joint UNDP/World Bank Energy Sector Management Assistance Program THAI LAND IMPACT OF LOWER AND UNCERTAIN OIL PRICES ON ENERGY SECTOR INVESTMENTS APRIL 1988 Energy Efficiency and Strategy Division Industry and Energy Department World Bank Report of the joint UNDP/%4Dwd Bank Energy Sector Management Assistance Program This document has a restricted distribution. Its contents may not be disclosed without authorization from the Government, the UNDP or the World Bank. Views expressed in this working paper do not necessarily reflect the views or official policy of the Government, the UNDP or the World Bank. THAILAND IMPACT OF LOWER AND UNCERTAIN OIL PRICES ON ENERGY SECTOR INVESTMENTS APRIL 1988 ACRONYMS DED Defense Energy Department DMR Department of Mineral Resources EGAT Electricity Generating Authority of Thailand E&P Exploration and Production ICB International Competitive Biddinz IOC International Oil Companies LPG Liquefied Petroleum Gas MEA Metropolitan Electricity Authori:y NEA National Energy Administration NESDB National Economic and Social Development Board PEA Provincial Electricity Authority PTT Petroleum Authority of Thailand RTG Royal Thai Government TP Texas Pacific Oil Company ABBREVIATIONS bbl Barrels (approximately 6.6 barrels per tonne fuel oil) bpd Barrels per day Bcf Billion cubic feet (109) BTU British Thermal Unit - a unit of heat equal to 0.25 k calories c.i.f. Cost, insurance, and freigh: M Thousand m (MM) Million mbbl Million barrels MMCFD Million cubic feet per day TCF Trillion cubic feet (1012) toe Tonnes oil equivalent (10.415 x 106) kcals km kilometers CURRENCIES 2QUIVALENTS (As of March 1987) 1 Baht = US$0.038 26 Baht = US$1.00 ENERGY CONVERSION FACTORS Energy Form Calor, ic va'ue TOE 'ton (m Iion K 0oca lor;es/ton) Petroleum Products LP'G 0.8 1.059 Gasol ine 10.5 1.029 jet Fuei 10,4 1.020 Kerosene 10.3 1.010 Gas Oil/Diesel 10.2 1.00 Fuel Oil 9.5 0.941 Crude 0;1 10.2 1.00 Electricity Calorific value = 860 kcal/kWh (0.084 TOE/MWh) Otrer Te'ric ton Df crude oil = 7.505 Darrets o; Nigerial Bonny Light impe,iai ga lon = '.2 U.S. g3,bons This report is based on the findings of a mission which visited Thailand ir. March 1987. The mission consisted of . Aleem (Mission Leader); L. Wijetilleke (Refinery Specialist); T. Fitzgerald (Geologist); S. Shum (Gas Specialist); I. Harlow (Gas Economist); F. Lecaros (Power Special- ist). Secretarial assistance was provided by Basharat Ahmad and Morrissa Young. TABLE OF CONTENTS Page SUMMARY AND RECOM1ENDATIONS .......................................... i I. COUNTRY BACKGROUND AND RATIONALE FOR STUDY . ........... 1 Introduction ..1....................... Purpose of Study and Report Structure .................... .Il Background/Rationale ....................................... 2 The Issues ................................................. 3 (i) Short-term consequences ............................. 3 (ii) Longer-term consequences ............................ 3 A Macroeconomic Perspective .................. 4 Recent Economic Developments .. 4 Energy-Economv interactions .. 5 Medium-Term Outlook .. 5 Overview of the Energy Sector .. 6 Composition of Supply .. 6 Composition of Demand .. 6 Energy Planning and Sector Strategy . . 6 II. ANALYTICAL FRA-MEWORK FOR EVALUATING IMPACT OF OIL PRICE UNCERTAINTY ON INVESTMENT .................................... 9 Uncertainly and the Use of Multiple-Variate Scenarios ...... 9 Choosing Between Investment Strategies in an Uncertain Environment ................................................ 11 III. STRATEGY FOR INVESTMENT IN POWER GENERATION .................. 13 Introduction ........ I...................................... 13 Existing Power Svstem ...................................... 14 Plans --or Future Investment ................................ 14 Anaiytical Approac ......................................... 15 Impact of Uncertalnty on Least Cost Investment Plans ....... 16 Evaluating Alternative Investment Strategies ...............1 8 Adapting .nitia' Investments to TOM or SQ ................ 18 Strategy Evalua-ion ......................................21 The Recommended Strategy: Problems and Options ............. 23 The Criteria for Selection and the Recommended Strategy.. 23 Adaptation Problems ...................................... 24 Linkages with the Gas Subsector .......................... 25 Conclusions and Recommendations ............................ 26 Summary of Results ....................................... 26 Recommendations .......................................... 26 IV. STRATEGY FOR INVESTMENTS IN THE GAS SUBSECTOR ................ 28 Introduction ............................................... 2; Options for Further Gas Development ........................ 29 (i) The U3 deve:opmen .................................... 29 (ii) Development of the TP field .......................... 30 (iii) Development of ESSO field at Namphong ................ 30 Methodology for Analysis ........................... 31 Impact of Uncertainty on the Development of Lhe Gas Subsector .................................................. 32 Sequence and Timing of Field Development . . 32 Supply-Demand Balances and Production Profiles. ....::- 35 ImDlied Least Cost Investment Plans for fOM and SQ ....... 38 Strategy Issues and Options .. ...................... . 40 Decision Agenda for 1987-89..........................42 Recommendations ......................... 43 V. STRATEGY FOR INVESTMENT IN THE REFINING SUBSECTOR ............ 45 Introduction ............. 45 Background ........... 46 Crude Oil and Petroleum Product Prices ................... 47 Demand Projections for Petroleum Products ................ 47 Impact of Uncertainty on Least Cost Investment Plans ..... 48 Investment Strategies and Risks ................ I ........... 50 Strategy Fvaluation ................... 53 Conclusions and Recommended Strategy ............. 53 VI. INSTITUTIONAL PLANNIING AND DECISION-MAKING UNDER UNCERTAINTY ...,........... , 55 Introduction ............. 55 Planning Framework and the Dependence on Forecasting .. 55 Changes in the Decision-Making Process . . 57 TABLES 2.1 Summarv of Scenarics................... .......... ........ 11 3.1 EGAT Installed Capacity, 1986 ........ 14 3.2 EGAT Power Development Plan, July 1986......... .......... 15 3.3 Least Cost Investmenr Plans Under TOM and SQ Scenarios .. 17 3.4 Summarv of Decisions for Alternative Strategies ............. . 19 3.5 Adaptation to the TOM Scenario . . ... I ......................... 20 3.6 Adaptation to the SQ Scenario ................................ 20 3.7 Regret Values for Individual Strategies . . 21 3.8 Investment Recommendation .. 27 4.1 Key Scenario Assumptions .. 31 4.2 Reserves Assumption .......................................... 32 4.3 Total Gas Reserves under Different Assumptions . . 34 4.4 Economic Cost of Natural Gas (NPV in 1987) . .................. 34 4.5 Demand for Gas ............................................... 35 4.6 Gas Netback Values (NPV in 1987) .. 35 4.7 Supply-Demand Balance, TOM Scenario .. 37 4.8 Supplv Demand Balance, SQ Scenario .. 38 4.9 SQ and TOM Scenarios; Investment Requirements, Gas subsector (1987-95) ... 39 4.10 Returns on Second LPG Plant .. 39 4.11 Regret Values for Individual Strategies . . 43 5.1 Thailand: Total Refinir. Capaciy . .46 5.2 Crude Oil and Petrolem Product Prices ..47 5.3 Projected Demand for Diesel and Gasoline ..................... 48 5.4 Projected Demand for Fuel Oil ................................ 48 5.5 Least Cost Investment Plans under TOM and SQ Scenarios ....... 49 5.6 Adaptation to the TOM Scenario ............................... 5; 5.7 Adaptation to the SQ Scenario ................................ 52 5.8 Regret Values for Individual Strategies ...................... 53 FIGURES 1 Development of Crude Oil Prices. 2 ANNEXES 1 Scenario Assumptions .59 2 Energy Balances. 6 3 Basic Power Sector Data and Assumptions .63 4 Results for TOM Scenario .67 5 Results for SQ Scenario ...................................... 75 6 Pipeline and Compressor Investments .83 7 Marginal Gas Netbacks .84 8 Geological and Associated Production Information .86 9 Netback Values (NPV in 1987) .87 10 Gas Availability .88 11 Indicative Cost of Production froan New Contract Areas .91 12 Decision Agenda 1: Evaluation of Alternative Strategies .92 13 Scenario Implications for Gas Subsecor .94 14 A Combined Gas/Power Subsector Strategy .99 15 Petroleum Products Demand Projection .102 16 Crude Oil and Petroleum Product Prices ....................... 103 SUMMARY AND RECOMMENDATIONS Introduction 1. This report evaluates tne impact of lower and uncertain oil prices on erergy sector investments in Thailand and attempts to identify policies and investment options that will give the country the flexibility to adapt to a range of circumstances at least cost. 2. The past two years have seen another major change in international oil prices. Oil prices have collapsed from about $28 per barrel in January 1986 and have since been fluctuating in the $10-20 per barrel range. 1/ The prices of oil products also have declined with crude oil prices, although the extent of the decline has varied according to product thus altering differentials. As a consequence of these changes, differentials between the prices of petroleum products and other fuels such as coal and gas have also been significantly altered. 3. The decline in oil prices has had a significant impact on the Thai economy which depends on imported petroleum to satisfy more thqn half of its commercial energy needs. In the past oil imports have been a major drain on foreign exchange earnings; over the period 1980-85 the cost of oil imports has absorbed over a third of export earnings. The fall in prices in 1986 has saved the country close to $1 billion (2% of GDP) and helped to achieve its first current ccount surplus in more than 20 vears. Only a small part of this windtall gain was passed on to consumers in the form of lower retail prices of petroleum produc.s. The rest was retained by the Government through higher excise taxes, and the resulting increase in fiscal revenue has enabled national savings to rise from 18% GDP in 1985 to 20% in 1986. The developments on the oil price front have taken place against the background of an ongoing program of economic reforms introduced by the Government during the Fifth Plan period (1982-86). An objective of the program is to reduce, and avoid a recurrence of, serious imbalances that had emerged and persisted since the late i970s in the form of the savings-investment gap and the associated current account deficit. By directly helping to reduce these imbalances, the fall in oil prices has provided Thailand with an unusual opportunity to restructure its economy without sacrificing its growth momentum. 1/ For a countrv like Thailand, the fall in oil prices has been even greater in real terms because the US dollar, the currency in which oil prices are denominated, has depreciatec sharply since 1986, thereby reducing the price cf oi: relative to trat of non-d;llar denominated imports. 4. Within the energy sector important implications of the collapse in oil prices arise because previous expectations about the future movement of oil prices have been altered. No firm views exist about the possible evolution of future oil prices except, perhaps, for the widely held opinion that in the medium to longer term they will start rising again. The change in perception about future energy prices requires a reassessment of energy policy and :-vestment strategy for the next decade. In the short run, lower oil prices have opened up possibilities for gains from interfuel substitution, while they may also require that incentives for petroleum exploration be strengthened. For the long run, an investment planning framework is needed which allows policy makers to make decisions with confidence in an uncertain environment. Within this framework, Thailand must identify specific strategies for investing in power generation, gas projects, and refineries that will allow the country to benefit from lower oil prices without jeopardizing longer term options for the development of its energy sector. 5. Underlying the concerns about investment strategy is the desire of the Government to evaluate the possibilities afforded by the change in oil prices to achieve its broader objectives for the energy sector. The Government's goals for the sector during the Sixth Plan period (1986-91) are: (a) to satisfy the growth in demand so that energy supply facilitates, rather than constraints, economic development; (b) to reduce the country's reliance on imported energy; and (c) to reduce the share of public sector investment allocated to energy. During the Fifth Plan period (1982-86) the Government was able to sharply reduce the country's dependence on imported energy (mainly petroleum). As a result of increased domestic production of natural gas and lignite, the share of imported petroleum in commercial energy consumption declined from about 90% in 1980 to 55% in 1986. However, the reduced dependence on imported energy was achieved at a high cost, requiring almost a third of total public capital expenditure during the period 1982-86. 6. Consistent with its goals for the Sixth Plan period, the Government had in 1985 decided to reduce the share of energy in public investment to 26%. To resolve the apparent inconsistencies between the desire to reduce public investment in energy and what it concluded was an economically justifiable objective of reduced dependence on imported energy, the 1985 Energy Assessment Report proposed an expanded role for the private sector in developing domestic energy supplies, especially petroleum. 2/ However, the decline in oil prices has made it more difficult to implement the solution proposed by the Energy Assessment Report regarding the extent to which reliance can be placed on private resources for investment in the sector. Lower oil prices have given rise to a view within the Government that public investment should be further reoriented towards other sectors of the economy especially at a time of 2/ See "Thailand: Issues and Options in the Energy Sector". (Report No.5313-TH, September 1985). - iii. - macroeconomic restraints on borrowing by the public sector. Yet, in the current economic environment where lower oil prices have created disincentives for private investment in petroleum exploration and recent upsurge in economic activity has generated greater demand for energy investment, a sharp reduction in public energy investment may be difficult to achieve, This Report attempts to throw more light on this problem. 7. Within the above context the Report proposes an investment planning framework to assist policy makers in evaluating investment options and strategies in an uncertain environment. The sn..rt-term questions have already been addressed in briefing papers completed in an earlier pLase of the study (see paras. 1.6-1.7 for main findings). Impact of Oil Price Uncertainty on Investment: Suggested analytical framework 8. The planning and development of energy sector facilities in Thailand, such as power generating plants and petroleum refineries, is based on a set of forecasts that encompass demand, prices, and fuel availability. These forecasts provide the foundation for long-terra. expdnsion plans which are used by policy-makers as a framework for optimising near-term investment decisions. Once these decisions have been taken the long lead times associated with most energy sector projects, as well as their relatively large-size or "lumpiness", limit the flexibility with which investment can be adjusted under conditions of uncertainty. Indeed, if projects could be cancelled or scaled down at no cost and if construction lags were negligible the energy supply system could be kept to a near optimal development path. 9. Because of the aforementioned rigidities the economic costs of uncertainty materialise when actual developments deviate from the underlying forecasts of earlier investment decisions. The risks associated with any set of investment decisions is defined by: (a) the extent and nature of uncertainty in the future environment facing policy- makers; (b) sensitivity of investment choices to forecasts of future developments; and (c) rigidity of the investment and the costs associated with adapting it to an environment different from one originally forecast. The approach used in this analysis takes account of all these three factors in evaluating the risks of alternative investment strategies in the wake of lower and uncertain oil prices. Defining uncertainty in a complex environment: the scenario approach 10. There are three major interrelated uncertainties that will influence the level, composition and timing of energy investments in Thailand over the next 10 to 15 years: (a) the price of oil; (b) the availability of domestic gas supplies; and (c) the growth in domestic demand for energy products. The price of oil will influence both the - iv - supply of gas and the demand for energy. The linkages between oil prices, gas supplies, and energy demand imply that oil prices impact on energy investments in Thailand through a variety of channels, both direct and indirect, including: (a) the impact r oil prices on energy demand and, through it, on the level of in, ants in refining and power generation capacity; (b) the impact of re-cive fuel prices on investment choices and composition; and (c) the impact of oil price levels on the search for, and availability, of natural gas, and the need for associated investments in infrastructure (pipelines, etc.). To add to the complexity of the environment facing policy-makers, the three variables-- oil prices, gas supply, and energy demand--share some common features, but they are not rigidly linked. Thus, geological uncertainties regarding gas reserves could more than offset the effect of changing oil prices on the supply of gas. Hiven the limited resources available for the exercise, the mission has adopted a scenario approach for evaluating investment decisions under uncertainty. Two multiple variate scenarios were developed in collaboration with the NESDB and the line agencies to encapsulate the uncertainty in oil prices, as well as the related uncertainties in gas supply and energy demand: (a) Tightening Oil Market (TOM) scenario, where rising oil prices are associated with increases in gas supplies and a lower rate of growth in energy demand. Higher oil prices offer producers increased incentives to explore for, and develop, oil and gas resources, in Thailand as well as abroad. While supplies of energy, particularly domestic gas, are plentiful, economic growth, and with it, energy demand, is tempered by rising oil prices. (b) Status Quo (SQ) scenario, where the current softness in oil prices is maintained for an extended period, thus reducing the attractiveness of oil and gas exploration and development. In this scenario, where gas supplies are limited, stagnant oil T. ces encourage a higher rate of growth of the economy and of t.rgy demand. 11. The TOM and SQ scenarios are intended to be plausible descriptions of alternative states of the world; they are not forecasts. The scenarios are designed to cover a credible range of uncertainty and this is reflected in the values of the key variables which were agreed with the NESDB and the line agencies. The scenario values, which are consistent with the results of macroeconomics and sectoral models used by the NESDB, are summarized in the table below. SUMMARY OF SCENARIOS Tightening Oil MarKet (TOM) Status Quo (SQ) 1986 1990 1995 2000 1986 1990 1995 2000 Oil Prices 15 20 25 30 15 15 '7 20 (S/barrel, 1986 prices) GDP Growth 3.51 per annum 6% per annum Gas Avaia1Dility 470 570 1,100 1,200 470 500 570 500 (MMCFD) Source: Projections agreeo Dy the mission with NESDB and line agencies (See Annex 2). Sensit: ity of investment plans to oil price uncertainty. 12. Usir.g the. ro scenarios outlined above, the study was carried out in two stages. In the first stage, least cost investment plans were estimated for the energy sector under each scenario, taking into account the prices of altc.native fuels and constraints in the supply of domestic gas and lignite. The differences between the least cost investment plans in terms of the level, composition, and timing of investment provide a measure of the magnitude of the problem created by oil price uncertainty. Strategies to reduce the economic costs of uncertainty 13. The two least cost investment plans derived in the first stage of the Study also serve as 'benchmarks", representing investment plans with perfect foresight (i.e., they reflect optimal development of energy sector facilities if planners in 1988 knew with certainty that the assumed environment, TOM or SO , would emerge over the next 10 to 15 years). In practice policy-`,nakers will be less certain about future developments and the second stage of the Study evaluated the costs and benefits of alternative strategies to cope with uncertainty. For example, one strategy in the power sector might be to install triple- fired power generation facilities which would allow plants to run on the most economic fuel (coal, gas, or fuel oil) at any time. This flexibility, however, would be gained at a certain cost, which may be viewed as an insurance premium. An alternative way of coping with uncertainty is to gain the flexibility upstream by ensuring (e.g., through incentives to gas producers), that a minimum amount of gas supplies (higher than in the SQ scenario) is always available. Such a strategy, requires careful consideration to ensure that gas producers are not excessively rewarded. 14. To decide between competing strategies, the analysis focuses on the short-term decision agenda (1988-89). It identifies the set of initial investment decisions that need to be made during 1988-89 as near-term components of each strategy and evaluates the cost - vi - consequences of subsequently adapting that strategy to widely differing circumstances, as illustrated by the TOM and SQ scenarios. This is done by assuming that: (a) initial investment decisions made during 1988-89 as part of each strategy are irreversible; and (b) taking the initial investment as given, policy makers, beyond 1989, will adapt (re- optimise) the investment program for 1990-2001 to the scenario or environment which the) expect to emerge at that time. The efficiency with which a given investment strategy adapts to each scenario is assessed by comparing the net present value (NPV) of the costs of meeting future demand of the initial investment component of the strategy and subsequent adjustments to longer term investments, with the NPV of the "benchmark" investment plan for that scenario. This difference between the two programs represents the regret associated with a given strategy, i.e., the cost of being wrong. A final recommendation on the optimal strategy to cope with uncertainty is based on a comparison of regret values. Sensitivity of Future Investments to Oil Price Uncertainty 15. A comparison of least cost investment plans estimated by the mission for the two scenarios reveals that the level, timing, and composition of future investments in all the three subsectors (power, gas, and refining) are highly sensitive to the evolution of oil prices. 16. Investments in power generation. In terms of the level of investment, capital expenditures in power under the SQ scenario are about twice as high as under the TOM scenario; total undiscounted investment for 1987-2001 is $5.5 billion under SQ and $2.7 billion under TOM in 1986 prices (see Table 3.3). These results suggest that the indirect effects of oil prices on power investment--through their influence on fuel choices and power demand--dominate generation expansion plans, and that investment levels are inversely related to oil price changes. The two least cost investment plans also differ substantially in terms of timing and composition. Under -TOM, new investments in power are essentially gas based (combined cycle plants) and the development of large steam units such as those at Bang Pakong (coal-fired) and Mae Moh (lignite) are delayed to 1999 and beyond. SQ, on the other hand, calls for these units to be advanced--they are commissioned in 1995. Finally, under the SQ scenario, Thailand would have a much greater dependence on imported fuels for power generation (fuel oil and coal) than under the TOM scenario. In SQ, fuel oil acts as the swing fuel in the first half of the 1990s: as the limited gas available is concentrated in combined cycle plants, existing dual-fired thermal plants are progressively switched from gas to fuel oil. Beyond 1995, both coal and fuel oil would progressively compensate for the decline in gas supplies. 17. Investments in the gas subsector. In gas, total undiscounted investments over 1987-95 under a regime of rising oil prices (TOM) were, -vii - at $260 million, about twice the level under the scenario of stagnant prices (SQ) (see Table 4.9). The composition and timing of investment also differ between the two scenarios, reflecting, in part, corresponding variations in gas demand for power generation. Under TOM where gas production is higher to fuel a greater number of combined cycle power plants, investment is needed in the Namphong-Bangkok pipeline by 1993 and in an offshore gas compressor (to increase the capacity of the Erawan-Bangkok pipeline) by 1991. Neither investment is needed under the SQ scenario because of the lower level of economically recoverable reserves. The capacity of the LPG plant in TOM (200 MmCFD) is also greater than in SQ (150 "MCFD). Lastly, because of the shortage of gas under SQ, the TP-Erawan pipeline is planned for commissioning by 1995, while under TOM this investment can be delayed until 1999. It should be noted that investment in the gas subsector react differently from power in the two scenarios; higher oil and gas prices increase investments in gas due to incentive effects while in the power sector they lead to lower investment by tempering demand and enabling increased use of less expensive gas combined cycle plants. 18. Investments in refining and oil supply infrastructure. In the refining subsector as well, the pattern of investment under the two scenarios differs substantially. Despite a lower level of demand, the TOM scenario calls for significantly higher investments than SQ; total undiscounted investment for 1987-2001 is about $340 million under TOM and $200 million under SQ, in 1986 prices (see Table 5.5) The lower level of investment in SQ arises because product price differentials and margins do not justify investment in refinery facilities. The increase in demand is met through increased imports; $200 million are required for infrastructural investments--increasing storage and improving port facilities--to accommodate much higher imports. Under TOM, the pattern of product prices justifies the rehabilitation and upgrading of facilities at Bangchak, including the installation of 14,000 barrels per day (bpd) of cracking facilities. To meet growing demand, the least cost plan under TOM also calls for the commissioning of a new refinery by 1996 with 100,000 bpd of additional crude distillation capacity and 21,000 bpd of secondary conversion facilities. Product balances reflect the different pattern of investments under the two scenarios: under SQ, large imbalances (shortfalls) emerge between demand and domestic supply for all products by 1997 reflecting corresponding levels of imports. In contrast, under TOM demand and supply are approximately in balance for most of the products, although a significant amount of middle distillates (kerosene and diesel) will still need to be imported. 