The improvements in this program are largely a result of his experience. Finally, Mr. Peter Heinrich of IAEA's Computing Centre Staff has carried much of the burden of debugging and implementation on the IAEA computer.

REFERENCES

[1] BALERIAUX, H., SAMOULLE, E., LINARDDeGUERTECHIN, Fr., "Establishmentof aMathematicalModelSimulatingOperation of Thermal Electricity - Generating Units Combined with Pumped-Storage Plant," Review E (edition SRBE), S. A. EBES, Brussels, Belgium 5_ 7, pp. 1-24. [2] BOOTH, R. R., "Power system simulation model based on probability analysis", Proceedings of the 1971 IEEE PICA Conf., 71-C26-PWR, pp. 285-291. [3] SAGER, MA,, RINGLEE, R.J., WOOD, A.J., A New Generation Production Cost Program to Recognize Forced Outages, IEEE paper T72 159-7, pp. 2114-2124. [ 4] JOY, D. S., JENKINS, R. T., A Probabilistic Model for Estimating the Operating Cost of an Electric Power Generating System, ORNL-TM-3549 (1971).

A-8 APPENDIX B

GENERATING PLANT CAPITAL COSTS (ORCOST)

STEAM-ELECTRIC POWER PLANTS

In order to carry out the very large number of capital cost estimates for the thermal generating units being considered as expansion alternatives, it was necessary to make use of a digital computer program, ORCOST. This program was prepared specifically to provide estimates of the capital costs of steam-electric power plant in the United States of America for use in studies conducted by the Oak Ridge National Laboratory for the USAEC Division of Reactor Development and Technology. The code includes cost models for PWR, BWR, HTGR nuclear plants and coal, oil, and gas-fired plants which were developed from ORCOST's "big brother" CONCEPT II [1-7 ]. In developing both CONCEPT II and ORCOST the assumption was made that, for a given type and size of power plant and irrespective of its geographical location, the sizes of individual items of equipment, the amounts of construction materials, and the number of man-hours of construction labour remain the same for each of the nine major direct plant cost accounts shown in Table B-l. (Accounts 21-26/91-93 of the USAEC uniform system of accounting.) Such an assumption permits one to start with a base model in which costs for each of the major direct plant cost accounts are identified and to adjust these costs to conditions prevailing at different site locations by applying appropriate indices for equipment, material and labour cost. These indices reflect the unit costs of these items relative to the unit costs used in the base model. In the case of plant equipment costs the index to be used includes both cost escalation factors and cost factors specific to the site. In CONCEPT II these indices are calculated within the program from input data on the actual unit costs of equipment, materials and labour, whereas in ORCOST the indices are calculated separately. After applying the specific indices, the computer program sums up the adjusted total direct cost of the physical plant. In order to estimate these direct plant costs as a function of plant size, a second assumption is made, namely that the exponential scaling laws developed for the base model (to reflect the variation in costs of each of the major accounts with plant size) are independent of the indices used for equipment, materials, and labour costs.

TABLE B-l. 2-DIGIT ACCOUNTS USED IN THE USAEC SYSTEM OP ACCOUNTING

Account No. Item

Direct costs

21 Structures and site facilities 22 Reactor/boiler plant equipment 23 Turbine plant equipment 24 Electric plant equipment 25 Miscellaneous plant equipment 26 Special materials Indirect costs

91 Construction facilities, equipment and services 92 Engineering and construction management services 93 Other costs

B-l Having found the direct physical cost of the plant for a given size and site location, the program adds allowances for contingencies and spare parts and then computes the indirect costs by applying appropriate percentages to the physical plant costs. The technique of separating the plant cost into individual components, applying appropriate cost indices, and summing the adjusted components is the basic tool used in ORCOST. The procedure is illustrated schematically in Fig.B-1.

BASE COST ADJUST FOR SIZE

r DIVIDE INTO EQUIPMENT, SITE MATERIALS, AND SITE LABOR COMPONENTS

r r PLANT SITE SITE EQUIPMENT COST MATERIAL COST LABOR COST

t r f ADJUST BY ADJUST BY ADJUST BY APPROPRIATE COST APPROPRIATE COST APPRPRIATE COST INDEX RATIO INDEX RATIO INDEX RATIO

I SUM TO ADJUSTED COST OF PHYSICAL PLANT

r ADD ALLOWANCES FOR CONTINGENCIES AND SPARE PARTS

) I COMPUTE AND ADD INDIRECT COSTS

ORCOST

PRINT-OUT DISTRIBUTION OF COSTS FOR EACH 2 DIGIT ACCOUNT

FIG.B-1. SCHEMATIC ILLUSTRATION OF ORCOST (AND CONCEPT II) PROCEDURE.

B-2 Selection of nuclear reactor type

It should be noted here that in view of the diversity of reactor types now available commercially and because of the limited scope of the Survey, it was decided to base the evaluation of nuclear versus conventional power plants on a single reactor type, the PWR. Such a selection is not intended to imply a preference for this particular type of nuclear plant, but merely to provide an illustration which is believed to be representative of nuclear power in general. Other types of power reactors which have already been constructed and could be considered for developing countries in their future plans include AGR, BWR, HTGR, PHWR, and SGHWR. It is believed that breeder reactors will not be developed to the point of being useful in planning systems in developing countries within the study decade. To date, the following reactor types have been purchased or committed by the countries listed:

Type Gross electricity output (MW)

Argentina PHWR 340 CANDU-PHWR 600

Brazil PWR 657 Bulgaria PWR 2x440 Czechoslovakia HWGCR 144 PWR 2x440

India BWR 2x210 CANDU-PHWR lx 220 CANDU-PHWR 3x 220

Korea PWR 595 Pakistan CANDU-PHWR 137

The base cost model

The base cost model for each type of plant was established from a detailed cost estimate for a reference 1000 MW plant assumed to be located at "Middletown", USA, the standard hypothetical site described in Ref. [3]. Since the base cost models in the original OR COST program were developed in 1971, these were updated to the end of 1972 by applying appropriate escalation rates on equipment, materials and labour. These costs are referred to in the Survey as ORCOST-1. However, recent construction experience in the USA indicated that some adjustments should be made in the scope of work, particularly as it affects the construction costs of nuclear power plants. These adjustments were made and the resulting costs are referred to in the Survey work as ORCOST-3.1 The ORCOST-3 data are used as the reference case data in the Survey analyses. Table B-2 shows the ORCOST-3 total plant base cost models used for the Survey. Table B-3 shows a comparison of ORCOST-1 and ORCOST-3 total plant costs for 300, 600 and 1000 MW PWR and oil-fired plants. It also shows the modified costs (see below for discussion of country cost indices) for the participating country having the maximum cost levels and the one having the minimum cost levels. It is to be noted here that the adjustments made to obtain ORCOST-3 costs (from the ORCOST-1 values) resulted in essentially no change in the oil-fired (or other fossil-fired) plants, but there were substantial increases in the costs of nuclear plants of the order of 21-22% on all sizes. This resulted in the ratio of nuclear to oil-fired plant costs increasing from values of about 1.5 - 1.8 for ORCOST-1 to about 1.9 - 2.2 for ORCOST-3. ORCOST-1 costs were used to make a few sensitivity studies in selected countries in order to indicate the possible effect on Survey results if the ratio of nuclear to fossil-plant costs reverted to their pre-1972 levels.

ORCOST-2 referred to data not used for Survey analyses.

B-3 TABLE B-2 . OR COST - 3 BASE COST MODELS USED IN THE MARKET SURVEY (all 1000 MW capacity) PWR Coal-fired Oil-fired Gas-fired Account

No. 6 6 6 10 US $ Scaling exponent 10 US $ Scaling exponent 10 US $ Scaling exponent 10$ US $ Scaling exponent

21 52.03a 0.80a 29.18 0.75 26.67 0.75 26.67 0.75 22 77.20 0.60 67.91 0.90 56.00 0.90 36.50 0.90 23 74.95 0.80 53.21 0.80 53.00 0.80 53.00 0.80 24 27.84 0.60 18.52 0.45 14.15 0.45 13.40 0.45 25 5.39 0.30 4.35 0.30 4.08 0.30 4.08 0.30 26 0 0 0 0 0 0 0 0

Total 237.41 173.17 153.9 133.65

a For plant sizes below 800 MW, these figures become US $ 47.75 x 106 and 0.40 respectively.

TABLE B-3. COMPARISON OF CAPITAL COSTS FOR NUCLEAR AND OIL-FIRED PLANTS ORCOST-1 ORCOST-3 i Size Type (MW) Maximum country Minimum countty USA Maximum country Minimum country USA

300 PWR 490 378 517 593 442 624 Capital costs (US $/kW) Oil 272 210 316 268 206 315 Cost difference (US $/kW) 218 168 201 325 236 309 Cost ratio PWR/Oil 1.8 1.8 1.63 2.21 2.15 1.98

600 PWR 358 275 377 439 322 460 Capital costs (US $/kW) Oil 216 171 249 216 170 253 Cost difference (US $/kW) 142 104 128 223 152 207 Cost ratio PWR/Oil 1.64 1.61 1.51 2.03 1.89 1.82

1000 PWR 296 225 312 365 266 382 Capital costs (US $/kW) Oil 187 145 218 189 146 223 Cost difference (US $/kW) 109 80 94 176 120 159 Cost ratio PWR/Oil 1.58 1.55 1.43 1.93 1.82 1.71 The base model plant costs include, in all oil and coal-fired plants, electrostatic precipitators. However, these costs do not include any of the other so-called environmental control equipment such as SO2 removal systems, cooling towers/lakes or near-zero radiation release systems. It was felt that environmental considerations which have caused designs of almost all future plants in industrialized countries to include such equipment, or provision to add it at later dates, would not generally apply during the study period in the developing countries included in the Survey. It is recognized, however, that in certain countries these considerations might possibly have to be faced and coped with during the study decade. Therefore, the following should be noted when considering the capital costs of future plants.

(a) High-efficiency (99.5 + %) electrostatic precipitators to remove particulate matter from stacks of oil or coal/ lignite -fired plants cost of the order of US $8-10/kW of installed capacity. Thus, if precipitators are not required in any given instance, this amount may be omitted from the appropriate costs in Tables B-2 and B-3.

(b) Although there is no known proven process for the effective economic removal of SO2 from the stack gases of fossil-fired plants, it is at present estimated that such equipment, when commercially applicable, could involve an additional equivalent investment cost of the order of US $50/kW for a 1000 MW plant burning coal containing 3.0% sulphur. This would include both the initial investment (about US $35-40/kW) and the capitalized operating cost and capacity penalty (about US $10-15/kW). The actual final costs would, of course, depend on the original sulphur content of the fuel being used, the size of plant, the ability to dispose of the recovered sulphur etc.

(c) Cooling towers, of various designs, are presently in use in many power plants and they can be considered fully developed technically. Their costs are reasonably well known for installations under a wide variety of conditions. The initial investment for a 1000 MW plant would be of the order of US $5-10/kW for fossil-fired plants depending on whether a mechanical draft or natural draft design is used. For nuclear plants, these values should be increased by about 50%. The costs of cooling lakes, ponds or equivalent methods of disposing of thermal discharges will vary quite widely, but they can be generally considered as less expensive overall than cooling towers if the amount of land required is available at a reasonable price. An upper limit of their cost can be considered as the cost of equivalent cooling towers.

(d) The addition of equipment to light-water nuclear plants to accomplish near-zero radiation release will be likely to cost about US $5-10/kW for larger sizes of plants, depending on the type of reactor plant involved.

It is quite possible, therefore, that costs for future fossil-fired plants could increase substantially more than for nuclear plants if precipitators, SO2 removal systems and cooling towers or the equivalent were required for the fossil-fired plants and cooling towers or the equivalent and near-zero radiation release systems were required for nuclear plants. On a comparable basis, therefore, for large plants of the order of 1000 MW, the possible future incremental penalty against fossil-fired plants would appear to be of the order of US $40/kW when precipitators are not required and US $50/kW if precipitators are required for the coal-fired plants. These US $/kW values could increase by as much as 50% for the smaller sizes of units considered in the study. It should be noted that, in addition to the increases in capital cost for environmental control equipment, the operating and maintenance costs of the plants, as discussed in Appendix E, will be increased.

Modifications of indirect costs

Indirect costs in the base model (construction facilities, equipment and services, engineering and construction management services, taxes, insurance and owner's general and administrative expenses) are estimated as percentages of the direct physical plant cost based on experience in the USA. It was recognized that this experience would not be directly

B-5 applicable to conditions prevailing in the countries being studied; therefore, the indirect cost percentages in the base model were adjusted to reflect such conditions. Such adjustments to the base model are easily made by changing the indirect cost indices applicable to Accounts No. 91, 92 and 93. The indices actually used are shown in Table B-4. These indirect cost indices were derived for the Survey as follows: Firstly, it was assumed that the plants being considered would be two-unit plants; therefore, the costs of temporary facilities which would be common to both units were divided by two. Secondly, it was assumed that the costs of local labour and materials associated with account 91 would be about 75% of the costs used in the base model. These assumptions decreased account 91 from 6.6% of the physical plant costs to 5.3%, resulting in an index of 0.8 for account 91. For account 92, engineering services were taken to be the same as for the USA based on the assumption that all design and engineering for the nuclear plant would be done by an architect-engineering firm from outside the country being studied. Costs of construction management services, moreover, were increased by US $ 5 million in the base model for overseas support of personnel supervising the construction. This increased the percentage of physical plant costs from 11.6% in the base model to 13.6% resulting in an index of 1.17 for account 92. Account 93 was adjusted to remove the local taxes assumed for the base model resulting in an index of 0. 71 for account 93. Indirect cost indices for conventional plants were derived in a similar manner, to give the values: account 91 = 0. 72, account 92 = 1. 06, account 93 = 0. 65. In the cost model, indirect costs are calculated using a hyperbolic function. This results in abnormally high indirect costs for unit sizes below 300 MW both in terms of total dollar costs and the ratio of the indirect costs to total plant costs. Therefore, the calcula-

TABLE B-4. ADJUSTMENT OF THE INDIRECT COSTS OF THE BASE MODEL (1000 MW PWR) TO MARKET SURVEY CONDITIONS

Percentage of physical plant cost Account No. Base model Market survey

91 Construction facilities, equipment and services

911 Temporary facilities 2.0 1.5 912 Construction equipment 3.3 3.0 913 Construction services 1.3 0.8

Total for account 91 5.3 Ratio — Market survey/base model 0.80

92 Engineering and construction management services 921 Engineering services 5.8 5.8 922 Construction management services 5.8 7.8

Total for account 92 11.6 13.6 Ratio ~ Market survey ^ase model 1.17

93 Other costs 931 Taxes and insurance 2.7 1.5 932 Staff training and plant start-up 0.3 0.3 933 Owner's general and administrative expenses 1.2 1.2

Total for account 93 4.2 3.0 Ratio — Market survey/base model 0.71

B-6 tion of indirect costs for the smaller sizes of plants was made by taking a linear approximation. It should be noted that although the percentages applied to the physical plant costs to obtain the indirect costs vary with size of plant, the indirect cost indices remain constant for all sizes of plants.

Derivation of country cost indices

Specific cost indices for equipment, materials and labour were used for each participating country. These cost indices are stated as a ratio of the effective foreign costs to the US-based costs and thus allow the determination of total construction costs of the various types and sizes of plants in each country based on equipment, materials and labour cost indices and interest rates unique to each country. The following paragraphs explain how the cost indices were obtained and used to modify the US-based costs:

(a) Equipment cost index

The equipment cost indices were determined after giving consideration to international sources for the items of equipment, the location of the country relative to those sources, the transport costs from likely sources to the country, the competitive nature of the international market, known country preferences for equipment types and sources and the likely location of the power plants within the country, i.e. inland or on the seashore. On balance, the equipment cost index, for an "ideal" plant site in an "average" country, was established as 1.0 for nuclear plants and 0.9 for fossil plants relative to the US values in the ORCOST models. A specific index was then established for each country relative to these values, considering the above factors as they were known to apply or as best they could be approximated.

(b) Materials cost index

The materials cost indices were determined either from detailed costs of completed power plants provided by the countries or from specific prices in the country for construction materials such as structural steel, re-inforcing steel, concrete (ready-mix), ply-form and lumber. In some cases where such data were not available the indices were estimated based on a comparison with known data for a neighbouring country or for the general area.

The labour cost indices were calculated from the wages for different types of craft usually available in the country, such as common labour, bricklayer, carpenter, ironworker, electrician, steam-fitter, operating engineer, and other classifications as available. These wages were weighted by the amount of man-hours to be spent in the construction of a power plant. For this purpose a labour efficiency was estimated. Where no detailed information about wages was available, the labour cost indices were calculated from detailed costs of constructed power plants, or it was estimated by comparison with other countries.