19. A conclusion that can be derived from these results is that least cost investment plans for energy subsectors, which implicitly assume that the future is known with certainty, may not be the optimal strategy for the country. An optimal investment strategy or plan in these circumstances is one which is sufficiently flexible to adapt quickly to circumstances different from those for which it was designed; a premium has to be paid to achieve this flexibility and the key question is which investment option can provide this flexibility at the - viii - lowest premium. The investment decision process should therefore consider how robust a particular investment plan is against alternative scenarios which cover the range of uncertainty. 20. The above results also suggest that the Government should carefully review the premises underlying suggestions for cut-backs in public investments in energy, especially in the power subsector, in the wake of lower oil prices (see para. 6 above and para 1.18 in main text): a prolongation of the current environment of soft oil prices, as depicted in the SQ scenario, calls for higher, rather than lower, investment in power which acounts for almost three-quarters of public investment in energy in the Sixth Five-Year Plan (1987-1991). Investment Strategy In an Uncertain Environment 21. To help the Government make decisions in this uncertain environment, investment strategy needs to focus on identifying the policies and investment options in the near-term, which will allow the economy the possibility to adapt to a range of circumstances at least cost. The importance of selecting an appropriate strategy is underlined by the results of the analysis which suggest that the costs to the country of making the wrong decision in present circumstances could be significant; for the energy sector as a whole, the incremental cost of switching from the optimal to even the second best strategy is estimated to exceed US$100 m, in present value terms. To help identify an optimal investment plan, existing and alternative strategies, representing the main feasible options open to the Government, were considered by the mission for each subsector. Strategy for Investment in Power Generation 22. An examination of alternative strategies reveals that future development of the power system in Thailand would best be based on first developing combined cycle units to be commissioned during the early 1990s while delaying larger (lignite and coal based) steam units until the mid-1990s. 3/ This strategy can easily cope with the emergence of a 3/ The maximum cost or regret (as defined in para 14 above) associated with adapting the recommended strategy to a wide range of circumstances is estimated to be above $60 million (at 1986 prices) and is incurred in adjusting it to the SQ scenario. The recommended strategy involves the lowest ievel of regret (to the economy) of all the feasible options considered by the mission. In comparison, the maximum regret associated with a strategy based on the early development of large and costly triple fired steam units is estimated to be $121 million; adapting this strategy to the SQ scenario would incur costs to the economy exceeding $70 million. - ix - low demand, high oil price scenario by merely delaying decisions related to future power plants and rescheduling power plants under construction for one year. Coping with a scenario where gas supplies would become relatively scarce and demand would grow strongly due to low oil and gas prices would possibly call for some load shedding in the early 1990s. This strategy, which depends for its implementation on the availability of an appropriate level of gas supplies, would call for the conclusion of at least one additional gas production contract with the private sector gas producers by 1988. The recommended strategy (of an early development of combined cycle gas units) would also fit well in an environment where gas supplies were abundant and demand grew strongly under the influence of high GDP growth and low oil prices (i.e., a combination of TOM and SQ environments, which the Government currently considers a distinct possibility). Recommendations 23. Specific recommendations are as follows: (a) Consistent with the recommended strategy, short-term investment decisions are needed during 1988-89 for developing gas turbines and combined cycle plants as shown in the table below. INVESTMENT RECOMMENDATION FOR POWER Cost Year in whicn Commissioning Investment 1986 $ aecision needs Date to be made Gas turbine (2xl0O MW) $ 32m 1988 1991-92 Combined Cycle Unit 1 (300 MW) $185m a/ 1988-89 1992-94 Combined Cycle Unit 2 (300 MW) S185m a/ 1988-89 1992-94 a! Assumed location, Bang Pakong. Source: (b) At least one additional gas production contract should be in place before the decision is taken to invest in new combined cycle power plants. (c) In order to prepare for the possibility of a higher demand materializing in the next decade, EGAT should proceed to make advance preparations for gas and lignite fueled power plants including completion of preliminary designs and feasibility studies. The latter would include identification of sites for locating plants and assess the need for complementary invest- ments in infrastructure. (d) Another measure to prepare for higher demand would be to explore demand management possibilities. For example, agreements could be made with large industrial customers for non-disruptive peak load reductions, possibly by providing appropriate incentives. (e) Interconnection agreements with Malaysia and Laos should be sought for the purpose of obtaining support during peak load periods. Strategy for Investment in the Gas Subsector 24. The results of the study indicate that the lowest cost to the economy is associated with a strategy that calls for signing up gas production contracts with Unocal (for the third area it is developing off-shore and referred to as U3) and Esso (for its on-shore field at Namphong) as soon as possible and delaying decisions both on the development of the Texas Pacific (TP) field and on the installation of the off-shore gas compressor. This strategy provides the flexibility to adapt to either TOM or the SQ scenario at least cost. It is based on the results of the scenario analysis which suggest that developing the U3 and Esso fields in sequence (with TP last) would be part of the least cost development of gas resources under both scenarios. The cost of the recommended strategy is significantly lower than the option (also analyzed) of developing the TP and Esso fields before U3 in order to diversify the sources of gas supply; the higher cost of a multiple- sourcing strategy is also a reflection of under-utilized infrastructure capacity. 25. The evaluation of alternative strategies and hence the recommended strategy was based on the assumption that new production contracts can be concluded based on price levels similar to those under existing contracts at the time of the mission. However, a major problem to be considered by decision-makers is the need for additional measutes to reduce the economic losses associated with the high level of demand rationing in the SQ scenario, should it emerge. One possibility, in these circumstances, would be to increase producer prices, net of taxes and royalties, for future contracts above the level of prices based on previous contracts. While the TOM scenario also assumes generally more optimistic conditions, gas supplies under an SQ environment could be increased by providing incentives through higher gas prices. 26. Before any major review of gas prices, the Government needs to make a judgement about which environment it is in or moving towards. If the environment is going to be similar to TOM any substantial premium would prove unnecessary. Also, if an SQ scenario prevails, but with a significantly higher level of reserves than assumed in the main scenario, then shortages will be less of a problem and only a modest premium may be suffficient and justifiable. If, however, future developments are expected to be close to the main SQ scenario, with its attendant high- level of shortages, then the case for considering a significant increase 4 I~ ~ ~ ~~~~~~x in prices is strengthened. The mission's discussions in Thailand reveal that at least one new production contract can be concluded in the near future even at prices no higher than those based on existing gas contracts; higher prices will expedite, but were rnot viewed as a necessary condition for, the conclusion of these contracts. As indicated in the evaluation of power strategy above, an additional gas production contract is needed in ;988 to avoid delaying the power investment program. In these circumstances the RTG can delay the initiation of any strategy for meeting demand uncertainty through significantly increased prices for gas supplies, at least for a few years, until the likely environment becomes clearer. In general, the analysis carried out above suggests that raising prices need not be part of the decision agenda for 1988-89. Recommendations 27. Consistent with the least cost strategy the specific recommendations are as follows: In early 1988: Finalize U3 contract for a minimum production level from 1990 onwards of 550 MMCFD as a combined total from Ul, U2, and U3 (paras. 4.33 and 4.37). Include in the contract a provision for increased outputs in the future at mutually agreed prices (paras. 4.27- 4.29). Complete and facilitate testing production agreement with Esso. Initiate investment in a second LPG plant of 200 MM'CFD capacity (para. 4.33). Work out proposal to develop the TP field based on a detailed economic evaluation of gas production there. In 1988: Monitor Esso testing; if initial assessment of reserves proves disappointing (i.e., close to or less than SQ levels of 0.5 TCF, lower end of expected range), accelerate and complete arrangements for development of TP gas. In 1989: Check Esso results; if satisfactory (i.e., proven reserves within the 0.5-1.7 TCF range), determine need for pipeline to Bangkok; if Esso volumes are low (less than 0.5 TCF), complete arrangements for development of TP gas. In 1990: If demand and price follow TOM scenario, make decision to install offshore compressor. If demand is high and an SQ-type environment prevails, consider the possibility of raising prices for increased output (paras. 4.27-4.29) and complete negotiations for the development of TP, if not already completed. - xii - It is to be noted that no investments are part of the recommended decision agenda for 1988-89 apart from the LPG plant ($42m). Strategy for investment in the refining subsector 28. The impact of the volatility in oil prices is particularly felt in the refining industry. Refining investments are closely linked to the viability of secondary conversion (i.e. facilities for converting fuel oil to lighter products such as gasoline, kerosene, and diesel), since in the absence of investments in secondary conversion, the rehabilitation or extension of primary distillation capacity is not expected to be viable in any credible scenario. Based on the range of prices and differentials covered by the two scenarios, the best near-term investment strategy is to make a modest investment (of about $50 million) in conversion capacity and to rehabilitate primary distillation capacity at the Bangchak Refinery. This strategy gives the refining industry the flexibility to adapt to future circumstances at least cost. 29. Domestic demand for fuel oil will depend on future oil prices as well as ECAT's own specific requirements. The considerable uncertainty associated with both these factors does not make it economically justifiable to aim at meeting domestic demand in full at all times. Rather, the recommended strategy will add facilities to meet a limited increase in product requirements in the expectation of moderately profitable margins, and the RTG should monitor the situation over the near one to two years, particularly the movement of demand and/or profit margins. Given available refining capacity, and recommended additions to it, decisions for further expansion (over and above recommended additions) can be delayed until 1989. By all estimates, fuel oil is projected to be readily available at competitive prices both regionally and globally, requirements therefore could be imported to meet the demand by power and industrial sectors. The recommended strategy, which includes facilities for upgrading fuel oil to lighter products together with the upgrading of the TORC refinery already underway will help to structure the industry to meet transportation requirements and middle distillate demand to the extent economically justifiable and allow fuel oil to be imported or exported as dictated by market conditions. Institutional Planning and Decision Making under Uncertainty 30. To increase the flexibility of the energy sector to adapt to an uncertain environment quickly and cost effectively, and to support the subsector strategies recommended above, the institutional decision making process should also be streamlined to: (a) improve the coordination and information flows between line agencies; (b) reduce the time lag involved in making investment decisions; and (c) shift the emphasis of sector planning away from an excessive dependence on single line forecasting to one using a range of values for key variables, such as oil prices and gas availability, and establishing linkages between these variables. To I I - xiii - speed up the decision making process the mission recommends that a "reserve" of pre-approved projects be established that would require only limited additional Government clearance before implementation. DAP/THAI-B/28-APR-88/ch I. COUNTRY BACKGROUND AND RATIONALE FOR STUDY Introduction Purpose of Study and Report Structure 1.1 The aim of the study is to assist the Government of Thailand to review its energy investment strategy so that Thailand can best exploit the recent changes in international oil markets without jeopardizing longer term options for the development of its energy sector. To achieve this objective the study has evolved as two interlinked components. The first part, already completed, has taken the form of background papers on the impact of lower and uncertain oil prices on petroleum exploration activities and the optimal mix of fuels to be used in existing power generation facilities. I/ These papers examined the case for a more flexible energy policy, which would permit short-term gains to the economy while retaining investment options consistent with the long-term strategy for the sector. The second and main component of the study, which is the focus of this report, proposes an investment planning framework to assist policymakers in evaluating investment options in an uncertain environment. Within this framework, the report attempts to identify strategies for investing in power generation, gas supply infrastructure, and refineries that will help give the country the flexibility to adapt to a range of circumstances at least cost. 1.2 The rest of this Chapter is divided into three sections. The first section looks at the background to, and rationale for, the study including the main findings of the briefing papers on short-term issues which were completed earlier. The last two sections outline recent developments in the macroeconomy and the energy sector that are of relevance to the analysis in subsequent chapters. Chapter II presents a long-term investment planning framework which takes into account not only the uncertainty in oil prices, but also the associated uncertainties in the availability of domestic gas supplies, and the growth in energy demand. Chapter III through V apply this framework to the development of specific strategies for investing in power generation, gas projects, and refineries. Finally, Chapter VI makes recommendations for institutional planning and decision-making under conditions of uncertainty. 1/ Background papers were produced on these subjects during July-August 1986. -2- Background/Rationale 1.3 The past two years have witnessed another major shift in oil prices. Crude oil prices have collapsed from about $28/barrel in January 1986 and have since been fluctuating in the $10-20/barrel range (see Figure 1). The prices of oil products also have declined in the wake of falling crude oil prices, although the extent of the decline varies according to the specific product; the prices of middle distillates (diesel and kerosene) have fallen less than those of fuel oil, thereby raising differentials between these products. In this process, price differentials between petroleum products and other fuels such as coal and gas also have been altered. =£igure1: DEVELOPMENT OF CRUDE OIL PRICES 12.0 - 4.0 - 28.0- 24.0 - 22 .0 - 20.0 - 16.0 - 14.0 - 12.0- 10.0- 8.0- 4.0 4.0 - 2.0 _ _ _ _ _ _ JAM APR JUL OCT JAN APR JUL OC JAM APt JUL OCT JA APBE JXL OC 1984 1 1985 1 198 1967 MICE I USD/ULI DA CIUV - - - ma? CRUD 1.4 Beyond the immediate, and largely beneficial, direct macro- economic impact through the balance of payments, discussed in the next section, the decline in oil prices has significant implications for the energy sector. These implications arise because expectations about the future movement of oil prices have `een altered with no firm views about their possible evolution except, perhaps, for the widely held opinion that in the medium- to longer-term they will start rising again. The -3- uncertainty and weakness in oil prices has raised issues in Thailand, as in other developing countries, relating to fuel substitution options in the short-term and investment priorities in the medium- to longer-term. A strategic consideration for the Government in circumstances where any specific price forecast runs the risk of soon being outdated, is the process of decision-making: what rules, criteria or strategy should the Government use to make key policy decisions, whether on pricing or investment at a time of fluctuating prices? Can the country benefit from these price changes without compromising long-term options? If so, how? The Issues 1.5 This study has focussed on two sets of issues. The first set has dealt with short-term policy implications of immediate concern to the RTG regarding incentives for petroleum exploration and interfuel sibstitution options. The second set of questions, representing the bulk of the exercise, has been concerned with the consequences of uncertainty for investment choices and priorities. 1.6 (i) Short-term consequences. In the case of short-term interfuel substitution, the study considered the prospects and constraints for using different types of fuels (lignite, fuel oil, or natural gas) in existing facilities for power generation. A methodology, based on the principle of avoided or opportunity cost was derived to assist the Government in making short-term decisions on this issue. The main conclusions of the analysis, outlined in earlier briefing papers, were: (a) lignite, which is a domestic resource, has the lowest short- run economic cost and the maximum available quantity of this fuel should be used for generation; (b) the marginal cost of using natural gas was about the same as for fuel oil--based on existing gas contracts with producers and the average price of fuel oil (about $9/barrel) during the period July-September of 1986--and as such, it would not have been advan- tageous to Thailand to cut gas production at that time; (c) the lagged adjustment of gas prices to oil price changes, as in the Government's contract with Unocal, can create significant economic losses for Thailand in a period of fluctuating prices by artificially introducing a gap between gas prices and the economic costs of substitutes (fuel oil) and it would be mutually beneficial to both the RTG and the producers to remove these rigidities; and (d) the presence of taxes and subsidies provides highly distorted signals to the power utility (EGAT) regarding economic costs of using different fuels for generation, and hence there is a need to review fiscal aspects of energy pricing to provide incentives for efficient production decisions. 1.7 The briefing paper on petroleum exploration concluded that the collapse of international oil prices and uncertainties about their future movement would have a negative impact on incentives for exploration and development activities in Thailand unless effective and early action was taken by the Government to offset the reduction in the expected profit- ability of these activities. A slowdown in oil and gas exploration will adversely affect oil and gas production proFiles in Thailand in the early -4- 1990s. The paper proposed specific modifications in the policy framework including a better incentive package, measures to help reduce the industry's costs, greater flexibility in work obligations, more efficient and speedier negotiations, and greater clarity in policies and objectives. 1.8 (ii) Longer-term Consequences. The key investment issues stem from questions about: (a) the optimal mix of power generation facilities for the future (i.e., the mix of gas, coal, fuel oil and hydro-based generation) taking account of the uncertainty in both the relative prices of fuels and the availability of domestic energy resources; (b) the timing and sequence of developing new gas fields and making associated investments; (c) optimal configuration of refineries and associated product import infrastructure at a time when refinery margins for the future remain uncertain; and (d) the effectiveness of institutional processes for making decisions in a period of uncertainty. 1.9 Underlying the above concerns on investment strategy are questions about the appropriate level of public investment in the energy sector in Thailand. With the energy sector expected to use up to 26% of public investment funds, oil price developments have given rise to expectations within Thailand, however incorrect, that the Government's investment program could be reoriented in the medium-term towards other sectors of the economy, particularly at a time of macroeconomic restraints on public foreign borrowing (see para. 1.19). A Macroeconomic Perspective 2/ Recent Economic Developments 1.10 Thailand's economy grew rapidly over the past two decades and developed successfully by most standards of international comparison. During this period, 1960-80, CDP growth averaged 7% per year and the country experienced a far-reaching transformation of its economic and social structure. However, like other middle income oil-importing countries, Thailand was adversely affected by the harsh conditions in the international economy between 1979 and 1982. During that time, the country experienced sharp increases in oil import prices and interest rates on foreign borrowing, while the demand and prices received for its exports were falling. These external difficulties were exacerbated by Thailand's heavy dependence on imported oil and an expansionary fiscal stance maintained since the late 1970s. Initially, the country was able to maintain growth, but at the expense of spiralling inflation, large 2/ For details of recent economic developments and a discussion of macroeconomic issues, see the most recent World Bank Country Economic Report on Thailand (Report 6036-TH, June 1986). -5- fiscal and external deficits, and growing dependence on foreign borrowing. 1.11 Recognizing the need for change, the Government in 1981 embarked on a major program to restructure the economy during the Fifth Five-Year Plan (1982-1986), supported by two structural adjustment loans (SALs) from the World Bank. While some progress was made, structural adjustment measures took longer to implement than originally expected and the serious internal and external imbalances in the form of the savings- investment gap and the attendant current account deficit proved difficult to control. To reduce the internal and external deficits, the Government implemented an additional adjustment program, supported by an International Monetary Fund (IMF) stand-by arrangement in 1985 and 1986. The adjustment program included depreciation of the baht, deceleration of monetary growth, and scaling down of the public investment program. At the same time, the ceiling on the public sector's external borrowing was reduced sharply from $2.6 billion to $1 billion per annum. Initially, these policies served to reduce GDP growth from 6.0% in t984 to 3.2% in 1985. However, as the policies took hold GDP increased by 3.5% in 1986 and is projected to have grown by 6% in 1987. There has also been a significant turnaround in the current account from a deficit of $2.1 billion in 1984 to a surplus of $0.2 billion in 1986. While the financial policies implemented by the RTG since 1984 have helped to achieve this improvement in the external account, a substantial contribution was made by the decline in oil prices. Energy-Economy interactions 1.12 The recent fall in oil prices has provided Thailand with windfall gains and an unusual opportunity for economic reform and adjustment. In the past, oil imports have been a major drain on foreign exchange earnings; in 1984 the oil bill of $1.8 billion absorbed over one-third of exports earnings. The 40% decline in oil import prices during 1986 saved the country close Lo $1 billion (2% of GDP), and helped it achieve its first current account surplus in more than 20 years. Only a small part of this oil windfall was passed on to consumers in the form of lower retail prices of petroleum products. The rest was retained by the Government through high excise taxes, and the resulting increase in Government revenue enabled national savings to rise from 18% of GDP in 1985 to 20% in 1986. The oil price decline has therefore also helped to reduce the savings-investment gap without any sacrifice in growth momentum. The fall in oil prices, equivalent to a 10% improvement in the terms of trade, has thus produced a significant macroeconomic impact. Medium-Term Outlook 1.13 The improvement in Thailand's current account now makes it possible for the Thai economy to resume high economic growth without straining the external balance. At the same time, many of the factors which contributed to Thailand's economic growth and improvements in the standard of Living over the preceding two decades are expected to -6- continue. These include the relatively equitable distribution of agricultural land, responsiveness of Thai farmers to improved technology, a good network of public infrastructure, and a dynamic private sector in industry, agriculture and services. Large reductions in population growth between the mid-1960s and 1970s and the discovery and exploitation of domestic oil and gas supplies also will support the country's continued development. These various factors provide Thailand's economy with the potential to grow at 5-6% per annum between 1987 and 1991. In the short run, reduced oil prices and the increased competitiveness of Thai exports would continue to stimulate economic activity. However, over the medium-term, the increase in the GDP growth rate can only be sustained by implementing macroeconomic reforms that would ensure an investment-savings gap consistent with a financially sustainable current deficit in an environment that could include rising oil prices. The economy's vulnerability to oil price rises is expected to increase in the mid 1990s as Thailand's dependence on imported energy begins to rise (see para 1.16). Overview of the Energy Sector Composition of Supply 1.14 In 1986 the gross supply of energy in Thailand was 28 million tons of oil equivalent (toe). 