ORCOST input and output

With the above modifications to the basic ORCOST program the actual input data required for each country include plant size and type, labour cost index, materials cost index, equipment cost index, cost escalation rates (if any), interest rates, construction period, length of working week (if different from 40 hours). From these input data total capital costs are obtained as the output, with the cost adjusted to the specific country's cost levels. Table B-5 shows a printout sheet from the ORCOST-3 program summarizing input data for a 600 MW PWR with equipment, materials and cost indices set at 1.0. Tables B-6 to B-9 show output data from ORCOST-3 for various fossil-fuelled 600 MW plants, again with the cost indices set at 1.0. It should be pointed out

B-7 TABLE B-5. ORCOST-3 PRINTOUT OF INPUT DATA FOR 600 MW PWR

PLANT SIZE, MW(E). 600.0 PLANT TYPE. Ts PWR YEAR CONSTRUCTION STARTED. YS = 1973.00 YEAR OF COMMERCIAL OPERATION. YO = 1978.50 BASE YEAR FOR ESCALATION YBX = 1971.50 LENGTH OF WORKWEEK, HRS. HW = 40.0 ANNUAL INTEREST RATE, PERCENT. XIR = 8.0 INITIAL EQUIP. ESCAL. RATE, ANNUAL PERCENT EREB= 0.0 INITIAL MATLS. ESCAL. RATE, ANNUAL PERCENT ERMB= 0.0 INITIAL LABOR ESCAL. RATE, ANNUAL PERCENT ERLB= 0.0 EQUIPMENT ESCALATION RATE, ANNUAL PERCENT. ERE = 0.0 MATERIALS ESCALATION RATE, ANNUAL PERCENT. ERM = 0.0 LABOR ESCALATION RATE, ANNUAL PERCENT. ERL = 0.0 PROVEN DESIGN IFLAG = 0 SUBROUTINE NAMELIST OPTION NOT SELECTED JFLAG = 0 HEAT REMOVAL - RUN OF RIVER ICT = 0 UPGRADED RADWASTE SYSTEM NOT SPECIFIED I EC = 0

CONTINGENCY AND SPARE PARTS FACTORS, PERCENT DIVIDED BY 100 CONTINGENCY FACTORS SPARE PARTS FACTORS

EQUIPMENT £ MATERIALS LABOR EQUIPMENT £ MATERIALS F21CEM= 0.050 F21CL= O.IOO F21SEM= 0.010 F22CEM= 0.050 F22CL= 0.100 F22SEM= 0.010 F23CEM= 0.050 F23CL= 0.100 F23SEM= 0.010 F24CEM= 0.050 F24CL= 0.100 F24SEM= 0.010 F25CEM= 0.050 F25CL= 0.100 F25SEM= 0.010 F26CEM= 0.050 F2 6CL= 0.100 F26SEM= 0.010 FSOCEM= 0.050 FSOCL= 0.100 FSOSEM= 0.010 FHRCEM= 0.0 50 FHRCL= 0.100 FHRSEM= 0.010

EQUIPMENT COST INDEX. A(IN,1) = 1.coo MATERIALS COST INDEX. A(IN,2) = 1.000 LABOR COST INDEX. A(IN,3) = 1.000

BASE COST MODEL COST COST BREAKDOWN FACTORS $MILLION EXPONENT EQUI PMENT MATERIALS LABOR ACCT 21 C(l)= 47.75 N(l)=0.40 EF(1)= 0.03 MF(l)=0.35 LF(1)=O..62 ACCT 22 C(2)= 77.20 N(2)=0.60 EF{2)= 0.52 MF(2)=0.,21 LF(2)=0..27 ACCT 23 C(3)= 74.95 N(3)=0.80 EF(3)= 0.54 MF(3)=0. 17 LF(3)=0.,29 ACCT 24 C(4)= 27.84 N(4)=0.60 EF(4)= 0.23 MF(4)=0.34 LF(4)=0..43 ACCT 25 C(5)= 5.39 N(5)=0.30 EF(5)= 0.39 MF(5)=0.04 LF(5)=0..57 ACCT 26 C(6)= 0.0 N{6}=0.0 EF(6)= 0.0 MF(6)=0.0 LF(6)=0.,0 RAD. W. C17)= 0.0 N(7J=0.60 EF(7)= 0.69 MF(7)=0.13 LF(7)=0.18 C. TOW. C(8) = 0.0 N(8)=0.80 EF(8)= 0.47 MF(8)=0.04 LF<8)=0.49 INDIRECT COSTS F91= 0.80 F92= 1.17 F93= 0.71

B-8 TABLE B-6. ORCOST-3 PRINTOUT OF OUTPUT DATA ON THE CAPITAL COST OF A 600 MW PWR

PLANT CAPITAL INVESTMENT SUMMARY (SMILLIQN) MIDD 600.0 Mw(E) PWR 1973.00 - 1978.50

DIRECT COSTS 20 LAND AND LAND RIGHTS 0.1 PHYSICAL PLANT EQU. MAT. LABOUR TOTAL 21 STRUCTURES AND SITE FACILITIES 1.2 13.6 24.1 38.9 22 REACTOR/BOILER PLANT EQUIPMENT 29.5 11.9 15.3 56.8 23 TURBINE PLANT EQUIPMENT 26.9 8.5 14.4 49.8 24 ELECTRIC PLANT EQUIPMENT 4.7 7.0 8.8 20.5 25 MISCELLANEOUS PLANT EQUIPMENT 1.8 0.2 2.6 4.6 26 SPECIAL MATERIALS 0.0 0.0 0.0 0.0 INCREMENTAL ALLOWANCE FOR ENVIRONMENTAL ASSURANCE UPGRADED RADWASTE SYSTEM 0.0 . 0.0 0.0 0.0 COOLING TOWERS 0.0 0.0 0.0 0.0 SUBTOTAL (PHYSICAL PLANT) 64.1 41.2 65.4 170.7 CONTINGENCY ALLOWANCE — 11.8 SPARE PARTS ALLOWANCE 1.1 SUBTOTAL (PHYSICAL PLANT) 183.5 OVERTIME ALLOWANCE { 40.0 HR WORKWEEK) 0.0 SUBTOTAL (TOTAL PHYSICAL PLANT) 183.5

INDIRECT COSTS

91 CONSTRUCTION FACILITIES* EQUIPMENT, AND SERVICES - 10.9 92 ENGINEERING AND CONSTRUCTION MANAGEMENT SERVICES - 28.1 93 OTHER COSTS 6.1 94 INTEREST DURING CONSTRUCTION ( 8.0 PCT- 5.50 YRS) 47.3 SUBTOTAL (TOTAL INDIRECT COSTS) 92.5

SUBTOTAL (TOTAL PLANT COST) 276.1 CAPABILITY PENALTY ( 0.0 PCT- 0.0 MW(E)) 0.0 TOTAL PLANT CAPITAL COST (AT START OF PROJECT) 276.1 $ / KW(E) 460.

ESCALATION DURING CONSTRUCTION ( 0.0 PCT ) 0.0 TOTAL PLANT CAPITAL COST (AT COMMERCIAL OPERATION) 276.1 $ / KW(E) 460.

B-9 TABLE B-7. OR COST-3 PRINTOUT OF OUTPUT DATA ON A 600 MW COAL-FIRED PLANT

PLANT CAPITAL INVESTMENT SUMMARY ($MILLION) MIOD 600.0 MW(E) COAL 1973.00 - 1977.00

DIRECT COSTS

20 LAND AND LAND RIGHTS 0.1

PHYSICAL PLANT EQU. MAT. LABOUR TOTAL 21 STRUCTURES AND SITE FACILITIES 0.6 7.8 11.5 19.9 22 REACTOR/BOILER PLANT EQUIPMENT 22.7 5.1 15.0 42.9 23 TURBINE PLANT EQUIPMENT 19.1 6.0 10.3 35.4 24 ELECTRIC PLANT EQUIPMENT 4.9 2.4 7.5 14.7 25 MISCELLANEOUS PLANT EQUIPMENT 1.0 0.7 2.0 3.7 26 SPECIAL MATERIALS 0.0 0.0 0.0 0.0 INCREMENTAL ALLOWANCE FOR ENVIRONMENTAL ASSURANCE SO-2 REMOVAL SYSTEM 0.0 0.0 0.0 0.0 COOLING TOWERS 0.0 0.0 0.0 0.0 SUBTOTAL (PHYSICAL PLANT) 48.3 22.0 46.3 116.6 CONTINGENCY ALLOWANCE 8.1 SPARE PARTS ALLOWANCE 0.7 SUBTOTAL (PHYSICAL PLANT) 125.4 OVERTIME ALLOWANCE ( 40.0 HR WORKWEEK) 0.0 SUBTOTAL (TOTAL PHYSICAL PLANT) — 125.4

INDIRECT COSTS

91 CONSTRUCTION FACILITIES, EQUIPMENT, AND SERVICES - 8.0 92 ENGINEERING AND CONSTRUCTION MANAGEMENT SERVICES - 13.1 93 OTHER COSTS 3.6 94 INTEREST DURING CONSTRUCTION ( 8.0 PCT- 4.00 YRS) 21.9 SUBTOTAL (TOTAL INDIRECT COSTS) •— 46.6

SUBTOTAL (TOTAL PLANT COST) 172.1 CAPABILITY PENALTY ( 0.0 PCT- 0.0 MW(E)) 0.0 TOTAL PLANT CAPITAL COST (AT START OF PROJECT) 172.1 $ / KW(E) • 287.

ESCALATION DURING CONSTRUCTION ( 0.0 PCT ) 0.0

TOTAL PLANT CAPITAL COST (AT COMMERCIAL OPERATION) 172.1 $ / KWIE) 287.

B-10 TABLE B-8. ORCOST-3 PRINTOUT OF OUTPUT DATA ON A 600 MW OIL-FIRED PLANT

PLANT CAPITAL INVESTMENT SUMMARY (JMILLION) M.IDD 600.0 MW(E) OIL 1973.00 - 1976.50

DIRECT COSTS

20 LAND AND LAND RIGHTS 0.1

PHYSICAL PLANT EQU. MAT. LABOUR TOTAL

21 STRUCTURES AND SITE FACILITIES 0.5 6.9 10.7 18.2 22 REACTOR/BOILER PLANT EQUIPMENT 18.0 4.6 12.7 35.4 23 TURBINE PLANT EQUIPMENT 19.0 6.0 10.2 35.2 24 ELECTRIC PLANT EQUIPMENT 4.4 1.7 5.2 11.2 25 MISCELLANEOUS PLANT EQUIPMENT 1.0 0.7 1.8 3.5 26 SPECIAL MATERIALS 0.0 0.0 0.0 0.0 INCREMENTAL ALLOWANCE FOR ENVIRONMENTAL ASSURANCE SO-2 REMOVAL SYSTEM 0.0 0.0 0.0 0.0 COOLING TOWERS 0.0 0.0 0.0 0.0 SUBTOTAL (PHYSICAL PLANT) 43.0 19.9 40.6 103.5 CONTINGENCY ALLOWANCE 7.2 SPARE PARTS ALLOWANCE 0.6 SUBTOTAL (PHYSICAL PLANT) 111.3 OVERTIME ALLOWANCE ( 40.0 HR WORKWEEK) 0.0 SUBTOTAL {TOTAL PHYSICAL PLANT) 111.3

INDIRECT COSTS

91 CONSTRUCTION FACILITIES, EQUIPMENT, AND SERVICES - 7.6 92 ENGINEERING AND CONSTRUCTION MANAGEMENT SERVICES - 12.3 93 OTHER COSTS 3.4 94 INTEREST DURING CONSTRUCTION { 8.0 PCT- 3.50 YRS) 17.0 SUBTOTAL (TOTAL INDIRECT COSTS) 40.3

SUBTOTAL (TOTAL PLANT COST) 151.8 CAPABILITY PENALTY { 0.0 PCT- 0.0 MW(E)) 0.0 TOTAL PLANT CAPITAL COST (AT START OF PROJECT) 151.8 $ / KW(E) 253.

ESCALATION DURING CONSTRUCTION ( 0.0 PCT ) 0.0

TOTAL PLANT CAPITAL COST (AT COMMERCIAL OPERATION) 151.8 $ / KW(E) 253.

B-ll TABLE B-9. ORCOST-3 PRINTOUT OF OUTPUT DATA ON A 600 MW GAS-FIRED PLANT

PLANT CAPITAL INVESTMENT SUMMARY ($MILLION) MIDD 600.0 MW(E) GAS 1973.00 - 1976.50

DIRECT COSTS 20 LAND AND LAND RIGHTS 0.1 PHYSICAL PLANT EQU. MAT. LABOUR TOTAL 21 STRUCTURES AND SITE FACILITIES 0.7 7.1 10.4 18.2 22 REACTOR/BOILER PLANT EQUIPMENT 12.7 2.3 8.1 23.0 23 TURBINE PLANT EQUIPMENT 19.0 6.0 10.2 35.2 24 ELECTRIC PLANT EQUIPMENT 4.6 1.1 5.0 10.6 25 MISCELLANEOUS PLANT EQUIPMENT 0.9 0.8 1.8 3.5 26 SPECIAL MATERIALS 0.0 0.0 0.0 0.0 INCREMENTAL ALLOWANCE FOR ENVIRONMENTAL ASSURANCE SO-2 REMOVAL SYSTEM 0.0 0.0 0.0 0.0 COOLING TOWERS 0.0 0.0 0.0 0.0 SUBTOTAL (PHYSICAL PLANT) 37.9 17.2 35.4 90.6 CONTINGENCY ALLOWANCE • 6.3 SPARE PARTS ALLOWANCE 0.6 SUBTOTAL {PHYSICAL PLANT) 97.5 OVERTIME ALLOWANCE ( 40.0 HR WORKWEEK) 0.0 SUBTOTAL (TOTAL PHYSICAL PLANT) 97.5

INDIRECT COSTS

91 CONSTRUCTION FACILITIES, EQUIPMENT, AND SERVICES - 7.2 92 ENGINEERING AND CONSTRUCTION MANAGEMENT SERVICES - 11.6 93 OTHER COSTS 3.2 94 INTEREST DURING CONSTRUCTION ( 8.0 PCT- 3.50 YRS) 15.1 SUBTOTAL (TOTAL INDIRECT COSTS) 37.1

SUBTOTAL (TOTAL PLANT COST) 134.6 CAPABILITY PENALTY ( 0.0 PCT- 0.0 MW{E)) 0.0

TOTAL PLANT CAPITAL COST (AT START OF PROJECT) 134.6 $ / KW(E) 224. ESCALATION DURING CONSTRUCTION ( 0.0 PCT ) 0.0

TOTAL PLANT CAPITAL COST (AT COMMERCIAL OPERATION) 134.6 $ / KW(E) 224.

B-12 that these costs do not represent costs of plants built in the USA, but costs of plants in a hypothetical developing country with equipment costs, materials costs and labour rates equal to those in the north-east of the USA.

Land costs

Land costs are treated as a separate item in both ORCOST programs. To reflect the lower cost of land in the Survey countries relative to the USA, land costs were assumed to amount to US $100 000 instead of US $1 million assumed in the original program.

GAS TURBINE PLANTS

Only 50 MW gas turbine plants were considered in the studies. Their installed cost was assumed to be US $125/kW at 1 January 1973 price levels. The costs were assumed to escalate at the same general inflation rate used for the other types of plants and equipment. Where more than 50 MW of capacity of this type was required, multiples of this 50 MW unit size were assumed with installed costs constant at US $125/kW.

HYDROELECTRIC PLANTS

As explained in Appendix A, all hydro or pumped-storage capacity, at any point in time, is merged in the WASP program with the then existing hydro or pumped storage into one equivalent hydro or pumped-storage plant. The costs of each hydro or pumped storage plant added to the system during the study period was taken as given by the country. In a few cases where costs of individual hydro projects were given, but no schedule was provided as to the order in which the projects would be constructed, average costs in US $/kW were determined for all projects in the group for which costs were given, and these average costs then used to obtain the installed costs of the required hydro capacity. Where known hydro potential was identified, but no costs were available, estimates were made of the installed costs based on known costs of existing projects in the same area or based on average conditions.

REFERENCES

[1] DeLOZIER, R.C., REYNOLDS, L.D., BOWERS, H.I., CONCEPT - Computerized Conceptual Cost Estimates for Steam- Electric Power Plants — Phase 1 User's Manual, ORNL-TM-3276 (1971). [2] DeLOZIER, R.C.. SRITE, B.E.. REYNOLDS, L.D., BOWERS, H.I., CONCEPT II - Computerized Conceptual Cost Estimates for Steam-Electric Power Plants — Phase II User's Manual, ORNL-4809 (1972). [3] USAEC Guide for Economic Evaluation of Nuclear Reactor Plant Designs, NUS-531 (1969). [4] United Engineers and Constructors Inc., 1000 MW Central Station Power Plants Investment Cost Study, UEC-AEC-720630 (1972). [5] BOWERS, N.I., MYERS, M.L., Estimated Capital Costs of Nuclear and Fossil Power Plants, ORNL-TM-3243 (1971). [6] United Engineers and Constructors Inc. (Jackson and Moreland Division), Cost of Large Fossil Fuel-Fired Power Plants, J + M No.636 (1966). [7] USAEC, CONCEPT - A Computer Code for Conceptual Cost Estimates of Steam-Electric Power Plants, WASH-1180 (1971).

B-13

APPENDIX C

LOAD DESCRIPTION DATA FOR WASP PROGRAM

REQUIRED DATA

The load description data required for the WASP program are as follows:

(1) Study increment, in MW. (2) Peak load demand for each year of study period, in MW. (3) Seasonal (quarterly) peak load demands expressed as a percentage of the annual peak load. (4) Coefficients of a polynomial describing the shape of the load duration curves for each of the four seasons of the year. The program will thus calculate the corresponding annual load factor for each year of the study. The following describes how these data were obtained.

Study increment

In carrying out the computations associated with the load duration curves, these are divided into blocks of capacity (MW) equal to a selected study increment. To avoid on the one hand a too rough approximation of the load curve and on the other hand a waste of computer time, the study increment was selected in accordance with the following rules: (a) It must be greater than the largest value of system installed capacity, during the entire study period, divided by 590. (b) It should be less than 2% of the smallest value of system installed capacity during the entire study period. (c) It should be less than approximately three times the capacity of the smallest generating unit in the system.

Peak load demands for each year of study

Peak load demands for each year of the study were derived from data provided by the country or by mathematical or graphical interpolation of the five-year interval forecasts developed by the method described in Appendix F.

Seasonal peak load demands

The seasonal variation of peak load demand in each case was obtained from historical data for representative years provided by the country. To simplify preparation of input data, the seasonal peak loads measured as a percentage of the annual peak load were assumed to remain constant throughout the study period.

Coefficients of a polynomial describing shape of load duration curves

Coefficients of a fifth order polynomial were used to represent the shape of the load duration curves. This fifth order polynomial gave a satisfactory fit in virtually all cases. The curve fitting was done by a standard polynomial regression program (No. 1001G/ST3 in the WANG 700 series program library) on a WANG Model 700 computer with plotter. This program calculates the coefficients bi in the expression

2 5 L = b0 + bxX + b2X + + b5X where L = fraction of peak load, X = fraction of total time.

C-l The computer then plots the fitting curve as shown in Fig. C-l. Examples of the coefficients b0 to b5 are shown in Table C-l under the heading "Load coefficients in force this year". In addition, a special program calculates both the slope of the curve at the point X=l and also the load factor which is given by

It is important that the polynomial should not have a negative slope at any point. It follows therefore that

L = bx + 2b2 + 3b3 5b5 has to be less than 0 for 0 § X § 1.

The value of bo is forced near to unity by entering the point (0, 1) a number of times. An additional program on the WANG forces it exactly to 1.

1UU - 100 H •-4- *+.

80- 80- o z z UJ 60- o z 2. +* X Z 40- a) SPRING 40- b) SUMMER U- o

20- 20-

1 1 —T 20 40 60 80 10C 20 40 60 80 100 7oOF TIME 7.0F TIME 100- 100-H

'•f •.+

80- ++* 80-

Q z s UJ Q 60- 60-

Z X •I 40- c} AUTUMN d) WINTER u. + o

20- 20-

~1 1" 20 40 60 80 100 20 40 60 80 100 % OF TIME 7= OF TIME FIG. C-l. EXAMPLES OF THE FITTING OF A FIFTH ORDER POLYNOMIAL TO LOAD DURATION CURVES.

C-2 TABLE C-l. SAMPLE OUTPUT OF COMPUTER CALCULATIONS OF LOAD DURATION DATA.

PERIOD PEAK LOADS IN PU OF ANNUAL PEAK 0.867000 0.989000 1.000000 0.971000

PERIOD PEAK LOADS IN MW 25143.0 28681.0 29000.0 28159.0 LOAD COEFFICINTS IN FORCE THIS YEAR ARE 1.000000 -2.958504 11.891810-23.599838 20.824448 -6.759686 1.000000 -3.193929 12.838108-25.477798 22.481552 -7.297591 1.000000 -3.131148 12.5 85763-24.977005 22.039658 -7.154149 1.000000 -2.974198 11.954898-23.72 5037 20.934921 -6.795546

PERIOD 1 PEAK LOAD 25143.0 MW MIN LOAD 10012 MW ENERGY UNDER LOAD DURATION CURVE 34304.1 GWH PERIOD LOAD FACTORS) 62.30

PERIOD 2 PEAK LOAD 28681.0 MW MIN LOAD 10048 MW ENERGY UNDER LOAD DURATION CURVE 37246.9 GWH PERIOD LOAD FACTORm 59.30

PERIOD 3 PEAK LOAD 29000.0 MW MIN LOAD 10530 MW ENERGY UNDER LOAD DURATION CURVE 38169.4 GWH PERIOD LOAD FACTORm 60.10

PERIOD 4 PEAK LOAD 28159.0 MW MIN LOAD 11123 MW ENERGY UNDER LOAD DURATION CURVE 38295.7 GWH PERTOD LOAD FACTORS) 62.10 ANNUAL LOAD FACTORm 58.26 ENERGY 148016.1 GWH

END OF DATA FOR YEAR 2000 ************

ANNUAL LOAD FACTORS

The following equations must hold: 4 AE = > PE = 2190 ) (PLF ) (PP )

AE = 8760 (AP) (ALF) = 2190 AP \ (PPFn ) (PLFn) where AE = annual energy forecast, AP = annual peak load, ALF = annual load factor, PLF = period load factor, PP = period peak load, PPF = period peak as a fraction of annual peak, PE = period energy forecast.

From PLF, AP and PPF the WASP program will calculate an annual load factor (see Table C-l). If this calculated annual load factor (ALFca) is not equal to the projected annual load factor (ALFpr ) the values of PLF are modified by the quotient ALF /ALFca. A code is available for the WANG 700 calculator which modifies the coefficients corresponding to a given PLF to give new coefficients corresponding to the projected PLF. This is done by calculating and applying a factor, a, as follows:

2 5 L = bQ + a (bxX + b2X b&X ) = 1 + Thus the shape of the curve is conserved.

C-3 This program was also used when the load factor varied during the time of the study. Figure C-2 shows an example of varying the load factor while conserving the shape of the load duration curve. In some cases, seasonal load curves and load factors were not available but only one annual load curve and the seasonal minima and maxima. In these cases the following approximation for the load curve was used:

L = 1 - (1-LF2) XLF

From this expression the load factor LF can be shown to be

/minimum load maximum load

:.•:.. 60 DEMAN D

Zi 7115 2 t- LOAD "::::::* X |.6882 FACTOR 40 UL o o

1 20 40 60 80 100 %> OF TIME

FIG. C-2. ILLUSTRATION OF THE EFFECT OF LOAD FACTOR ON A LOAD DURATION CURVE.

C-4 APPENDIX D

ECONOMIC METHODOLOGY AND PARAMETERS

The purpose of the Survey was to estimate the possible role of nuclear power in meeting the electric energy requirements of the countries over ten years from 1980 to 1989. Ideally the performance of this task would require estimating and comparing benefits and costs, both direct and indirect, arising from alternative development patterns, in order to determine in each case the power expansion plan yielding maximum total net benefits. The above requirement has seldom been met in full even in analyses of a single project in one country. To fulfil it for the comparison of chains of projects extending over ten years and covering 14 countries would have been theoretically questionable and practically impossible. A series of simplifying assumptions affecting both input data and the procedures for their aggregation, treatment and comparison was therefore unavoidable. The methodology described in the following sections represents an attempt at achieving a compromise between practical constraints and theoretical consistency. The main components of this methodology involved: (1) A definition of costs and benefits to be considered and the development of methods for estimating their quantitative values. (2) A selection of criteria for comparing benefits and cost streams extending over time and containing domestic and foreign currency components in variable proportions. (3) A choice of an optimization procedure and of a time horizon. These three major components are reviewed in the following paragraphs.

DEFINITION AND ESTIMATES OF COSTS AND BENEFITS It was assumed that costs rather than net benefits would be the only yardstick. This is tantamount to assuming that all programs of electric power expansion meeting projected demand with the imposed constraints on reliability offer the same total benefits and that the least cost program consequently yields maximum benefits to the ultimate consumers. In the case of comparing alternative ways of producing the same commodity, in this case electric power, this is a less questionable alternative than it would be in the general case of comparing alternative projects with different outputs. It does, however, ignore such indirect effects as, for instance, different employment levels arising from different power programs and their consequent effects on savings and investment or the future value of acquiring a pool of labour skilled in constructing and operating nuclear stations. Further- more, it can lead to serious distortions where multi-purpose hydro plants are involved in the comparisons. Consequently in the latter case the share of costs assignable to power production was estimated. Only costs directly connected with electricity production through a particular type of plant were taken into account. In particular such external or social costs as those arising from increasing environmental pollution in the case of fossil-fuelled stations or from the relatively larger thermal pollution by nuclear stations were disregarded in the basic analysis. The imposition of strict environmental controls by industrial countries leading to higher capital and fuel costs for thermal power stations shows that "external" costs may easily become "internal" over time. For the purpose of a basic analysis, however, and in spite of the recognition that the major industrial urban areas of some developing countries may well enact quantitative pollution controls, the effect of this assumption for the period under review does not appear to be decisive. In all basic cases costs were defined as costs to the economy rather than costs to the electricity producers. A major consequence of this criterion was to eliminate taxes on all types of fuel and equipment from all cost inputs. This was a particularly critical assumption in the case of countries imposing a heavy fiscal burden on some types of fuel and in particular on fuel oil. It was felt, however, that the basic purpose of the Market Survey was

D-l to advise countries on the total costs of alternative power programs estimated at the national level and that in this approach taxes represented internal transfers whose impact might distort the selection of power equipment which is most economic for the country as a whole. However, since the countries concerned are best judges of their tax policies which may involve items of social benefits disregarded by the Survey, since the electric utilities certainly view taxes on fuel and equipment as elements of costs, and since the Market Survey is addressed not only to the countries, but also to the potential equipment suppliers, alternative computations treating taxes as elements of costs were carried out for the cases which were expected to show critical differences in the results. Finally, the actual data used as bases for capital and fuel costs of power stations and their extrapolation to varying local conditions are discussed in the relevant sections of the report.

SELECTION OF CRITERIA

The aggregation of domestic and foreign currency costs was carried out on the basis of the official rates of exchange prevailing on 1 January 1973. It is recognized that in many of the countries surveyed, the official rates do not reflect the relative values of foreign and domestic capital resources to the economy. Nor do they always represent values which achieve equilibrium between the supply of and the demand for foreign capital as evidenced by foreign exchange rationing and control, as well as by the existence of parallel markets. The only defence of this approach which may substantially underestimate the true value of the ratio of foreign to domestic costs rests on its comparison with possible alternatives. The procedure of estimating "shadow" foreign exchange rates from 1980 till 1990 is dependent on political and economic forecasting and involves such a degree of uncertainty as to make its use unrealistic and its results highly doubtful. An estimate based on prevailing parallel rates would on the other hand rely on figures based on transitory trends and subject to large and rapid fluctuations. The theoretical inaccuracies of using official rates of foreign exchange were somewhat reduced by the practices followed by some of the countries where the problem of instability was most acute. In some of these all domestic cost items of future projects were converted into hard currency equivalents on the basis of experience on past similar projects completed during periods when foreign exchange rates were more stable and more representative of the relative values of domestic and foreign capital resources. As to the selection of the hard currency serving as common denominator, the US dollar was chosen for purposes of convenience and not because of any expectations of particular stability. Increases of costs over time were assumed to take place at a rate identical for all countries and remaining constant over time. This rule involves three assumptions:

(a) The recognition of inflation as a permanent feature of the future economic development of both industrial and developing countries, an assumption which can hardly be questioned in the light of past experience. (b) The assumption of an identical rate of inflation for all countries, which is admittedly wrong both on theoretical and empirical grounds but practically justifiable in view of the impossibility of realistic individual forecasts. The difficulty was, however, partially met by the combination of a single inflation rate with a series of alternative present-worth discount rates, a procedure more fully explained in the next section, thus giving each country the opportunity of basing its decisions on the values which it considers most relevant to its own case. (c) The assumption of a rate constant over time is also based on considerations of practical expediency.

Finally the selection of 4% as the numerical value of expected annual price growth is a compromise between the much higher values recorded by most countries in the past and the somewhat lower targets set by their governments for the future.1

1 The major exception was the rate of escalation for fuel oil which was taken at 6% for reasons explained at length in Appendix I.

D-2 The aggregation and comparison of time flows of costs was done through a discounting of their present-worth values and in all basic cases at a rate identical for all countries and assumed to remain constant in time. As in the previous case, this principle implies three decisions:

(a) The selection of present worth as a criterion. This decision must again be assessed against its possible alternative, which would have been to rank different patterns by their internal rate of return. The latter was, however, clearly ruled out since, apart from its theoretical flaws in the comparison of mutually exclusive projects, it requires estimates of benefits which the Survey deliberately refrained from making. (b) The choice of an identical rate for all countries although the time value of money and resources is likely to be different for each of them. An objection to this choice is entirely valid and it was therefore decided to use a range of discount rates, computing for each country the corresponding present-worth values and consequent rankings of alternative expansion patterns and leaving to its discretion the decision which rate appears most suitable to its own conditions. (c) The decision to assume that the rate of discount would remain constant in time may be open to theoretical objections since its value should in principle slowly decrease with higher levels of economic development and larger stocks of capital equipment. It was felt, however, that in the countries surveyed the practical difficulties involved in estimating, and in using, variable rates of discount far outweighed the possible advantages.

Finally the rates of discount and of inflation were combined into a single rate of discount equal to their difference. This considerably simplified the computational work since it was then possible to proceed on the basis of constant prices.2 For the basic case the rate of present-worth discount was chosen as 12% annual compound which was felt to be a representative average of the cost of money in most countries surveyed. Since, as was noted above, the rate of inflation was chosen as 4% annual compound, the corresponding constant price discount rate was 8%. For sensitivity studies constant price discount rates of 6% and 10% were used. The time origin for discounting was taken to be 1 January 1973.

METHODS OF OPTIMIZATION AND TIME HORIZON

In theory the selection of a lowest costs pattern of development for an electric power system requires:

(a) The choice of a method for a simultaneous optimization of the construction and operation of power plants expected to be available. (b) The choice of a time horizon or cut-off date beyond which the differences of future costs arising from alternative decisions taken during the period under review may be considered negligible when reduced to their present-worth values at the date of origin for discounting.

Among the several methods of optimization, linear, non-linear and dynamic programming, the last was originally selected as offering the best combination of theoretical consistency and realistic system description. It became apparent, however, that the amount of computer time and man-power which the systematic application of this method would require were exceeding the limited resources of the IAEA computer made available for the Market Survey. Furthermore, the margins of uncertainty affecting some of the major input data did not always warrant the costs of applying a procedure based on such a comprehensive, detailed and exhaustive approach. It was therefore decided, except for a few cases, to proceed along more empirical lines, thus achieving a substantial saving in time and man-power without an undue sacrifice

2 This procedure of using a rate of constant costs discount r' = r - i, where r is the real rate and i the rate of inflation, is strictly valid only in continuous discounting, but the errors involved in discreet discounting are negligible.

D-3 of accuracy. For each country numerous plausible patterns of power system expansion of generating capacity for the 1980 to 1989 period were developed, their operation simulated under imposed constraints and the corresponding values of total present-worth costs computed for each pattern to find the minimum cost configuration. In each system, special attention was paid to determine in advance the system configurations which past trends and future constraints made particularly plausible. The theoretical flaws inherent in this empirical search were felt to be of relatively minor importance provided sound judgement was exercised in the selection of the alternative patterns used for simulation. The selection of a time horizon was also based on compromise between theoretical accuracy and practical possibilities with the final decision substantially constrained by the latter factor. Consequently, while recognizing that a full analysis of the costs of power expansion patterns during the 1980-1989 period should theoretically extend up to a point in time when the economic consequences of alternative decisions lead to insignificant differences in present-worth values, it was also felt that detailed forecasts of development beyond the year 2000, and even beyond 1990, would not in most cases be realistic. Consequently, it was decided to take some, but not full account of future consequences by establishing for each system a single expansion plan for the 1990 - 2000 period which was then attached to each alternative plan for the 1980 - 1989 decade in the simulation and present-worth computation procedures. Furthermore, salvage values based on linear depreciation were factored in for all plants at the end of the Survey period. The use of salvage values based on straight line depreciation, a practice current in most electric utilities accounting, involves a slight departure from strict economic accounting which should be based on sinking fund depreciation. It should be noted, however, that this procedure errs on the conservative side with regard to nuclear power stations since it leads to the use of higher present-worth coefficients than those of the sinking fund method. As an example, for a power plant with a capital cost C commissioned j years before the cut-off date of the study and which is expected to have a useful life of 9 years, the present- worth values of the capital cost of the plant net of salvage value discounted at the interest rate i would be given by Vi = c[l-(l-i)(l+i)-j] according to the straight line method used in the survey, and

Vo - C *- -J—,• according to strict sinking fund depreciation. For a plant built in 1985 or 15 years before the cut-off date set at year 2000, these formulae would yield the following capital cost charges to the objective function:

\ = 0.84 C and V2 = 0.76 C

Appendix A gives a comprehensive presentation of the WASP program used for simulating system operation and, in some selected cases, for dynamic optimization.

D-4 APPENDIX E

STANDARDIZED DATA FOR GENERATING UNITS CONSIDERED AS EXPANSION ALTERNATIVES

In order to facilitate preparation of input data for the WASP program, it was decided to standardize the characteristics of the various alternative types of thermal plants which might be used to expand the power system of each of the countries being studied. It was recognized that in some countries these standardized data might not be representative of units which would actually be considered as expansion alternatives and in such cases provision was made for modifying the data as necessary. The following paragraphs describe the methodology used to develop the characteristics of the standardized alternative generating plants and the actual data used in the studies.

CHOICE OF UNIT SIZES, TYPES OF PLANTS AND NOMENCLATURE

Table E-l shows the unit sizes, types of plants and standard nomenclature used for expansion alternatives. These choices were fixed in order to achieve comparable computer outputs.

TABLE E-l. SIZES, TYPES AND STANDARD NOMENCLATURE FOR EXPANSION ALTERNATIVES

Type of plant Size Gas Nuclear Lignite Oil Coal Gas (MW) turbine

50 GT50

100 N100 L100 O100 C100 G100

150 L150 O150 C150 G150

200 N200 L200 O200 C200 G200

300 N300 L300 O300 C300 G300

400 N400 L400 O400 C400 G400

600 N600 L600 O600 C600 G600

800 N800 L800 O800 C800 G800 1000 N1T0 L1T0 O1T0 C1T0 G1T0

MINIMUM OPERATING CAPACITIES

It was recognized that thermal power plants can be designed to operate at as low as 25% of their rated capacity; for the purpose of the Survey, however, the minimum operating capacity of the standard plants was set at 50% of full load. Gas turbines were assumed to be operated at full load or not at all. Units in the fixed system (i. e. plants in the system at the start of the study period) with capacities below 50 MW were also assumed to operate only at full load and, for units of 50 MW and larger, the minimum operating capacity was taken to be that stated by the country.

HEAT RATES

Full load and half load heat rates for the standard alternative generating plants were derived from data provided by Bechtel Corporation and Lahmeyer International GmbH (see Appendix G for details of these).