3/ Of this amount, biomass fuels provided 40% of this supply, and non-traditional commercial fuels the remaining 60%. Of the latter, petroleum products accounted for the largest share (64%); followed by gas (19%); lignite (9%); hydro (7%); and coal (1%). After netting out 30% of energy supply for transformation losses (losses in generation, transmission, and distribution of electricity), about 20 million toe were made available for final consumption. Composition of Demand 1.15 Between 1960 and 1986, total final energy consumption in Thailand grew at a rate of about 10% per year, well above the GDP growth rate over the same period of about 7% per annum. Electricity consumption over the 1960-80 period grew an average of 15% per annum, but slowed to 9.5% between 1980 and 1985. During the 1970s, petroleum product consumption expanded at about 10% per year; however, during 1980-1984 growth has been negligible because of rising domestic prices and substitutions of gas for fuel oil. In 1984, average oil consumption was 224,000 barrels per day. The composition of final energy consumption in 1986 was dominated by the transportation, residential, and industrial 3/ For 1986 energy balance, see Annex 1. -7- sectors, accounting for 35%, 32%, and 28% of final consumption, respectively. Most of the energy used in the household sector is provided by traditional fuels--fuelwood, charcoal, and agricultural residues. Electricity supplies 10% of household energy needs and 16% of industrial energy demand. Energy Planning and Sector Strategy 1.16 The Government's goals in the sector are to: (a) satisfy the growth in demand so that energy supplies facilitate, rather than hinder, economic development; (b) reduce the country's reliance on imported energy; and (c) reduce the share of public sector investment allotted to energy. The 1985 energy assessment report 4/ concluded that these long- term goals for the energy sector might be difficult to achieve unless specific policy measures were implemented. At the heart of the problem is the large share of petroleum imports in total or gross commercial energy consumption. The share of imported petroleum in total commercial energy consumption has declined sharply from about 90% in 1980 to an estimated 55% in 1986 due to increased reliance on domestic gas and lignite for power generation and industry. However, the supply of domestic gas and lignite is expected to level off in the early to mid- 90s, allowing the share of energy (mostly oil) imports to rise again. In the absence of additional steps to sustain oil and gas exploration efforts, imported energy could, by estimates made in the energy assessment, again account for as much as 70% of total commercial energy consumption by the turn of the century, with coal taking a 15% share. The increase in energy imports is likely to lead to a deterioration in the balance of payments and reduced GDP growth: even under the most optimistic scenario, where oil prices were assamed to remain at their 1985 levels to the end of the century, the energy assessment report projected the current account deficit to widen rapidly and curtail economic growth during the 1990s. Given the long lead time between exploration and production efforts, the disincentive effects of the recent (1986) decline in oil prices will be more acutely felt in the 1990s especially if oil prices return to their 1985 levels by that time. 1.17 To maintain a healthy balance of payments and steady CDP growth beyond 1990, while at the same time reducing the share of energy in the public investment program, the assessment report proposed an expanded role for the private sector in exploring, diversifying and developing domestic energy supplies. To provide the necessary incentives to the private sector to bring this about, the report recommended that: (a) the regulatory process governing private involvement in energy activities be streamlined, and (b) a pricing system be established which more accurately reflects the opportunity costs of energy resources. In 4/ "Thailand: Issues and Options in the Energy Sector", Report 5393-TH, September 1985. This was a joint project of the NESDB and the World Bank, funded by the JNDP. -8- addition, the report recommended the following priorities for energy sector development: (a) expanding the uses of natural gas; (b) formulating a program for expanding the exploration and development of lignite for power, industrial, and domestic uses; (c) increasing electricity tariffs to augment internal cash generation for the utilities and to discourage non-economic uses of power; (d) deregulating petroleum product prices and eliminating the oil stabilization fund; and (e) improving operational efficiency and sector planning in the energy agencies. The RTG has endorsed this strategy. 1.18 Even with a greater role taken on by the private sector, the proposed strategy will still require a substantial investment on the part of the public sector. The tentative program for the Sixth Five-Year Plan (1987-1991) calls for $5.6 billion in public investment, equivalent to 26% of the public investment program. About 71% of the planned investment is in the power subsector, with the remainder spread over petroleum (20%); lignite (6%); rural energy (2%), and energy efficiencv (1%). Both the level and composition of this investment plan have been questioned. In part this reflects the decline in, and current uncertainty surrounding, oil prices, which has raised some basic questions about investment strategy. Both at the time of, and just prior to, the mission there were discussions within the RTG about the need to cut back the power generation program because of (a) lower expected growth in power demand, and (b) the perceived need to curtail overall investment levels in the public sector. This study attempts to assess the justification for these perceptions in the current environment. II. ANALYTICAL FRAMEWORK FOR EVALUATINC IMPACT OF OIL PRICE UNCERTAINTY ON INVESTMENT Uncertainty and the Use of MultiDle-Variate Scenarios 2.1 The central aspect of the framework being proposed is the identification of the main uncertainties (including oil prices) that will influence energy investment decisions in Thailand and the linkages between them. There are three main uncertainties that will influence the level, composition and timing of energy investments in Thailand over the next 10 to 15 years. These uncertainties relate to: (a) the price of oil; (b) the availability of domestic gas suppLies; and (c) the growth in domestic demand for energy products. The price of oil influences both the supply of gas and the growth in energy demand. The uncertainty in gas supplies is in part due to the unknown outcome of negotiations between oil companies and the Government for contracts to develop fields that have already been discovered. Lower oil prices, in addition to their disincentive effects on petroleum exploration, reduce the likeli- hood of additional contracts being successfully concluded over the next few years; without new contracts, the supply of gas is expected to start declining over the next five to seven years. Growth in energy demand is also influenced by changes in oil prices both directly, through pricing effects, and indirectly, through balance of payments constraints on domestic economic activity. 2.2 The influence of oil price uncertainty on investment choices is a complex phenomenon. The linkages, indicated above, between oil prices, gas supplies, and energy demand imply that oil prices impact on energy investments in Thailand through a variety of channels, both direct and indirect, including: (a) the impact of oil prices on energy demand and through it on the level of investments in refining and power generation capacity; (b) the impact of relative fuel prices on investment choices and composition; and (c) the likely impa-t of oil price levels on the availability of natural gas and associated investments in infrastructure (pipelines, etc.). To add to the complexity of the problem, the three variables--oil prices, gas supply, and energy demand--share some common features but they are not rigidly linked. Thus, geological uncertainties regarding gas reserves could more than offset the effect of changing oil prices on the supply of gas. To overcome some of these difficulties, given the limited resources available for this exercise, the mission has adopted a scenario approach. Two multiple-variate scenarios were developed in collaboration with the NESDB and the line agencies, which are intended to encapsulate the uncertainty in oil prices, as well as the related uncertainties in gas supply and energy demand. (a) Tightening Oil Market (TOM) scenario, where rising oil prices are associated with increases in gas supplies and a lower rate of growth in econcmic activity, and with it, in energy I - 10 - demand. Higher oil prices provide incentives to producers to increase exploration efforts for both oil and gas, in Thailand as well as abroad. While supplies of energy, particularly domestic gas, are plentiful, energy demand is tempered by rising prices. (b) Status Quo (SQ) scenario, where the current softness in oil prices is maintained for an extended period, thus limiting the attractiveness of oil and gas exploration and development. Lower prices, however, encourage higher rates of economic growth and energy demand. 5/ 2.3 These two scenarios are designed to cover a credible range of uncertainty. They should be seen, however, not as forecasts but rather as plausible descriptions of alternative worlds. The mission was able to reach a consensus with the NESDB and the line agencies on the values to be used for the key variables in the two scenarios. These are presented in Annex 2 and summarized in Table 2.1. While the values used were derived largely on the basis of qualitative discussions, their consistency, in particular the linkages between variables at the macroeconomic level (GDP growth) and those relevant to the energy sector (such as oil prices), was established by comparison with the results of previous studies and the outputs of macroeconomic and sectoral models used by the NESDB. 2.4 The framework proposed and used by the mission assumes that actual developments will, barring exceptional circumstances, fall within the range of values used to define the TOM and SQ scenarios (Table 2.1). However, no assumptions have been made about the most 5/ An additional uncertainty that affects refinery investment is the differential that exists between fuel oil and gas oil prices. This differential is determined by a number of factors which are largely outside the scope of this study. In theory, therefore, a high or low price differential between gas oil and fuel oil could be associated with either scenario. However, in order to keep the analysis manageable, a relatively higher price differential might be associated with higher oil prices (TOM scenario): higher oil prices are likely to lead to relatively greater increases in gas oil rather than fuel oil prices because the possibilities for substituting alternative fuels for gas oil are more limited compared with fuel oil, which competes with oil, nuclear, and gas in the under-boiler market. The reverse would apply in the SQ scenario: the differential would be lower, reducing the incentives for investment in refinery conversion facilities. Empirically, this appears to be plausible; the evolution of the gas oil-fuel oil differential during the 1960s and 1970s is onsistent with these assumptions, i.e., that the differential has bcen relatively higher in times of increasing prices. - !1 - likely location of future values within the range, or the sequence or time-path they will follow. Actual developments could well oscillate beween a TOM and an SQ environment: for example the present circum- stances facing the Thai economy of relativelv low oil prices and strong CDP growth, circumstances somewhat akin to SQ, could well be replaced in the l990s by a TOM-type environment as oil markets tighten and economic growth is tempered by balance of payment difficulties. 2.5 To avoid, making a subjective assessment, probabilities have not been attributed or estimated for the occurrence of TOM or SQ. However, in comparing alternative investment strategies estimates have been made of the probability of occurrence of TOM and SQ that would make competing strategies equally attractive (see discussions in Chapters III-V). Table 2.1: SUMMARY OF SCENARIOS Tightening Oil MarKet (TOM) Status Quo (SQ) 1986 1990 1995 2000 1986 1990 1995 2000 Oil Prices 15 20 25 30 15 15 !7 20 (W/barrel, 1986 prices) GDP Growth 3.51 per annum 61 per annum Gas AvaiiaD- ;iy 470 570 1,100 1,20e 470 500 570 500 (MMCFD) Source: P-oject'ons agreed Dy the mission with NESDB and m,ne agencies. Choosing Between Investment Strategies in an Uncertain Environment 2.6 Using the two scenarios outlined above, the study was carried out in two stages. In the first stage, least cost investment plans were estimated for each scenario, taking into account the prices of alternative fuels and constraints in the supply of domestic gas and lignite. This first stage required the active participation of counter- parts from the line agencies in Thailand. It also made use of the results of ongoing studies such as the Chem Systems Refinery Study. The differences between the two scenarios in terms of composition and timing of investment provide a measure of the magnitude of the problem created by the oil price uncertainty. 2.7 In the second stage, the study evaluated possible investment strategies to cope with the problem of uncertainty. In essence, the approach has been to evaluate the costs and benefits of alternative strategies in terms of the flexibility they provide the economy to adapt to widely differing circumstances, such as those defined bv the TOM and SQ scenarios. For example, one strategy in the power sector might be to - 12 - install triple-fired power generation facilities which would allow plants to run on the most economic fuel at any time. This flexibility, however, would be gained at a certain cost, which may be viewed as an insurance premium. An alternative way of coping with uncertainty is to gain the flexibility upstream by ensuring (e.g., through incentives to gas producers), that a minimum amount of gas supplies (higher than in the SQ scenario) is always available. Such a strategy, while appealing, requires careful consideration to ensure that gas producers are not excessively rewarded. Another strategy is to focus on "core" projects, which are economic in the sense that they yield rates of return above a minimum rate for a range of oil prices, and then consider additional projects as oil prices rise above certain threshold levels. The problem with this strategy is that total investment costs may turn out to be higher than the least cost expansion plan under either scenario since an economic core project is not necessarily a least cost project. A fourth strategy would be to follow an investment plan which minimizes total investment even though, again, it may not be least cost. These and other strategies which have been idencified with the RTG have been evaluated in the second and final stage of this analysis and are discussed in more detail in the following chapters, focusing on the decisions to be made over the next few years. In each of the three subsectors (power, gas, and refining) alternative short-term investment plans have been evaluated to identify policies and options that will give the economy the flexibility to adapt to a range of circumstances at least cost. DAP/THAI-B/29-APR-88/ch III. STRATEGY FOR INVESTMENT IN POWER GENERATION Introduction 3.1 The planning of generating facilities in Thailand is based on a set of forecasts that encompasses demand, prices, and fuel availability. These forecasts provide the foundation for a long-term Power Development Plan which is used as the framework for making near-term investment decisions. Given the long lead times characterising most power sector investments, once these decisions are taken the power system becomes at least partially locked onto a medium-term expansion path regarding new power plants to be commissioned. The effects of uncertainty manifest themselves when actual developments deviate from the underlying forecasts of earlier investment decisions. In this analysis, one of two possible states of the world, or scenarios, as described in Chapter II, is assumed to emerge: either demand grows quickly, energy prices remain low, and gas availability is limited--the Status Quo (SQ) Scenario--or demand grows slowly, energy prices rise quickly, and gas supplies are abundant-- the Tightening Oil Market (TOM) Scenario. 3.2 Under the above conditions, the main options open to the power sector are either to develop large, capital-intensive power plants with long lead times and low fuel costs (e.g., lignite, or coal-based steam units) in the expectation of high demand or low gas availability, or to postpone these projects and rely more on low cost gas-based units which involve shorter lead times, if demand grows more slowly or gas is abun- dant. If, for example, a decision is taken in the short-term to develop plans based on high demand expectations and demand in fact grows slowly, the system will run into excess capacity and associated financial problems. On the other hand, if such decisions are delayed, load growth may increase above expectations, preventing the power system from satisfying demand or requiring the implementation of a hasty and costly emergency program. 3.3 The purpose of this analysis is to icentify a strategy that will allow policymakers to make short-term decisions which will minimize economic losses in case either scenario materializes. Ultimately, this approach calls for: (a) adopting a flexible plan where decisions can be adapted to a changing environment; and (b) identifying and taking advance measures to ensure that unexpected developments can be dealt with. 3.4 The remainder of this Chapter is devoted to: (a) a brief over- view of the existing power system and generation expansion plans; (b) an outline of the method of analysis used in this Chapter and its relation- ship to the scenarios; (c) a comparison of least cost investment plans under the two main scenarios; (d) an evaluation of short-term investment strategies that would give the Electricity Generating Authority of - 14 - Thailand (EGAT) the flexibility to cope with an uncertain environment; and (e) a discussion of the problems and options related to the recommended strategy. Existing Power System 3.5 EGAT's peak load in 1986 was around 4200 SW, with a total generation requirement of 24.7 TWh. Although the demand for power grew at a high rate during the 1970s, recently it has slowed down as some markets have been saturated and economic growth has fallen off. During 1985-86 demand grew at 5.9%, compared with 9.6% over 1980-85 and about 13% during 1974-80. According to EGAT's latest estimates, future demand will grow at an average rate of 5.5% for the period 1987-97, with peak load reaching 7800 MW in 1997. EGAT now has an installed capacity of 6850 KW. Table 3.1 presents breakdown of this capacity, according to the primary fuel used. Table 3.1; EGAT INSTALLED CAPACITY, 1986 (MW) Hydro Thermal a/ Gas Combined Cycle *Gas Turbines Total Installed MW 2,240 3,630 720 260 6,860 % of capacity 33% 53% 10% 4% 100% a/ Capacity of plants operating on lignite, fuel oil, or gas, of which 2,400 MW are dual- fired (fuel oil-gas). Source: Plans for Future Investment 3.6 There are a number of alternatives for increasing power system capacity in the future. These include the use of hydro units, conventional thermal plants (burning domestic lignite, natural gas, imported coal or fuel oil), combined cycles plant, and gas turbines (not necessarily gas-fuelled). The potential for new hydro developments is limited to about 1800 MW, distributed among 10 projects. Large conventional triple-fired steam units could be installed at Bang Pakong (2x600 MW) and Ao Phai (6x600 MW) where port facilities for the transport of coal are available. The extent to which gas-based units can be developed is limited only by the availability of natural gas supplies since the economics of power generation indicate that, within thf range of fuel prices covered by the two scenarios, combined cycle plants constitute the most attractive alternative (except for a few low-cost hydro schemes). Two combined cycle plants of 300 MW each can be located at Namphong, where gas has bee. discovered by ESSO, and two of similar capacity at Bang Pakong, to be supplied either by ESSO or by offshore gas from the Erawan-Bangkok pipeline. With greater gas supplies at least six units could easily be installed by the year 2000 (with two additional plants being located at Bang Pakong); more than six additional units can probably be installed but feasibility studies would have first to be carried out, including the identification of suitable site locations. 3.7 Committed additions to the system in the 1988-1990 period consist of conventional thermal units based on lignite (Mae Moh 8 and Mae Moh 9, each with a capacity of 300 MW) and a Second Power Plant Barge in Khanom (75 MW). In the official power development plan submitted by EGAT to the NESDB in July 1986, and made available to the mission, these and other additions to the power system were expected to cost EGAT about $4 billion in investments for the Sixth Plan Period, covering 1987-91. However this Power Development Plan (PDP), whose details and underlying assumptions are given in Table 3.2, as they provide a useful context for discussions later in this Chapter, is now out of date: many of the original commissioning dates are no longer feasible and two of its key projects, the 580 MW Nam Choan hydro scheme and the installation of gas combined cycle plants at Khanom in the south of the country, have been abandoned; the former for environmental reasons and the latter due to the high cost of supplying gas to that location. Table 3.2: EGAT POWER DEVELOPMENT PLAN, JULY 1986: ASSUMPT,ONS AND PROJECTED ADDITIONS TO CAPACITY Assumptions (1) Load Growth 1987-2000 5.91 (2) Gas suDplies (MMCFD) 1987 260 1990 468 1995 656 2000 235 Summary of Addit ions to CaDacity (MW) 1991-1994 1995-1998 1999-2001 Total Hydro 470 416 0 886 Thermal-Lignite 75 300 0 375 Thermal-Coal 0 1,800 1,950 3750 Combined Cycle 1,050 150 0 :100 6111 Source: Analytical Approach 3.8 At any particular time, policyvnakers in the power sector need to take decisions on investements that must be initiated in that year. Long-run plans are a background against which to optimize immediate - 16 - investments rather than providing a rigid blue-print for later invest- ent. The long lead time involved in most power projects limits the flexibility with which investment can be adjusted under conditions of uncertainty. Indeed, if projects could be cancelled at no cost and if construction lags were negligible, the power system could be kept to a near optimal development path. The approach used in this analysis is designed to take account of the above mentioned rigidities. First, the two least cost development plans under each of the two scenarios (whose key assumptions are given in Annex 2) are estimated and compared in order to evaluate the impact of oil price uncertainties on future investment. These optimal plans are subsequently referred to as "benchmark" Power Development Plans (PDPs), reflecting alternative investment plans with perfect foresight. Second, the analysis focuses on the short-term decision agenda (1987-89) in order to identify alternative sets of initial investment decisions. Once taken, these decisions are in effect irreversible, and are included in alternative strategies to cope with uncertainty. Third, the analysis evaluates the relative efficiency with which each set of initial investment decisions adjusts to each possible outcome or scenario. This is done by assuming that, by 1989, planners will realize which scenario, will actually emerge and then adapt the investment program to it. 6/ The efficiency with which a set of initial investment decisions taken in 1987-89 adapts to each scenario is assessed by comparing the net present value (NPV) of the costs of meeting future demand of the initial investment program and subsequent adjustments to longer term investment, with the NPV of the benchmark PDP for that scenario. This difference between the two programs represents the regret associated with a given set of initial decisions, i.e., the cost of being wrong. A final recommendation on the set of initial investment decisions is based on a comparison of the regret estimates. Impact of Uncertainty on Least Cost Investment Plans 3.9 Least cost generation plans under TOM and SQ were estimated using the WASP III model program. These plans, which serve as benchmarks are referred to henceforth as TOM-O and SQ-O, and implicitly assume, as mentioned earlier, that planners know with perfect foresight in 1987 which scenario will emerge, over the next twenty years. 6/ This assumption, which is arbitrary, has important implications given that the later the system reacts to a scenario, the more off- course it is likely to find itself. In the 1989 adaptation rule, it is assumed that power sector decision-makers will recognize the scenario that is emerging, using such information as variations in demand, GDP growth rates, oil price development, and other macro- economic signals. - 17 - 3.10 A comparison of these two plans reveals that the level, timing, and composition of future investments in the power subsector are highly sensitive to the evolution of oil prices. Details of TOM-0 and SQ-O plans are provided in Annexes 4 and 5 and summarized in Table 3.3. According to the Table, the two PDPs differ significantly in a number of respects. In terms of the level of investment, capital expenditure under the SQ Scenario is about twice as high as under the TOM Scenario; total undiscounted investment for 1987-2001 is $5.5 billion under SQ and $2.7 billion under TOM in 1986 prices. These results suggest that the indirect effects of oil prices on investment--through their influence on fuel choices and power demand--dominate generation expansion plans, and that investment levels are inversely related to oil price changes. The two least cost plans also differ substantially in terms of timing and composition. Under TOM, new investments in power are essentially gas based and the development of large steam units such as those at Bang Pakong (coal fired) and Mae Moh (lignite) are delayed to 1999 and beyond. SQ, on the other hand, calls for these units to be advanced-- they are commissioned in 1995. Finally under the SQ scenario, Thailand could have a much greater dependence on imported fuels (fuel oil and coal) than under the TOM Scenario. In SQ, fuel oil acts as the swing fuel in the first half of the 1990s: as the limited gas available is concentrated in combined cycle plants, dual-fired thermal plants are progressively switched from gas to fuel oil. Beyond 1995, both coal and fuel oil would progressively compensate for the decline in gas supplies. Tatle 3.3: LEAST COST INVESTMENT PLANS UNDER TOM AND SQ SCEN.ARIOS TOM-O SQ-0 Cumulative Investments, 1987-2001 2,660 5,480 (1986 $ million undiscounted) a/ Additions to Capacity (Commissioning dates) Total MW 3,835 (1991-2001) 7,385 (1991-2001) of which coal/lignite 1,575 (1999-2001) 5,325 (1995-2001) gas 1,800 (1991-1998) 1,200 (1991-1994) Fuel mix (g of total supply) 1987 1997 1987 1997 Hydro 18 12 18 10 Gas 48 60 48 24 Lignite 21 21 21 24 Total domestic 87 93 87 58 Fuel oil 11 6 11 27 Coal 0 0 0 13 Other 2 1 2 2 Total imports 13 7 3 42 Fuel Consumption (barrels/day) 8,000 12,000 8,000 64,000 Gas consumPtion (MMCFD) 330 580 330 280 a/ Excluding investmerts already commit'ed. Source: - 18 - Evaluating Alternative Investment Strategies 3.1L The large differences between the two least cost investment plans discussed above presents a major problem for planners and decision- makers. Least cost investment plans based on a single set of forecasts about oil prices, demand growth, etc., may simply not be the optimal strategy. A better strategy would be flexible enough to adapt auickly to circumstances different from those for which it was initially designed. Since a premium has to be paid to achieve such flexibility, short-term investment should be chosen to provide this flexibility at the lowest premium. 3.12 To help identify an optirnum investment plan for the sector, four alternative strategies representing the following main feasible options open to EGAT were considered. These strategies which commit investments for power plants to be commissioned over 1991-1996, are as follows: Strategy 1: This strategy aims at investment in low cost plants, and postponing heavy steam units to a later date. Investment decisions over 1987/89 would comprise gas turbines (200 MW), a combined-cycle plant (300 MW), and the relatively low-cost installation of the fifth 180-MW unit at the Srinagarind hydrc scheme. In essence, it is the strategy with the iowest initial investment commitment. Strategy 2: The decisions defined in this strategy for 1987-89 are a modified version of EGAT's July 1986 Power Development Plan with the two 150-MW combined cycle plants originally intended for Khanom now assumed to be located at Bang Pakong. In particular, this strategy attempts to quantify the effect of including the 580-MW Nam Choan hydro project in ECAT's expansion plan. Strategy 3: This would aim at providing flexibility by the early (1993) commissioring of a triple-fired 600 HW steam plant at Bang Pakong. This alternative is also designed to assess the consequences of an early investment based on high load growth expectations. Strategy 4: This alternative involves the early deployment of gas-fueled combined cycle plants (600 MW) so as to cope with uncertainty by making greater use of gas resources. (The scope for a more comprehensive strategy combining increased incentives for gas production with a substantially higher use of gas for power generation are discussed in Chapter IV). Adapting Initial Investments to TOM or SQ 3.13 Corresponding to each strategy, an optimal adapted Power Development Plan (starting in 1990 onwards) was prepared for both the TOM and SQ scenarios. The corresponding investment plans are referred to as - 19 - TOM-1, TOM-2, TOM-3, TOM-4, and SQ-1, SQ-2, etc. The results of adapting each of the four initial strategies (in 1989 or after) to the TOM and SQ scenarios are summarized in Tables 3.5 and 3.6 respectively, and explained in greater detail in Annexes 4 and 5. Table 3.4: SUMMARY OF DECISIONS FOR ALTERNATAvE STRATEGIES a/ S*rateqy ' Strategy 3 Year of lear o' Com- Year of Year of Con- Power Plan, mW Decis-on rr'ssioning Power Plant MW Dec,sion mission ng Gas Turoines 2xl00 1988 1992 Bang Pakong 3 600 1987 1993 CC Uni* 1 300 1988 1993 Gas Turbines 2x100 1988 1992 Srinagarind 5 180 1989 1992 Srinagar;nd 5 180 1989 1992 Strategy 2 Strategy 4 NamChoan hvaro 580 1987 1996 Gas Turbine 2x100 1987 1991 CC Unit 1 150 1987 1991 CC Unit 1 300 1987 1992 R3 Lignite 75 1987 1991 CC Unit 2 300 1987 1992 Sr.nagarind 5 180 1988 1991 CC Uni, 2 150 1988 1992 CC Unit 3 300 1988 1993 CC Unit 41 300 1989 1994 a/ A one-year delay between decision making and commenceme^+ of construction is built into the schedules shown in the TaDle. Source: 3.14 If a TOM scenario emerges, load growth will be relatively low (compared with EGAT's July 1986 PDP forecasts) and the initial decisions over 1987-89 as outlined above would tend to commit the system to power plants that could be delayed. Consequently, there would be some years of excess capacity (as can be seen bv comparing TOM-1, TOM-2. etc., with the benchmark TOM-0) until the system adapts to the lower demand and reaches the optimal benchmark value (around 1994/95 for strategies 1, 3 and 4.) Strategy 2, which is committed to developing the 580 KW Nam Choan project, does not fully adapt to the lower growth scenarios until a later date. Beyond 1995, the investment plans for Strategies 1, 3, and 4 are identical to the benchmark PDP, TOM-0. 20 - TaDle 3.5: ADAPTATiON TO THE TOM SCENARIO Commulative Additions to Caoacity (MW) TOM-0 a/ TOM-I 'OM-2 TOM-3 TOM-4 1991 0 0 405 0 200 1992 0 380 555 380 600 1993 280 680 855 980 800 1994 580 680 1,155 980 800 1995 980 980 1,155 980 980 Total costs (1986 Sm) o/ d,446 4.478 4,324 4,567 4,512 Commission'g Dates Mae Moh 10 (I;grite) 1,999 1,999 2,000 2,000 1,999 Bang PaKong 3 (3-fired) 2,001 2,001 2,001 1,993 2,001 Combined Cycle Plants Added c/ 1991-1994 1 1 3 0 2 1995-1998 5 5 2 4 4 a/ Benchmark PDP. b/ Sum of discounted annual fixed and variable costs incurred over 1990- 2001. c/ Each of 300 MW. Source: TabDe 3.6: ADAPTATION 'O THE SQ SCENARIO Cumulative Add;tions to Capac,*ty (MW) SQ-0 a/ SQ-1 SQ-2 SQ-3 SQ-4 199' 300 300 405 300 300 1992 880 880 855 880 900 1993 1,380 1,380 1,455 1,480 1,380 1994 1,980 1,980 2,055 2,080 1,980 1995 2,660 2,880 2,655 2,680 2,660 Total costs (1986 Sm) o/ 5,788 5,839 5,648 5,860 5,803 Commission;ng Dates Mae MOh 10 (lignite) 1,995 1,995 1,995 1,996 1,995 Bang Pakong 3 (coal) 1,996 1,996 1,997 1,993 1,996 ComDined Cycle Plants Added c/ 1991-1994 4 3 4 2 4 1995-1998 0 1 0 2 0 Gas Turbine Units Added 1991-1994 6 9 6 7 6 a! BencrmarK PDP. 0/ Sum of discounted an-ua ftxec ana var;aole costs ,ncurrea ove, '99C- 2001. c/ Each of 300 MW. Source: - 21 - 3.15 In contrast to TOM, the emergence of SQ requires that the initial decision adapt to a higher demand outlook by advancing the commissioning dates of combined cycle and large steam units and the installation of gas turbines (because more efficient units are ruled out by the lead time involved in their construction). The main problem is to meet short-term demand: strategies 1 and 3 would require the installa- tion of 300 MW in gas turbines by 1991, but this might prove difficult unless procurement were accelerated. Otherwise, the system would have to operate with a reserve margin of around 15% of capacity in 1991 and an unacceptably high loss of load probability (LOLP) in excess of 3 days per year. This problem also occurs, albeit to a lesser extent, in the case of strategy 4 which would require one additional 100 MW unit by 1991. Alternative means for dealing with this problem, including demand management, are considered in the next section. Strategy Evaluation 3.16 The total discounted costs associated with meeting demand over 1990-2001 for the four adapted PDPs as well as for the benchmark PDPs are given in Tables 3.5 and 3.6. The efficiency with which a set of initial decisions adapts to say the TOM scenario can be assessed by comparing the present value of its generation costs with those of the benchmark plan TMO-0. The difference between the two programs represents the regret associated with these initial strategy. The regret values for the four strategies are summarized in Table 3.7. Table 3.7: REGRET VALUES FOR INDIVIDUAL STRATEGIES (1986 S million) Scenario Maximum Regret Strategy TOM SQ 1 32 51 51 2 -'22 -140 -122 3 12l 72 121 4 66 15 66 Source: 3.17 The negative regret values of Strategy 2 indicate that the power development plans corresponding to TOM-2 and SQ-2 have a lower cost than the benchmark Power Development Plans. This is because Strategy 2 involves investment in the low cost Nam Choan hydro project ($609/kW). However, because of the questionable environmental impacts of this power plant, the mission did not include it as a candidate in the other strategies or in the benchmark power development plan. The regret values corresponding to Strategy 2 therefore are not comparable with those of - 22 - the other strategies. Instead, this evaluation suggests that the environmental costs of Nam Choan should be carefully reviewed and compared with the potential savings from the scheme. 3.18 The regret values associated with Strategies 1, 3, and 4 are interpreted as follows: Under TOM, the lowest regret corresponds to the strategy with the lowest initial capital investment, Strategy 1, followed by Strategy 4 and, finally, Strategy 3, which imposes the highest capital costs on the system due to the early commissioning of the large Bang Pakong 3 power plant. Under SQ, Strategy 4 is the most efficient in that it is able to meet the higher demand at lower cost than Strategy 1, since the former requires fewer emergency gas turbines for the years 1991-93. Finally, it should be noted that the regret values do not take into account adjustment costs involved in embarking on an emergency plan or the cost to customers of any load shedding. 3.19 Minimizing greatest regret. A number of criteria can be used to choose between strategies 1, 3, and 4 (assuming that Strategy 2 is at present deemed not viable). A risk averse, or conservative approach would be to choose a strategy which avoids the worst consequences, independent of the likelihood of their occurrence. The last column in Table 3.7 defines the maximum regret values for the four strategies. The lowest regret is associated with Strategy 1 which limits the regret value to a ceiling of $51 million. 7/ 3.20 Minimizing expected regret. An alternative approach would be to assign probabilities to the occurrence of the TOM and SQ scenarios and choose the strategy with the minimum expected regret. In this case, Strategy 3 can be eliminated because its regret value is higher than those of Strategies 1 and 4 in either scenario. Strategy 1 shows the lowest regret value in a TOM environment, while Strategy 4 would be preferred in an SQ world. Both strategies, however, are approximately equivalent (i.e., they yield the same average regret value of about $41 million) if TOM and SQ were equally likely. If, however, decision-makers felt that an SQ type environment was even slightly more likely, then 7/ The relatively small difference in reget values is due to the arbitrary assumption that the system will adapt to the emergence of one scenario in 1989. If the adaptation process takes longer, the corresponding regret values would be larger. However, the ranking of strategies would probably remain the same. - 23 - Strategy 4 would have the advantage. Similarly if they felt that a TOM scenario nas greater than even odds to emerge, then Strategy 1 would be preferred. 8/ The Recommended Strategy: Problems and Options 3.21 The following issues arise in utilizing the preceding analysis to design a practical power generation investment strategy: (a) the criteria for selecting the most favorable strategy when the decision environment is uncertain; (b) how to deal with future adaptation problems, especially in adjusting to an SQ environment (i.e., a faster than expected growth in demand); and (c) integrating the proposed power investment strategy with policy and investment decisions in the gas and refining subsectors. The Criteria for Selection and the Recommended Strategy 3.22 Two main criteria for selecting an investment strategy under uncertainty were examined in the previous section: (a) .Minimizing Greatest Regret, aimed at minimizing the highest possible level of economic losses, and (b) Minimizing Expected Regret, designed to ascer- tain the probability of the occurrence of the scenarios and minimize the weighted average level of economic losses. 3.23 The first criterion, as already mentioned, avoids the worst possible outcomes and reflects a highly risk averse attitude. The second criterion, on the other hand, is "risk neutral": it does not take into acount the riskiness of alternative strategies. It implies, for example, that managers will be indifferent between a strategy that leads to guaranteed profits of $50 million and another where the average return is the same, $50 million, ignoring the 50% probability that the gains could be zero or $100 million. In reality, the degree of risk avoidance practised by decision-makers is likely to lie somewhere in between these 8/ An alternative approach for choosing between strategies involves focusing on investment levels for the near term. A planning agency could arguably take this approach when facing competition for scarce public funds from different subsectors. For the strategies under consideration, the mission examined regret values based on undis- counted cumulative investments over 1987-94. The choice of strategies turned out to be similar to the main analysis. It should be noted, however, that minimizing financial exposure is not a good decision rule. It is better to go to the heart of the problem, i.e., ensure macroeconomic-subsector consistency and raise tariffs to manage demand if t riffs are below economic costs. Energy investments should be !Uwered by reducing energy demand and not by artificiallv constraining investment levels. - 24 - two criteria. It is possible to derive a criterion which gives some weight to the riskiness of a strategy but not as much as in the Minimizing Greatest Regret rule. However, this approach would require a subjective assessment about how to weight the risk. The advantage of the two criteria used in Lhe analysis is that they do not require such a subjective assessment and define the limits to the problem facing decision-makers. 3.24 Given the above considerations, the analysis of the previous section would suggest the following choices: if decision-makers are inclined to avoid risk, or if the TOM Scenario appears more likely to emerge than SQ, the RTG should adopt the strategy with the lowest initial investment, i.e., Strategy 1. This strategy consists of investments in peak load capacity (2x100 MW. gas turbines), the relatively inexpensive Srinagarind hydro unit 5, followed by a 300-SW combined cycle plant to be commissioned in 1993. If on the other hand, government policy is risk neutral and the SQ Scenario appears more likely, Strategy 4 would be the preferred course of action. This strategy involves early development of (2x100 MW) gas turbines and (two 300 MW) combined cycle plants. Strategy 4 is the recommended strategy under more restrictive conditions than Strategy 1. 3.25 While there is no 'tbest" strategy under all conditions, a case can be made for combining the common features -of Strategy 1 and Strategy 4. This would involve the development of combined cycle units as a first priority during the early 1990s, while delaying larger steam units until the mid-to-late 1990s as being the preferred medium-term strategy for the power subsector. Consistent with this broad strategy, investment decisions need to be made during 1988-89 for developing (2x100 MW) gas turbines to be commissioned in 1991-92 and for combined cycle plants (2x300 MW) to be commissioned in 1992-94. Adaptation Problems 3.26 Adaptation problems would depend on which scenario emerged after initial decisions had been made. Under TOM, adjustment would be required to reduce excess capacity. Adapting to this scenario involves finding ways to deLay committed investments. Although the analysis assumed the decisions made in 1987-89 and the corresponding commissioning dates were irreversible, in practice there is usually some flexibility for rescheduling a project's in-service date. Projects could be delayed by agreement with equipment suppliers or to more relaxed monitoring of procurement dates. This would stretch investment schedules and allow the system to achieve some modest capital savings in present value terms. 3.27 Adapting to the SQ Scenario poses more difficult problems, involving EGAT's capacity to meet unexpectedly high demand in the short run. These problems arise because of the lag between commissioning and decision dates and are especially acute for Strategy 1 sinc: it invests in lower base load capacity than Strategy 4. in the exampie used here for Strategy 1, where planners recognize the emergence of the SQ scenario - 25 - by 1989 and reassess generation capacity needs accordingly, 300 MW of gas turbines have to be installed in an emergency program by 1991 and another 400 MW by 1992. Even given the short lead time for gas turbines, it appears unlikely that the 1991 target could be met. As a result, reserve margins would drop below the minimum acceptable level in the early 1990s, reducing power system reliability. These problems would be exacerbated if the commissioning of gas turbines took longer or if planners identi- fied the emergence of SQ conditions later than 1989. To cope with the problem of meeting higher than expected demand, the following actions are suggested: (a) use load management techniques on the demand side to reduce peak loads through appropriate incentives and direct load controls for reducing consumption in major industries, since load management may be cheaper, under these circumstances, than installing new capacity; (b) purchase power from Malaysia and Laos through interconnections which at the moment are at the planning stage; (c) advance the completion of designs, site identification, feasi- bility studies, and tender documentation for gas and lignite based power plants; and (d) speed up the decision-making process for power projects pos- sibly by developing a "reserve" of pre-approved projects that would require only limited additional Government clearance before implementation. Linkages with the Gas Subsector 3.28 Gas-based investments in the power subsector are dependent on the availability of future gas supplies and associated infrastructure. Supplies from existing concessions are expected to decline sharply in the early 1990s and the fields under production cannot therefore justify the installation of new combined cycle plants as envisaged in both Strategies 1 and 4. Thus to implement either Strategies 1 or 4, it is essential that at least one additional gas production contract be in place before the decision is taken to install combined cycle power plants. 9/ 3.29 SQ demand and TOM gas supplies. A possibility considered by the mission, but not discussed above, was that actual developments may 9/ It should be noted that infrastructure investments necessary to develop additional gas supplies (i.e., pipelines and compressors) do not affect the results of the strategy evaluation carried out in this Chapter. The reason is that these investments, which are discussed in che next chapter, occur after 1990 and are common to all strategies as they adapt to a given scenario. - 26 - turn out to be a combination of TOM and SQ scenarios. More specifically, the geology may turn out to be better than expected and significantly higher gas supplies (close to TOM levels) may emerge in an SQ environ- ment. The performance of alternative strategies in such an environment was evaluated. The main conclusions of the evaluation are: (a) adding this third scenario (a combination of TOM and SQ) does not alter the earlier ranking of strategies and in fact further reinforces the recommendation to develop combined cycle units as a first priority during the early 1990s and delaying larger steam units until the mid-to-late 1990s; (b) the availability of TOM level gas supplies under SQ implies that at least two additional combined cycle units could be put in service in 1995, thus delaying the Mae Moh 10 lignite units as well as the Bang Pakong coal plant by 1-2 years (under SQ these olants are expected to be commissioned in 1995 and 1996 respectively; see Table 3.6). Conclusions and Recommendations Summary of Results 3.30 The level, timing, and composition of future investments in the power 3ubsector are highly sensitive to the evolution of future oil prices. The wide differences between the least cost investment pians under the two scenarios examined presents a major problem for planners and decision-makers: least cost investment plans drawn up under a single set of forecasts (about energy prices, demand, etc.) may not be the optimal course of action for the country. An optimal strategy under conditions of uncertainty would need to be sufficiently flexible to adapt quickly and cost-effectively to circumstances different from those for which it was initially designed. 3.31 An examination of alternative strategies reveals that future development of the power system in Thailand would best be based on first developing combined cycle units to be commissioned during the early 1990s while delaying larger (lignite-, gas-, and coal-based) steam units until the mid-1990s. This strategy can easily cope with the emergence of a low demand, high oil price scenario by merely delaying decisions related to future power plants and rescheduling power plants under construction for one year. Coping with a scenario where gas supplies would become relatively scarce and demand would grow strongly due to low oil and gas prices would possibly call for some load shedding in the early 1990s. This strategy would of course call for the conclusion of at least one additional gas contract by 1988. The recommended strategy (of an early development of combined cycle gas units) would also fit well in an environment where gas supplies were abundant and demand grew strongly under the influence of high GDP growth and low oil prices. Recommendations 3.32 Specific recommendations are as follows: - 27 - (a) Consistent with the recommended strategy, short-term investment decisions are needed during 1988-89 for developing gas turbines and combined cycle plants as shown in Table 3.8. Table 3.8: INVESTMEN' RECOMMENDATION Cost Year in wnich Commiss5oning Investment 1986 i decision needs Date to De made Gas turoine (2xl0O MW) S 32m 1988 1991-92 ComDined Cycle Unit 1 (300 MW) 1185m a/ 1988-89 1992-94 Combined Cycle Unit 2 (300 MW) $185m a! 1988-89 1992-94 a/ Assumed location, Bang Pakong. Source: (b) At least one additional gas production contract should be in place before the decision is taken to invest in new combined cycle power plants. (c) In order to prepare for the possibility of a higher demand materializing in the next decade, EGAT should proceed to make advance preparations for gas and lignite fueled power plants including completion of preliminary designs and feasibility studies. The latter would include identification of sites for locating plants and assess the need for complementary invest- ments in infrastructure. (d) Another measure to prepare for higher demand would be to explore demand management possibilities. For example, agree- ments could be made with large industrial customers for non- disruptive peak load reductions, possibly by providing appro- priate incentives. (e) Interconnection agreements with Malaysia and Laos should be sought for the purpose of obtaining support during peak load periods. (f) EGAT and the RTG should seek to minimize the approval process for new projects. This could be done by creating a catalog of pre-approved power plants for which all issues would be resolved in advance, n order to minimize the subsequent decision-maki.g time. DAP/THAI-B/29-APR-88/ch IV. STRATEGY FOR INVESTMENTS IN THE GAS SUBSECTOR Introduction 4.1 The major issue in the gas subsector is that present contracted gas supplies may not be sufficient to atisfy market demand much beyond 1990. Given the lead time necessary for field development there is a need to complete negotiations for additional gas from at least one new area in the near future. Depending on the sequence in which the areas are developed, associated decisions also need to be made regarding investments in pipelines, compressors, and additional LPG extraction plants. 4.2 Thailand has substantial gas resources. Current estimates of proven and probable recoverable reserves range between 4-13 TCF, reflect- ing the uncertainty regarding both geology and future price. Some 2 TCF of these reserves, which were discovered offshore, are already under contract with Unocal and are being produced; average production from this contract is expected to rise from about 400 MMCFD in 1987 to a peak of 500 MMFCD by 1990, before declining progressively after 1992. Another three areas with known reserves have been discovered but production contracts have not yet been concluded. Two of these are offshore dis- coveries made by Unocal and Texas Pacific (TP), the third on-shore by ESSO. The Petroleum Authority of Thailand (PTT), which owns and operates the pipeline linking the Unocal fields to Bangkok (together with an LPG extraction plant), and which is the purchaser of gas from Unocal and is negotiating for the purchase. of gas from the three other areas not yet under contract. 4.3 The main problem facing decision-makers is how to schedule the development of new gas areas, taking into account uncertainties in oil prices and in economically recoverable reserves. These issues are examined within the context of the Status Quo (SQ) and Tight Oil Market (TOM) scenarios. 4.4 Under the SQ Scenario, there is a potential shortage of gas as oil and gas prices stagnate, demand accelerates, and gas availability is limited by reduced incentives for production. In contrast, under the TOM Scenario there is a potential surplus of gas as oil and gas prices rise, demand is restricted, and gas is available in abundance due to increased incentives for exploration and production. Even in the absence of a competitive gas market where price signals effectively balance supply and demand, such potential surpluses ut shortages can be managed in an orderly, non-disruptive manner through appropriate sequencing of fields and allocation of demand, provided the future environment is correctly predicted. However, uncertainty about the future environment raises the possibility of large and costly imbalances between supply and demand due to: (a) the long lead times associated with gas field developments and - 29 - infrastructure investments; and (b) rigidities in gas contracts with producers (such as "take-or-pay clauses"). If an accelerated program of field development is initiated based on low gas reserve expectations (i.e., SQ) and a high gas supply scenario emerges, excess gas supply capacity will develop over time; contrarily, if such decisions are delayed, supply shortages may develop if availability is less than expected, which may require implementation of a costly rationing program for gas users (or restricted access to new consumers, or substitution of oil for gas). The analysis in this Chapter attempts to identify a strategy that would enable the gas subsector to make short-term decisions which would help avoid large economic losses in the case of unexpected developments. 4.5 The next section provides an overview of the main contractual and geological features of the three new areas available for further gas development. The method of analysis used in this Chapter is then outlined and is followed by an assessment of the implications of the two scenarios for gas demand, reserves, production profiles, and associated investment requirements. In the next section, the consequences of alternate short-term policy and investment strategies are evaluated. The last section makes recommendations regarding key policy and investment decisions that need to be made over the next few years. Options for Further Gas Development 4.6 Existing supplies of gas in Thailand are almost all produced by Unocal from a combination of four fields--Erawan, Baanput, Platong, and Satun--all located in the Culf of Thailand. 10/ With the output from these fields expected to peak in the early 1990s, there are three main options open to the RTG for further gas development: (a) producing from additional fields within the Unocal license, a development referred to as U3; (b) developing the large offshore gas fields discovered by TP further south in the Gulf; and (c) concluding a contract with ESSO for the development of the on-shore field at Namphong in the Khorat Basin. The main features of these three options, including estimates of reserve size and status of negotiations with the RTG are discussed briefly below. A comparison of geological feature is given in Annex 8. 