E-l OPERATING AND MAINTENANCE COSTS

Operating and maintenance costs of PWR and oil-fired plants were taken from data in the open literature [1, 2] adjusted to "end of 1972 dollars" by escalating at 4%/yr. Assuming that power stations would on an average have two units per station, operating costs for single unit plants were reduced by 15% to allow for the second unit. Property damage insurance was added to these costs. In the case of nuclear plants, this was assumed to amount to 0. 25% of the capital cost and in the case of oil-fired plants to 0.1% of the capital cost. Tables E-2 and E-3 show the breakdown of operating and maintenance costs for PWRs and oil-fired plants. Gas-fired plants were assumed to have the same operating and maintenance costs as oil-fired plants, coal-fired plants were assumed to be 7% higher and lignite-fired plants 10% higher. These costs were adjusted to local conditions (i. e. lower staffing costs etc.) when warranted.

TABLE E-2. BREAKDOWN OF UNADJUSTED OPERATING AND MAINTENANCE COSTS FOR PWRs (103 US $/yr)a

Capacity (MW)

100 200 300 400 600 800 1 000

Staffing 750 800 850 860 910 960 970 Maintenance supplies and services 260 330 410 465 580 680 760 Insurance13 500 570 610 690 810 940 1 070

Total 1 510 1 700 1 870 2 015 2 300 2 580 2 800

US $/kW per month 1.26 0.71 0.52 0.42 0.32 0.27 0.23

Includes property damage and third party liability insurance.

TABLE E-3. BREAKDOWN OF UNADJUSTED OPERATING AND MAINTENANCE COSTS FOR OIL-FIRED PLANTS (103 US $/yr)a

Capacity (MW) Item 100 150 200 300 400 600 800 1 000

Staffing 500 520 540 580 630 700 780 870 Maintenance supplies and services 170 200 240 300 360 500 620 760 Insurance 60 80 95 120 150 180 240 290

Total 730 800 875 1 000 1 140 1 380 1 640 1 920

US $/kW per month 0.61 0.45 0.36 0.28 0.24 0.19 0.17 0.16

a Based on US conditions.

E-2 SCHEDULED MAINTENANCE TIMES AND FORCED OUTAGE RATES

The scheduled maintenance times and forced outage rates assumed for the alternative generating plants are shown in Table E-4, These data result in the unavailability percentages given in Table E-5. They are essentially the same as the unavailabilities experienced on plants in the USA. These figures were also used for existing plants when actual data were unavailable. It is recognized that at the present time plant availabilities in some of the developing countries are substantially lower than these values. In addition, as nuclear units and much larger sizes of conventional plant are introduced, it is likely that total (forced and

TABLE E-4. SCHEDULED MAINTENANCE TIMES AND FORCED OUTAGE RATES OF ALTERNATIVE GENERATING PLANTS

Scheduled maintenance Forced outage rate (days/yr) cy°)

Unit size Oil/Gas, Coal, Conventional Nuclear (MW) Nuclear Lignite

50 21 - 7.5 9.6 100 21 28 6,5 8.6 150 21 - 5.3 7.5 200 21 28 5.4 7.5 300 28 28 6.5 8.7 400 28 28 9,8 12.0 600 28 28 12.0 14.1 800 35 35 12.2 14.5 1 000 35 35 12.2 14.5

TABLE E-5, PERCENTAGE UNAVAILABILITY OF ALTERNATIVE GENERATING PLANTS

Unavailability (<7o) Unit size (MW) Nuclear Oil/Gas Coal/Lignite Electrical Worlda

50 - 13 15 13 + 100 14 12 14 10-13 150 - 11 13 10-11 200 13 11 13 11 300 14 14 16 11-17 400 17 17 19 17 600 19 19 21 21 800 21 21 23 21 1 000 21 21 23 21

Average for US plants as reported in Ref [3],

E-3 maintenance) outage times will be greater. This, however, is considered to be a transitory situation and it is expected that plant availabilities in the developing countries will improve with time as experience is gained with more sophisticated units until they approach those of the industrialized countries. This improvement is expected to occur within the study period of the Survey.

PLANT LIFETIME

Plant lifetimes were assumed to be 30 years for both nuclear and conventional plants. Linear depreciation of the plant investment cost was taken over this period. Since the levelized working capital component of the nuclear fuel cycle cost is treated as an addition to the plant investment cost, two years were added to the nuclear plant lifetime to correct for the fact that this working capital does not depreciate.

STUDY HORIZON

Although the time period of interest to the Survey is 1980 to 1989, the study horizon was extended to the year 2000 to allow for the influence of plants built in the second decade on the load factor of those introduced up to the end of 1989. Extension of the study horizon also results in a better approximation of the effect of escalation on the generating costs of oil- fired plants introduced in the 1980-1989 period (see also Appendix D).

TRANSMISSION COSTS

Transmission costs were not treated explicitly in the study, based on the assumption that they would be essentially the same for the alternative generating units being considered. In cases where extra transmission costs were required for the installation of a specific plant, such as a remote hydro plant, these were added to the capital costs of the plant and the available energy of the hydro plants was discounted by appropriate amounts to correct for transmission line losses.

REFERENCES

[1] KHAN, M.A., ROBERTS, J.T., "Small and medium power reactors - technical and economic status", 4th Int. Conf. peaceful Uses atom. Energy (Proc. Conf. Geneva, 1971) 6, IAEA, Vienna (1971) 57. [2] NUS Corporation, Guide for Economic Evaluation of Nuclear Reactor Plant Designs, USA EC Rep. NUS-531 (1969). [3] Electrical World (1 Nov. 1971) 47.

E-4 APPENDIX F

LONG RANGE FORECASTING OF THE DEMAND FOR ELECTRICAL ENERGY

H. Aoki

The basic objective of an electric power program is to provide sufficient power to meet the demand and to do so as economically as possible. In view of the time required for planning and constructing power plants, a plan for installing new power generation, transmission and distribution facilities should be established at least ten years in advance of the actual required date. The formulation of a reasonably reliable method for long range forecasting of the likely demand for electrical energy is therefore of vital importance. A number of methods have been used and these are briefly reviewed below. The particular method used for providing forecasts for the countries covered by the Market Survey is described in detail.

VARIOUS METHODS The methods used fall into two groups. In the first the country is considered in isolation, and the forecast is based upon past trends in that country.

(a) Simple extrapolation

The average growth rate of the demand for electrical energy over the past years is determined. A factor, usually less than or equal to 1, is applied to the historical growth rate, and this modified growth rate is assumed for the future. Clearly the difficulty with this method lies in the determination of the modifying factor to be used for a particular country, particularly if it is a developing country.

(b) Correlation between the national economy and the energy demand

This involves taking some measure of the national economy, such as GNP or GDP, and comparing its historical growth with that for the demand for electrical energy. The past relationship between the two is then extrapolated into the future. Again this method is not particularly useful in the case of developing countries which are usually in a transitional stage of development in respect of their national economies and of their electrical energy demand. Both methods can be useful for comparatively short range forecasts.

In this method various sectors of the country's economy and specific industries in the country are studied and estimates made of the likely individual future demands for electrical energy. These separate estimates are then added in order to give a complete forecast for the country. Again, this method is useful for short range forecasting but for long range it involves the making of sweeping assumptions about the long term development of particular industries and, whilst giving the appearance of accuracy, is in the end no more reliable than the first two methods. The next three methods depend upon comparisons with one or more other countries.

(d) Sentiment method

This involves basing the forecast for a particular country upon either the forecast for what is believed to be a closely comparable country, or upon the recent experience of a country believed to be similar but rather more developed. Clearly the accuracy of this

F-l 160- FOR THE YEARS 1961 TO 1968

U0-

120

< 100- z UJ

80- o a: o u. o 60 UJ i— <

40- o a: o 20-

-20-

-10 -5 0 5 10 15 20 25 GROWTH RATE OF GNP - %yr

INDIVIDUAL YEARS

X 80 - 1965 60 - 1968

60- 40- X

40- 20 - X X X X ^ 20- XxX 0 - <"• ^.—"— X

*KXx 0- x x x*P^i X XXX —I— —I— —r~ -10 10 20 30 -10 10 20

FIG. F-l. CORRELATION BETWEEN GROWTH RATE OF GROSS GENERATION AND OF GNP.

F-2 FOR THE YEARS 1961 TO 1968

<1— o (L •r O OL UJ o Q. o ID 2 CM o

O O :RA T UJ 00 Z LU o D O ici r a: o UJ O UJ O CM t/) o ce O o O LL (O o 1

Q: »-

oOr O

I it $THb\tf I 3 iss^eg^cf 2 3 4 GROSS ELECTRICITY GENERATION PER CAPITA -kWh

INDIVIDUAL YEARS

1965 1968 CD

CO 43-2 1 X

O o X u> D X X X X > X ™ x x X M CD X XX X x x y X x - X x x X x x, # x x y y v •> x^xxxx X X X K- X />- A .8 0 1 X > X X XX^X CO 1 X . X 5 X X X - X K X xX ^ o o 00 X o- CsJ 1 1 X X o" XX X o" to CO to itf \

FIG. F-2. CORRELATION BETWEEN GROWTH RATE OF GROSS GENERATION AND OF GROSS GENERATION PER CAPITA. .

F-3 method is completely dependent upon how comparable the reference country (or countries) really is. In this comparison it is necessary to take into account, for instance, the kind of energy resources available in the two countries since they might be similar in all respects except that one has a great deal of potential hydroelectric power which can be developed cheaply and the other has little potential or potential that would be costly to develop. The method is superficially attractive, but for the reasons stated cannot be recommended.

(e) World-wide correlation between growth rate of GNP and of energy generation

In this method the growth rate of GNP is plotted against the growth rate of electrical energy generation for as many countries as possible. If a correlation is seen to exist, and given that a reliable forecast of the future GNP can be made for the country being studied, this correlation can be used to forecast the future energy demand. Such data are plotted in Fig. F-l for 111 countries, for the years 1961 to 1968 and for the two individual years 1965 and 1968. It will be seen from this figure that the correlation is very poor and this fact is confirmed by statistical analysis of the data. As a result this method cannot give reliable forecasts of electrical energy demand.

(f) World-wide correlation between the per-capita generation of electrical energy and the rate of growth of per-capita generation

This method would be used in a similar fashion to (e). The data for 111 countries are plotted in Fig. F~2. Clearly the correlation is a little better than that obtained for (e), but it is still inadequate for obtaining accurate forecasts of electricity demand.

THE AOKI METHOD USED FOR THE MARKET SURVEY

This method is similar to the last two described in that it is based upon data from a large number of countries. It is similar to method (e) in that it assumes that there must be a connection between generation of electrical energy and the state of the national economy. But it introduces the concept that the per-capita values of these variables, rather than the absolute values should be correlated. Figure F-3 shows a plot of electricity generation per capita against GNP per capita for 111 countries. The historical GNP data used in this plot were obtained from the IBRD World Table, January 1971, and are expressed in terms of constant prices (1964 US $). The correlation between these two quantities is clearly much better than the one achieved in either method (e) or (f) and the correlation coefficient of the straight line fit shown in Fig. F-3 is remarkably high. Since the data at the upper and lower end of the figure tend to fall below this line, it is obvious that a better fit could be obtained by using a polynomial. This has been done in effect by determining the best straight line fit over a series of intervals of per-capita GNP as shown by Fig. F-4 for the 1968 data. It is important to note that both the single correlation lines and the curves obtained from the series of straight linesare virtually the same whether determined for any single year in the period 1961 to 1968 or determined from the data for all eight years grouped together (see Fig. F-5). Thus there is evidence that the relationship is stable and can be accepted as "universal". The consequent recommended relationship is plotted in Fig. F-6. Close examination of the individual country lines in Fig. F-3 shows that, in general, if the initial point representing a particular country falls above or below the line, subsequent points at higher values of GNP per capita approach more closely to the trend line. It is therefore possible to draw a number of "indicative" lines on each side of the main trend line which will indicate the likely path that will be followed by countries whose present state does not lie exactly on the line. Such indicative lines are drawn in Fig. F-6. The use of the Aoki method has essentially been indicated above. A copy of the master trend curve is taken. The available historical data for the country being studied are plotted

F-4 10 000 HISTORICAL ELECTRICITY DEMAND IN THE WORLD 5 000 (111 COUNTRIES) 1961 TO 1968 3 000

2000 2000

< 1000 o JC 500 1 400

1— 300 a. o 200 PE R 100 DO N

50 GENERA "

"BEST" STRAIGHT r*o 30 oc LINE FIT I— o UJ _l UJ (/) en o cc o

2 f 30 40 50 100 200 300 500 1000 2000 3000 5000

PER CAPITA GNP AT FACTOR COST (1964 US|)

FIG. F-3. CORRELATION BETWEEN GROSS GENERATION PER CAPITA AND GNP PER CAPITA FOR 111 COUNTRIES FOR 1961 - 1968.

F-5 en- en- =r<\l- 1961 1963 o y O X >/ 05- x)f x x x^'^ x in- a>" X MJ/X tn- x><^ n- x^y oJx#< * ^ X x xx xj^x* O x X ,x ^x^ 3- h-- yftx x (D- xx J-- in- X JXXK n- ^5 X m- -rj- x y\ y O yf XX X O X z>fx x X x x "y. x X 1 'x 1 X v x X * x x x x rj- X <"V- —r I i i I | 1 1 1—I—I : : : 1 \ ' ' r 1 1 1 TTT , 1 1 1 1 1 1 1 1 1 " J 1 5 S 7 3 3 'l 0 2 3 H 557-5 3 10 2 3 U 1 5 6 7 3 3 'l 0 2 3 156783'lO 2 3 0 D

rnTJ- X x v x#.y> •X x x **>^ X X O o x SK^ x x*x*^f X ll x^ in- xx |/^ -•3-- x cn- v JfXx =r- X x^ ^^ X x X xx^Tx xJJy: O O x x * J^ X< X

(O$-i L">- x 1 •*- ^ J* X o- LT- <"- y* JS x x/Tx X O o X xx X X X ^to1- i.i -* x xx £: jr — xX ro- i—i—i i i i 1 1—i—i—i i i i 5 S 7 y 3 'j Q ! 2 3 5S7S31Q 2 3 '4 567-3 3 '10' i 3 2 3 i S 5703'10- 0 D

CO- =•- 1967 1968 o X . O X p= y- £'- r-- vpf X m- ^>- X zr- 31- tn- >x> X X *>* x xx/4. x^ O o W X VJ** x (D- x xx^ l/l- v X jg/^ £^V o f.'-J- ,« J?*xx O X o **$\ X xx yfCx X X<^< X la- X ^ X xx JJx * x >K <"- %'- x y^ x •J* x X Jffix x O a Xxx • x x x ?: X in- X

! •15S753 'lO* ^ 3 it 06759 'lO' 2 3 5 6 733'l0 2 3^5 S^^o' 2 3 ll 0 0 FIG. F-4. REGRESSION ANALYSIS OF THE PLOT OF GROSS GENERATION PER CAPITA VERSUS GNP PER CAPITA FOR 1968.

F-6 10 000

5 000 /

3 000 //

^ 2 000 a. // // // // 1000 // - // I f/ o g 500 UJ - "Z / Ul 300 o 200 / CIT Y on - /

100 /• ELEC 1 //

- //

GROS S 50 < i— / 30 J CAP I cc / UJ 20

10 /

1 1 1 1 1 1 1 1 1 1 30 50 100 200 300 500 1000 200 300 5000 PER CAPITA GNP AT FACTOR COST (1964 US$1

FIG. F-5. CORRELATION FOR INDIVIDUAL YEARS.

F-7 15 000

10 000

< r— 5Z 5 000 o .C 3 000

1 2 000 Z o I— QC UJ Z 1 000 UJ o > o 500 o; 400 »o— UJ 300 UJ RECOMMENDED 200 UNIVERSAL CURVE o a: o < 100 o oc UJ OL 50 40 30

20 50 100 200 300 500 1000 2000 3000 5000 PER CAPITA GNP AT FACTOR COST (1964 US $)

FIG. F-6. RECOMMENDED RELATION BETWEEN GROSS GENERATION PER CAPITA AND GNP PER CAPITA.

F-8 on the diagram. The future is then forecast by extrapolating this line following the main trend line or one of the indicative lines as appropriate. Given that a forecast of the future growth of GNP per capita is available, the future demand for electrical energy is then calculated from this extrapolation. This is done for the Survey countries in Figs F-7 and F-8.

10000

5 000 Q. 5 4 000 2001 3000 7/ I 2000 g 1000 W UJ o 500 400 71/ 300 HI CO CO 200 O a: o < 100

< o ID Q. 50 40 H.A< K»

40 50 100 200 300 400500 1000 2000 3000 5000 PER CAPITA GNP AT FACTOR COST (1964 US $)

FIG. F-7. SURVEY COUNTRIES PLOTTED ON RECOMMENDED CURVE.

F-9 10000

£L 5 000 4 000 •* 3000 z o 2000

LLl 1000 o on t— 500 o 400 UJ 300

200 GROS S E l < i— 100 CAP! ' PE R 50 40 30

40 50 100 200 300 400 500 1000 2000 3000 5000 PER CAPITA GNP AT FACTOR COST (1964 US $)

FIG. F-8. SURVEY COUNTRIES PLOTTED ON RECOMMENDED CURVE.