4.7 Ci) The U3 development. From the RTG's point of view, the attractiveness of this option is that, by making use of the gas trans- mission system already in place, it involves very little investment in new infrastructure by the PTT. It will also produce gas rich in LPG and liquids that could be extracted with a limited expansion of the existing LPG plant. On the other hand, the reserves are currently less well 10/ Shell is producing a limited amount of associated gas from its on- shore oil operations at Sirikit. - 30 - defined than in the other areas and this option would not diversify .he source of supply, which is one of RTG's objectives. Estimates of feasible recoverable reserves, made by Unocal, range from 0.8-1.8 TCF; a 3-D seismic survey is planned by Unocal for a better definition of reserves, to be followed by appraisal of wells. Gas can be produced from U3 in two modes: either independently of the fields covered by earlier production contracts, or as a swing field to sustain total supply from all Unocal contracts at a level of 500-550 >!MCFD well into the 1990s. In the latter mode, the free flow limit (550 M"CFD) of the existing pipeline from Erawan to Rayong (on the mainland) would not be exceeded, thus obviating the need for immediate investments in compressors. In all, three years are expected to elapse between the conclusion of a contract, for which negotiations are in progress, and the start of production; the earliest production date, assuming contracts are concluded by end 1987, is therefore 1991. 4.8 (ii) Development of the TP field. This option has the advantage of being based on well-defined levels of recoverable reserves; reserve estimates range from 1.2 to 2.0 TCF for areas where a 3-D seismic survey has already been completed and there is a possibility of an additional 3.0 TCF in areas covered by a more limited (2-D) survey. In comparison with U3, however, TP gas is drier, thus offering a reduced potential for extracting LPG and liquids and development of the field would require substantial investment by PTT (about $110 million) for the construction of a submarine pipeline to link the TP field to the existing offshore pipeline at Erawan. In common with U3, additional investments may be required in compressors to raise the capacity of the existing Erawan-Rayong pipeline. 4.9 The development of the TP option has been considerably delayed because in spite of protracted negotiations, RTG and TP had been unable to reach agreement on mutually acceptable conditions for the development of the field. The key obstacle has been the price of gas in the production contract. The Government recently ended an eight-year stalemate over the development of the field by agreeing to an outright purchase of TP's production rights for $83.8 million. The field can now be developed either by PTT, or a new contractor if agreement can be reached on the terms of the contract, including the price of gas and the risks to be shared. Finally, pipeline extension and installation of compressors (if necessary) would take at least four years. Hence, even if the negotiations between the RTG and a new operator can be completed by end 1988, the earliest that TP can come on-stream is 1993. 4.10 (iii) Development of ESSO field at Namphong. Negotiation of an intitial contract for long-term testing of the field and provisions for future production have recently been completed. Assuming that a production contract can b' concluded by 1990, based on tests during 1988 and 1989, full production could begin by 1992-93, allowing about three years for field development. As in the case of the other two fields reserve estimates are subject to considerable uncertainty, and currently range between 0.5 and 1.7 TCF. Depending _n the eventual size of - 31 - reserves, investment in a Namphong-Bangkok pipeline may be necessary (see para 4.30). 4.11 Once the the Government has decided on the sequence of gas field development, a number of associated decisions will be required including the choice of pipeline size for the rwo possible pipeline connections (TP-Erawan, Namphong-Bangkok), the need for an offshore compressor, and the desirability and size of a second LPG plant. Methodology for Analysis 4.12 Investment choices in the gas subsector have to consider uncertainty in the future demand, supply, and price of gas. The analysis presented here aims at developing a strategy which allows the gas subsector to adapt to a variety of conditions while minimizing the costs of adjustment to unexpected developments. First, an evaluation is made of the implications for gas development of the scenario assumptions, which are the same as used for analyzing power investment (Table 4.1). Since the demand for gas is highly dependent upon the demand for electricity, the two scenarios, which assume different levels of gas availability, are associated with two different derivations of gas demand. A comparison of gas availability with demand allows supply- demand balances to be defined, which in turn leads to a set of least cost investment requirements for each scenario. Table 4.1: KRY SCENARIO ASSUMPTIONS Key Scenario Assumptions TOM Scenario SQ Scenario (1) GDP Growth 3.5% 6% (2) Oil Prices 1986 15 15 19865/bbl 1990 20 15 1995 25 17 2000 30 20 Gas Availability, MMCFD 1986 470 470 1990 570 470 1995 1100 570 2000 1200 500 Source: 4.13 The two least cost gas investment plans (under the TOM and SQ scenarios, respectively), which include sequences of field development, reflect the development of the gas subsector in a world of perfect foresight. In the next stage, the analysis focuses on the short-term - 32 - decision agenda (1987-89) to identify alternative sets of initial investment decisions. The analysis then evaluates the efficiency with which each set of decisions adapts to each possible outcome or scenario, assuming planners will know by 1989 which state of the world will emerge and react accordingly. An attempt is made to identify a strategy which minimizes regret, defined as either revenue foregone or the opportunity costs of over-investment (see para. 3.10). Impact of Uncertaintv on the Development of the Gas Subsector Sequence and Timing of Field Development 4.14 As a first step towards determining the timing and sequence of developing new fields, projections of gas reserves and the economic costs of supplying gas from the various fields were prepared for each scenario. The supply potential from the new areas at various delivered costs was compared with the demand for gas and its value to consumers, for establishing supply-demand balances over time for each scenario. More details of the analysis are provided in Annex 13. 4.15 Supply. Most hydrocarbon fields in Thailand are typically made of small, lenticular and often highly faulted sandstone reservoirs, which make for high production costs of gas. The ESSO Namphong field in the onshore Khorat Basin produces from highly fractured limestone reservoir and production costs there will probably also be on the high side. Consequently, economically recoverable reserves are highly price sensitive. Moreover, future exploration and development in new areas will also likely be highly dependent on future oil prices. Projections of future reserves have been prepared by the mission in consultation with PTT, NESDB, and the oil companies. These projections take account of both geological and price uncertainties, and are summarized in Table 4.2. Table 4.2: RESERVES ASSUMPTION (TCF) TOM SQ Field (high reserves) (low reserves) Ul and U2 2.3 1.8 U3 1.8 0.8 TP3D a/ 2.0 1.2 TP2D b/ 3.0 - Other Offshore 2.0 Total Offshore 11.1 3.8 ESSO 1.7 0.5 Total 12.8 4.3 a/ The TP3D a-ea is that part of -re TP f ieid aw ch has already onde-go-e 3-D seismic sjrvey. b/ The TP29 area has onty had a 2-D seismic survey. Source: miss-on es-mates of fie'! s,ze, based upon aiscuSS O-s with PT7, Unocal, and ESSO. - 33 - 4.16 Overall, there is an 8.5 TCF difference in the Level of reserve estimates between the two scenarios. Of this difference, 2.0 TCF is due to new offshore discoveries deemed to be encouraged by higher oil prices, 3.1 TCF reflects the effect of oil prices on existing discoveries, and the remaining 3.4 TCF is due to reservoir uncertainty, which is indepen- dent of oil price changes. The comparison of reserves (see Annex 13 for details) reveals that, apart from the effect of prices on new dis- coveries, the assumed variation in reserves attributed to geological uncertainties is of the same order as the difference in gas availability arising out of price effects. The upper and lower limits to future reserves are referred to as high and low reserves. Price and geological uncertainty effects have been combined to provide four different scenarios: two main scenarios (SQ with low reserves, TOM with high reserves), with two intermediate scenarios (SQ with high reserves, TOM with low reserves). The reserve levels associated with each of the four cases is shown in Table 4.3. The figures imply that even with economically low prices, recoverable reserves could be substantially above SQ levels of 4.3 TCF if the country was lucky in geological terms. Table 4.3: TOTAL GAS RESERVES UNDER D;FERENT ASSUMPTIONS TOM Main Scenario with high oil prices and good geological prospects 12.8 TCF (Suo-Scenario with high oil prices and poor geological prosDects) 8.5 TCF SQ (Sub-Scenario with low oil prices and good geological prospects) 7.3 TCF Main Scenario with low Oil prices and poor geological prospects 4.3 TCF Source: M;ssion estimates. 4.17 Schedule of Field Development. The sequencing of field development needs to take account of both the relative economic cost of supply of each option and the lead time needed in developing it. The mission prepared tentative estimates of economic costs of gas delivered at Bangkok for the purpose of comparison (Table 4.4). Since gas is developed by foreign companies and the cost to the country is essentially remitted profits (wellhead prices less taxes and royalties), the economic cost of gas at the wellhead is better represented by wellhead prices based on existing contracts with producers, rather than the latter's production costs. The figures in Table 4.4 indicate that under both scenarios, U3 and ESSO provide a cheaper supply alternative compared to TP. In the case of U3, this is because no pipeline investment would be required except for the installation of a compressor under TO.M. For - 34 - ESSO, the delivered costs are lower than TP under SQ because the gas would be used locally for power generation and no pipeline would be required. Even with substantial pipeline investments, as in TOM, ESSO gas would be cheaper since compression costs are not incurred, as is the case with TP. This indicates that it would be to the best interest of Thailand to give priority to U3 and ESSO in meeting early demand growth, scheduling the development of TP to meet long-term growth in an optimal manner. The production profiles shown later in this Chapter reflect such optimization under the TOM and SQ scenarios. Table 4.4: ECONOMIC COSTS OF NATURAL GAS (NPV IN 1987) (1986 S/mBTU) TOM SQ U3 TP ESSO U3 TP ESSO indicative production costs (well head) a! 1.6 2.0 1.2 1.2 1.8 1.9 WelIhead prices based on existing contracts b/ 2.4 2.4 2.4 2.2 2.2 2.2 Pipeline costs c/ - 0.3 0.3 - 0.3 - Compressor costs 0.1 0.1 - - - - Total delivered costs at BangKok d/ 2.5 2.8 2.7 2.2 2.5 2.2 a/ Estimated as NPV of cos' of production divided oY NPi of gas output. Cost of production includes expioral;on, appraisal, and prodUction expenditures but excludes royalties and taxes. b/ Excluding royalties, and taxes. No account has beer. taken of condensate content. c/ Pipeline unit costs based on estimated through-put (see Tables 4.7 and 4.8). d/ Sum of wellhead prices, pipe ine costs, and compressor costs. Operating costs are assumed to be negligible. Source: 4.18 Demand for Natural Gas 11/. Demand projections, together with related price assumptions, are shown in Table 4.5. The main sources of demand for gas at present arise from power (77%); LPG extraction (20%); and process heat for cement production (3%). As detailed in Annex 13, the domination of gas use by the power subsector is expected to continue in both scer.,rios. 11/ Demand for natural gas is defined as the total of all potential uses which would provide a netback value of gas higher than itz opportunity cost. - 35 - Table 4.5: DEMAND FOR GAS (MMCFD) 1986 1990 1995 2000 (Actual) SO Scenario 30e 583 1003 1423 (Price a/ in 1986 S/mb,U) (2.1) (2.1) (2.2) (2.41 TOM Scenario 308 573 843 953 (Price a/ in 1986 $,mBTU) (2.1) (2.3) (2.6) (2.8) a/ Wellhc3d prices net of royalties and taxes. Source: 4.19 A comparison of netback values and delivered costs affords an assessment of the economic viability of gas supplies from the three areas. Power generation is the marginal gas user; netback values are estimated to be higher in non-power uses (see Annex 9). Netback values of gas in power generation (Table 4.6) exceed delivered costs of gas (Table 4.4) in all cLses, thus indicating the viability of gas supplies from each area. Table 4.6: GAS NETBACK VALUES (NPV IN 19871 a/ (1986 S mBTU) Scenarios TOM SQ Use Power Generation a/ 2.9 2.7 a/ Net present value derived oy discounting marginal netbacK values covering the perioa 1987-2000. The marglnai net- back value in power reDresents the replacement cost of substitute tuels together with a fuel credit due to increased effic,ency of gas cormoned cycle p'an*s and a capacity credit for the lower cost of these units (see Annex 7 for details). Source: Supply-Demand Balances and Production Profiles 4.20 The demand for natural gas can he met from existing and new contract areas in different combinations. The gas production irofiles shown in Tables 4.7 and 4.8 are based on the reserve estimates and least cost saquence of developing new fields discussed in the previous - 36 - section. They also assume that: (a) investments associated with field development are utilized at planned capacity under each scenario; (b) since the dominant use of gas is in power generation where investments have a long economic life (15-25 years), the depletion policy is such as to provide a steady supply of gas over a prolonged period from each contract area, rather than reflect the peaked production profiles which would maximize producers' revenue 12/; and (c) as already discussed, the earliest U3 can come into production is 1991, while in the case of ESSO and TP, production in either scenario does not come on- stream before 1993. 4.21 While the sequence of entry of new fields is the same in both scenarios to ensure least cost development (para. 4.17), the timing of decisions differs significantly between TOM and SQ. For the TOM scenario, there is a potential surplus of gas and imbalances between supply and demand are avoided by delaying the development of the higher cost field (i.e., TP), thereby obviating the need for either rationing demand or prorating production. Production of gas under TOM is therefore demand constrained. It should be noted that the surpluses shown in Table 4.7 are potential and not actual. The potential surplus in TOM emerges because supply exceeds demand at prices which are based on existing contracts. If prices were flexible they would adjust downward to equate supply and demand in this scenario. As prices are in fact rigid the balancing is achieved through quantity, as opposed to price, adjustment: by delays in the development of TP. PTT would need to reach agreement on a contract for the third Unocal development (in addition to the recent contract for the test production of the ESSO field) within the next twelve months. This would allow U3 to come on-stream by 1991, followed by ESSO in 1993. The development of the TP field would be delayed and actually would not be justified to commence full production before the turn of the centurv. 12! Proven reserves in each new find are assumed, in the absence of pipeline or other ccnstraints, to be depleted approximately at a constant rate c.er a 15-vear period. - 37 - Table 4.7: SUPPLY-DEMAND BALANCE - TOM SCENARIO a/ DEMAND C0NSTRAINED PRODUCTION (MMCFD) Potential Offshore Product,on Onshore Total Potential Surplus/ U1+U2 U3 3DTP Otner Total ESSO Supp'v Demand supply (shortfalls) 1987 400 400 400 400 450 50 1988 490 490 490 490 500 10 1989 550 550 550 553 550 (3) 1990 570 570 570 573 570 (3) 1991 470 163 633 633 633 650 17 1992 450 223 673 673 673 700 27 1993 430 218 648 125 773 773 825 52 1994 400 213 613 200 813 813 950 137 1995 360 233 593 250 843 843 1,100 257 1996 310 203 573 300 873 873 1,150 277 1997 240 383 623 300 923 923 1,200 277 1998 170 463 633 300 933 933 1,200 267 1999 120 470 83 673 300 973 973 1,200 227 2000 80 470 103 653 300 953 953 1,200 247 200' 60 470 103 633 300 933 933 1,200 267 RESERVES (TCF) Total 2.3 1.8 2.0 5.0 '1.1 1.7 12.8 Remaining in 2010 0.0 0.1 1.0 4.0 5.1 0.2 5.3 a/ The total output from the Unocal fields (Ul , U2, and U3) Should be taken as an upper iimit, and Could be lower even if tne TOM scenar,o emerged. Source: 4.22 Under the SQ scenario, early agreement is also needed for the development of U3 as well as for the test production of the ESSO field. As in TOM, they would come on-stream in 1991 and 1993 respectively. However, because of higher demand and lower reserve levels, gas from TP would be needed by 1995. While TP could be made ready earlier, producing before 1995 would exceed the free-flow capacity limit of the existing pipeline and, given the more limited size of offshore reserves in SQ, investment in a compressor to increase pipeline capacity would not be justified. For the same reason, output from U3 is restricted in the 1991-93 period. As shown in Table 4.8, gas demand would be sharply constrained by the availability of supplies. The shortfall in supplies would be largelv absorbed by the power subsector through appropriate curtailment of new investment in combined cvcle plants thus limiting gas consumption in power uses in the 1990s to about 400 !1CFD compared to a potential demand of 1,160 vMCFD bv 2000 (see Annex 13). - 38 - Table 4.8: SUPPLY-DEMAND BALANCE - SQ SCENARIO SUPPLY CONSTRAINED PRODUCTION (MMCFD) Potential Offshore Product;on Onshore Total Potential Surplus/ UI+U2 u3 3DTP Total ESSO Supply supply (shortfalls) 1987 400 400 400 400 - 1988 450 450 450 465 (15) 1989 450 450 450 543 (93) 1990 500 500 500 583 (83) 1991 450 50 500 500 693 (193) 1992 385 115 500 500 743 (243) 1993 325 175 500 70 570 873 (303) 1994 275 225 500 70 570 933 (363) 1995 195 250 55 500 70 570 1,003 (433) 1996 135 250 115 500 70 570 1,073 (503) 1997 95 250 155 500 70 570 1,153 (583) 1998 45 215 240 500 70 570 1,233 (663) 1999 30 185 250 465 70 535 1,323 (788) 2000 20 160 250 430 70 500 1,423 (923) 2001 I5 110 250 375 70 445 1,533 (1,088) Reserves (TCF) Total 1.8 0.8 1.2 3.8 0.5 4.3 Remaining in 2010 0.0 0.0 0.1 0.1 0.5 0.6 Source: Implied Least Cost Investment Plans for TOM and SQ 4.23 PTT investments for each scenario are shown in Table 4.9 and reveal that both the level and pattern of investment differ significantly between TOM and SQ. Cumulative undiscounted investments over 1987-95 under a regime of rising oil prices (TOM) were, at $260 million, about twice the level under a scenario of stagnant prices (SQ). The composi- tion of investment also differs between the two scenarios. Under TOM where gas production is higher, investments are needed in the Namphong- Bangkok pipeline by 1993 13/; and in the offshore compressor by 1991, neither of which are needed under the SQ scenario because of the lower level of economically recoverable reserves. The capacity of the LPG plant in TOM (200 XMCFD) is also greater than in SQ (150 YeMCFD); the 13/ There have been a number of studies as to whether a pipeline or electrical transmission line would be liast cost from Namphong, these have shown the relative cost to be quite close. - 39 - rationale for its size is discussed below. Lastly, because of the shortage of gas under SQ, the TP-Erawan pipeline is planned for commissioning by 1995, while under TOM this investment can be delayed until a later date. -aoie 4.9: SQ AND TOM SCENARIOS !N,ESTMENT REQU'REMENgS JAS SUBSECTOR (1987-95) SQ SCENARIO TOM SCENARIO Nompnong- Second Year Texas-Erawan Second Texas-Erawan Offshore Bangkok LPG Pipeline LPG Plant Total Pipeline a/ Compressor Pipeline Plant To+al 1988 2 2 1 2 2 2 1989 10 10 12 17 17 17 1990 3 11 14 54 3 19 21 1991 20 2 22 10 26 4 31 1992 75 75 90 101 1993 13 13 24 78 1994 0 10 1995 0 0 Ill 25 136 0 76 143 42 260 a/ Investments after 1995. Source: 4.24 LPG Plant. A second LPC plant can be installed with a throughput capacity of either 150 "MCFD or 200 'tMCFD. An analysis of the returns on these two investments for both a low price of products (LPG = $180/ton, NGL = $167/ton) and for a relatively high price (LPG = $210/ton, NGL $195/ton) confirms that construction of a second LPG plant is attractive in most circumstances but that, for the SQ scenario, the 150 MMCFD plant would offer the higher returns, while for the TOM scenario the 200 "MCFD plant would be preferable. These r-esults are summarized in Table 4.10. Taoie 4.'0: RETURNS ON SECOND LPG PLANT a/ (NPV in US$ million) SQ TOM Assumptions 150 MMCFD 200 MMCFD 150 MmCFD 200 MMCFD (a) Low product price 13.8 8.85 26.0 58.5 ('8) ('6) (21, (29) (b) H,gt product price 23.3 '8.4 34*4 74.2 (22) (20) (25) (334 a2 Figures in bDac4e-s are inee-.ai rates of reijr'. Souce: - 40 - Strategy Issues and Options 4.25 The analysis of the previous section reveals that there are large differences between the least cost development plans for each of the two scenarios. These differences apply in particular to: (a) the timing of development of the fields and the contractual aspects of such development; (b) the timing and size of a second LPG plant; (c) the need for an offshore compressor, (d) the timing and size of investment in a TP-Erawan pipeline; and (e) the need for a Namphong-Bangkok pipeline. Another major problem to be considered by decision-makers is the need for a contingency program to cope with the economic losses associated with the high level of demand rationing in the SQ scenario, should it emerge. This section focuses on short-term policy and investment strategies to deal with these problems. 4.26 Contractual terms. The degree to which a gas supply-demand balance can be achieved depends on contractual arrangements with the international oil companies (IOCs). A major issue is whether producer prices, net of taxes and royalties, for future contracts should broadly parallel previous contracts, or improve on the latter to avoid the severe shortages that would emerge under an SQ scenario. While the TOM scenario also assumes generally more optimistic geological conditions, gas supplies under an SQ environment could be increased by providing incentives through higher gas prices. Gas pricing is a complex issue which this report does not pretend to have looked into in any depth, in part because it has already been comprehensively addressed in the 1985 Energy Assessment Report and Government policy has not changed in the interim except as recommended in that Report. However, some tentative observations are offered below on the pricing issue to provide a context for the discussion on gas strategy. 4.27 Firstly, before any major review of gas prices, the Government needs to make a judgement about which environment it is in or moving towards. If the environment is going to be similar to TOM any substantial premium would prove unnecessary. Also, if an SQ scenario prevails, but with a significantly higher level of reserves than &ssumed in the main scenario, then shortages will be less of a problem and only a modest premium may be suffficient and justifiable. If, however, future developments are expected to be close to the main SQ scenario, with its attendant high-level of shortages, then the case for considering a significant increase in prices is strengthened. The mission's discussions in Thailand reveal that at least one new production contract can be concluded in the near future even at prices no higher than those based on existing gas contracts; higher prices will expedite, but were not viewed as a necessary condition for, the conclusion of these contracts. As indicated in Chapter III (para. 3.28) one additional gas production contract is needed in 1988 to avoid delaying the power investment program. In these circumstances the RTC can delay the initiation of any strategy for meeting demand uncertainty through significantly increased prices for gas supplies, at least for a few - 41 - years, until the likely environment becomes clearer. In general, the analysis carried out above suggests that raising prices need not be part of the decision agenda for 1988-89. 4.28 Secondly, the mission examined the possibility of obtaining greater gas supplies under the SQ scenario by analysing a gas policy where prices were raised to provide increased incentives for exploration and development of natural gas reserves. The assumptions used in the analysis were: (a) producers were paid TOM-level wellhead prices; (b) TOM -level gas supplies were obtained; and (c) the same price was paid for all gas supplied, not just for incremental supplies. The results of the simulation are shown in Annex 14. The additional price paid to producers on future gas supplies is offset by the savings generated through the use of two additional 300 MW combined cycle units which have a lower capital cost and are more efficient than the coal-based steam units they displace from the original least cost expansion plan under SQ. 4.29 The conclusions from the simulation are that the incremental cost of gas supplies is higher than the benefits accruing from the use of two additional c.ombined cycle plants. Approximate calculations show that eight, rather than two, additional combined cycle plants of 300 MW capacity would nleed to be installed to justify paying TOM prices under SQ. This combined gas/power subsector strategy would absorb an additional 280 MMCFD of gas. To enable this strategy to be viable with fewer additional power plants, PTT would need to design and implement two-tier pricing contracLs which wouid permit the premium to be limited to incremental supplies. A further probLem in implementing such a strategy is the uncertainty surrounding the supply response to higher prices. The mission estimated (see Table 4.3) that a substantial supply response is possible based on the experience of gas production profiles from existing fields. However these estimates are at best tentative. A more reliable estimate of supply elasticity would be needed if the RTC were to give serious consideration to such a strategy. 4.30 TP-Erawan and Namphong-Bangkok pipelines. Under either scenario no decision or commitment needs to be made about developing the TP field before 1990 at the earliest and therefore consideration of the need for a pipeline can be deferred. However, if the ESSO field does not turn out to be commercial (considered highly unlikely), an early program for the design and construction of the TP-Erawan pipelinn would need to be implemented in association with TP field development. Nevertheless, given that production testing at Namphong is to be started in the near future, the magnitude of economically recoverable reserves should be much clearer bv the end of 1988. This would allow time for the construction of the TP associated infrastructure to be completed by 1993 if required. Decision on the Namphong-Banekok pipeline can also be delayed until the supply of gas available for long-term transmission becomes clearer. 4.31 Offshore Compressor. The offshore compressor is not required under the SQ scenario w;:- :ts sow leve of economicaily recoverable - 42 - reserves and pronounced production plateaus. For the TOM scenario. installation of a compressor is required in 1990, which means that a decision would have to be made by 1988. Decision Agenda for 1987-89 4.32 While a number of investment decisions in the gas sector can be deferred by at least a couple of years, several will require the RTG's early attention. As discussed in Chapter IHI, gas contract and investment decisions are closely linked to key investments in the power subsector to be decided around the same time. As discussed above, the main questions facing the Government include: which of the fields in the three new areas should be produced first, and whether or not to install an offshore compressor. The analysis is based on the assumption that production contracts can be concluded within the gas pricing policy existing at the time of the mission (see para. 4.27). 4.33 Three alternative strategies for the main items contained in the decision agenda are identified and analyzed below. They are: (a) Finalize contracts for gas supplies from U3 and ESSO, the latter conditional on a successful testing program. This strategy is based on the results of the scenario analysis which suggested that developing these two fields in sequence would be part of a least cost development program under both scenarios. (b) As in (a), plus construct offshore compressor by 1990. (c) Finalize contracts now for gas supply from TP and ESSO. This implies investment in both the TP-Erawan pipeline and the compressor. In all the strategies, it is assumed that a second LPG plant of 200 .4MFCD will be constructed, instead of the smaller (150 "CFD) plant. Analysis under the TOM and SQ scenarios carried out by the mission, showed that potential gains from the larger plant far outweigh the capital savings that the smaller plant entails. 4.34 In the first strategy the contract with U3 could be concluded by the end of 1987. A second LPG plant, with a capacity of 200 MMCFD, would be constructed during 1988 and 1989. No further decisions would be made until 1990. Decisions made beyond 1990 would depend on which of the two scenarios, TOM or SQ (or their sub-components, TOM with low reserves or SQ with high reserves), evolves. These decisions include the installation of the offshore compressor for the TOM scenario, which would be two years late (as it should have started in 1988 for the TOM scenario) thus incurring a loss of gas sales valued at the difference between the forecast price ot gas and netback value for the incremental quantities that would De foregone. Other investments, including the TP- Erawan pipeline, would be made as required and there would be no penalty in either scenario for la-e or early deve.opment. The regrets associated - 43 - with this strategy, including the opportunity cost of installing an oversize LPC plant, are given below in Table 4.11, with the details in Annex 12. Taole 4.11: REGRET VALUES FOR INDIVIDUAL STRATEGIES (1986S million) Scenarios Average Max-mum TOM SQ Regret Regret Strategy 1 31.6 7.5 19.5 31.6 Strategy 2 0 48.1 24.1 48.1 Strategy 3 47.6 64.3 55.9 64.3 Source: 4.35 The second strategy is identical to the first one except that the offshore compressor is installed in 1989. There are thus no regrets in the TOM case since demand can be fully satisfied. Regret in the SQ case consists of the cost of installing the offshore compressor unnecessarily p.us the opportunity cost of installing an oversize LPG plant. 4.36 The third strategy specifically aims at promoting the diversification of supplies. The regret is evaluated by comparing RTC's opportunity costs with the least cost solution; it represents the cost of multiple sourcing with the TP field. In this strategy ESSO is given a production testing license, and both the offshore compressor and the TP pipeline are installed by 1990. 4.37 Based on the above, the lowest regret is associated with Strategy 1 which calls for signing up contracts with ESSO and U3 as soon as possible, and delaying decisions both on TP and the compressor. It is important also to note that Strategy 3 is inferior to Strategies l and 2 under both scenarios indicating the high costs of multiple-sourcing in terms of under-utilised infrastructure capacity. Recommendations In 1987 (or early 1988): Finalize U3 contract for a minimum production level from 1990 onwards of 550 4MCFD as a combined total from Ul, U2, and U3 (paras. 4.33 and 4.37). Include in the contract a provision for increased outputs in future at mutually agreed prices (paras. 4.27-4.29). Complete and facilitate testing production agreement with - 44 - ESSO. Initiate investment investment in second LPG plant of 200 "MCFD capacity (para. 4.33). Work out proposal to develop the TP field based on a detailed economic evaluation of gas production there. In 1988: Monitor ESSO testing; if initial assessment of reserves proves disappointing (i.e., close to or less than SQ levels of 0.5 TCF, lower end of expected range), accelerate and complete arrangements fur development of TP3D gas. In 1989: Check ESSO results; if satisfactory (i.e., proven reserves within the 0.5-1.7 TCF range), determine need for pipeline to Bangkok; if ESSO volumes are low (less than 0.5 TCF), complete arrangements for development of TP3D gas. In 1990: If demand and price follow TOM scenario, make decision to install offshore compressor. If demand is high and an SQ-type environment prevails, consider the possibility of raising prices for increased output (paras. 4.27-4.29) and complete negotiations for the development of TP, if not already completed. 4.38 It is to be noted that no investments are part of the recommended decision agenda for 1988-89 apart from the LPG plant ($41.7m). - 45 - V. STRATEGY FOR INVESTMENT IN THE REFINING SUBSECTOR Introduction 5.1 The impact of the volatility in oil prices on investment decisions is particularly felt in the refining industry. The economics of hydroskimming refinery operations depend on the margin between the weighted average price of products and that of crude oil; those of conversion facilities on the differential between fuel oil and middle distillate prices. The effect of the changes in crude oil prices on the spread between product prices is unclear. Generally, as discussed in Chapter II, one would expect a high crude oil price environment (TOM Scenario) to be associated with higher differentials and increased incentives for investment in conversion facilities. Certainly, stagnant crude oil prices (SQ Scenario) are likely to encourage comparatively lower differentials, thus reducing incentives for refinery investments. 5.2 A related issue is the role of heavy fuel oil as a swing fuel for power generation; if oil prices remain low, the demand for fuel oil over the next ten years could be substantially greater than in a high price environment which would tend to increase the availability of natural gas for power generation. Three broad options are available in this sector: (a) the capacity and configuration of the refining facility could be expanded so as to approximately match the pattern of domestic product demand, including that for fuel oil; (b) facilities could be expanded so as to meet the demand for transportation fuels and other middle distillates (kerosene, diesel, and gasoline) while other products wouLd be imported or exported as dictated by market conditions; or (c) investment risks could be minimized by avoiding any new additions to existing capacity and importing products to meet future increases in demand. 5.3 Using the results of a recently completed study 14/ for rationalizing the Thai refining industry, this Chapter examines some of the factors that influence investment decisions in the sector, i.e., demand projections for petroleum products, pricing of crude oil and products, supply options, and investment risks. Finally, it proposes a strategy for making investment decisions in the sector. 14/ Chem Systems Refining Study for Bangchak Refinery. The assumptions used in the refinery study regarding prices, CDP growth, and product demand are broadly similar to the scenarios (SQ and TOM) developed for this Report. - 46 - Background 5.4 Industry Structure. Thailand has a total refining capacity of 9.8 million tons per year which is distributed among four refining companies as shown in table 5.1: TaDle 5.!: THAILAND: 7OTAL REFiNING CAPAC Ty a/ (Tons Der Year) Refinery Capacity g o# Total Bangchak Petroleum Co. Ltd. (BCP) 3.25 33.2 Thai Oil Refinery Co. Ltd. (TORC) 3.25 33.2 ESSO Standard Thailand Ltd. 3.25 33.2 Pang Retinery 0.05 0.4 rotai 9.80 100.0 a! Nameplate capacity. Source: 5.5 TORC is owned jointly bv PTT, Crown Properties, and private investors (including Shell and Caltex). PTT and the group of private share-holders each own 49% of TORC's equity, with the remaining 2% held by Crown Properties. TORC is operated under an operating and service contract with Shell. The ESSO refinery is fully owned by ESSO Eastern Inc., a subsidiary of Exxon (U.S.), while the Fang Refinery, which is operated by the Defense Energy Department (DED) primarily as a pilot plant, is owned by RTC. BCP (formerly known as Bangchak Oil Refinery - BOR) was originally owned by the Ministry of Industries and operated by the DED. Following its recent restructuring, it now operates as an autonomous commercial enterprise. 5.6 Approved investments include the rehabilitation of parts of the Bangchak refinery and installation of hydro-cracking facilities at the TORC refinery. New investment options under discussion at the time of the mission included the rehabilitation of the crude distillation unit No. 2 at Bangchak, installation of secondary conversion facilities also at Bangchak, additional capacity at any of the three main refineries, or the construction of a new "grass roots" refinery with conversion facilities. In the absence of significant additions to refining capacity over the next five to ten vears, investments of (estimated at about $200 million) would be required to increase storage and improve port facilities so as to accomrodate the required imports of products. - 47 - Crude Oil and Petroleum Product Prices 15/ 5.7 Projections of prices for crude oil and products under the two scenarios are summarized in Table 5.2 and given in more detail in Annex 3. Table 5.2: CRUDE OIL AND PETROLEUM PRODUCT PRICES (19861/barre ) Crude O;l Gasoline Diesel Fuel Oil rear TOM SQ TOM SQ TOM SQ TOM SQ 1990 20.0 15.0 26.0 19.5 24.3 19.8 17.0 13.0 1995 25.0 17.0 32.5 22.; 3 .5 21.0 21.0 15.6 2000 30.0 20.0 39.0 26.0 33.3 23.0 25.0 18.1 Source: Chem Systems Refinery Study. 5.8 Under the SQ Scenario, gross refinery margins, as measured by the difference between the price of crude oil and the weighted average of product prices, are substantially lower than under the TOM Scenario. Although this implies that investments in simple crude distillation facilities would be more profitable in TOM, despite relatively weak demand, under either scenario, the margin would not be sufficient to justify investment in new simple crude distillation facilities. For these refineries to be viable (in TOM or SQ, and as is the case in 1987) they must be complemented by facilities for upgrading their fuel oil output to lighter products (gasoline, kerosene, diesel, and LPG), assuming that such upgrading facilities are economic. The differential between middle distillate and fuel oil prices would be higher under TOM than under SQ. Under SQ, the expected differential would average $5.7 per barrel ($44 per ton) during the 1990s, which would be insufficient to justify most types of conversion investments let alone distribution capacity. In contrast, under TOM the average yield would be $8.7 per barrel ($67 per ton), which would normally be sufficient to justify moderately expensive conversion investments over and above distillation capacity investments. Demand Projections for Petroleum Products 5.9 A summary of projected demand is given in Table 5.3. More details are provided in Annex 15. Under both scenarios demand for gaso'ine is expected to increase at a higher rate than in the past, in 15/ The section on pricing (paras. 5.8-5.9) and demand (paras. 5.10- 5.11) are derived directly from the Chem Systems Study. - 48 - part because prices under either scenario are not projected to grow at as rapid a pace as they did during the 1970s. Gasoline demand is also expected to grow at a faster rate than diesel due to the RTG's recent decision to reduce differentials between gasoline and diesel prices and to increase taxes and custom duties on diesel engine vehicles. Table 5.3: PROJECTED DEMAND OR DIESEL AND GASOLINE (Thousands ot Dar'e's Der cay' Diesel a! 1990 1995 2000 TOM '18 144 176 SQ i23 i57 200 Gasoline D/ ToM 47 60 77 SQ 50 68 91 a/ Implied Dr;ce eiasticity, -0.35. 0/ ImDliea Dprce e;astcity, -0.41. Source: Chem Systems Refinery Study. 5.10 The demand for heavy fuel oil varies widelv between the two scenarios, as can be seen from the figures in Table 5.4, due mostly to fluctuations in gas availability: fuel oil accounts, in 1995, for about about 7% of the requirements of the power sector under TOM but for nearly 35% under SQ. An important issue for the refining industry is how to cope with such a large range of possible outcomes. aD'e 5.4: PROJECTED DEMAND POR luEL OIL 3/ (Thousands ot barrels/day) 1990 1995 2000 Scenar.o Power Industry Total Power noaustry Total Power Industry Total TOM 8 27 35 13 32 45 '2 39 51 SQ 50 27 77 72 32 104 64 39 103 a! The mpl led price elasticity of Totai fuel o0I demand Ts -2.18, -e+lecting the large potential for substitution between fuel o i and other products. Source: Chem Systems Refinery Stucy ano r ss on estimates. impact of Uncertainty on Least Cost Ir.vestment Plans 5.11 The least cost `nvestmer.t plans made under TOM and SQ were estimated using the results of Linear Programming planni..g models. These - 49 - plans, which are shown in Table 5.5, serve as benchmarks for strategy evaluation and assume that planners know with perfect foresight which scenarios will emerge over the next 20 years. Table 5.5: LEAST COS INVESTMENT PqANS UNDER TOM AND SQ SCENARIOS TOM SQ Cumulative lnvestmen*s (1986 S miliion undiscounted) a/ 1987-95 48.0 200.0 1996-2005 289.0 0.0 Total (1987-2005) 337.0 200.0 Present value of total costs b/ (1986 S mill on) 38,300 43,000 AddiTions to Capacity (commissioning dates) Rehabilitation, bpd 20,000 ('990' _ New crude distillation, bpc 100,000 (996i - New secondary conversion capaci'y, c/ bpd 14,000 ('990, 21,000 ('9961 Importat;on/storage 'ac ' t'e5, tOnS 200,000 (1990) Supply-demand balances (bpC) 1987 _997 1987 1997 Gasoline (3,800) (2,3001 (3,800) (24,000) Middle distillates (42,000) (44,000) (42,000) (97,000) Fuel oil (2,6001 (2,500) (2,600) (6,500) Total (48,400) (48,300) (48,400) (174,500) Fuel oil consumption (tpd) EGAT & other industr,es 35,500 48,300 35,000 104,000 Refine'y conve51of1 d/ '0,800 52,000 '0,800 - 46,300 ;02,300 46,300 '04,000 a, Figures exclude investments alreac comm tted. b/ Sum of discounted fixed and va'iable cos-s ncu,rea over '900-2005. c/ Investment in Res,dual Cata )t c CracKe, (RCC), a relat'vely nexpe^s.ve form of conversicn investment. d/ Including existing cracKing faci tes a, tne TORC -efinery. Source: 5.12 As shown in Tanle 5.5, the pattern of investment under the two scenarios differs substantiallY. Despite a lower level of demand, the TOM scenario calls for s:gr.ificantly higher investments than SQ; total undiscounted investment fcr 1987-2005 is about S340 million under TOM and $200 million under SQ, in 19F6 ,rizes. 7he cwer leve' of investment in - 50 - SQ arises because product price differentials and margins do not justify investment in refinery facilities. The increase in demand is met through increased imports; $200 million are required for infrastructural investments--increasing storage and improving rail and port facilities-- to accommodate much higher imports. Under TOM, the pattern of product prices justifies the rehabilitation and upgrading of facilities at Bangchak, including the installation of 14,000 bpd cracking facilities at a cost of $48 million. To meet growing demand, the least cost plan under TOM also calls for the installation of 100,000 bpd of additional crude distillation capacity and 21,000 bpd of secondary conversion facilities. Finally, the two plans also differ significantly in terms of product balances. Under SQ, very large imbalances emerge by 1997. In TOM, demand and supply are approximately in balance for most of the products, but a significant amount of middle distillates will still need to be imported. Investment Strategies and Risks 5.13 The above discussion shows that the optimal configuration of future refinery investments is highly sensitive to the evolution of crude oil and product prices. Least cost investment plans based on a single set of assumptions about demand growth and product prices thus carry a significant risk. The focus at present is to identify a near-term investment strategy that would minimize the cost of wrong decisions (i.e., minimize regret) in the long-run. 5.14 There are basically two main options open to the Government in terms of decisions to be made on the near term (1987-89): Strategy 1: This strategy calls for an increase in the refining industry's capacity, and a change in its process configuration, both on a modest scale consistent with the front-end portion of the least cost investment under TOM. Investment decisions over 1987-89 would include the rehabilitation of crude unit No. 2 at Bangchak and the installation of a 14,000 bpd residual catalytic cracker (RCC), both to be commissioned by 1990. In essence, this strategy assumes that product prices will continue to support investments in limited refining expansion and modification, or at the very least, that this option will not lead to unacceptably high losses. Strategy 2: This risk averse strategy would postpone any commitments for investments until later vears and allow imports to meet any .ncrease in deficits. This strategy is consistent with the front-end pcrtion of the least cost plan under SQ but onlv partly so becaase it would also defer any investment in storage and port facilities. - 51 - 5.15 The efficiency with which each of these two alternative strategies copes with an uncertain environment and adapts to the TOM and SQ scenarios is shown in Tables 5.6 and 5.7. Table 5.6: ADAPTATION TO THE TOM SCENARIO Strategy 1 Strategy 2 Short-term investment - Rehaoii tat;on of - Zero investment; Bangcnak Unit No.2 increased imports. (Decision 1987, Commiss,on in 1990) - Install new 14,000 bpd RCO (Decision 1987, Commission in 1990) investment aecisions - install new refinery with - Bring forward new 1989 arc beyond 100,000 bpd distillation refinery, with 120,000 capacity and 21,000 bpd of bpd capacity, to 1993 secondary conversion capacity (Decision in 19891 (Decision 1992, Comm,ss,on '9961 - install new 35,000 bpd RCC in 1993 (Decision in 1990) Cumulative Addit ors to Capacity (Qmd) Distillation Conversion Dis- Ilation Conversion '990 20,000 14,000 _ _ 1991 20,000 '4,000 _ _ 1992 20,000 i4,000 - - 1993 20,000 14,000 120,000 35,000 1994 20,000 '4,000 120,000 35,000 1995 '20,000 35,000 120,000 35,000 Present value of total costs (19863 ml lion) 38,300 1 38,430 Source: if the TOM Scenario emerges and this becomes obvious to planners in 1989. Strategy . will require no adaptation as the initial decisions are cons.stent wlth the least cost investment plans under TOM. However, the initial decisions under Strategy 2 (zero investment) would need to adapt to the hi4her than expected margins by embarking on an immediate program of refinery investmen-s. The regZrew ith Strategy 2 is the value of * 2 foregone profits. As investments decided upon in 1989 cannot be commissioned before 1993 it would be more economic to bring forward the commissioning date of the new refinery and raise its capacity to 120,000 bpd rather than invest additionally in rehabilitating Bangchak. Tabte 5.7: ADAPTATION TO THE- SQ SCENARiO Strategy ' Strategy 2 Short-term investment - Rehab'litation of Zero investmen1; BangchaK Unit No.2 increased imports. (Decision 1987, Comm,ssion in 1990) - instal new 14,000 5pd RCC (Decis on 1987, Commission in 1990) Investment devisions - inves-men in 200,000 tonls Investment in 200,000 tonis in 1989 and Devond *mpo''>'storage facili ies import/storage facilities (Decision in 1989) (Decision in 1989) Cumur'at ye Acd *ions .C. aac'vY D :+ _Inf-a- DOstil- Infra- *at on Converson strwcture iation Conversion structure )Dpd bpa tons bpd bpa tons 1990 20,000 14,000 - - 1991 20,000 14,000 - - 1992 20,000 14,000 200,000 - - 200,000 1993 20,000 14,000 200,000 - - 200,000 1994 20,000 14,000 200,000 - - 200,000 1995 20,000 14,000 200,000 - - 200,000 O-esen+ va'ue of total costs 11986$ m,lion) 43,120 43,070 Sour ce: In contrast to TOM, the emergence of SQ would reduce the profitability of the investments in refinery expansion and rehabilitation that would have been committed under Strategy 1, these investments would have turned out to be are less profitable than planned. No further investments would be made under SQ but $200 million worth of investments would be required in infrastructure. As neither strategy has made anv commitment towards this investment, the earliesz it can be commissioned is '992 if a decision is made in 1989. Additional nandling and storage costs over 1990-92 due to the !ate commisscr.ing of the new facilities is estimated at about 570 miu ! nn. - 53 - Strategy Evaluation 5.16 The total discounted costs of meeting demand over 1990-2005 for the two strategies are also given in Tables 5.6 and 5.7. The efficiency with which a set of initial decisions adapts to say the TOM scenario can be assessed by comparing the present value of total supply costs with those of the least cost plan under TOM (Table 5.5). The difference between the two plans represents the regret associated with the initial strategy. The regret values for the two strategies are summarized in Table 5.8. Table 5.8: REGRET VALUES FOR NDIV'DUAL STRATEGIES (1986S m;lion' Scenarios Average Maximum Regret TOM SQ Regret Strategy I 0 120 60 120 Strategy 2 130 70 100 130 Source: 5.17 Under -OM, the lowest regret is associated with Strategy 1 as it is consistent with the least cost investment plan in that scenario. There is a regret of $130 million associated with Strategy 2 due to the loss of foregone profits. Under SQ, where investments in refining would be unprofitable, the lowest regret corresponds to the strategy with the lowest initial capital investment, i.e., Strategy 2. 5.18 A number of criteria can be used to choose between the two strategies. If decision-makers intend to avcid the worst outcome (i.e., minimize maximum regret) they would prefer Strategy 1, which limits the regret value to a ceiling of $120 million. If, on the other hand decision-makers give equal probabilities to the occurrence of the TOM and SQ scenario, they may want to concern themselves with average regret. Here again Strategy 1 would be preferred. It is only if decision-makers give a high probability to the occurrence of SQ that they would opt for a zero investment strategy (Strategy 2). Conclusion and Recommended Strategy 5.19 Investments in refining are clear:y highly sensitive to oil price uncertainty, The kev differential between fuel oil and middle distillate prices, however, may or may not be sufficient to support the necessary investments r. secondary conversior. Refinery investments are seen to be closely linked to the viaDil:ty of secondary conversion, since in the absence of :n.estments an secondary convers:on, the rehabilitation or extens.Cn of pr:marv disti..aticn capac:-y is no: expected to be - 54 - viable in any credible scenario. Based on the range of prices and differentials covered by the two scenarios, the best near-term investment strategy in the circumstances is seen to include a modest investment (of about $50 million) in conversion capacity and to rehabilitate primary distillation capacity at Bangchak (Strategy 1, para. 5.15). This strategy gives the refining industry the flexibility to adapt to future circumstances at least cost. 5.20 Domestic demand for fuel oil will depend on future oil prices as well as EGAT's own specific requirements. The considerable uncertainty associated with both these factors does not make it economically justifiable to aim at meeting domestic demand in full at all times. Rather, the recommended strategy will add facilities to meet a limited increase in product requirements in the expectation of moderately profitable margins, and the RTG should review the situation in about two- years time in view of updated demand and/or profit margins. Given available refining capacity, and recommended additions to it, decisions for further expansion can be delayed until 1989. By all estimates, fuel oil is projected to be readily available at competitive prices both regionally and globally, requirements therefore could be imported to meet the demand by power and industrial sectors. The recommended strategy, which includes facilities for upgrading fuel oil to lighter products (gasoline, kerosene, and diesel), together with the upgrading of the TORC refinery already underway will help to structure the industry to meet transportatiQn requirements and middle distillate demand to the extent economically justifiab.e and allow fuel oil to be imported or exported as dictated by market conditions. 5 55 VI. INSTITUTIONAL PLANNING AND DECISION-MAKING UNDER UNCERTAINTY Introduction 6.1 The analysis and decision-making processes at the line agencies and the NESDB have largely evolved in an environment of relatively steady growth and stable prices. Planning in a less predictable environment, which Thailand currently faces, may require changes in these processes. The analysis presented in the preceding chapters points to the need to increase the flexibility of the energy sector as a whole to adapt to a range of circumstances quickly and cost effectively. To implement such a strategy, sector planners should concern themselves with the following issues: (a) the absence of a framework for evaluating the consequences of uncertainty and the excessive dependence on single line forecasts for the key variables involved (e.g., oil prices); and (b) delays involved in reaching decisions on policies and projects and criteria for their selection. These issues are discussed below. Planning Framework and the Dependence on Forecasting 6.2 At present there is no formal framework or process available to energy sector institutions to take account of uncertainty in drawing up policies and investment options. Sector planning is presently based on single line forecasts of key variables with individual subsectors evaluating the sensitivity of their plans to variables which directly affect them. While this approach worked reasonably well in the first half of the 1980s, forecasting errors have recently become increasingly frequent (as for example, the greater than expected decline in electricity demand during 1985-86) and have occasionally taken on a dramatic and unprecedented magnitude (such as shifts in the price of oil). An additional problem with this approach is that evaluating sensitivities of investment plans to individual variables, assuming everything else remains constant, can give misleading results as it is incorrect to assume that variables such as oil prices can change withouat altering demand, or fuel availabilities. 6.3 One method to tackle the problem presented by the increased uncertainty and greater difficulty of preparing accurate forecasts may not be to look for better forecasts through perfecting new techniques (as appears to be the approach being adopted by ECAT and NESDB for power demand), but to accept the uncertainty as a structural feature of the environment and attempt to understand its implications and find ways to cope with them. To nstititionalize planning under uncertainty the Government shculd consider: - 56 - (a) Shifting the emphasis of sector planning away from its excessive dependence on single line forecasting to one using a range of values for key variables such as oil prices, gas availability, etc. In practical terms, this implies that investment plans should be derived for both the upper and lower end of an agreed range for such key variables. (b) Establishing linkages between the key variables involved. For example, in deriving investment plans for the refining subsector, a qualitative and quantitative link (e.g., a correlation) needs to be established between future oil prices and demand with a corresponding linkage between the range of product prices (high and low), and that of future demand. This would allow investment plans to be derived for two cases or scenarios: high price, low demand; and low price, high demand. (c) Since some judgement is required in establishing realistic scenarios, key decision-makers in the line agencies and NESDB should participate in the preparation of scenarios, before the process of identifying invest,ment programs and strategies is initiated. 