F-10 APPENDIX G

BASIS FOR HEAT RATE DETERMINATION

To permit an evaluation of the performance of various types of thermal power plants, the heat rates for energy conversion are required. Experienced power plant designers were requested to supply heat rates for modern plants of the type and size used in the expansion program for the various systems studied. The most detailed response was received from the Bechtel Corporation and the heat rates used in the study are based on the Bechtel data. These data were confirmed by information received from Lahmeyer International and also by data on existing plants collected in the participating countries. The net and gross heat rates for pressurized water reactors (PWR) of capacity from 100 to 1500 MW and for coal, lignite, gas and oil stations from 100 to 1000 MW are listed in Tables G-l to G-4. To be consistent with the country data on the fixed systems and on load forecasts, the gross heat rates were used in the study. The net heat rates are given to permit people familiar with design data to appreciate to more easily the values used. The net heat rates for light water PWRs are calculated on the following bases:

(1) The use of a seven-heater cycle utilizing a two-reheat turbine is assumed. There are two high pressure heaters whose cascaded drains, combined with those of the third heater, are pumped into the reactor feed pump suction. Reactor feed pumps are driven in all cases by auxiliary turbines. All data on nuclear steam supply systems (NSSS) are based on information obtained from the Combustion Engineering Company (CE). This NSSS generates saturated steam at 70 kg/cm2 (a 1. 5 kg/cm2 pressure drop to the turbine stop valve was assumed in all cases). Final feed-water temperature is 230°C.

(2) Auxiliary power requirements for reactor sizes'of 800 MW and above are based on information obtained from CE. Auxiliary power requirements for reactor sizes below 800 MW are assumed to be 1. 75% of output at the generator terminals at rated power and condenser pressure of 3. 0 in Hg abs. In all cases, auxiliary power for the balance of plant is broken down in the following fashion:

Rated load 50% load

Main transformer losses 0.40% 0. 70% Circulating water system (once through) auxiliary power 0. 30% 0. 60% Balance of plant exclusive of main transformer & circulating pumps

Total balance of plant auxiliary power

(3) It should be noted that all heat rates assume that steam is generated at 70 kg/cm2. Historically, the smaller units in the range 400 to 800 MW generated steam at 55 kg/cm2 (770 lb/in2abs. ); later steam pressures for larger units were increased to 58 kg/cm2 (815 lb/in2abs.), and then to 63 kg/cm2 (900 lb/in2abs.). Thus the heat rates in this study would appear better in comparison. However, CE states that were they to offer any of these smaller units today, they would quote them all on the basis of steam generated at 70 kg/cm2 (1000 lb/in2abs. ). Heat rates were computed on the basis of using in all cases the smallest turbine exhaust consistent with turbine exhaust loading limits as specified by the two US turbine manufacturers.

G-l TABLE G-l. NET HEAT RATES FOR FOSSIL-FUELLED PLANT'

Full load Half load Incremental Type Power Heat rate Power Heat rate energy rate° (MW) (kcal/kWh) (MW) (kcal/kWh) (kcal/kWh)

Coal 100 2 443 50 2 592 2 294 150 2 421 75 2 551 2 291 200 2 378 100 2 501 2 255 300 2 360 150 2 474 2 246 400 2 358 200 2 463 2 253 600 2 350 300 2 467 2 233 800 2 352 400 2 47-2 2 232 1 000 2 348 500 2 483 2 213

Lignite 100 2 666 50 2 832 2 500 150 2 642 75 2 787 2 497 200 2 595 100 2 732 2 458 300 2 574 150 2 702 2 446 400 2 573 200 2 690 2 456 600 2 565 300 2 694 2 436 800 2 567 400 2 701 2 433 1 000 2 561 500 2 712 2 410

Gas 100 2 529 50 2 671 2 388 150 2 506 75 2 629 2 383 200 2 461 100 2 577 2 345 300 2 443 150 2 551 2 335 400 2 441 200 2 539 2 343 600 2 433 300 2 593 2 323 800 2 435 400 2 549 2 321 1 000 2 431 500 2 560 2 342

Oil 100 2 390 50 2 528 2 252 150 2 368 75 2 487 2 249 200 2 327 100 2 438 2 216 300 2 309 150 2 413 2 205 400 2 307 200 2 403 2 211 600 2 300 300 2 406 2 194 800 2 302 400 2 412 2 192 1 000 2 297 500 2 422 2 172

a Based on information received from Bechtel Corporation. (Full load heat rate) (Full load power) - (Half load heat rate) (Half load power) Incremental energy rate == (Ful; l load;— powe* r - Half load; power): *

G-2 TABLE G-2. GROSS HEAT RATES FOR FOSSIL-FUELLED PLANTS2

Full load Half load Incremental Type Size heat rate heat rate energy rate (MW) (kcal/kWh) (kcal/kWh) (kcal/kWh)

Coal 100 2 311 2 411 2 211 150 2 290 2 374 2 206 200 2 233 2 306 2 160 300 2 280 2 361 2 199 400 2 233 2 351 2 115 600 2 270 2 354 2 186 800 2 272 2 360 2 184 1 000 2 268 2 370 2 166

Lignite 100 2 512 2 615 2 409

150 2 490 2 574 2 406

200 2 427 2 500 2 354

300 2 478 2 560 2 396

400 2 427 2 549 2 305

600 2 468 2 553 2 383

800 2 470 2 559 2 381

1 000 2 465 2 570 2 360

Gas 100 2 420 2 526 2 314

150 2 404 2 486 2 322

200 2 344 2 415 2 273

300 2 393 2 473 2 213

400 2 344 2 461 2 227

600 2 383 2 465 2 301

800 2 385 2 471 2 299

1 000 2 381 2 482 2 280

Oil 100 2 290 2 388 2 192 150 2 270 2 347 2 193 200 2 213 2 280 2 146 300 2 259 2 335 2 183 400 2 213 2 324 2 098 600 2 250 2 328 2 172 800 2 252 2 334 2 170 1 000 2 248 2 344 2 152

Based on information received from Bechtel Corporation.

The net station heat rates for fossil-fired units are based on the following assumptions: (1) Steam generator efficiencies are based on 144°C exit gas temperature at full load, and on the following fuels: (a) bituminous coal at 5544 kcal/kg (10 000 Btu/lb), (b) lignite at 3465 kcal/kg (6250 Btu/lb), (c) low sulphur or "bunker C" fuel oil, (d) natural gas.

G-3 TABLE G-3. PWR NET HEAT RATES

Full load Half load Net generator output Heat rate Net generator output Heat rate Incremental energy rate (MW) (kcal/kWh) (MW) (kcal/kWh) (kcal/kWh)

100 2 591 50 2 840 2 342

200 2 590 100 2 834 2 346

300 2 589 150 2 828 2 350

400 2 589 200 2 822 2 355

600 2 587 300 2 811 2 363

800 2 585 400 2 799 2 371 1 000 2 583 500 2 786 2 380

a Based on information received from Bechtel Corporation.

TABLE G-4. PWR GROSS HEAT RATES

Full load Half load Size heat rate Size heat rate Incremental energy rate (MW) (kcal/kWh) (MW) (kcal/kWh) (kcal/kWh)

100 2 504 50 2 651 2 357

200 2 503 100 2 648 2 359

300 2 502 150 2 645 2 361

400 2 502 200 2 643 2 362

600 2 501 300 2 637 2 365

800 2 500 400 2 632 2 368

1 000 2 499 500 2 627 2 372

a Based on information received from Bechtel Corporation.

(2) All steam generators are balanced draft, with both forced and induced draft fans. (3) Flue gas electrostatic precipitators are included for coal and lignite steam generators only. Precipitator power requirements are assumed to be 0. 20% of rated generator load at full load, and 0.40% of generator load at half load. Flue gas SO2 scrubbers and associated auxiliary power have not been included. (4) Turbine throttle conditions are assumed to be 125 kg/cm2 and 537°C with reheat to 537°C for the 100 and 150 MW units; and 168 kg/cm2 and 537°C with reheat to 537°C for the 200 MW to 1000 MW units. (5) All turbines are tandem compound, with the low-pressure turbine frame-size chosen for the closest possible approach to maximum allowable exhaust-steam flow loading, to obtain the required unit generator load rating. (6) Boiler feed pumps are motor driven for the 100 to 200 MW units and steam turbine driven for the 300 to 1000 MW units. (7) A once-through condenser cooling water system has been assumed (no cooling towers), with the circulating water pumping power assumed to be 0. 25% of the rated generator load at full load, and 0. 50% of the generator load at half load. (8) The main transformer loss has been assumed to be 0. 40% of rated generator load at full load, and 0. 70% of generator load at half load. The net station heat rates are at the high voltage side of the main transformer. (9) All full load heat rates are 3. 0 in Hg abs. condenser pressure and all half load heat rates are at 2. 0 in Hg abs. condenser pressure. Note: Assumptions 8 and 9 apply also to the PWR units.

G-4 APPENDIX H

GENERALIZED POWER SYSTEM ANALYSIS APPROACH TO DETERMINE SYSTEM LIMITATIONS

Associated Nuclear Services Ltd(ANS)*

Power system anaysis plays an important role in determining the technical constraints to be taken into account in system design and planning studies and powerful and sophisticated techniques are available for evaluating such aspects as power flows, short-circuit levels, transient stability and frequency stability. However, the limited extent and wide tolerances associated with system data normally available for long-term planning studies of the present nature often contrast considerably with the sophistication and accuracy of these analysis techniques. Fortunately, in a study involving the comparison of a number of expansion plans, the optimization process is relatively insensitive to system data over the typical range encountered on present-day networks. A simplified approach to system analysis is thus sufficient for the Market Survey purposes, provided this is applied consistently. The technical constraints of major interest to the Survey are transmission limitations and limits to generator unit size. This appendix describes the generalized methods adopted for the assessment of these constraints in the majority of countries. In one or two countries either or both aspects had been studied in sufficient depth by the supply authority or their consultants over the study period (1980 to 1989) and only a comparative checkis necessary. Details of the application of the methods (where necessary) and results are given in Section 11 of the Country Reports.

TRANSMISSION LIMITATIONS

The main functions of transmission may be categorized as follows:

(i) Bulk distribution/collection within a load/generation region, (ii) Point-to-point bulk transmission from a 'remote1 power station to a load centre (may be long or short distance), (iii) Inter-regional bulk transmission (i.e. an extension of (ii) to a group of remote power stations). (iv) Inter-regional interconnection, (v) International interconnection.

The normal transmission limitations encountered are excessive short-circuit levels, thermal ratings and transient stability limits. The varying importance and generalized approach to the assessment of these limits with reference to the above categories is discussed below.

Short-circuit levels

Where possible the short-cricuit rating(s) of grid switchgear for the various categories above are generally chosen with sufficient margin to cover system development into the foreseeable future taking into account average transmission distances, load density and the relative expected proportion of local and remote power generation. Excessive short-circuit levels are most commonly encountered in very high load density areas (category (i)) particularly where the grid system is predominantly cabled (small transmission impedances) and it has been found necessary to employ switchgear of the maximum commercially available short-circuit rating. Also, increasing the proportion of load fed from generation connected at local grid voltage level will aggravate the grid short-circuit problem.

London, United Kingdom.

H-l The normal eventuality of excessive short-circuit levels is the introduction of a higher voltage grid, other measures such as system segregation merely introducing a time delay which will be approximately equal for all plans. Hence the timing of a higher grid voltage in a particular system as dictated by short-circuit ratings will tend to be a common factor in all practical plant programs and will generally have little influence on the economic comparison of programs. Thus it was only necessary to check grid switchgear ratings against normal practice and where applicable to identify any special limitations or requirements.

Load flow transient stability

To achieve a reasonable standard of supply security the transmission grid should be capable of meeting the normal and 1st contingency power flow requirements throughout each plan without exceeding cricuit thermal ratings, loss of system stability (system splitting) or recourse to load shedding. Information on standard grid circuit thermal ratings was generally available from each country. Transient stability limits were estimated using the 30° transmission angle criterion. This is a guiding criterion which, for the typical fault types and fault clearance times encountered on present-day systems, will ensure the retention of transient stability in the majority of cases. In the few cases where unforeseen difficulties arise, it is usually possible to retrieve the situation by introducing or increasing shunt and/or series compensation. With transmission costs of typically 15% to 20% of total plant costs and compensation costs at 10% to 15% maximum of transmission costs, the rare maximum error thus involved in this approach is of the order of 2% of total plant costs. This is regarded as being well within the accuracy of the capital cost data available to the Survey and there is no justification for a more elaborate approach to transient stability assessment, barring perhaps some well recognized exceptions. The most common restriction to power flows in category (i) transmission are the thermal capabilities of circuits. However, this will tend to be a common factor in all generating plant programs considered for a particular country and detailed load flow studies within major load or generation regions were not necessary for the Market Survey. For category (ii) transmission, the power flow requirement was simply estimated from the capacity of the power station less any local load to be supplied. Inter-regional power flow requirements (categories (ii) and (iii)) were determined by a simple regional plant/load balance tabulation taking into account generating set size and outage criteria and varying hydrological conditions. The number of transmission circuits at grid voltage to meet the power flow requirements so determined for categories (ii), (iii) or (iv) was then estimated to sufficient accuracy, taking into account thermal ratings, transient stability limits and transmission security criteria. If the number of circuits was excessive, then a higher voltage was considered and first establishment costs and also step-down transformer capacity were taken into account. A further consideration in determining the capacity of category (iv) transmission is the integrity of the interconnected system following faults or a sudden loss of load or generation. Experience of interconnected systems in particular in the USA and the Scandinavian countries [1, 2] indicates that for a reasonable stability performance the capacity of system intercon- nectors should be at least 10% of the installed generating capacity of the smallest of the two systems interconnected. This was used as a guiding criterion for analysis purposes. Details of any existing or proposed international interconnections (category (v)) were obtained from the Survey countries. In all cases these were found to be of insufficient capacity to have any noticeable influence on the Survey results.

LIMITS TO SET SIZE

The economics of scale play a major role in reducing the specific cost of installed generation and this is particularly so for nuclear power generation. On the other hand, increased unit size has associated penalties in system requirements such as generation and transmission reserve capacity. Thus there exists an optimum size for overall minimum cost of power delivered to the consumer [ 1 ].

H-2 The effects of increased unit size on the transmission system are taken into account in the network analysis described in this appendix. Any special transmission requirements can be allowed for by adjusting plant input data to the WASP computer program as described in Appendix E. The effect of increased unit size on non-availability rates and generation reserves can be directly allowed for in the corresponding input data items of the WASP computer program as required by the loss-of-load probability routine described in Appendix A. In this manner the 'economic optimum' set size can be determined. However, in addition to the economic optimum set size there is what may be termed a 'technical limit1 set size (or reactor size in the case of nuclear stations) dictated by the permissible disturbance effects following the sudden loss of the largest generating unit. In cases where this technical limit is less than the economic optimum, (which is highly probable in smaller systems) this can have a dominant influence on the economics of introducing large units into such systems. The system frequency transient following sudden loss of a large generation unit has been found of prime interest in the assessment of this technical limit. The complete representation of this transient, termed 'frequency stability', is very complex, but a simplified analysis method and computer program was developed by ANS for the sudy of typical system response to sudden loss of generation. Although approximate, the analysis technique is regarded as adequate for the Market Survey purposes, bearing in mind the relatively large tolerances in data inherent in a forecasting exercise. The technique and computer program are described in the following paragraphs.

The average system frequency model

The dynamic response of a power system to a sudden loss of generation is generally characterized by two distinct components of power variation in the period of 10 to 20 seconds immediately following the disturbance. These are the faster transient oscillations in synchronizing power (time period typically 1 - 2 s) which arise due to angular disturbances from the steady state and the slower variation in prime mover power (time period typically 10- 20 s) due to the primary regulation effects of the governor/turbine response to frequency change. The ability of a system to remain in synchronism following a given angular disturbance is mainly dependent on the transfer impedances between sources, i. e. on the transmission network. System faults will usually give rise to much larger angular deviations than loss of generation and will thus dictate the requirements of the transmission network for retention of transient stability. Thus, provided the transmission network has been designed with due regard to transient fault studies and the emergency redistribution of power flow resulting from plant outages, it is reasonable to assume that synchronous stability will be retained following a sudden loss of generation. (A possible exception to this premise is the case of a sudden loss of generation immediately following a severe system fault. However, such second contingency events are not considered here.) Assuming that the system remains in synchronism then, neglecting losses (which may be assumed constant throughout the disturbance), the rate of change of stored kinetic energy (i. e. frequency) at any instant is equal to the difference between power input to the system (i.e. prime mover power) and power output (i.e. load),

(2HT)(f8 where HT is the total inertia constant of connected machines including rotating loads (typically 3.0 to 5.0), EPmk is the sum of prime mover input power of connected generators, PL is the total connected load, fa is the average system frequency. All quantities are in p. u. on the base of nominal system frequency and total nominal power of connected generation. Since the system is assumed to remain in synchronism the transmission network may be neglected and Eq. (1) may be modelled by a number of prime movers and their generating units feeding a single block load as indicated in Fig. 1 and referred to as 'the average system frequency model' [3] . Simplified equations modelling the variation of prime mover power and load are described in the next section.

H-3 rm1 m2 tn3

RATE OF CHANGE OF AVERAGE

SYSTEM FREQUENCY fa IS GIVEN BY:

df (2HT)(fa) a - dt

FIG. H-l. AVERAGE SYSTEM FREQUENCY MODEL.

Prime mover and load regulation

Maximum frequency dip before recovery (if it occurs), the time of maximum dip and the amount of load shed (if load shedding is permitted) are the main items of interest and thus the following assumptions can be made:

(i) Non-regulating base load units are assumed to have constant power output. (ii) Only the governor/turbine response of regulating units is considered. Boiler response is neglected in thermal plants. (iii) Secondary regulation is neglected. (iv) Governor response is based on average system frequency. (The oscillating component due to synchronizing swings is generally at a much shorter time period than the governor/turbine response time and does not appreciably affect the prime mover output.) (v) The total load PL is assumed to depend only on average system frequency. Variations due to the oscillating component arising from synchronizing swings are neglected. Load variation with voltage, if desired, can be sufficiently represented by conversion to an equivalent variation with frequency.

Three types of regulating units are modelled:

(a) Thermal — non-reheat (b) Thermal — reheat (c) Hydro including pumped storage

For the time period of interest (about 10 s) thermal units will generally permit faster power change rates than hydro units, but with a limit on sustained change (typically up to 15%

H-4 of nominal power). Hydro units on the other hand can give much larger sustained variations in output approaching their nominal rating with total response times of typically 10 to 20 seconds.

(a) Thermal — Non-reheat model

It is assumed that the disturbance is of sufficient magnitude to drive the steam valve to its limiting position at constant rate. The time constant of a non-reheat turbine may be neglected and thus the change in power output of this type of unit may be represented to a first approximation by the equation

with limit of Plc

where Plc is the maximum permissible power change, T-^ is the time for the valve to move to its limiting position., t is the time from loss of generator

(b) Thermal— Reheat model

As for the previous type the movement of the steam valve may be approximated by the equation

= V2 |-7p (t) with limit of P2c (3)

where P2c is the maximum permissible power change, T2 is the time for the valve to move to its limiting position. The change in power output of this type of regulating unit may thus be represented by

_ 1 + (m)(Th)(p) . ) (A) *2 " l + (TJ(p) {^> W where m is the proportion of power developed by the high pressure turbine Th is the reheat time constant p is the Laplace operator

The maximum permissible power change for both reheat and non-reheat type generation will depend on the allocation of spinning reserve but will be typically about 10% of the nominal power of the generation block and may lie in the range 5% to 20%. The valve motion time is typically one second and may vary between 0. 5 and 1. 5 seconds. The factor m is typically 0. 3 and the reheat time constant Th may lie in the range 5 to 12 seconds.

H-5 (c) Hydro model

In Ref. [4] a simplified transfer function is derived which gives a very good approximation to the response of a hydro governor with dashpot. From this the change in gate opening may be represented by the equation

T + Td (6 + St) where T, = —K& j-^

Td is the dashpot time constant (typically 5 s, range 2. 5 - 25 s), Tg is the governor response time or the inverse of governor open loop gain (typically 0.2 s, range 0.2 - 0.4 s), 6 is the permanent droop (typically 0. 04 p. u., range 0. 03 - 0. 06 p. u.), 6t is the temporary droop (typically 0. 31 p. u., range 0. 2 - 1. 0 p. u,), CTa is the average frequency deviation ( = fa - f0), Pn3 is the nominal rated power output of regulating hydro generation, P3c is the maximum available change in power output (hydro spinning reserve).

Thus the change in power output for this type of regulating unit is given by

P 1 - (Tw)(p) .. 3 {KJ) " l + 0.5(Tw)(p)

where Tw is water starting time and is inversely proportional to water head and directly proportional to penstock length. Typical values of Tw lie in the range 0.5 to 5.0 seconds. The above model was also used to represent pumped storage plant operating in the generating mode.

Load regulation model

The variation of load with frequency may be represented by an equation of the type

PL = (l + («)(

where P is the total connected load at t = 0 and fa = fQ, Ps is the load shed as function of frequency and time,, a is the load frequency regulation coefficient.

In those countries where load shed schemes are in existence, frequency settings and the amount of load shed for each stage were based accordingly. In other cases typical values were assumed. The determination of whether or not load shedding occurs is generally the prime factor of interest and thus the first stage frequency setting is the major item of load shed data. This is typically 48. 5 to 49. 0 Hz for 50 Hz systems and 58. 5 to 59. 0 Hz for 60 Hz systems.

H-6 The special case of pumping load being shed by under-frequency detection can be included in the load shedding scheme. InRef. [5] a range of values for the load/frequency regulation coefficient from 0 to 2. 5 is given. The effects of load/voltage regulation can generally be adequately represented by increasing a. Thus a typical value for a of 2.0 was used except where more accurate information was available from the country studied.

Total regulation

The total prime mover power of connected units at instant t is given by

Pm (8

where £Pmk0 is the pre-disturbance power output of connected generating units excluding the lost generator, and Pm = Pj + P2 + P3 is the total change in prime mover outputs of connected regulating units. 8 Let the loss of generation be AP (= P. - EPmk0) and since • • j - = —-&-, Eq. (1) becomes

(2H1)(fa) &*• = Pm - AP - («)(aa)(PL0 - P ) + P (9)

The effect of variations in f a on the solution of Eq. (9) is small and may be neglected, hence

Ps ) dO) «(PL0 - P.) + (2HT) (p) ^ - AP + Ps

The computer program

The computer program AVSYF (Average system frequency) for the step-by-step solution of Eq. (10) has been obtained by appropriately "patching" an existing digital program representation of an analogue simulator. Transfer functions of the type of Eqs (4-6), integral functions and limit functions exist as standard routines. Integration is performed by a simple three-step method, but provided a small enough time step is used, accuracy is sufficient. The program also includes a plot routine which permits an immediate plot of the frequency variation to be obtained as output.

REFERENCES

[1] CASAZZA, C.A., HOFFMAN, Ch., "Relationship Between Pool Size, Unit Size and Transmission Requirements", CIGRE (1968) Paper 32 - 09. [2] OLWEGARD, A., "Improvement of System Stability in Interconnected Power Systems", CIGRE (1970) Paper 32 - 17. [3] CHAN, M.L., DUNLOP, R.D., SCHWEPPE, F., "Dynamic Equivalent for Average System Frequency Behaviour Following Major Disturbances", IEEE Winter Meeting, New York (1972) 1637. [4] RAMEY, D.G., SKOOGLUND, J.W., "Detailed Hydro Governor Representation for System Stability Studies", IEEE PAS-89 No.l, (1970) 106. [5] ASHMOLE, P.H., FARMER, E.D., MYERSCOUGH, C.J., "The Dynamic Response of the Load to Changes in System Frequency", SUPEC (1972) 36.

H-7

APPENDIX I

FUTURE FOSSIL FUEL PRICES R. Krymm

INTRODUCTION

Although practically all countries covered by the Market Survey possess and exploit domestic fossil fuel resources, fuel oil either imported or derived from imported crude remains in most cases the main competitor of nuclear fuels for future electric power production. This fact alone suggests the use of fuel oil as the "reference fuel" and the validity of this assumption is further strengthened by the tight supply and demand relationship which is expected to prevail for oil products in the foreseeable future. The latter consideration suggests that the few Market Survey countries which are domestic producers of oil and gas in substantial quantities would be perfectly justified in pricing these resources on the basis of opportunity uses; that is, on the basis of thermal costs parity with imported fuel oil with due correction for transportation expenses. Also, prices of coal and lignite are dependent on local conditions and must be considered separately in each specific case. It is, therefore, not surprising that the bulk of this section is devoted to the problem of costs and prices of crude and fuel oils entering international trade. It was, however, clear from the beginning that the fuel oil picture in developing countries could not be seriously studied without reviewing the world-wide structure of the oil industry and its rapidly changing trends. It was, therefore, decided to consider in turn:

(1) The present and expected demand and supply structure of crude oil and the major producing and consuming areas. (2) The changing cost and price structure of crude oil and its future trends. (3) The cost of transport of oil by tanker and pipelines. (4) The relationship between crude and oil product prices. (5) The treatment of domestically produced fossil fuels.

DEMAND AND SUPPLY OF CRUDE OIL

Table 1-1 shows the actual 1970 and estimated 1980 demands for oil in major areas of the world. The forecast is based on conservative rates of growth and the average annual rate of 5.4% for the world should be viewed against the 7. 8% rate which prevailed during the 1950-1970 period.

TABLE 1-1. PAST AND ESTIMATED DEMAND FOR CRUDE OIL (106 t)

Rate of growth 1970 1980 6 of demand (106 t) (10 t) C7») USA 750 1 160 4.5 Western Europe 600 980 5 USSR and Eastern Europe 390 700 6 Japan 200 400 7 China 20 80 15 Rest of world 300 500 5

Total world 2 260 3 820 5.4

1-1 TABLE 1-2. WORLD ESTIMATED CRUDE OIL PRODUCTION3

1970 1971 1972 "Jo Change 1972: Countries 103 t 1971/72

NORTH AMERICA0 USA 533 677 530 385 532 000 +12.3 Canada 69 954 75 025 87 500 +16.6

603 631 603 410 619 500 +2.7 23.9

CARIBBEAN AREA Venezuela 193 209 184 921 167 400 -9.5 Colombia 11 071 11 127 10 400 Trinidad 7 225 6 690 7 400

211 505 202 738 185 200 -8.7 7.9

OTHER LATIN AMERICA

Mexico .. 21 877 21 920 22 600 +3.0 Argentina 19 969 21 494 22 150 +3.0 Brazil 8 009 8 376 8 400 Ecuador 191 174 3 500 Peru 3 450 3 048 3 300 Bolivia 1 124 1 714 1900 Chile 1 620 1 652 1 700

56 240 58 378 63 550 +8.9 2.4

MIDDLE EAST Saudi Arabia .. .. 176 851 223 515 285 500 +27.7 Iran 191 663 227 346 254 000 + 11.7 Kuwait 137 398 146 787 152 000 +3.6 Iraq 76 550 84 000 67 000 -20.2 Abu Dhabi 33 288 44 797 50 000 +11.6 Kuwait/SA "Neutral Zone" .. 26 724 29 118 30 300 +3.9 Qatar 17 257 20 201 23 300 + 15.3 Oman .. .. 17 169 14 106 13 600 -3.6 Egypt 16 404 14 706 11 000 Dubai 4 306 6 252 7 500 Sinaic 4 500 6 000 6 000 Syria ., .. .. 4 353 5 254 5 300 Bahrain ,. .. ., 3 834 3 728 3 500 Turkey 3 461 3 253 3 350 Israel 77 62 50

713 835 829 125 912 400 +10.0 35.0

AFRICA (excluding Egypt) Libya 159 201 132 250 105 000 -20.5 Nigeria 53 420 75 306 89 500 +18.8 Algeria 47 253 36 346 52 000 +42.1 Angola 5 065 5 830 7 200 Gabon/Congo 5 442 5 794 6 600 Tunisia 4 151 4 097 4 100 Morocco ,. .. 46 22 30

274 578 259 645 264 430 + 1.8 10.2

1-2 TABLE 1-2. (cont.)

1970 1971 1972 °]o Change 1972: Countries 3 io t 1971/72 "h of Total

WESTERN EUROPE

West Germany 7 535 7 420 7 100 Austria ...... 2 798 2 516 2 500 Norway - 301 1 700 Netherlands 1 919 1715 1 630 France ...... 2 309 1 858 1 500 Italy 1 408 1 294 1 200 Spain ,. .. ., 156 120 250 Denmark ...... - - 100 UK 83 84 84

16 208 15 308 16 064 +4.9 0.6

FAR EAST Indonesia 42 102 44 521 54 000 +21.3 Australia 8 292 14 373 15 150 Brunei .. 6 916 6 528 9 200 India 6 809 7 191 7 500 Malaysia ...... 859 3 275 4 450 Burma 750 840 900 Japan ...... 750 751 730 Pakistan 486 487 450 Taiwan 90 112 100

67 054 78 078 92 480 + 18.4 3.6

Western Hemisphere 871 376 866 526 868 250 +0.2 33.4 Eastern Hemisphere 1 071 675 1 182 156 1 285 374 +8.8 49.4

1 943 051 2 048 682 2 153 624 +5.0 82.8

EASTERN EUROPE AND CHINA

USSR 352 574 376 992 394 000 +4.5 Romania ...... 13 377 13 794 14 000 Yugoslavia 2 854 2 953 3 100 Hungary 1 937 1 955 1 950 Albania ...... 1 199 1 350 1 575 Poland 424 395 370 Bulgaria 334 304 250 East Germany .. .. 200 200 250 Czechoslovakia ,, .. 203 193 195 Chinad 20 000 25 500 29 600 + 16.0

393 102 423 636 445 300 +5.1 17.2

World totals 2 336 153 2 472 319 2 598 924 +5.1 100.0

a Excluding small-scale production in Cuba, Thailand, New Zealand, Mongolia and Afghanistan, b Including natural gas liquids, in Canada also synthetic oils. c Under Israeli occupation. Including oil from shale and coal.

Even under these modest assumptions, Tables 1-1, 1-2, 1-3 and 1-4 demonstrate some striking developments, the most important being: (a) A growing dependence of the USA on imported oil and, in particular, on Middle Eastern oil even though allowance has been made for Alaskan production at the end of the decade.

1-3 (b) A growing Western European dependence on imported and Middle Eastern oil even though allowance has been made for maximum North Sea production and the percentage share of imports is expected to decrease. (c) A continuation of Japan's total dependence on oil imports. (d) A sharp rise in Middle Eastern production which is expected to double over the 1970-80 decade from 700 to 1500 million tons per year when it will represent close to 40% of total world production and more than 50% of that of the non-socialist countries while bringing to the countries of the region annual revenues of the order of 30X 109US $/yr.

TABLE 1-3. NATIONAL PRODUCTION AND IMPORTS IN THREE MAIN CONSUMING AREAS (106 t)

Imports from National production Total imports Middle East

1970 1980 1970 1980 1970 1980

USA 534 660 214 500 30 300 ("jo of consumption) (71) (57) (29) (43) (6) (26) Western Europe 16 160 584 820 300 600 C7o of consumption) (2.6) (24) (97.4) (76) (50) (61) Japan 1 2 199 398 170 300 ("lo of consumption) (0.5) (0.5) (99.5) (99.5) (85) (75)

Total 551 822 987 1 718 500 1 200

TABLE 1-4. PAST AND ESTIMATED PRODUCTION IN MAJOR EXPORTING AREASa (106 t)

Share of Share of 1970 world consumption 1980 world consumption

Middle East 714 31.6 1 500 39.3 Africa 274 12 330 8.6 Caribbean 212 9.3 220 6

Total 1 190 52.6 2 050 54

For exact definition of the geographical areas, see Table 1-2.

No mention is made at this stage of estimated world oil reserves, not because the subject is not important, but because the figures usually advanced are highly questionable and cover an extremely wide range. Thus, for instance, figures of the order of 60 X 109 tons are often advanced for proven oil reserves while ultimate potential reserves which were estimated at around 90 X 109 as late as 1960 are now quoted as exceeding 900 X 109 tons if account is taken of probable off-shore oil fields, secondary recovery methods, oil-bearing shales and tar sands. It thus appears that the question for the next few decades is not one of exhaustion, but of costs. It should, however, be noted that if demand continues to expand indefinitely at the 5.4% rate forecast for the next seven years, even the 900 X 109 tons of presently estimated ultimate reserves would only last 55 years instead of the 15 years assured by 50X 109 of proven fields. Consequently, the 15 to 1 ratio between the two reserve figures should not be construed too optimistically.

1-4 COST AND PRICE STRUCTURE OF OIL AND ITS FUTURE TRENDS

The question of cost and prices of oil is fraught with difficulties unparallelled in any other industry:

(a) Technical difficulties in accurately defining a particular type of crude. Oils of different characteristics have, of course, historically sold at different prices, but the problem, has become particularly acute recently because of environmental consideration which could restrict drastically the sulphur emissions from oil-fired stations in most industrial countries. Without going into the intricate problem of costs of desulphurization it should be noted that differentials of 50% and more can exist between prices of crudes in the same producing area depending on their sulphur content. (b) Accounting difficulties in ascertaining the real price of crude rooted in the structure of the international oil industry which has, up to now, controlled the production, distribution and marketing of petroleum through vertically integrated operations. As a result, most of the oil entering international trade was moved from producing to refining and marketing subsidiaries at accounting prices fixed internally by the integrated companies essentially in the light of fiscal considerations, while only small amounts of crude were sold to outsiders at what might have been considered market prices. (c) Political difficulties arising from the relatively small share of production costs in the total selling price. As Table 1-5 shows, the cost of production represents less than 10% of the price of crude in the Middle East, the remaining 90% being divided between revenues to host countries and profits to producing-companies. Historically, the split between two groups has been the result of a constant power struggle which has recently turned in favour of the countries which now collect more than three-fourths of the f. o.b. price of crude. The lateat steps of the struggle were marked by the Teheran Agreement which sharply increased the share of the host nations and provided for automatic increases every year until January 1975. A no less important step was taken at the beginning of 1973 with the Participation Agreement entered into by several of the Arab countries and, in particular, by Saudi-Arabia and Kuwait, providing for a 25% ownership of production by the countries with a final objective of 51% participation by 1981. While Iran and Libya may follow different approaches, there is an unmistakable trend towards control of production by the countries of origin. For the time being, the participating countries plan to re-sell their share of production to the international oil companies which control the necessary distribution and marketing channels, but the situation may well change over the present decade. (d) Economic difficulties arising from the theoretical impossibility of allocating costs of crude oil to the variety of oil products obtained as a result of refining. Gasoline, kerosene, naphtha, light fuel oil, and heavy fuel oil obtained from a single input of crude are priced separately by private companies according to market conditions in order to maximize total profits. There is no way in which the cost of producing, transporting and refining one ton of crude oil can actually be allocated to the different products derived from it.

TABLE 1-5. ILLUSTRATIVE BREAKDOWN OF PRICE OF HEAVY KUWAIT CRUDE IN PERSIAN GULF AND WESTERN EUROPEAN HARBOURS3 (US $ /t)

Production cost 1 Producing country royalties and taxes 10 Company profit 2

Total 13

Transport cost to Rotterdam by 130 000 t tanker 6 Delivered cost at harbour refinery 19 a Needless to say, this table and Table 1-6 are presented as illustrations rather than precise cost breakdowns which would require an analysis of the refining, distribution, marketing and fiscal situation in a specific country.

1-5 To this should be added another important consideration affecting the whole price structure of oil products. Table 1-6 illustrates two important and connected points: the heavy impact of indirect and direct taxes levied by oil importing countries on the total costs of oil products to the ultimate consumers and the wide gap between these total final costs paid by the users and the "technical production costs", however widely these may be defined. Although the values given in this table are approximate averages and although Western Europe is one of the areas with the heaviest burden of taxation on oil products, the conclusions are nevertheless generally valid.