6.4 Base on the experience gained on this STudy, the imission makes the following specific recomnmendations for developing and using scenarios: (a) If they are to be effective, scenarios should not be developed on an ad-hoc basis; rather they should be made an integral part of the planning review and decision-making process. Scenario identification should coincide with the relevant planning cycles (for example Five Year Plans). Also line agencies could be required, by the NESDB, to estimate the implications of their proposals against the agreed scenarios before thay are submitted to the NESDB for approval. (b) Scenarios reflect long-term development possibilices, and if they have been carefully constructed, there should not be any need to modify them frequently. The mission suggests that new scenarios be identified once every three years, with an annual review process to take account of surprise developments. (c) The use of scenarios as a planning tool is still evolving. As such there are no hard and fast rules but the following guide- lines are suggested for their construction. Firstly, the key variables, and the scenarios that are identified should cover the energy sector as a whole and not just a particular sub- sector or line agency. Secordly, for ease of communication it is preferable that scenarios be given brief names (or acronvms) that reflect their inherent characteristics, rather than being identified by numerical or alphabetical characters such as Scenario"A" or Scenrio "B". Thirdly, to avoid confusion, the - 57 - number of scenarios should be restricted to a maximum of three (or possibly four), one of which could reflect the Government's views of most likely developments, the cenral or base case. The scenarios developed should span the range of uncertainty and any investment strategy which can effectively cope with these main scenarios should also be able to adapt to any of the large number of intermediate scenarios or sub-cases that can be constructed. Changes in the Decision-Making Process 6.5 A more effective evaluation of the impact on uncertainty should be complemented by changes in the decision-making process itself particularly in respect of: (a) establishing criteria for making decisions in a period of uncertainty; and (b) reducing the time involved in reaching initial investment decisions. Specifically: (a) Project selection. Optimal investment planning, under conditions of uncertainty calls for sufficient flexibility to adapt quickly to circumstances which are different from the original assumptions: a premium has to be paid to achieve such flexibility and a key issue relates to the choice of invetment options to provide this flexibility at the lowest premium. A part of the investment decision process should therefore concern itself with an evaluation of the performance of a particular investment plan or project under alternative scenarios, in addition to the normal criteria for project selection. (b) Time Lags in Decision-Making. The mission found widespread agreement, both within and outside the Goevernment, that excessive time lags are involed in reaching decision on projects and policies. To increase the ability of the energy sector to provide a flexible response to uncertainty, actions would also be required to expedite the decision-making process. For instance, EGAT currently estimates an 18-month lag between project presentation and project approval (by its own Board and by the NESDB). This lag could probably be reduced through the establishment of a "reserve" of preapproved projects that would only require financing arrangements to be finalized before implementation. An example of this could be the Khanom power plant: if sufficient gas supplies were to become available to justify the project, it could be implemented quickly by using already completed studies and designs, including approval by PTT on pipeline design and costs. The same applies to the construction of new combined cycle plants at Bang Pakong and Nam Phoog. To make such a process of pre-approval effective would require coordination and agreement between the line agencies and the NESDB since the - 58 - latter reviews projects once they have been approved by the Board of the relevant agency. Annex 1 -59 - Page I of 2 Scenario 1: TIGHTENING OIL MARKET (TOM) a/ GDP Growth (1986-2000) 3.5% per annum 1986-90 1990-95 1995-2000 Electricity Demand Growth 7% 5% 4% Oil Prices 1986 1990 1995 2000 (Average OPEC, $/barrel, 15 20 25 30 constant 1986 $) Fuel Oil Prices (Bangkok) (Constant 1986 $/barrel) 13 17 21 25 (1986 $/mBTU) 2.2 2.8 3.5 4.2 Coal Prices (1986 $/ton at Bangkok) 44 58 62 65 (1986 $/mBTU) 1.7 1.9 2.4 2.46 Gas Availability (MMCFD) b/ 470 570 1,110 1,200 Product Price Differentials (Gas oil - Fuel oil, 1986 $/ton) 62 69 78 83 Gas Costs ($/mBTU) (Well head prices as per existing Unocal contract) 2.1 2.3 2.6 2.8 a/ Lignite prices, which are assumed to be the same under both Scenarios, are given in Annex 3 (Tables I and 2). b/ Recoverable reserves of 14.4 TCF. Source: - 60 - Annex 1 Page 2 of 2 Scenario 2: STATUS QUO (SQ) a/ CDP Growth (1986-2000) 6% per annum 1986-90 1990-95 1995-2000 Electricity Demand Growth 9% 7.5% 6% Oil Prices 1986 1990 1995 2000 (Average OPEC, $/barrel, 15 15 17 20 constant 1986 $) Fuel Oil Prices (Bangkok) (Constant 1986 $/barrel) 13 14 15.5 18 (1986 $/mBTU) 2.1 2.3 2.5 2.9 Coal Prices (1986 v/ton at Bangkok) 44 46 54 57 (1986 $/mBTU) 1.7 1.8 2.1 2.2 Gas Availability (MNCFD) b/ 470 500 570 500 Product Price Differentials (Gas oil - Fuel oil, 1986 $/ton) 62 65 72 75 Gas Costs ($/mBTU) (Well head prices as per existing Unocal contract) 2.1 2.1 2.2 2.4 a/ Lignite prices, which are assumed to be the same under both Scenarios, are given on Annex 3 (Tables 1 and 2). b/ Recoverable reserves of 5.6 TCF. Source: %''lId Fossil Fuel- Pafrolew, Nteloleu. P,,Adcts ,~ N .l,mnI (I. t,ICl' P. n,,e"Ie 1., Tota Co- en.,, I,,. "..CI 1lQnItS Cr-ude r.,,,.,,- Natural trrl~ Ottioll,, Jot Fust (e,-o- tOlesel Fuel Oil C. PrIntey f-I,-d 1,, h-,. ,7 Pe.-., it Oil .te. 0eAt,lle Aeletlon te. '( P-f- t,,t Wo,d (n.rqv Ite'n '..h - -- -- --01 1 ( I ns,( -- - - -e .. . II,-1.w I- Me ,.f PI-,, (,..n 1476 1,790 704m 0 0 0 (I n7 0 7 200 :7'- *.02, lap-..t. 2 27 0 S,,14? a 0 10o 8 214 0 2.70'M 87 I,s? q v '- 0 ~ 'IT It 0 0 0 (6501 0 0 0 ((5) 7)II I(,) (7211 0 7 0 (571) Me,le./AeIelI~nOunkrt4 0 0 0 n n 0 0 0 0 0 7 0 a 0 0 0) 0 '.t.~~l. tNnn~~e 9, 0 (9(1, 144 Ut?) r9% (II) 61 82z (5) (III) 1151 17 0 0 0 11 0 0 p'l-,~ F-..y -ry 6 2;'t 5.476 ,7 71 (ASi (I 89 69 221 (51 2, 061 221 (2,517 1277,79, 5,'1 41 21,111 17If,1 10,41? 51,122 F-, q,, I.-v.eo.*(4,681) (9.72(1 (FA.)) M4 1,029 2,700 1,149 U4S 5,660 1,323 (I9311122,749) (5, 513) 24.?16115.509) 3,0 (961) (11,361) 'dI.....,. ~It.,.. letA 00 (9,m71 (o81 (1) 246 2,200 1,149 149 1,61F4 2,189 (2011l 0 0 0 0 0 0 0 SC.. r,,.,.,I.,0 'n~ 7 0 0 0 109 781 0 0 0 0 0 "901114.4A91 0 0 0 0 0 0 1g~,.l- 7'.~.,, lie?. I (4.,695? 0 0 n 0 0 0 0 (14) (866? (880? t7.7611 (5,5-5) 274. 716 0 0 0 0 0 (III,. i.,e.., .7.,,. I F7 0 0 0) 0 0 0 0) 0 0 0 0 0 0(15,5051 3,101 (37A1 fI3,363P F In..- A Owe, In.. I/ 17 0 0 0 0 0 (I) 0 0 0 0 0 (7) 79.1167 0 (5.421) 0 0 0 0 C-q-Vfo F- k."(-Q,flis 1 0 0 0 0 fI1l 0 0 0 0 0 0) (71) 0 00 0 0 0 0 M., Fn4.,w Suyp(vf (4 2/1 741 0 0 0 I'll? 2, 769 '. 10 141 )1,777 1,5,44 12,168 0 /.0`14 ?,826 ', 0)t- 9.A' 70,`157 Stt.t0hller.-4. II 0 0 0 0 0 0 0 0 0 0 0n 0 0 0 0 0 Final r."uipf,q(,. I/ 16 271 741 0 0 0 1,117 2,269 1,370 143 1.121 1,544 12.168 0 0 722034 7,82A 9.071 9,456 70,115? Aar I- l(ure P? 0 0 0 0 0 4 70 0 I 941 5 1,025 0 0 57 0 0 0 0 ATInnq IA 0 0 0 0 0 0 2 0 I 16 20 59 0 0 0 0 0 0 0 M.neI..tw Ing 19 223 741 0 0 0 133 10 0 47 94 1,217 1,601 0 0 10,162 1,714 0 8,601 10.119 loe ~~~20 0 0 0 0 0 0 2 0 0 II? 21 140 0 0 0 0 0 0 0 Aesid.ntlal P. C,,aaalal 21 0 0 0) 0 0 6i70 0 0 90 4 12 776 0 0 11,645 6,112 1.071> 651 10,01A Ir-op.. tot Ion 27 I0 0 0 0 0 10 2,091 1.370 2 4,219 259 8,309 0 0 0 0 0 0 096.re .' ' 0 ~ ~~ ~~~~ n 'I ? 94 0 2 150 to 7189 0 172 0 n 7 pinoe M.$ h-k, hj7nt.4, eI, 2/Inisas II, andeoet-~ e dI*trIh,(I,e end ,t.d t,lv enerqy sector.; lbm 5.' Sour,*- M M~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1 ri,47). ,.7.I'. 1.(wi. CrI..0.)o N i1 1. a,In 3)TreI.,, I* Ito r i) '..,7 " 'r) .....r) 7. .'474 A 07) I i' ~ ctri AvI.r)i ..." ,,..S., ,1 4., 7'...4.r,)rri,, ..r.g..,,~~ ~ I A I7,9(4 "(7 '.an o o ni 0 a a ,t & 1.i' I A s n r'rs q%qo o '.429 is,) - )*.841 7 74) 0 I - A 0 A Al f 179 0 1,16 77 979 0 0 PA7 1,''? 0 9 7 0 9,999 * .r..*. 7~~~ ~ ~~ ~~~~~ ii A (49') a' WI 0 0 (Ii) (5607) 0 0(7) ("70) 0 ( 76) 0) (71) (9,6) 4 .i (7 0 7) 0 0 0 0 0 0 0 0 0 A 0 0 fl 0 0 * ~~~~~~~~~~~~~ ~~Ii (777 (74 finS) 13) 177 43 Al? 14) (06) (46 376 0 0 0 ¶04 a 0 0 (7 996 .74'1 74417A CA. VI III 71 -17 9I It) (4) 9,7*0n 70*1 in (76* 3.190 9.7I3M 99 16,14? 8.6130 (III 7,470 11.747 2,069(0 (¼,., 7* ' ill le'q).'17 7~~~ n 7 (A7 7 'r) 49) 27 7(4) (1799q 09 97) '."A 7, 249 7917) 13,0047 (1,7)0) 7,(, 79,g P.??)1 (3,698R) 7,970 I7117, (4,07,)76 1 7,646) 7a9 ft ii ii~~~~~~~, I i. 'ii,iI7 ('R - ''7I . I,990 %99 '77 7).71 2,7-0 7 I #.r 0 0` 77 (779) 0 0 0 0 (¶74 W.. P".-..).rt.n n7 't Al 4 0 0 0 0 0 I''7 17,4) 0 A 79)) 0 0 0 0 Jay) PI.,,.7 li "11)17) I' 7 n 0 0 0 nl 712 (A 79))') 7Q0) (1,.790) 7,709A i'07i- C 0 n 7 (3,9*7) 'i7Ii~~~~~i.I...ip 7) 1 ~~~~~~~0 0 7 0` n 0 ( 0PI1 n A. r77,PV. 7, t770 ))7,I7 4,7-A) 914,0Y)6 nip 1-i PINe. PI I 0 n 0 0 0` (7 0 (7 fl (9) (207) )4,'7 7) 0 9 7) (4769 *r,~~'.i,.q'*)~~ 'ip 74., 7.... ,~~~ Ii.eq 79 0 0 0 0~~ 7747 0 0 0 0 1 0, )."4) ) n 0) (704) 0 0 n nl (7049 fb .., 1v9,,' 4 la) 921 A 0) 0 704 1,1790 1 7,77 II? 4,094 9,449 70,)4n 0 0 ,*1* 97,447i 7,'67 7,103 (N0 , 9 0.73Q 4*.7.fll)9e,..,.e, I' 0 ~~ ~~ ~~~ ~~~0 0PI0 0 0 0 0 0 0 0 0 0 0 0 A00 PP1lr.. *f . 9 717 (4) 9179 0 0 I70 11,900) 9,970 9I7 4,934 9,433 10,0)6 0 0 .987* (7,447 2,962 72,903 7,90) 7,966 196,693 Ac)e.u-I,re It n 0 0 0i 0 3 972 0 I *7I 9 876 0 0 13 *8A) A 0 00 6*I 7*1 0 0 0 0 0 0 7 0 I II to 31 0 0 (0 99 0 0 0 0 3 N~~nr,9er7o'7.'o 70~~~~l 94) %29 i r 4 7 0 38 967 9,941 9,64) 0 0 MPA 7,A9* flaq 0 7.6' 2,"9,7 r,,.97r,i. Plo,. 70 7' I' ~~~~~~~~~~~~~~~~~~~~~~~~~11 7 A 0 90) 70 973 0 0i 't 7.- ( 072 7. .".r-'. 7.,' i . , . .* .*.r A~~~~~ ~~~~~~ ~~~~~~ ~ 77* 1,9-9? 7,770 2 3,6*8 244 17,4(n 0 'i A,4("5 0 O 'I , ii 1 7 7i7~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1 0 2 (70 9 "it ni ( 75 7n f flnV 17). ilqir.e. 7.. i..,~~~~~~~~~~~~~~~~~~~~~~~~~*'rtri* ~ ~ ~ ~ ~ ~ ~ 9/ )pol.r7a* psoop . 7.e.7r,.le. *7. - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~~J ¶/ 7,,.'Ii,47,rq It.., 4,4 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ -63 - Annex 3 Page 1 of 4 BASIC POWFR SECTOR DATA AND ASSUMPTIONS Ca.adidate Power Plants fnr Expansion Plan Formulation 1. Potential power l'ants included in the expansicr prograr are summarized in the follow ng tables. Hydro MW GWh. Cos- a C^- c I on Average Firm $m $kWh Years Srinagar:nd 5 180 132 112 29 162 3 Nam Chon 580 1,198 865 353 609 9 Kaeng Krung 80 178 '56 82 1,031 6 Sai Burl 46 119 78 76 1,650 5 a/ Includes IDC at '22' and associated transmission. Thermal. MW Cost a/ Eff.ciency Construct;on Sm S/kW BTU/kWh Fuel b' Years R3 Lingr,ite 75 100 i,329 10,700 L 5 R3 Coal :s5 192 1,278 10,300 C 6 Mae Mch 10 300 307 1,023 9,900 L 6 Mae Moh 11 300 290 967 9,900 L 6 Mae Moh 12 300 249 829 9,900 6 Namn PhoGg CCi 300 203 677 8,100 C 5 Narn Phong CC2 30C .212 707 8,100 G 5 New CC units 300 193 643 8,100 C 5 Peaking G'Cs :00 36 363 12,100 C 4 Bang PakonF 3 600 677 1,128 9,000 C-F-C 6 Ao Phai 1 bO) 68. ;,133 9,000 C 6 Ao Phai 2 ,O'l 653 :,0°8 9,000 C 6 Ao Phai 3 bO t?7 ,ocL6 9,000 C 6 Ao Phai 4 6 C 6?> 1,O.1 9,000 C 6 a' Inc:& o) as a . - *ra s-:ssion. b/ G = Ca-, E !i.zote, C - oa:, F =Fue Oil. Load Evol u'. i, ar- Sc-. ar. o Demar.ds 2. Historical tcta: generation requirements *cr the EGAT svster snow the follow.nr pat:err,: Year Ml h ,974 1.256 7,25 '980 2.417 14,754 :985 3.87E 23,3C7 198*? (es:.) L ,'77 24,730 - 64- Annex 3 Page 2 of 4 Samirarv of Demand Data EGAI Base C;;- TOM Scenario SQ Scenario Year Mh -WIn MW TWh fW TWh 1987 4,560 27.1 4.450 26.5 4,540 27.0 1988 4,950 29.4 4,770 28.3 4,950 29.4 1989 5,310 31.8 5,070 30.4 5,360 32.1 1990 5,610 33.7 5,410 32.5 5,820 35.0 1991 5,940 35.7 5,670 34., 6,250 37.6 1992 6,240 37.7 5,930 35.8 6,690 40.4 1993 6,530 39.7 6,180 37.6 7,1140 43.4 1994 6,820 4:.8 6,450 39.5 7,630 46.7 1995 7,130 43.9 6,740 41.5 8,160 50.? 1996 7,450 46.0 6,980 43.1 8,610 53.2 1997 7,780 48.2 7,200 44.8 9,090 56.4 1998 8,120 50.6 7,500 1'6.6 9,600 59.8 1999 8,480 53,1 7,760 8.5 10,100 63.4 2000 8,860 55.6 8,030 50.4 10,700 67.2 2001 9,260 58.4 8,320 52.5 11,300 71.2 Gas Availability Data 3. Poten.ial gas supplies used in the PDP runs are: Gas Ava::ability for the Power Sector (MMCFD) Year SQ Scer.ario TOM Scenario 1987 328 360 1988 375 400 989 3i8 420 1990 368 440 ;991 276 440 1942 278 480 1993 330 5:0 199'1 330 550 1995 330 580 .996 330 610 ;997 330 bb0 1998 330 670 1999 300 710 2000 260 690 2001 200 670 Pr:ce s 4. Tab.eu 1 and 2 s - -e -e Dr:ce a >_7 .or he 04 ar SQ scenarios. Tdbte 1: I UL i'lk CE Ii fit '',.I 1 IN [)I HIIN4 1G ',SI COST fXF'ANSION PIAN UNDtR TOM SCENARIO t( ,n-.t nt 106 pr (t 1 Year Fuel 0Il trO IIe dI ' _ Impttl CoUril i' e da/ Na i ur a Ga' rPrice i/ L.ignite rPrice b/ US$/MfI3U B/l. itre Incrfedae USS jon lIS1,'MHtU I Increa.e l,St/M9TU B/ME!U I Increa.e US$/MBITU USS/Ion S Increase 1986 2.11 2.21 - 44.00 1.91 - 2.10 56.70 c/ - 1.82 18.58 - 1987 2.28 2.38 7.14 45.50 I1.B 3.30 2.10 56.70 c/ 0.00 1.83 18.70 0.64 19FI8 2.4 1 2.55 6.67 47.no 2.04 3.19 2.10 56.70 0.0o) 1.85 18.83 0.69 1Q89 2.60 2.7? 6.25 48.50 2.11 3.09 2.20 59.40 4.55 1.86 18.97 0.74 1990 7.16 2.89 S.88 50.00 2.1 s.00 2.30 62.10 4.35 1.88 19.13 0.84 1991 2.R9 3.02 4.49 52.40 2.28 4.58 2.40 64.80 4.17 1.89 19.31 0.93 1992 3.02 3.16 .1.30 51.R0 2.38 4.38 2.40 64.80 0.00 1.91 19.50 0.97 T99q 3.*1 i.;q 4.1, $7.21 2.49 4.20 2.90 61.50 4.00 1.93 19.72 1.12 1994 3.29 3.1 x 3.Q6 59.9/n 2.59 4.03 2.50 67.50 0.00 1.96 19.95 1.15 1995 3.1?1 3.57 3.81 6.'.00 2.70 3.87 2.60 70.20 3.85 1.98 20.20 1.24 1996 . %1. 3.61 62.60 2 .72 0.96 2.60 70.20 0.00 2.01 20.49 1.42 1997 3.68 3.84 3.54 63.20 2.75 0.95 2.70 72.90 3.70 2.04 20.80 1.49 1998 3.81 3.97 3.47 63.80 2.77 0.94 2.70 12.90 0.00 2.07 21.14 1.61 1999 5.94 4.11 3.31 64.40 2.8(f 0.93 2.80 79.60 3.57 2.11 21.51 1.7? 2000 4.07 4.25 3.20 65.00 2.83 0.92 2.80 75.60 0.00 2.15 21.92 1.87 2001 4.20 4.38 3.20 65.60 2.85 0.92 2.80 75.60 0.00 2.19 22.33 1.87 2no? 4.33 4.92 3.20 66.21 2.88 0.92 2.R0 75.60 0.00 2.23 22.75 1.87 a/ fuel price-. ire ha-.ed on "1hallind: Impact of loIwer and llncerfain ()I I'riIc- on Inergy Investment" Pre-mission I)isruw,,ion Paper. b/ Bdsed on Mitie f(mlpneer inq Oepdrlmielt memordudlum ddfed Jdfludciy 23. 1987 rird excluded royaIty of 15 Bahl/ton. ci At5.umncd pr icc i . ron,ltant at 70 HEiht,MHIU. w c d/ INS$1 =1.0) KIOO. * 0 T,.t?lc 2: fill I PW'iI ', 10 ] IN 114 RIlN( I 1IS I ;V1 XI'ANSION IPtN UNI)IR SQ SCINAR1IO (At Const-ini 1986 priceS) Year fijcl Oil Prices d/ Imrputed Cod? Price d, Naturdl Gas Price a! Lignite Price b/ IlS¶/Mi'lIII ll.'t tre % Inrrei-.e t ,S$ 'Tlor I4/Mill) I increlr s llS¶i/W1l11) II/MOTlI % Incrrase tlS$/M131iJ 11SS/Ton % Increa.e 1986 2.11 2.21 41.00 1.91 - 2.10 56.70 c/ - 1.82 18.58 - 198/ 15. I s I.P1 t.9o I.q* 1.12 2.10 56.70 c/ 0.0( 1.83 18.70 0.64 1988 2.20 2.29 1.85 15.00 1.96 1.11 2.10 96.70 0.00 1.85 18.83 0.69 1989 2.21 2.i4 1.8'8 '4'.,t 1.*8 1.1(0 2.10 56.70 0.00 1.86 18.97 0.74 T9(n i>2.2l ;'.S8 I YQ -If.0(l 0 Z. 1.09 2.10 56.70 0.00 1.88 19.13 0.84 1991 2.3t 2.43 .I.(0 17.60 2.07 3.36 2.20 59.40 4.55 1.89 19.31 0.93 1 1992 2.3 2.48 <.Uh 19.2(1 2.14 5.29 2.20 59.40 0.00 1.91 19.5(1 (0.97 crE 1993 2.42 2.9; 2.01 n0.80 2.21 3.15 2.20 59.40 0.00 1.93 19.72 1.12 1994 2.47 2.58 1.97 S2.40 2.28 3.05 2.20 59.40 0,00 1.96 19.95 1.15 199) 2.52 2.63 1.94 S4.(1( 2.i 2.9fi 2.20 59.40 0.00 1.98 ?0.20 1.24 1996 2.0 2.72 3.13 541.60 2.37 1.10 2.30 62.10 4.35 2.01 20.49 1.42 1997 2.68 2.80 3.05 95.U0 2.40 *If)9 2.30 62.10 0.00 2.04 20.80 1.49 1998 i2.70 2.89 2.94 95.80 2.45 I.(8 2.30 62.10 0.00 2.07 21.14 1.61 1999 2.88 3.01 3.95 56.40 2.45 1.06 2.40 64.80 4.17 2.11 21.51 1.72 2000 2.95 3.06 1.67 57.00 2.48 1.0S 2.40 64.80 0.00 2.15 21.92 1.87 2001 2.98 3.11 1.67 57.60 2.50 1.05 2.40 64.80 0.00 2.19 22.33 1.87 2002 3.03 3.16 1.67 58.21 2.53 1.05 2.40 64.80 0.00 2.23 22.75 1.87 d/ Fuel pr ice, are based on "Tdld ldnd: InTdaci of Lower and Uncertain 0 il Pr ices on Enerqy Invesiment", Pre-mission Discussion Paper. b/ Hawed on Mine I ngincerlng Departmcnt mcmor-3ndum dated January 23, 1987 and excluded royalty of 15 Baht/ton. c/ Ass-imerd li ice is 4 onstant at 70 Rat, 1MlAll. I OD d/ IJSSI -BlRt 27.0. 0 _ W~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- .: 67 - Annex L& Page 1 of 8 RESULTS FOR TOM SCENARIO Definitions 1. According to the cases structured in Annex 2, the following PDPs were defined: Case Decision Base Adiustment Year TOM-0 Optimal a priori 1987 TOM-1 February 1987 PDP 1989 TOM-2 July 1986 PDP (adjusted) 1989 TOM-3 Bang Pakong in 1993 1989 TOM-4 Early Development of CC 1989 Results 2. The PDPs corresponding to the different cases are summarized in Table 1. Qualitatively they show the following characteristics: (a) In all cases there is a development of gas-based combined cycle units, with installation of thermal units (starting by the Mae Moh lignite plants) delayed until the late 1990s. (b) The least cost PDP corresponds to case TOM-2; the Nam Chon hydro plant is the project that makes this case least cost, subject to the environmental debate surrounding it. (c) Of those PDPs not involving Nam Chon, the next best plan nearest TOM-O (the latter being the optimal a priori PDP without Nam Chon) corresponds to TOM-1 and TOM-4, i.e., the decisions of the base case PDP and the case of accelerated gas develcpment. 3. The objective functions associated to the TOM cases show the following values in $m: TOM-0 TOM-1 TOM-2 TOM-3 TOM-4 Fixed Costs 966 990 941 1,074 1,049 Variable Costs 3,480 3,488 3,383 3,493 3,463 Total 4,446 4,478 4,324 4,567 4,512 4. The regrets associated to the TOM cases, relative to TOM-0, are: TOM-1 TOM-2 TOM-3 TOM-4 Regret (Sm) 32 -122 121 66 Z of TOM-0 0.7 2.7 2.7 1.5 - 68 - Annex 4 Page 2 of 8 Investments 5. Tables 2 to 6 show the investment programs associated to the different PDPs. Total investments for TOM are summarized below ($m): TOM-O TOM-1 TOM-2 TOM-3 TOM-4 Total 1987-2001 2,567 2,567 2,578 2,657 2,657 NPV (1987) 892 906 977 991 949 Total 1987-1994 760 753 987 947 766 The total investment values are approximately the same in all cases but they differ substantially in relation to their timing, especially in the short-term (1987-94). Decision Agendas 6. The resulting decision schedules under TOM are summarized in the following table: tOM Decis-on Agendas sear TON4-O TOA-2 TOM-3 TOM-4 '9e7 CC '5O Bang Pakong 3 GI 1-2 R3 Lign'te CC 1-2 Nam Chon '988 Sr ngarlnd 5 CCi CCI5O GT 1-2 GT 1-2 CCI (300) '989 CC1 Srinag3r rd 5 CC2 Sr-nagarnd 5 GT1 Srnaga-na 5 '990 CC2 CC2 1991 CC3 CC3 Cc1 CC3 GT 2 Kaeng Krung Kaeng Krung Kaeng Krung I 1,IeI: COMPARATI Vi ppr C,AS( STUDIFS (TOM Scenr,rrio) i'~cdf Cd'e IOM-O Cad,.cL TOM- I Ca5e TOM-2 Cdse TOM-3 Case TOM-4 year Project MW P'ropeCt MW Plroject MIW PIrojeci MW P'roject mw 1991 - -- CC 150 #1 150 GI 6 *1-2 2x100 R3 (Liqntte #1 75 Sr irigar i id #5 100 1992 - - (,P it-I 2x!0O CC 150 #2 150 CT #1-3 2xlOO CC 300 #1-2 2x300 Sr I ii ~~~~~180 Sr-i nagar i md #5 190 1993 Srin,iijrind #5 180 cc 3ou ' 1 300 CC 300 *1 300 Bang Pakong #3 600 - 1994 CC 300 *1 300 -- CC 300 #2 300 - 19011 G I2 #(0 ((SI 2'O ' '- ~ 5 180 cc 50( #2 30') I996 Ur VI(( #5 itlo UC 501 300 Nam rThon 580 CC 300 #1 300 CC 0oO 13 300 1997 C(, 300 #4-5 ?ir3(1 Kaernq Kroriq #11-2 2x40 Kaeng Krang #1-2 2x40 Kaenq Krang #1-2 ?x40 0 CC 300 #1 300 CC 300 #2 300 CC 300 #4 300 1998 CC 500 46 300 CC 300 R5-6 ?x300 CC 300 *3-4 ?x300 CC 300 #3-4 ?x300 CC 300 #5-6 ?x300 KOLrn1 Kraun'j #1-? ?x40 1999 Make M(rll #10 300 Mde Moh #10 300 CC 300 WS 300 CC 300 #5 300 Mrie Moth #10 300 2000 Mae Moh, #11 12 2x300 Mae MWh #11- 12 ?K50(1 Mae Mo)h #10-Il 2xSOU R3 l ignite #1 75 Mae Moh #11-12 2x300 R3 I .ri,i le #1 75 R3 liquiite #1 75 CC 300 #6 300 R3 i (n ite #1 75 Mae Moh #10 300 21001 l3Ranq Pakonq' #3 600 8rar'.l Pdkonrl #5 6100 It,inq P'ikonq *3 600 Mae Mobi #11-1? 2x300 80ng P'dkorug #3 600 Capacity added from 1991-2(4101 (MW) 3,855 5,839 3,835 3,835 3,835 -Objective fun,t-on i'. tthe sum of di'.c",intea wrnu-i ftxed co',It- and variable costs over the 1991-2001 planning period. Discount rate 12%. fixe~d (o',t', fo- r (liven power plaint orre comrpr',.d~( nf the- dfnnijily corres,ponding to capital investment and fixed annual operating aind mainfen1irr,. c ,, -Variab!e coy,t', are compio.cd of fluel o' in the olijcrtive tunction. Variable operating and maninenance costs are negligible compared0 lo, the I it ter * (US$ millioni) I I ixtOd qhb) 941 1,074 1 ,049 -Vair li-te 3,480 3,4188 ,8 3,493 3,463 - bliP 4,446 4,4178 4,324 4,567 4,512 labie. ': CASE IOM-O INVtSTIMtNIS (In 1986 S mnIIior.) 1Q87 1988 1989 1990 1iVJ1 1992 1993 1904 1995 1996 1997 1998 1999 2000 2001 Total GTl 0.16 5.4 23 3.56 32.12 Sr ?nala r in d 7.h) 11 .1 7.2 25.85 CC' 2.2 20,9 73.8 51.1 10.2 158.17 CI2 n0.16 r,.4 23 3.56 32.1? CC2 2,2 20.9 74.0 59.3 28.9 185.22 t i 2.? 20,q 7,.0 )9r; 28.9 185.22 N)1tFl 2.2 20.9 74.0 53,3 17.9 168.24 1 NIrT2 2.2 20.9 74.8 57.0 21.6 176.45 ° K.menq Krunil I 2 4,2 3,? 10,9 14.6 19,9 13,3 66.63 Mde Moli 10 3.2 34.3 59.4 62.4 61.3 16.8 23P.39 Mae Moh I1 3.2 34.3 60.0 58.8 51.4 14.8 222.51 8(inq Pakortiq 3 4.6 47,1 177.2 182.1 63.0 46.9 520,81 Ri l ignite 1 26.5 50.2 14,5 8,5 79.5 Mac Moh 12 3.2 34.3 54.7 47,2 31.3 11.1 187.88 cr6 2.2 70,9 74,0 59,3 28,9 185.22 livlifnc in 21 78 13 114.7 ornpr eh,~,(i Ir ?2 56 10 79 lotdIl (1,1) 50,0 0() 6,0 h9.1 265,3 196.6 222,6 296.9 550,0 38716 416.9 301,8 97.4 46,9 2657,0 4X NPV (1987) 892.1057 liible 3: IOM-1 Investments (In 1986 i mil 1987 1988 1989 1990 1991 199? 1Q93 1994 1995 1996 1997 1998 1999 2000 2001 Total GIl 0.16 5.4 23 3.56 32.12 Srinapgr ird 5 1.6 t1 I 7.2 25.85 Ti 2.? 20.9 73.8 51.1 10.2 158.17 CI? 0.1f ".J 2' 3.56 3?.12 CC? 2.2 20.9 74.0 59.3 28.9 185.22 CC3 2.2 20.9 74,0 59.3 28.9 185.?? NPCCI 2.2 20.9 14.0 53.3 17.9 168.24 ' 1 IfC2 2.2 20.9 74.8 57.7 ?1.6 176.45 K loldil 0.0 47.4 96.9 155.8 42.6 455.4 420.0 571.6 591.1 707.1 742.5 687.5 485.4 153.8 69.2 5526.3 U z NPV (1987) 2,038 -n 0o Idble 6: SQ-3 INVISIMFNNS 1987 1988 19g3 1990 1991 1992 199 1°W94 1995 1996 1997 1998 1999 2000 2C01 Total GI 3xt00 0.9 16.? 69.0 10.7 96.36 Srindqarind 5 7.6 11.1 7.2 25.85 CC I 2.2 20.9 15.8 il. 10.2 158.17 G 4 0.2 5.4 23 3.6 32.12 2 2.? 20.9 74.0 59.3 28.9 185.22 (,I '-ti h* ) 1C.8 46.() 7.1 64.24 (31 0.2 5.4 23.0 3.6 32.12 141nC( 2.2 20.9 74.0 53.3 17.9 168.24 NiTU', 2.2 20.9 74.8 57.0 21.6 176.4; k detl1 rr ulni 1-? 4.2 3.7 10.9 14.6 19.9 13.3 66.63 M,,c Mon 10 3.? 34.3 59.4 62.4 61.3 16.8 257.39 Md- M,,)t I 1 3.2 54.3 60.0 58.8 51.4 14.8 222.51 '1dlIq r'dkQOfq 3 4.6 47.1 177.2 182.1 63.0 46.9 520.81 Q g I In te 10 A(. PhIta 1 4.7 47.6 180.9 187.4 61.0 38.8 520.46 Mae Moht 12 3.2 34.3 54.7 47.2 37.3 11.1 187.88 Ao lhai 2 3.58 44.5 171.81 184.54 65.58 30.39 500.4 Ao PhIdi 3 3.58 44.5 170.9 175.97 54.43 29.33 478.71 Ao Phat 4 3.58 44.5 170.9 175.97 53.67 27.22 475.84 Ao I hai 5 4.68 47.6 180.91 187.43 61.04 38.8 520.46 Ao Phai 6 3.58 44.5 171.81 184.54 65.58 30.39 500.4 R3 Thermal I 1.36 21.48 28.01 43.43 44.48 11.86 150.62 l i peI ite 2.7 21 78 13 114.7 lotal 0.0 5.1 64.0 219.7 348.4 355.1 503.4 624.7 667.6 638.8 770.8 675.6 340.4 131.6 30.4 5435.6 Id NVP (1987) 2036.877 X Ou 0 -75 - Annex 5 Page 1 of 8 RESULTS FOR SQ SCENARIO Definitions 1. According to the cases structured in Annex 2, the following PDPs were defined: Case Decision Base Adiustment Year SQ-0 Optimal a priori 1987 SQ-1 February 1987 PDP 1989 SQ-2 July 1986 PDP (adjusted) 1989 SQ-3 Bang Pakong in 1993 1989 SQ-4 Early Development of CC 1989 Results 2. The PDPs corresponding to the different cases are summarized in Table 1. Qualitatively they show the following characteristics: (a) An initial, rapid development of combined cycle units (4x300 MW of which 2 unics would be located at Nam Phong). (b) The Mae Moh 10-11 lignite plants are scheduled for 1995 at the earliest arnd 1996 at the latest. This calls for decisions to be taken -'I 1989 relative to the final go-ahead for these projects. (c) Unless forced, as in case SQ-3, Bang Pakong 3 is in service either in 1995 or 1996. In case SQ-3, an early commissioning date for Bang Pakong does not benefit the system as it diqplaces the first two CC units. 3. The objective functions associated to the SQ cases show the following values in Sm: SQ-0 SQ-1 SQ2 SQ-3 SQ-4 Fixed Costs 1,768 1,762 1,628 1,800 1,796 Variable Costs 4,020 4,077 4,020 4,060 4,007 Total 5,788 5,839 5,648 5,860 5,803 4. The regrets associated to the SQ cases, relative to SQ-0, are: SQ-1 SQ-2 SQ-3 SQ-4 Regret ($m) 51 -140 72 15 % of TOM-0 0.9 -2.4 1.2 0.3 - 76 - Annex 5 Page 2 of 8 Investments 5. Tables 2 to 6 show the investment programs associated to the different PDPs. Total investments for SQ are summarized below ($m): SQ-0 SQ-1 SQ-2 SQ-3 SQ-4 Total 1987-2001 5,483 5,510 5,526 5,436 5,494 NPV (1987) 2,023 2,011 2,038 2,037 2,043 Total 1987-1994 2,250 2,278 2,090 2,180 2,281 Investments for this scenario do not differ in NPV by more than 3% between extremes and less than 1% if the Nam Chon project is not included. Decision Agendas 6. The following table summarizes the decision schedules for the SQ cases: Taee 1: Decision Schedules for SQ Cases Year SQ-0 SQ-1 SQ-2 SQ-3 SQ-4 1987 GT 1-3 CC 150 Bang Pakong 3 GT 1-2 Cc I R3 Lignite CC 1 Nam Chon CC 2 1988 Srinagarind 5 GT1-2 Srinaga-ind 5 GT 1-2 GT4 CC 1 CC 150 CC 2 CC 1 1989 GT 5-6 Sringarind 2 GT 4-6 CC I GT 3-6 CC 3, CC 4 GT 3-9 CC 2 CC 2 CC 3, CC 4 Mae Moh 10 CC 2, CC 3 Mae Moh 10 Srinagar,nd 5 Mae Moh 10 Mae Moh 11 Mae Moh 10 Mae Moh 11 GT 3-7 Mae Moh 11 Kaeng Kr_ng Kaeng Krung 1990 Bang Pakong 3 CC4 CC 3, CC 4 Srinagarind 5 HVDC R1-R3 Mae Moh 10 Bang Pakong 3 Bang Pakong 3 Mae Moh 11 1991 Ao Phat I Ao Phai 1 HvDC RI-R3 Ao Phai I Ao Phai 1 Bang Pakong 3 Kaeng Krung Mae Mon 12 ioble 2: COMPARATIVE POP CASE STUDIES (SQ Scenario) I sc-ri SQO ~~~~~~~~~~SQ-i SQ-2 SO-3 S- ye~ar ProJeLlt MW Project MW Project MW Project r4IW Pr-oject MW 19 -- CI 3- 3x100 GI 1-3 30100 CC 150 1 150 GI 1-3 3x1OO Cl 1-3 3x100 R3 Lignite I 75 Srinagarind 5 180 iT). - S~r i nrc).r i nt 5 18 I474x100 CC 150 2 150 GI 4-7 3x100 CC 300 1-2 2x40 CC it0'l 300 Srinagarind 5 180 CT 100 1-3 3x100 Srinagarind #5 180 or 4 100 1Q9~~~ I 3(111 2 300~~i( CC. 300 1 500 CC 300 1 300 tlang Pakong #3 600 GI 4-6 3x100 r. ~~~~~2XI(O GI 8 -9 ?xIOO l flOT) 4-6 5Xl00 Srinagarind 5 180 '9941 F( (3f I 4 ?2300 CC. 300 2 300 CC 300 2-3 2x300 CC 300 1-2 2~300 CC' 300 3-4 2x300 I IVI)RI 0 -35 1 )(MW (1 50) ¶,rrr ru"q 1-? 2x40 CC 300 4 300 Mde Mohi 10-Il ?x300 CC 300 3-4 2x300 Mde Mott 10-11 2x300 Wi,, Mr,I 10-1 ?I x300 Mae Muli 10-11 ?x300 [WOOC R1-3 150 MW (150) Kdenq Krurrg 1-2 2x'00 99bpt lirrrq ('itrorrg 3 600 Banrg Pakong 3 600 Nam Chon 580 Mae Mob 10-11 2x300 [Ring Pakong 3 600 HVDC R1-R3 150MW (150) R3 Lignite 1 9497- V 03i1 '" I IrrIte I 7 5 Ao I'hai I 600 Mae Moh 1? 300 Ao Pha; 1 600 Ao Phai 1 600 Ao I 'har i 600 gang Pakong 3 600 Kaeng Krrrng 1-2 ?x40 HIVIC RT-3 150 MW (150) 1990 M,ire MuIl 1? 