TABLE 1-6. ILLUSTRATIVE AVERAGE COST STRUCTURE OF OIL PRODUCTS OBTAINED FROM ONE TON OF CRUDE IN WESTERN EUROPEAN COUNTRIES (US $/t)

Cost of crude at harbour 19 Cost of refining 3.50 Storage, inland transit, distribution and marketing 20 Profits of distributing companies 2.50 Taxes levied by consuming countries (excise taxes on products and corporate income taxes) 40

Total 85

With regard to the incidence of taxation by industrial countries, it will be seen that it represents close to 50% of the costs of the ultimate products, and about 4 times the amount of taxes levied by producing countries. True, these taxes fall mainly on gasoline (although several Western European and some developing countries also tax heavy fuel oil) and the fiscal revenues are used for highway maintenance, traffic control etc.; in other words, for tasks which actually make the use of oil products possible. Nevertheless, the fact remains that the impact on final costs is extremely heavy. This leads to the second point, i. e. the almost total divorce of costs of production from ultimate revenues derived from a given quantity of crude oil, a situation radically different from that of for instance coal for which the relationship is much more rigid. Production costs in the Middle East are less than 2% of the ultimate total (1. 2% of US $85/tinthe example given). If company profits, transportation and refining costs are added, the combined cost would still remain less than 20%. Finally, even if distributing and marketing costs are counted, the percentage would only increase to 41%, so that close to 60% of final outlay go to taxes levied by governments of either the producing or consuming countries. This cost structure has several consequences, one of the most important being the relative insensitivity of final product costs to variations in the costs of production at the oil field. In the example given, an increase of the cost of production of crude oil in the Middle East by a factor of 10, from US $1 to 10 per ton, would only lead to a 12% rise in the ultimate product costs to the consumers. This goes a long way towards explaining the wide disparity of actual oil production costs throughout the world. It also points to the probability that higher costs connected with off-shore production, shale oil recovery and other potential reserves will prove no serious obstacle to their future exploitation. Finally, it should be pointed out that taxes on heavy fuel oil may seriously affect its competitive position and lead to major distortions in the selection of power plants with a resultant economic loss for the country concerned. Taking these difficulties in turn, the following assumptions are made for the purpose of estimating prices of fuel oil for the Market Survey:

(a) Since none of the Survey countries had expressed special reservations on environmental constraints, one of the cheaper types of crude oil with no limitation on sulphur content was selected as the basis. This was Kuwait crude of 31° API.

1-6 (b) Its price was based on data available for transactions between producing companies and independent third parties to which this type of crude was sold in the Persian Gulf in 1972 and escalated to 1 January 19731. Transport costs to the major harbours of the countries concerned were estimated on the basis of data summarized in Table VII. (c) It was assumed that the strong position of the producing countries will permit them to maintain and probably increase the growing revenues already provided for by the Teheran and Participation Agreements. Consequently, an annual rate of growth of oil prices of 5% was considered minimal while 6% was viewed as probable. (d) The relationship between the prices of crude and heavy fuel oil was assumed on a basis explained at greater length in Section 4 of this Appendix.

COST OF TRANSPORT OF OIL BY TANKER AND BY PIPELINES

These costs are given in detail in Tables 1-7 and 1-8. The sensitivity of unit transport cost to size of tanker and pipeline must be stressed. Consequently, future transport costs will depend critically on the existence of harbour facilities capable of handling the largest type of tanker size compatible with the demand of the country.

RELATIONSHIP BETWEEN CRUDE AND OIL PRODUCT PRICES

As has already been pointed out, there is no generally valid relationship between the two products and the price of fuel oil is entirely dependent on supply and demand. There are, however, lower and upper limits imposed by the availability of substitutes. Regarding fuel oil for power plants, an immediate substitute is available in the form of crude oil itself which, subject to certain precautions, can and has been used as a fuel. Consequently, and except for short-lived special cases, the price of a given quality of crude in a specific location sets an upper limit to the price of heavy fuel oil of comparative sulphur content. With regard to a lower limit, the situation is much more complex since it depends on the availability of alternative fuels as well as on the possibility of altering the proportion of different refinery products, both in the short and long term. A historical study of the relationship between long term prices of fuel and crude oils of similar characteristics shows that the differential between them has seldom exceeded 10% (except in the special case of the US Eastern Seaboard and the Caribbean area). It was, therefore, decided to use as reference prices for heavy fuel oil landed in the major harbours of the countries covered by the Survey the price of landed crude as a maximum and 90% of the price of crude as a minimum. In fact, 95% of the price of crude was chosen as a representative single value.

CONCLUSIONS The procedure finally selected for estimating fuel oil prices for the countries of the Market Survey was based on four main assumptions each one being open to some objections:

(a) The price of crude in the Persian Gulf was used as the basis even though some of the countries covered, particularly in Latin America, are not importing crude from this

1 At the time these estimates were made, the impact of the 1973 devaluation of the US $ on the amount of taxes paid to the producing countries was still not officially agreed. It seems, however, that an increase of 10

1-7 TABLE 1-7. COMPARATIVE TRANSPORTATION COSTS FROM PERSIAN GULF TO ROTTERDAM4 IN VARIOUS SIZES OF TANKERS

Size of tanker (dwt) 50 000 70 000 90 000 130 000 250 000 500 000

Year of delivery Days at sea 58.2 58.2 56.4 58.2 58.2 58.2 Days in port 3.2 3.2 3.3 3.3 3.5 3.5 Trips per annum'' 5.7 5.7 5.9 5.7 5.7 5.7 Cargo (tons per trip) 47 200 66 300 85 000 123 400 240 000 480 700

Voyage costs (US $ x 103) 1971 Fixed direct costs 132.8 150.0 164.2 204.9 317.0 - Capital costs 121.0 154.7 175.0 246.7 396.7 - Bunkers c 50.5 70.7 98.2 126.6 163.1 - Port charges 13.0 17.2 19.5 24.8 43.6 -

Total 317.3 392.6 456.9 603.0 920.4 -

1973 Fixed direct costs 171.4 192.1 209.8 260.9 405.6 18.2 Capital costs 142.1 184.2 211.9 301.5 476.0 914.0 Bunkersc 45.5 63.7 88.4 114.0 146.8 276.8 Port charges 18.6 23.7 29.4 35.2 66.3 136.5

Total 377.6 463.7 539.5 711.6 1 094.7 2 045.5

1975 Fixed direct costs 195.1 218.0 237.4 294.3 441.1 748.5 Capital costs 173.7 228.4 276.3 397.5 740.4 1 269.2 Bunkers0 51.4 71.9 99.9 128.8 165.9 312.7 Port charges 20.5 26.1 32.4 38.8 73.0 150.7

Total 440.7 544.4 646.0 859.4 1 420.4 2 481.1

Costs (U S $/t of cargo) 1971 Direct costs 4.16 3.59 3.32 2.89 2.18 - Capital costs 2.56 2.33 2.06 2.00 1.65 -

Total costs 6.72 5.92 5.38 4.89 3.83 -

1973 Direct costs 4.99 4.22 3.85 3.32 2.58 2.35 Capital costs 3.01 2.78 2.49 2.44 1.98 1.90

Total costs 8.00 7.00 6.34 5.76 4.56 4.25

1975 Direct costs 5.66 4.77 4.35 3.74 2.83 2.52 Capital costs 3.68 3.44 3.25 3.22 3.09 2.64

Total costs 9.34 8.21 7.60 6.96 5.92 5.16

Costs (1972 world-scale equivalent)

1971 68 60 55 50 39 - 1973 81 71 64 59 46 43 1975 95 83 77 71 60 53

a Distance for round trip 22 338 miles. b All vessels in operation for 350 days each year. c Bunker prices (US $/t) 1971 Persian Gulf 13.50, North Europe 21.00 1973 Persian Gulf 13.00, North Europe 18. 00 1975 Persian Gulf 15.00, North Europe 20.00

I-t TABLE 1-8. ILLUSTRATIVE COSTS OF INLAND TRANSPORT BY PIPELINE

Throughput 2 x 106 t/yr 5 x 106 t/yr (10-in diam. pipeline) (16-in diam. pipeline)

Costs (US cents/t per 100 miles) Capital 60 39 Other fixed 18 10 Variable 4 8

Total 82 57

Total cost (US cents/106 kcal per 100 mile) 8.1 5. 6

Note: The table is restricted to pipeline sizes most likely to be encountered in oil-importing developing countries. The cost per ton of oil transported is, however, quite sensitive to size up to very large throughputs. Thus, for a pipeline with a transport capacity of 50 x 106 t/yr it would drop to less than 20 US cents/t per 100 miles, or to about 1/3 of the 5 x 106 t/yr figure. a Assumes: flat country, no major river crossing; capital cost of pipeline US $9000/in diameter per mile; fixed charge rate 13. based on an interest rate of 12% yr and on 20-yr sinking fund depreciation. £ Sufficient for supplying 1200 MW of oil-fired plants at 80<7o load factor.. Sufficient for supplying 3000 MW of oil-fired plants at 80<7o load factor.

source. This is not as serious a flaw as it may seem since the policy of pricing oil from various sources on the basis of equality of delivered cost, with the main producing region serving as a reference point, has been a recurring feature of past price policies. (b) An annual escalation rate of 6% was proposed for the 1973-1980 period, which is higher than the approximately 4% which the Teheran Agreement alone would imply, but takes into account the progressive impact of participation of the Arab countries in production and the sharp rise in oil demand. (c) A fixed relationship was assumed between the prices of crude and of heavy fuel oil while the actual connection is flexible and complex. As has been explained this is a simpli- fication but its impact on actual results is unlikely to involve errors of more than 5%. (d) Taxes levied on fuel oil by consuming countries were ignored since they are internal revenues to the governments and should not affect the economic selection of power plants. There is no question that even though from the standpoint of the electric utilities taxes levied by their own country on a particular type of fuel are an element of total costs, the same taxes appear as a revenue item in national accounting. Since the purpose of the Market Survey is to estimate national costs of alternative power programs, domestic taxes on fuel should be excluded, at least in the basic reference cases. (e) Estimated base prices, resulting from the above, for crude and heavy fuel oil in major harbours of the countries participating in the Market Survey are given in Table 1-9. (f) Gas turbine fuels were arbitrarily priced at 175% of fuel oil on the basis of an averaging of existing data. (g) Domestically produced oil and gas was priced on the basis of parity of thermal costs with imported fuel oil or fuel oil refined from imported crude. (h) Prices of domestically produced lignite and coal were estimated independently on the basis of the data supplied by the countries and escalated at the general rate of 4%/yr except in cases where there were convincing arguments to depart from this general procedure.

1-9 TABLE 1-9. ESTIMATED BASE PRICES FOR CRUDE AND HEAVY FUEL OIL IN MAJOR HARBOURS OF MARKET SURVEY COUNTRIES4, 1 January 1973

CIF Price Corresponding Sea trans- of crude in prices of 6 Harbour port costb US cents/10 kcal harbour fuel oil (US $/t) (US $/t) (US $/t)

Egypt Alexandria 3 16 15.2 150 Greece Piraeus 5 18 17.1 168 Turkey Izmet 5 18 17.1 168 Yugoslavia Trieste 6 19 18 177 Argentina Buenos Aires 6.5 19.5 18.5 182 La Plata

Chile Valparaiso 7 20 19 187 Quintero

Jamaica Kingston 6 19 18 177 Mexico Tampico 7 20 19 187 Vera Cruz Pakistan Karachi 1 14 13.3 131 Bangladesh Chittagong 2.5 15.5 14.7 145 Singapore 2 15 14.3 140 Thailand Bangkok 3 16 15.2 150 Philippines Bantangas 3 16 15.2 150 Korea Pusan-Ulsan 4 17 16.1 159

a Kuwait heavy crude 31° API with no sulphur restriction estimated at US $1.80/bbl or US $13/t f.o.b. in the Persian Gulf. 1 t crude = 7.2 bbl 1 t heavy fuel oil = 6.8 bbl 1 t heavy fuel oil = 40.3 X 106Btu. = 10.15 x 106kcal. Transport costs by sea estimated on the basis of journey by tankers of size suitable for country harbours except for Mediterranean countries where special allowances were made for possible transport by pipeline or canal through Suez in the future.

I-10 APPENDIX J

NUCLEAR FUEL CYCLE COST TREATMENT

James A. Lane

INTRODUCTION

Fuel cycle costs in a nuclear power plant depend on a wide variety of economic parameters, such as the costs of uranium, of separative work and of industrial operations which vary with time. It is likely that some of these costs, such as those for natural U3O8 and separative work will increase with time, while other cost components such as fuel fabrication and fuel recovery will decrease. To complicate the situation even more, the value of fissile plutonium recovered from spent fuel can go up or down depending on its marketability as recycle fuel. In addition to dependence on the above economic factors nuclear fuel costs also depend on engineering parameters such as the fuel burn-up per cycle, the fuel management scheme employed etc. , which the reactor designer or plant operator can vary to optimize overall generating costs. Because of this balancing of economic and engineering factors, total nuclear fuel cycle costs tend to remain relatively constant with time. In the case of light- water reactors, fuel costs lie within the rather narrow range 20 ± 5 US cents/106 Btu (80 ± 20 US cents/106 kcal) regardless of size or plant design. Unlike oil costs, moreover, nuclear fuel costs are not sensitive to where the plant is located in the world. In view of this situation, it was decided that it would be sufficient for the purpose of the Market Survey to base the economic evaluation on current nuclear fuel costs taken from studies published in the open literature. For the reference case, these fuel costs were assumed to follow the general inflation rate of 4%/yr, the same as all other capital costs (see Appendix D). Sensitivity studies were also carried out using a 6% escalation rate, the same as that used in the reference case for oil and gas.

FUEL CYCLE COSTS FOR A 400 MW PWR

In a paper by J. T. Roberts and R. Krymm [ 1], a variety of numerical examples of nuclear fuel cost calculations for a hypothetical 400 MW pressurized water reactor are presented and discussed in detail. Figure J-I shows a generalized schematic diagram of the LWR fuel cycle used as a basis for the calculations and Table J-l shows the assumed economic and engineering parameters. The data in Table J-l were used in a present-worth calculation to determine the levelized fuel cycle cost under steady state (equilibrium) conditions with one-third of the core being replaced each year. For this simplified equilibrium case, total fuel cycle costs and corresponding direct and indirect components are calculated by following a single batch of fuel throughout its three-year lifetime. Table J-2 shows the results of this calculation. Since the cost calculation for the equilibrium fuel does not take into consideration the higher unit costs associated with the first core, calculations were also carried out to find the first core cost and also the levelized 30-year average fuel cost for the first core plus the 29 equilibrium refuelling batches. Table J-3 compares the costs for the three cases considered. The levelized 30-year average fuel costs shown in the last column were taken as the reference case for the Survey; however, two adjustments were made for this purpose. Firstly costs were adjusted to reflect the increase in separative work costs to the US $36/kg announced by the USAEC on 14 February 1973, and secondly, indirect costs were based on the 8% interest rate taken as the reference case in the Survey. These two changes tended to balance one another with the result that levelized 30-year average fuel cycle costs amount to 1. 78 US mill/kWh for a 400 MW PWR.

J-l LOSSES & CASH FLOW FUEL CYCLE STEP PRODUCTION

Pay for U3O8 U Mining & Milling (U loss)

r Pay for Conversion Conversion of U3O8 to UF6 (U loss)

\f Pay for Enrichment Isotopic Enrichment

Preparation of UO2 —>• (Scrap recovery Pay for Fabrication and Fabrication •<— and recycle) of Fuel Elements (Uloss)

r Receive power (Energy & Pu Reactor Irradiation sale revenue produced)

( Recovery of U and Pu (including spent fuel shipment, Pay for Recovery reprocessing and waste (U & Pu losses) disposal, and reconversion of U to UF6)

Receive credit Sale or recycle of recovered for U & Pu Pu (nitrate) and UF6 recycled or sold

FIG.J-1. GENERALIZED SCHEMATIC DIAGRAM OF LWR FUEL CYCLE.

J-2 TABLE J-l. BASIS FOR FUEL CYCLE COST CALCULATIONS CARRIED OUT IN REF. [1]

1. Cost of natural uranium ore concentrate: US $7.00/lb U.O.

2. Losses (not economically recoverable) in processing:

Conversion - 0. 5<7o Enrichment - 0. 0% Fabrication - 1.0% Reprocessing (U and Pu) - 1. 0%

Reconversion, U nitrate to UF6 - 0.3°/o

3. Uranium enrichment: Tails assay: 0.25%U-235

Cost of separative work: US $32.00/SWU (kg)

4. Cost of converting U3O8 to UF6: US $2.60/kg U (product)

5. Fabrication cost (including cost of scrap recovery): First core - US $110/kg U (product) Equilibrium core - US $ 80/kg U (product)

6. Recovery cost (including spent fuel shipment, reprocessing, reconversion of recovered uranium to UF6): First core - US $44/kg U (feed)

Equilibrium core - US $40/kg U (feed)

7. Plutonium credit: US $10.00/g (fissile)

8. Times at which pre-irradiation payments are made: First core Equilibrium core U 15 months 12 months 3°8 Conversion 12 months 9 months Enrichment 9 months 6 months Fabrication 6 months 3 months Times at which post-irradiation payments or credits are made: Recovery + 6 months U and Pu credits + 9 months

9. Reactor power: 1222.5 MW(th) gross 400 MW(e) net Plant capacity factor: 80%

10. Irradiation history:

First core Batch "A" Batch "B" Batch "•C" Burn-up (MWd/t) 13 176 23 912 31 531 Initial enrichment (°!o U-235) 2.41 3.04 3 .48 Final enrichment 1.24 1.17 1.08 Final fissile Pu (%) (based on U) 0.46 0.61 0 .72 kg U charged to reactor 11 321 11 321 11 321 kg U discharged from reactor 11 100 10 949 10 846 In-core life at 80% load factor (yr) 1.00 2.00 3 .00

Equilibrium Batch: Same as Batch "C" above.

Power production (% of total): Outer region 24.16 Intermediate region 34.05 Inner region 41.79 100.00

J-3 TABLE J-2. FUEL COST ESTIMATE FOR THE EQUILIBRIUM CORE LIGHT WATER REACTORS [1]

Cost category and components Unit fuel cost (US mill/kWh) Direct Indirect Total

I. Fertile and fissile materials

(a) UjO8 purchase, gross 0.523 0.158 0.681 (b) Credit for equivalent U3OS in recovered U -0.126 0.022 -0.104 (c) Credit for recovered plutonium -0.276 0.048 -0.228 Subtotal I 0.121 0.228 0.349

II. Industrial operations (a) Conversion, gross 0.074 0.020 0.094 (b) Credit for conversion equivalent in recovered U -0.018 0.003 -0.015 (c) Enrichment, gross 0.623 0.150 0.773 (d) Credit for enrichment equivalent in recovered U -0.052 0.009 -0.043 (e) Fabrication 0.323 0.069 0.392 (0 Recovery 0.155 0.024 0.131

Subtotal II l— i .105 0.227 1.332

Total 1.226 0.455 1.681

TABLE J-3. LEVELIZED FUEL CYCLE COSTS FOR 400 MW PWR (US mill/kWh) [1] First core Equilibrium core 30-year average

Direct 1. 59 1,,23 1.32 Indirect 0. 51 0,,45 0.46

Total 2. 10 1,,68 1.78

FUEL COSTS FOR SMALL AND MEDIUM POWER REACTORS A paper by M. A. Khan and J. T. Roberts [2] presents information on fuel cycle costs for light water nuclear plants in the size range 100 to 600 MW. These costs adjusted to the conditions described above (8% interest rate, US $36/kg separative work) are summarized in Table J-4. Note that, due to different assumptions which are explained in the references, the fuel cycle costs for the two 400 MW cases (Tables J-3 and J-4) are slightly different.

TABLE J-4. FUEL COSTS IN SMALL AND MEDIUM POWER REACTORS [2] Levelized total Reactor size fuel cycle costs (MW) (US mill/kWh)

"100 2.10

200 1.85

300 1.75

400 1.65

500 1.60

600 1.60

J-4 FUEL COSTS FOR OTHER PWR SIZES

Total fuel cycle costs for other sizes of PWRs taken from Refs [3-5] are plotted in Fig. J-2 along with the costs from the IAEA studies previously described. All costs were adjusted to an 8% interest rate, 80% plant factor and US $36/kg separative work. A linear relationship between nuclear plant capacity and total fuel cycle costs was adopted for the Survey as shown in Fig. J-2.

US MILLS/kWh

2.2 REF. 3

2.1

EF.2 2.0

1.9 V FiEF.l 1.8 acc c < 7 ADOPTED FOR MARKED SURVEY 1.7 J

1.6

1.5

1.4 100 200 300 400 500 600 700 800 900 1000 PLANT CAPACITY - MW

FIG. J-2. TOTAL FUEL CYCLE COSTS.

FUEL CYCLE WORKING CAPITAL COSTS

For the purpose of the WASP computer program, it was necessary to separate total fuel cycle costs into a "fixed" component which varies with the assumed interest rate and a "variable" component which varies with the amount of energy generated. The "fixed" component of nuclear fuel costs represents the levelized value of all outstanding investments associated with the fuel cycle over the life of the plant. Figure J-3 shows values of this fixed component taken from the previously mentioned references. As in the case of the total fuel cycle costs, a linear relationship between fixed costs and plant capacity was assumed as shown in Fig. J-3. It should be noted that the fixed component of nuclear fuel costs varies by only US $18/kW over the entire range of plant capacities, which is equivalent to about 2 US cents/106 Btu.

US$/kW 50

REE 1 40 < J— —=A ADOPTED FOR NMARKED,SURVEY 30 - \ ^^ (- —* — \ REF. 2- / S /_ REF. 4 } — - >< > 20 - REE~1 10

100 200 300 400 500 600 700 800 900 1000 PLANT CAPACITY-MW

FIG. J-3. LEVELIZED FUEL CYCLE CAPITAL COSTS.

J-5 VARIABLE FUEL CYCLE COSTS

The difference between the total fuel cycle costs and the fixed component gives the variable fuel cycle costs. For the purpose of the WASP program, it was necessary to express the variable component in terms of US cents/106 kcal. For this purpose the full load gross heat rates estimated by the Bechtel Corporation (see Appendix E) were used. The resulting variable nuclear fuel costs are shown in Table J-5 along with total fuel cycle costs and the fixed component (calculated at 80% plant factor and 8% interest).

TABLE J-5. FUEL CYCLE COSTS ADOPTED FOR MARKET SURVEY

Levelized fuel cycle costs Fuel load gross Plant capacity Variable fuel cycle costs (US mill/kWh) heat ratea 6 (MW) Total Fixed Variable (kcal/kWh) (US cents/10 kcal)

100 1.93 0.43 1.50 2 504 59.8 200 1.89 0.41 1.48 2 503 58.9 300 1.84 0.39 1.45 2 503 57.9 400 1.79 0.37 1.43 2 502 57.0 600 1.70 0.32 1.38 2 501 55.1 800 1.60 0.27 1.33 2 500 53.2 1 000 1.51 0.23 1.28 2 499 51.3

a Gross heat rates were used to be consistent with the use of such heat rates in calculating conventional fuel costs.

REFERENCES

[ 1] ROBERTS, J.T., KRYMM, R., IAEA, "Fuel cost evaluation for light water reactors", Regional Training Course on Bid Evaluation and Implementation of Nuclear Power Projects, Tokyo, 29 November - 10 December 1971, IAEA-151, unpublished document, p. 191. [2] KHAN, M.A., ROBERTS, J.T., "Small and medium power reactors: Technical and economic status, potential demand and financing requirements", 4th Int. Conf. peaceful Uses atom. Energy (Proc. Conf. Geneva, 1971) 6, IAEA, Vienna (1972) 57. [3] STORRER, J., GALLOIS, G., KAFKA, P., " Belgonucleaire and Siemens PWRs for small and medium power reactors", Small and Medium Power Reactors 1970 (Proc. Symp. Oslo, 1970), IAEA, Vienna (1971) 131. [4] USAEC, Current Status and Future Technical and Economic Potential of Light Water Reactors, WASH 1082, March 1968. [ 5] Electrowatt Engineering Services and Sargent and Lundy, Feasibility Study for a Nuclear Power Plant in Luzon, Philippines, February 1973.

J-6 APPENDIX K

SENSITIVITY STUDIES

In order that the results of the analyses of the participating countries could be compared and summarized, it was deemed desirable to analyse each country using the same basic values of the parameters and then to perform other analyses using different values of these parameters in order to determine the sensitivity of the results of the base case to such variations. This was done so that each country would have results available using parameter values which might more nearly represent its unique values. Also, since the base values are forecasts determined from historical information and a consideration of present and future trends, it was considered important to check the sensitivity of the selected system expansion plans to possible variations in these parameters. The technique of using the WASP program to analyse predetermined system expansion plans allowed the addition of a number of sensitivity alternatives to each analysis at the expense of very little additional computer time. The parameters selected for sensitivity studies and the values used are:

(a) Economic parameters

Base case Other cases

Approximate Approximate Study Study equivalent equivalent valuesa values a "real" values "real" values

Discount rate (°?c 8 12 6 & 10 10 & 14 Oil & gas price escalation (°Jc 2 6 0 & 2 4& 8 Nuclear fuel price escalation (°/c 0 4 2b 6 Capital cost of plants c ORCOST-3 ORCOST-1

a General inflation rate was assumed constant at 4%/yr. b This value was used for sensitivity studies in only a few selected cases. c ORCOST-3 values are as of 1 January 1973 and show a ratio of PWR to oil-fired plant costs ranging from about 1.8 to 2.2 (depending on MW rating) whereas ORCOST-1 values show a corresponding range from about 1.6 to 1.8. For a complete discussion of these costs refer to Appendix B.

(b) Load forecasts

The basic load forecast for each country was prepared on a common basis by Aoki as described in Appendix F. For several countries his forecast compared closely with that provided by the country itself; in those cases only one forecast was used. For most countries, however, the country forecast was appreciably higher than the Aoki forecast and in these cases both were used as the basis for analysis.

An additional sensitivity study was carried out, in effect, on the variation in the loss-of- load probability. For a definition and further discussion of loss-of-load probability refer to Appendix A. The value of the loss-of-load probability for any given system is related to the amount of system reserve generating capacity and to the number, sizes and types of plants and this is also related to the degree of load shedding to be permitted at times of forced outage of generating capacity. Obviously, reducing the loss-of-load probability will increase the system cost to supply a given load and increasing it decreases system costs. Thus specific values, or a range of acceptable values, needed to be established for purposes of the

K-l studies, since any specific system expansion plan is optimum only for a specific loss-of-load probability. Therefore, it was decided to use an average of the yearly values over the study period, as close as possible to 0.005 with a maximum of 0.010. It is considered that these values are representative of the values acceptable to developing countries, although they are substantially higher than the acceptable values for the industrialized countries. The actual loss-of-load probability value can be expected to vary from year to year depending on the amount and timing of generating capacity additions. In a number of cases the loss-of-load probability value for a country's existing system was substantially higher than the maximum quoted above. The technique used in these cases was to bring the loss-of-load probability gradually down to the levels indicated above by adding more generating capacity. To achieve this generally required a number of attempts to determine the exact size of unit and the point in time when it should be added. A study of the results of these numerous analyses, involving varying values of loss-of-load probability, shows that although the value of the objective function (present worth) could vary considerably, the size and number of nuclear power units called for in the optimum (lowest present-worth value) case would vary only slightly. In this connection it should be pointed out that the probabilistic model used in deriving the loss-of-load probability values is limited in its handling of hydro power plants and, for systems with large proportions of hydro power, it tends to show unrealistically low loss-of-load probability values.

(d) Foreign exchange rates (shadow exchange)

In a few instances, studies were carried out to determine the sensitivity of the optimum case to variations in the rates of exchange between local and foreign currencies. This is intended to show the effect on capital-intensive projects of scarcity of foreign capital to finance such projects.

(e) Salvage values based on sinking fund depreciation

In the reference case, salvage values based on linear depreciation were factored in for all plants at the end of the study period (i.e. 2000). Although this practice is current in most electric utilities accounting, it involves a slight departure from strict economic accounting which should be based on sinking fund depreciation. Since the use of straight line depreciation gives a higher value of the objective function than sinking fund depreciation, its use tends to penalize capital intensive projects, i.e. nuclear plants. For this reason, the effect of using salvage values based on sinking fund depreciation was considered in some instances.

(f) Duties and taxes

Duties and taxes were not considered in the reference case; however, in some countries they might have an important influence on the market for nuclear power by increasing oil prices, on the one hand, and nuclear plant capital costs on the other. Sensitivity studies to evaluate the influence of duties and taxes were carried out for countries where their effect might be important.

(g) Environmental effects

It is not clear whether environmental considerations will play an important role in the participating countries; therefore, no allowance was made for these in the reference cases. If future environmental considerations require the use of fuels of low sulphur content or equipment to alleviate deleterious effects, capital and/or operating costs would increase and thereby influence the competition between fossil and nuclear plants. This factor was not treated in a finite quantitative manner in these studies; however, a qualitative and approximate quantitative discussion can be found in Appendix B.

K-2 APPENDIX L

IAEA SERVICES AND ASSISTANCE IN CONNECTION WITH NUCLEAR POWER

The International Atomic Energy Agency provides services and assistance to its Member States and to non-Member States under the United Nations Development Program (UNDP) in any technical field involving the peaceful application of nuclear energy permitted by its Statute. Information about the services and assistance available from and through the Agency is given in the publication "IAEA Services and Assistance"1 . This booklet also explains who is eligible to receive services and assistance from the Agency and how these may be obtained. In general, four stages can be identified in the initial introduction of nuclear power in a given country: Stage 1. Preliminary survey Stage 2. Preliminary study Stage 3. Feasibility study Stage 4. Construction and commissioning of power reactors. Stages 1 and 2 are the most likely suitable subjects for technical assistance and during Stage 3 assistance could be requested from UNDP. The activities in respect of which the Agency can assist or provide services related to nuclear power and the kinds of assistance possible are briefly summarized below. Neither this summary nor the "IAEA Services and Assistance" booklet can be exhaustive in coverage; therefore, if further information is required, it should be sought directly from the Agency's headquarters.

FIELDS OF ACTIVITY

(a) Activities connected with the development of nuclear power

Applications: Use of nuclear energy for the generation of electricity and possible other associated processes.

Economics of nuclear power: Comparison with other sources of power; economics of various fuel cycles; feasibility studies.

Nuclear power program: Planning of a nuclear power program; integration into a system; choice of reactor type; siting of reactors; training of sta'ff; auxiliary services.

Fuels and fuel cycles: Fabrication, testing and inspection of reactor fuel elements and related processes; technical problems of fuel cycles.

Nuclear materials management: Establishment of methods.

Raw materials: Prospecting, mining, processing.

(b) Activities related to safety in atomic energy

Safety standards, regulations and procedures: Standards, regulations, codes of practice and recommendations and their application to specific operations and related procedures.

Radiological protection: Design of installations and laboratories; shielding; protective devices; personnel, area and environmental monitoring; instrumentation; decontamination; medical examinations; diagnosis and treatment of radiation injury and internal contamination.

1 This publication is presently being revised.

L-l Safety of reactors and nuclear materials: Safety aspects in the siting, design, construction and operation of power reactors and related facilities; management of radioactive wastes.

Safety evaluations: Safety evaluations of nuclear installations in respect of their design and siting, operational procedures, associated environmental monitoring and emergency planning.

Framing legislation in establishing national atomic energy authorities; legislation on third-party liability and on the licensing of nuclear facilities; provisions for insurance and other adequate financial protection of nuclear installations; legal problems in connection with the production, transport, use and storage of radioactive materials.

KINDS OF SERVICES AND ASSISTANCE

(a) Technical cooperation programs

Resources made available so that the Agency can provide technical and pre-investment assistance are used to implement projects under the Agency's regular program of technical assistance and under UNDP. Under these programs assistance may include one or more of the following elements:

Expert services: Experts can be sent individually or in teams to advise on or assist in general or specific fields of activity within the Agency's competence.

Equipment and supplies; These are usually provided in association with an internationally recruited expert.

Fellowships: Fellowships can be awarded as part of a comprehensive project or on an individual basis as a direct contribution to projects in the country's atomic energy program. These fellowships are available to qualified applicants at all educational levels and are not restricted to university graduates.

Intercountry projects: The Agency organizes a number of regional and interregional training courses and study tours every year in cooperation with its Member States and other United Nations organizations. Some of them deal with nuclear power. Large-scale projects of significant economic importance to countries in a region can be accommodated under the UNDP.

(b) Advisory and field services

The Agency provides, on request, information and advice on a number of subjects relating, among others, to nuclear power, as outlined above. If requested, missions may also be organized.

The Agency also assists its Member States by means of a program of information services, including the International Nuclear Information System (INIS). Many of these activities relate to nuclear power.

(d) Supply of nuclear materials

Nuclear materials, such as uranium enriched in uranium-235 and plutonium, may be supplied to Member States by or through the Agency in accordance with Article XI of the Agency's Statute. The materials can also be supplied as fuel for power reactors.

L-2 APPENDIX M

ABBREVATIONS USED IN THE MARKET SURVEY REPORTS ampere A approximately approx. barrels bbl billion 109 board feet bd. ft. British thermal unit Btu calorie cal centimetre cm cubic foot ft3 cubic metre m3 cubic yard yd3 cycles per second Hz degree centigrade °C degree Fahrenheit °F direct current DC feet ft figure(s) Fig., Figs. foot ft Gigawatt GW Gigawatt-hour GWh Hertz (cycles per second) Hz horse-power hp hour h hundredweight cwt kilocalorie kcal kilogram kg kilometre km kilovolt kV kilovolt-ampere kVA kilowatt kW kilowatt-hour kWh litre 1 maximum max. megawatt MW megawatt-hour MWh metre m normal cubic metre Nm3 million 106 number No. per annum p.a. per cent % pound (weight) lb pounds per square inch lb/in2 square foot/feet ft2 square metre m2 thousand 103 ton t (always metric, unless specified otherwise as ton (UK) — long ton or ton (USA) — short ton. tons of coal equivalent TEC volt V volt-ampere VA watt W yard yd

M-l

APPENDIX N

COOPERATING ORGANIZATIONS AND PARTICIPANTS IN THE MARKET SURVEY MISSION

13-17 November 1972

Ministry of Science and Technology Public Utilities Board Department of Geography, University of Singapore Economic Development Board Ministry of Finance Jurong Town Corporation Petroleum Sub-Committee, International Chamber of Commerce

Liaison officer: Mr. Tay Sin Yan, Public Utilities Board

Country Status

H. Aoki, Utility expert, Electric Power Development Co. Ltd, Tokyo Japan J. von Bruchhausen, Nuclear Power expert, Lahmeyer International GmbH, Frankfurt FRG A. P. Coleman, Electric Utility Systems Planning expert, Associated Nuclear Services, London UK J. T. Roberts, Economist, IAEA USA

Status 1 = Cost-free expert with salary, travel and per diem paid by the sponsoring country Status 2 = Expert provided by contract with Engineering Consulting firm, the firm having the status of an independent contractor Status 3 = IAEA staff member

N-l