300 M,ai MoIt 12 300 Au Phai 1 600 Mico Moh 12 300 Mdte Motb 12 300 Aro Pai',r 2 600 Art I'hiai 2 600 Ao Phai 2 60(1 Au I'thai 2 600 13 therniill I ISO 19Q9 Ao Ph' Ir 1 600 Au Plhdi 3 600 Au, Plhd 2-3 2x600 Ao Plhai 3 600 Ao Pl'-r - 600 R3 Iherinal 1 600 2000 AoII 4 '15 2x600 Kaenq Krung 1-2 2x40 Ao Plha~ 4 600 Ao Phai 4-5 2x600 Ao f1ha '14 5 2x40 Au Plldi 4 600 R3 lhiermd'l 150 2-001 13 'Ilehcrmdl 1 150 Ao I'hai 5-6 ?x600 Ao 'ha' 5- 2x(600 Ao Pha i 6 600 60o Plha' 60 Au Pitt ri 6 600> Cahpor t y addel. f ( trom 19r() ~')00 (MW) /,385 7,610 1,155 1,111 7,580 w> OLuJef 1v I r*. upl it 2001cr/ 5,788 5,872? I I xed ~1,168 1,762 1,628 1,800 1,049 0 yar.o 'Iric 41,020 4,077 4,020 4,060 5,163 - 'yIn ~~~~~~5,788 5.839 5,648 5,860 4.512 1 (If dt l'~cr Ipt ion 'wer Annex 4, pg. 3. rfb I e 3: SQ-(0 INVI SIM[ NIS (In 1986 $ in' I I oil 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 199R 1999 ;000 2001 Totdl ix. l() 0.5 16.2 69.0 10.7 96.36 '" Id' . l sl t .l 7.6 11.1 7.2 25.85 II ' 2.2 20.9 73.8 51.1 10.? 158.17 a 0.2 5.4 23.0 3.6 32.12 2.2 20.9 74.0 59.3 28.9 185.22 > w 0.3 10.8 46.0 7.1 64.24 2.2 20.9 74.0 53.3 17.9 168.24 NPI.I 2.2 20.9 74.8 57.0 21.6 1 16.45 Kate,1 Nr '',nql I 2 4.2 3.7 10.9 14.6 19.9 15.3 66.65 t 0.1t Mtil, II) 3.2 34.3 59.4 62.4 61.3 16.8 237.39 t M,ae* Molb 11 5.2 34.3 60.0 58.8 51.4 14.8 222.51 Hoinl. Ii ',j1 i 4.6 47.1 177.2 182.1 63.0 46.9 520.81 R I tiqni t: 1 26.5 30.2 14.3 8.5 79.5 Ao 1'h I 1 4.7 41.6 180.9 187.4 61.0 38.8 520.56 Mde MoI 12 3.2 54.3 'J1.7 47.? 37.3 11.1 187.88 Ao Pthd 2 3.58 44.5 171.81 184.54 65.58 30.39 500.4 Ao 1(1ha i i. 8 44.5 170.9 175.97 54.43 29.53 478.71 Au Pl 1 4 3.58 44.5 170.9 175.97 53.67 27.22 475.84 AC' Ow, > 4.68 47.6 180.91 187.43 tI.0(A 8.8 520.46 A I P'ita 6 3.58 44.5 1)11 *II I14.4 4; 'v# 5(0.59 500.4 R3 TllhenidI 1 1.36 21.48 28.01 43.J3 44.18 11.86 150.62 ('Ipel III' 2.7 21 78 13 114.7 loltifi 0.0 2.7 39.4 192.? 299.4 459.7 534.7 721.9 617.8 621.9 744.0 659.1 372.0 176.1 42.3 5483.0 > Nl'V 1198)1 2, S23.515 D t _____________________________________________ ______________________________ 1-~~~~~~~~~~~~~X Table 4: SQ-1 INV(SIMINTS (In 1986 S millions) 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 lotal Gli 3x100 0.5 16.2 69.0 10.7 96.36 Sr ' naq,dr i nml 1 1.6 11.1 1.2 25.85 2.2 20.9 73.8 51.1 10.2 158.17 ,;r t 0.2 5.4 23.0 3.6 32.12 /.2 2(0.9 /4.0 S9.3 28.9 185.22 0. 1A.0 46.0 7.1 64.24 0.2 5.'1 23.0 3.6 32.12 4i 9 0.. 8IM 46.,) 1.1 64.24 NIT11 '?2.2 20.9 74.8 55.3 17.9 168.24 1 It'.( .1 '2.2 20.9 74.8 57.0 21.6 176.45 I VO)C 0.6 S 4.5 2.5 10.6 Kdeti'4 Kr uflt 1-2 4.2 3.7 10.9 14.6 19.9 13,3 66.63 Mae Molh Il) 5.2 54.3 59.4 b2.4 h1.5 Ib.8 251.39 MdeW M0lI 11 3.2 34.3 60.0 5U.8 51,4 14.8 222.51 HsR,q P1akomtj 3 4.6 47.1 177.2 18. 63.0 46.9 520.81 <3 Ii ijg I tto I 0 A(, P'I1,ji 1 4.7 47.6 180.9 187.4 61.0 38.8 520.46 M,e Mot 12 3.2 34.5 54.1 47.2 37.3 11.1 187.88 AO l ha-i i7 23.78 44.5 I 11.81 184.54 65.58 30.39 500.4 Au Plidi 3 3.58 44.5 170.9 175.97 54.43 29.33 478.71 Ao Phai 4 5.58 44.5 170.9 175.97 55.61 27.22 475.84 Ao Plrhi S 4.68 47.6 180.91 187.43 61.04 38.8 520.46 Ao Phdi 6 3.58 44.5 171.81 184.54 65.58 30.39 500.4 > R5 Ihermal 1 1.36 21.48 28.01 43.43 44.48 11.86 150.62 3 P i peIife 2.7 21 78 13 114.7 D X Total 0.0 U0, 19.0 1 50.2 31b.1 404.3 584.3 163.8 623.4 610.5 636.1 651.1 4/4.8 161.2 69.2 5510.4 U NP V (1987) 2,011.052 O? Tdble 5: SQ-2 INVESIMLNIS 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Total CC 1'0 #1 12.9 37 22.6 6.5 79.0 PS I iqn. 1 26.5 30.2 14.5 8.5 79,5 1,r n .1,i r ,,,I I ^. 0 1 1.1 7,2 25.9 #2 12.9 S6.9 22.5 6 78.3 0 t)(1 0)5 Ib,.2 69.0 10.7 96.4 4 1 0.2 5.4 23.0 3.6 32.1 0.3 10.8 46.0 7.1 t4.2 ?.2 20.9 74.(0 S9.5 28.9 185.2 *j'J 'T w2.2 20.9 74.0 53.3 17.9 168.2 Nl" ' .' 2.2 20.9 74.8 5/.0 21.6 176.5 1 M,i,. M0h \to 3.2 34.3 59.4 62.4 61.3 16.8 237.4 co Maie Muh 11 3.2 34.3 60.0 58.8 51.4 14.8 222.5 1'viIt RIi N?$ 0.6 3 4.5 2.5 10.6 IIVOII 1N? R 0.6 3 4.5 2.9 10.6 Ndmi Chon 8 6.5 22.6 28.2 31.3 38.2 47.3 36.8 21.4 240.3 Hlang l'.kon.q i 4.6 47,1 177.2 182.1 65.( 46.9 520.8 M,e Mull 12 3.2 34.3 54.7 47.2 37.3 11.1 187.88 Ao Phai 1 4.7 47.6 180.9 187.4 01.0 38.8 520.46 Ao Ph,i' 2 3.58 44.5 1/1.81 IM4.54 6f5.8 50.59 500.4 Au Plcdi 3 3.58 44.5 170.9 175.97 54.43 29.53 478.71 Ao Plha i 4 5.S8 44.5 1 10.9 115.9) 55.67 21.22 475.84 A P'hI,)i 5 4.68 47.6 180.91 1871.43 61.04 38.8 520.46 Ao Phai 6 5.58 44.5 171.81 184.54 65.58 30.39 500.4 P' pci 2.7 21 78 13 114.7 Ir8dl 0.0 47.4 96.9 155.8 42.6 455.4 420.0 571.6 591.1 707.1 142.5 687.5 485.4 153.8 69.2 5526.3 x ?'JIV ( 19fF? 1 2*(6 a'n 0r) lable 6: SQ- i INVI SIMI NIS 198/ 1988 19i39 1990 1991 1992 1991 1994 1995 1996 199/ 1998 1999 2000 2001 Total 1,1 Sx1(0ii .S5 Ih.2 69.0 10.7 96.36 '. "'ila 1'1 7 .6 1 1. 7.2 25.85 2.2 20.9 73.8 51.1 10.2 158.17 . 1 0.2 5.4 25 S.6 32.12 (1 72.2 20.9 74.0 59.3 28.9 185.22 I -b 0.3 10.8 46.0 7.1 64.24 ! ().2 .4 23.0 5.6 32.1? NPC'1 2.2 20.9 74.0 53.3 17.9 168.24 Nil t l. 2.2 20.9 14.8 57.0 21.6 176.45 K,icm1 K 1 4.2 3.7 10.9 14.6 19.9 13.3 66.63 1 Mde Mott to 3.2 34.3 59.4 62.4 61.3 16.8 237.39 OD MKw Moh II 5.2 54. S 60.0 58.8 51.4 14.8 222.51 Banal Pokill 4.6 47.1 177.2 182.1 61.0 46.9 520.81 R3 1 iti Ic 1 0 Ao Pha i 1 4.1 41.6 180.9 181.4 61.0 58.8 5211.46 Mdw Mob 12 3.2 34.3 54.7 47.2 37.3 11.1 187.88 Ao 1hi' 2 1.58 44.5 171.81 184.54 65.58 30.39 500.4 Ao Pia ' i 5.58 44.5 1 (1. 175.91 54.43 29.53 478i.71 Ao Pci, 1 3.58 44.5 170.9 175.97 53.67 27.22 475.84 Ao Phai S 4.68 41.6 180.91 187.43 61.04 s8.8 5 0.46 Acy Pht i t 3.58 44.5 171.81 184.54 65.58 5().59 500.4 R3 Tilei nli 1 1.36 21.48 28.01 43.43 44.48 11.86 150.62 I ipcl nc 2.7 21 18 IS 114.1 TlIdl 0.0 5.1 61.0 279.7 348.4 355.1 503.4 624.7 667.6 638.8 770.8 675.6 340.4 1I1.6 50.4 5435.6 Id > NVIt (19871) 2%56.877 fD X ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Got ldble 7: SQ-4 INVFSIMiNTS 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Tolal C,T 3x100 0.5 16.? 69.0 10.7 96.36 Srinaqarind 5 7.6 11.1 7.2 25.85 C 1 2.2 20.9 73.8 51.1 U..2 158.17 or 4 0.2 5.4 23.0 3.6 32.12 i(. 2.2 20.9 74.0 59.3 28.9 185.22 1 f 0.3 10.8 46.0 7.1 64.24 Ni" f 2.2 20.9 74.0 53.3 17.9 168.24 Nsf1, I .. 2.2 20.9 74.8 57.0 21.6 176.45 w . . ; Kr 'i 4.2 3.1 10.9 14.6 19.9 13.3 66.63 I M(jl) 10 3.2 34.3 59.4 62.4 61.3 16.8 237.39 1 M,o Moll It 3.2 34.3 60.0 58.8 51.4 14.8 222.51 or Hdil llrikoi,,1 S 4.6 47.1 177.2 182.1 63.0 46.9 520.81 R3 tqnite 1 26.5 30.2 14.3 8.5 79.5 IVI)C 0.6 3 4.5 2.5 10.6 Au Plidi 1 4.7 47.6 180.9 187.4 61.0 38.8 520.46 Mdc Moh 1? 3.2 34.3 54.7 47.2 57.3 11.1 187.88 Ao l'tia.i 2 3.58 44.5 111.181 184.54 65.58 50.39 500.4 Ao Plldi 3 3.58 44.5 170.9 175.97 54.43 29.33 478.7i AO lhai: 4 5.56 44.5 I/(1.9 175.91 5.61 21.22 475.84 Ao Flh 'i 5 4.68 47.6 180.91 '87.4 1,1 .04 38.8 520.46 Ao Ptia 6 3.58 44.5 171.81 184.54 65.58 30.39 500.4 R3 Ihernn71 1 1.36 21.48 28.01 45.4S 44.48 11.86 150.6? P i pcIinc 2.7 21 78 13 114.7 iotal U.0 4.8 58.0 232.2 263.7 452.6 545.2 /26.2 606.3 640.7 144.5 674.5 315.1 145.5 30.4 5493.6 b > 0 : NPIV (7987) 2043.085 D r,D 8 63 _ Annex 6 PIPELINE AND COMPRESSOR INVESTMENTS 1. In the TOM scenario, large amounts of gas are channelled to the power sector that require: (i) the commissioning of the TP-Erawan pipeline; and (ii) a compressor in the existing pipeline. The costs of this equipment are not taken into account in the power subsector optimization process and the prupose of the following analysis is to show that these investments are indeed justified. 2. The volumes of gas associated to the TP-Erawan pipeline are: Year 1993 1994 1995 1996 1997 1998 1999 2000 MMCFD 150 170 200 230 270 390 320 280 3. Investment costs, assuming a commissioning in 1993 are ($m): Year 1990 1991 1992 1993 Pipeline 2.7 21 78 13 Compressor 1 12 56 10 Total 3.7 33 134 23 The total cost at commissioning with a 12% discount rate is $220m. 4. The marginal use of gas in TOM lies in the South Bengkok and Bang Pakong thermal power plants until year 2001. The benefits associated with it are the substitution of heavy oil (HFO). The net benefits corresponding to these volumes of gas, based on case TOM-1 are: Year 1993 1994 1995 1996 1997 1998 1999 2000 MMCFD l50.0 170.0 220.00 230.0 270.0 350.0 320.0 280.0 Gas cost (S/mBTU) 2.5 2.5 2.6 2.6 2.7 2.7 2.8 2.8 Total cost 129.3 146.6 179.4 206.3 251.5 326.0 309.1 270.4 HFO cost (S/mBTU) 3.2 3.3 3.4 3.6 3.7 3.8 3.9 4.1 Total HFO cost (Sm) 165.6 193.5 234.5 285.6 344.6 458.8 430.5 396.0 Gross benefit (Sm) 36.2 46.3 55.2 79.3 93.1 132.8 121.4 '25.6 Minus: Compressor O&M 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 Net 33.7 44.4 52.7 76.8 90.6 130.3 118.9 '23.1 Present value of benefits at 12.: $370m. Even for a limited time period, the power sector benefits far outweigh the costs (B/C ratio of 1.7) and the investment is justified. 5. In order to check the timing, the first year cost for the pipeline and compressor, based on $220m and a 20 year life, is $29m. With first year benefits of $34m, the timing is also justified. -84 - Annex 7 Page 1 of 2 MARGINAL GAS NETBACKS 1. The development of the ECAT system goes through two distinct phases: a "gas" phase in the medium term and a "coal" phase in the long- term. The availability of an incremental amount of gas has two possible effects on the power system: (a) in the short-term it can substitute for fuel oil; (b) in the long-term it can delay che commissioning of a coal plant through the installation of combined cycle gas-based units. 2. The marginal netback for gas is therefore a function of two values: (a) a coal-based parity involving an energy parity and a capacity credit due to the lower cost of combined cvcle units; and (b) a fuel-oil parity based solely on its substitution value. The marginal netback for gas is therefore the greater of (0) the coal-based parity discounted from the date the first coal unit is commissioned in an optimized plan or (ii) the fuel oil parity. 3. The coal-based parity has the following value: Capacitv credit based on $690/kW for CC units ($92/kW/year) and $1100/kW for coal units ($140/kW!year) together with an O&M credit ($14/kW/year for CC, S20/kW/year for coal). The net capacity credit is therefore $54/kW/year. Assuming a CC operation of 5300 hours/year at 8100 BTU/kWh, the capacity credit translates into $1.26/mBTU. Assuming that CC availability is 77% and coal-unit availability is 83%, the final capacity credit becomes $1.17/mBTU. The fuel credit is just the coal parity corrected by the heat rate advantage of CC units (8100 BTU/kW vs. 9000 BTU/kWh for coal), i.e., 9/8.1 times the fuel cost of coal. The same calculation applied to fuel oil provides the gas parity for short-term substitution. 4. Netback values based on coal parity: using SQ and TOM prices, the coal-based parity, which includes the caparity and fuel credits, has the following values: Year 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 SQ 3.4 3.5 3.5 3.6 3.7 3.8 3.8 3.9 3.8 3.9 3.9 3.9 TOM 3.6 3.7 3.8 3.9 4.0 4.2 4.2 4.2 4.2 4.3 4.3 4.3 - 85 - Annex 7 Page 2 of 2 5. Netback values based on HFO fuel parity: Year 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 SQ 2.28 2.33 2.37 2.42 2.47 2.52 2.60 2.68 2.76 2.88 2.93 2.98 TOM 2.76 2.89 3.02 3.16 ,.29 3.42 3.55 3.68 3.81 3.94 4.07 4.2 6. Marginal Gas Netback for SQ scenario: the first coal-based unit (Bang Pakong 3) is commissioned in 1996. The gas netback is obtained as follows: Year 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 200_ Coal-based value 3.8 3.8 3.9 3.9 3.9 4.0 Discounted 2.4 2.7 3.0 3.4 Fuel Oil 2.3 2.3 2.4 2.4 2.5 2.5 Marginal Netback 2.3 2.3 2.4 2.7 3.0 3.4 3.8 3.8 3.9 3.9 3.9 4.0 7. Marginal Gas Netback for TOM scenario: Year 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Coal-based value 4.3 Discounted 3.8 Fuel Oil 2.8 2.9 3.0 3.2 3.3 3.4 3.6 3.7 3.8 3.9 4.1 4.2 Marginal NetDack 2.8 2.9 3.0 3.2 3.3 3.4 3.6 3.7 3.8 3.9 4.1 4.3 8. The netback values of gas in 1990 and 1987 are: 1990 1987 TOM: $3.1/mBTU $2.9/mBTU SQ: $3.0/mBTU $2.7/mBTU These values have been obtained by discounting marginal netback values at a 12% interest rate. The marginal netback values have been weighted by the gas used for power generation in each year. -86 - Annex 8 GEOLOGICAL AND ASSOCIATED PRODUCTION INFORMATION UNION AREAS I,II,III TP3D ESSO Distance from base (km) 260 225 35 Water depth (m) 60-70 60-70 Onshore Drill depth (m) 3000 3000 3000 Av. pay thickness (m) 44 35 8+ Prod. interval (m) 1220-3050 1220-2500 2900 Porosity (Z) 20 20 4-15 Top abnormal pressure (m) 2600 n/a .-00 BHP at TD (psi) 5800 4300 6500 Av. prod/well (MMCFD) 4.5 4;5 20 Conden3atc Production 225 90 nil barrels/day barrels/day Gas composition 70% methane 65% methane 96% methane 16% C2 11.75% C2 2% C2 + + 4+ 14% C02 22.5% C02 2% C02 ~~~~~+ Tr N2 Tr H2S - 87 - Annex 9 NETBACK VALUES (NPV in 1987) a/ $/mBTU _gQ TOM Power 2.7 2.9 LPG 2 7 Cement 2.9 3.0 Fertilizer 2.8 3.1 Source: Mission estimates. a! Netback value for power derived in Annex 7. For cement and fertilizers, it is based on an analysis of fuel substitution possibili- ties. LPG values based on import prices of $180/ton for SQ and $220/ton for TOM. _8 - _Annex 10 Page 1 of 3 GAS AVAILABILITY SQ (High Reserves) MMCFD …-------------- Offshore -------------- -- Onshore -- U1+U2 U3 3DTP 20TP Other Total ESSO Total 1987 450 450 450 1988 500 500 500 1989 500 500 500 1990 500 500 500 1991 430 120 500 500 1992 410 140 550 550 1993 390 160 150 700 125 825 1994 320 230 200 750 200 950 1995 220 330 200 750 250 1,000 1996 160 390 200 750 300 1,050 1997 120 430 200 750 300 1,050 1998 90 380 280 750 300 1,050 1999 70 360 320 750 300 1,050 2000 60 270 420 750 300 1,J50 2001 50 180 520 750 300 1,050 2002 40 140 430 140 750 300 1,050 2003 30 100 380 240 750 300 1,050 2004 20 80 320 330 750 260 1,010 2005 15 60 220 455 750 225 975 2006 50 160 400 610 195 805 2007 40 120 380 540 170 710 2008 30 90 350 470 150 620 2009 20 70 280 370 130 500 2010 15 50 200 265 110 375 RESERVES (TCF) Total 2.0 1.3 1.6 1.3 6.2 1.7 7.9 Remaining in 2010 0.0 0.0 0.0 0.3 0.3 0.2 0.5 - 89 - Annex 10 Page 2 of 3 CAS AVAILABILITY SQ (Low Reserves) 4MF,FD -- Offshore -------------- -- Onshore -- U1+U2 U3 3DTP 2OTP Other Total ESSO Total 1987 400 450 450 1988 450 500 500 1989 450 500 500 1990 500 500 500 1991 450 50 500 500 1992 385 115 550 550 1993 325 175 700 125 825 1994 275 225 750 200 950 1995 195 250 55 750 250 1,000 1996 135 250 115 750 300 1,050 1997 95 250 155 750 300 1,050 1998 45 215 240 750 300 1,050 1999 30 185 250 750 300 1,050 2000 20 160 250 750 300 1,050 2001 15 110 250 750 300 1,050 2002 80 250 750 300 1,050 2003 60 250 750 300 1,050 2004 40 250 750 260 1,010 2005 30 250 750 225 975 2006 20 210 610 195 805 2007 15 170 540 170 710 2008 140 470 150 620 2009 120 370 130 500 2010 100 100 70 170 RESERVES (TCF) Total 1.8 0.8 1.2 3.8 0.5 4.3 Remaining in 2010 0.0 0.0 0.1 0.1 0.0 0.1 90 - Annex 10 Page 3 of 3 GAS AVAILABILITY TOM (Low Reserves) MMCFD --------- Offshore -------------- -- Onshore -- U1+U2 U3 3DTP 2OTP Other Total ESSO Total 1987 450 450 450 1988 500 500 500 1989 500 500 500 1990 500 500 500 1991 430 120 550 550 19;'2 410 140 550 550 15,93 390 160 150 700 70 770 1994 320 230 200 750 120 870 1995 220 330 200 750 120 870 1996 160 390 200 750 120 870 1997 120 430 240 750 120 870 1998 90 340 320 750 120 870 1999 70 240 440 750 120 870 2000 60 180 51d /50 120 870 2001 50 130 440 130 750 120 870 2002 40 90 390 230 750 120 870 2003 30 70 330 320 750 120 870 2004 20 60 230 440 750 120 870 2005 15 50 170 350 165 750 120 870 2006 40 120 300 290 750 120 870 2007 30 80 230 430 750 120 870 2008 20 60 170 520 750 120 870 2009 15 50 120 550 735 100 835 2010 40 90 550 680 80 760 RESERVES (TCF) Total 2.0 1.1 1.5 1 2 7.6 0.9 8.5 Remaining in 2010 0.0 0.0 0.0 0.1 1.1 1.2 0.2 1.4 -91 - Annex 1 INDICATIVE COST OF PRODUCTION FROM NEW CONTRACT AREAS (5/mBTU) UNOCAL III T-P ESSO SQ Scenario 1.2 1.8 1.9 TOM Scenario 1.6 2.0 1.2 Calculated as NPV cost of Production/NPV of Gas Cost of production includes exploration, ap- praisal, and production expenditures. It en- cludes royalty and income tax payments to the RTG. - 92 - Annex 12 Page 1 of 2 DECISION AGENDA 1: Evaluation of Alternative Strategies Strategy 1 (a) TOM scenario Compressor is installed late; compressor costs saved but production lowered or lost, from offshore fields during early 90s. Economic Prodn. Losses Netback Gas Value Compressor (MMCFD) (M4MCFD) value price of lost Costs ($/mBTU) ($/mBTU) sales a/ (Sm) ($m) 1990 573 23 2.8 2.3 3.9 0.82 1991 633 83 2.9 2.4 14.1 11.73 1992 673 123 3.0 2.4 25.1 53.60 1993 773 223 3.2 2.5 53.1 9.47 a/ Estimated as losses times 365 (number of days) times the difference between netback value and price paid to producers divided by 1.076 (the calorific value of the gas in mBTU/MMCFD). Present value of lost sales $47.2m Savings from delay in compressor (NPV) $15.6m Total regret (NPV) $31.6m (b) SQ Scenario Regret for uneconomic size of LPG Plant = $4.9m STRATEGY 2 As in Strategy 1 but with compressor installed by 1991 (a) TOM No regret (b) SQ Regret for compressor capital costs + oversize LPG plant. NPV (87) = $43.2m + $4.9m = $48.lm - 93 - Annex 12 Page 2 of 2 STRATEGY 3 (a) TOM Regret Advancing TP Erawan pipeline: Original Schedule Strategy 3 87 2.57 88 20.17 89 74.85 90 12.91 91 92 93 94 95 2.57 96 20.17 97 74.85 98 12.91 99 PV(87) $32.25m $79.83m Regret =$47.6m (PV of advancing capital expenditure unneces- sarily). (b) SQ Regret Advancing Erawan Pipeline by 2 years 79.8 - 63.6 = $16.2m Unnecessary compressor = 43.2m Oversized LPG plant = 4.9m Total Regret 64.3m 9 94 - Annex 13 Page 1 of 5 SCENARIO IMPLICATIONS FOR THE GAS SUBSECTOR 1. As in the case of power, the development of the gas subsector differs significantly between the two scenarios: under TOM there is a large surplus of gas, and some investments can be delayed since demand is tempered by higher prices. Under SQ there is a severe gas shortage due to a combination of strong demand and lower reserves, which are deemed to be price-sensitive. Available supplies would have to be subject to price-rationing. Demand projection 2. The demand for gas is a derived demand and is dependent upon the general level of economic activity and the price of gas vis-a-vis alternate fuels. The main sources of demand for gas arise ac present from power, LPG extraction, and process heat for cement production. .aDie 1: Demand for Gas in 1986 -,se MMCFD Power 237 .PG 61 Cement 10 Total 308 3. In projecting demand, the main differences between the two scenarios are their associated rates of economic growth and the level of oil prices. These two factors impinge directly upon the demand for gas leading to a considerable variation in gas use under the two scenarios. The demand for gas in the power sector has been derived, as already discussed in Chapter III, from the least cost development programs for the two scenarios. The input prices for gas, assumed while deriving the least cost power development program, are given in Table 2. These price projections are equivalent to the economic cost of natural gas to the country and correspond to the wellhead price paid, as per existing contract, to the foreign operating companies that extract it, less royalties and taxes. It has been assumed that future prices on new contracts will be the same as those based on the existing contract with Unocal. - 95 - Annex '3 Page 2 of 5 Table 2: Natural Gas Prices a/ (1986 S/mBTU) Year TOM SQ 1986 2.1 2.1 1990 2.3 2.1 1995 2.6 2.2 2000 2.8 2.4 a/ These are the welIhead prices of gas, corresponding to the existing contrast with Unical, less royalties and taxes. Future prices have been calculated to take account of changes in oil Drices, as per contract. The differences between the prices are due to the differences in the evolution of oil prices in tne two scenarios. 4. The demand for gas under the two scenarios is given in Tables 3 and 4. For production of LPG there is sufficient demand to absorb additional supplies of gas if the existing 350 MMCFD plant is augmented by a second plant of 200 Y*MCFD capacity in 1991. Cement demand is also expected to grow to 40 MMCFD by 1991 and a petrochemical demand of 43 "MCFD will appear in 1989 for the already committed ethylene plant. Finally by 1993 the fertilizer plant will need 30 MMCFD and small users are expected to use 40 "MCFD. Demand in all these categories (which exclude power) is assumed to be invariant between the two scenarios since: (a) netback values ih these uses (see Annex 9) are greater than the likely economic cost of gas; and (b) they are associated with large discrete capital investments which are relatively insentitive to short- term differences in growth. The demand for gas in the power sector is scenario-dependent. The high levels of gas demand under SQ reflects hypothetical or unconstrained demand that would emerge if supplies were available. -96 - Annex 13 Page 3 of 5 Table 3: Demano for Natural Gas - SQ Scenario (MMCFD) Petro- Fertil- Small Year Power LPG Cement Chem. izers Users Total 1987 310 70 20 400 1988 375 70 20 465 1989 410 70 20 43 543 1990 450 70 20 43 583 1991 500 110 40 43 693 1992 550 110 40 43 743 1993 610 110 40 43 30 40 873 1994 670 110 40 43 30 40 933 1995 740 110 40 43 30 40 1,003 1996 810 110 40 43 30 40 1,073 1997 890 110 40 43 30 40 1,153 1998 970 110 40 43 30 40 1,233 1999 1,060 110 40 43 30 40 1,323 2000 1,160 110 40 43 30 40 1,423 2001 1,270 110 40 43 30 40 1,533 Table 4: Demand for Natural Gas - TOM Scenario (MMCFD) Petro- Ferti- Small Year Power LPG Cement Chem. izers Users Total 1987 310 70 20 400 1988 400 70 20 490 1989 420 70 20 43 553 1990 440 70 20 43 573 1991 440 110 40 43 633 1992 480 110 40 43 673 1993 510 110 40 43 30 40 773 1994 550 110 40 43 30 40 813 1995 580 110 40 43 30 40 843 1996 610 110 40 43 30 40 873 1997 660 110 40 43 30 40 923 1998 670 110 40 43 30 40 933 1999 710 110 40 43 30 40 973 2000 690 110 40 43 30 40 953 2001 670 110 40 43 30 40 933 Reserves 5. Most hydrocarbon fields in Thailand are typically made of small, lenticular sabdstone reservoirs which are further compartmental- - 97 - Annex 13 Page 4 of 5 ized by faulting. The ESSO Namphong field in the on-shore Khorat Basin produces from a highly factured limestone reservoir and production costs will probably also be on the high side. This type of reservoir structure leads to rapid reservoir depletion, requiring a high density of wells and frequent workovers, thereby resulting in high cost production. 6. Economically recoverable reserves are therefore highly price- sensitive. Also, expLoration and development in new areas will also likely depend upon future oil prices. In establishing forecasts of future reserve levels, an attempc has been made to distinguish between the effects of price movements and geological uncertainty. The upper and lower limits to future reserves from a merely geological standpoint are referred to as high and low reserves. Price and geological uncertainty effects are then combined to provide four different scenarios: two main scenarios (SQ with low reserves, TOM with high reserves); and two inter- mediate subscenarios (SQ with high reserves, TOM with low reserves). 7. The reserve assumptions for each the two main scenarios are shown in Table 5. Taole 5: Reserves Assumptions (TCF ) a/ Field TOM (high reserves) SQ (low reserves) Ul and U2 2.3 1.8 U3 1.8 0.8 TP3D O/ 2.0 1.2 TP2D c/ 3.0 - Other Offshore 2.0 Total Offshore 11.1 3.8 ESSO 1-7 0.5 TOTAL 12.8 4.3 a/ These estimates of field size are based upon discussions with PTT and Unocal. b/ The TP3D area is that part of the TP field which has already undergone 3-D seismic survey. c/ The TP2D area hias only had a 2-D seismic survey. 8. Both levels of reserves assume that future gas prices for new contracts will be at the same level as prices based on existing contracts (see Table 2). These prices, under the SQ scenario, are summed to be insufficient to encourage further exploration and no unexplored fields are included. - 98 - Annex 13 Page 5 of 5 9. As mentioned above the reserve estimates urder the two scenarios include the effects of both price and geological variations. Based upon available knowledge on field structures and similarities with fields in the Gulf (see Annex 8), the following table attempts tc separate variations in reserves attributable to each factor. TaDle 6: Economicaliv Recoveraole Reserves (TCF) Reserves Reserves Total Reserves from Ef4ects of prices from new from existing Reserves existing ais- on reserves from discoveries discoveries coveres at ex,sting Scenario common prSces a/ discoveries (A) (B) (C) (D) (B)-(D) TOM (high reserves) 2.0 10.8 12.8 8.6 2.2 SQ (low reserves) 0.0 4.3 4.3 5.2 -0.9 Difference 2.0 6.5 8.5 3.4 3.1 a/ Obtained by reducing column (B) by 20% in case of TOM and increasing it by 20% in case o4 SQ to offset the effects of higher prices in TOM and vice versa under SQ. 10. Thus, of the 8.5 TCF overall reserve difference between the scenarios, 2.0 TCF is due to new offshore discoveries deemed to be encouraged by higher prices, 3.1 TCF reflects the effect of oil prices on existing discoveries, and the remaining 3.4 TCF is due to reservoir uncertainty. The comparison of reserves reveals that, apart from the effect of prices on new discoveries, the assumed variation in reserves in the scenarios due to geological uncertainties is of the same order as the differences in gas availability arising purely out of price effects. 11. Two subscenarios may now be defined which combine the effects of low oil prices in the SQ scenario with the higher reserve case, SQ- high, and the effects of high oil prices in the TOM scenario with the low reserve case, TOM-low. The reserves for SQ-high are 8.6/1.2 = 7.3 TCF, while those for TOM-low are 5.2/0.8 + 2.0 = 8.5 TCF. These two sub- scenarios represent intermediate cases discussed in the main text. -99 - Annex 14 Page 1 of 3 A COMBINED GAS/POWER SUBSECTOR STRATEGY For the purpose of obtaining greater gas supplies under scenario SQ, a gas policy consisting of raising the price in order to provide incentives for more exploration and development was analyzed. The assumptions used for the analysis are that TOM price levels are paid to producers in order to obtain TOM-level gas supplies: Year 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 S/mBTU 2.1 2.1 2.2 2.3 2.4 2.4 2.5 2.5 2.6 2.6 2.7 2.7 2.8 2.8 MMCFD 360 400 420 440 440 480 510 550 580 610 660 670 710 690 Within the power subsector, the assumption concerning future generation facilities consists of supposing that aditional gas supplies would sustain two additional combined cycle units of 300 MW as compared to the other SQ cases. These units would therefore displace investments in lignite and coal starting 1995. This case, designated as SQ-5, shows the following installation schedule: (a) For years 1991-1994: same as SQ-1. (b) For years 1995-2001: 1995 CC 300 #4 CC 300 #5 CC 300 #6 1996 Mae Moh 10-11 1997 Bank Pakong #3 1998 Ao Phai #1 Mae Moh #12 R3 Thermal #1 1999 Ao Phai #2 2000 Kaeng Krung #1-2 Ao Phai #3 2001 Ao Phai #4-5 The objective function associated to case SQ-5 shows the following results: Fixed Costs: $1,704m Variable Costs: $4,244m Total: $5,948m For comparison with Strategy 1, i.e., case SQ-1, the variable costs show the following breakdown by fuel types: - 100 - Annex 14 Page 2 of 3 Fuel Costs ($m) 1991-2001 SQ-1 SQ-5 Total NPV Total NPV Gas 2,748 1,059 6,326 2,265 Fuel Oil 3,936 1,487 1,661 674 Lignite 2,843 1,001 2,757 968 Coal 1,885 466 1,135 273 Diesel 278 64 278 64 Total 11,690 4,077 12,157 4,244 Objection function for SQ-1 is: Fixed costs $1,762m Variable costs $4,077m Total S5,839m The decision agenda for SQ-5 shows the following aspect: 1987 GT 1-3 1988 CT 4-7 Cc1 Srinagarind 5 1989 CT 8-9 CC 2-3 1990 CC 4-6 :4ae Moh 10-11 HVDC R1-R3 1991 Bang Pakong 3 1992 Mae Moh 12 Ao Phai 1 The investment schedule for SQ-5 is shown on the next page. SQ-5 INVESTMI.NTS 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Total GT 3xOO 0.5 16.2 69.0 10.7 96.36 Sronajnrond 5 7.6 11.1 7.2 25.85 CCi 2.2 ?0.9 73.8 51.1 10.2 158.17 GT4 0.2 5.4 23.0 3.6 32.12 t n . 2.? 20.9 74.0 59.3 28.9 185. 2? GI"-(> 0.3 10.8 46.0 7.1 64.24 CT7 0.2 5.4 23.0 3.6 32.12 (1 IH -'-s 0.3 10.8 46.0 7.1 64.24 Nf'(:Cl 2.2 20.9 74.0 53.3 17.9 168.24 NP(HC2 2.2 20.9 74.8 57.0 21.6 116.45 11VOC 0.6 3 4.5 2.5 10.6 a Kadcnq Kr ulq 1-2 4.2 3.7 10.9 14.6 19.9 13.3 66.63 CC3 2.2 20.9 13.8 51.1 10.2 158.17 CC4 2.? 20.9 74.0 59.3 28.9 185.22 Mde Moll 10 3.2 34.3 59.4 62.4 61.3 16.8 237.39 Mae Moll 11 3.2 54.3 60.0 58.8 51.4 14.8 222.51 Pdlik PdkoIIq 3 4.6 47.1 177.2 182.1 63.0 46.9 520.81 Ao Phai 4.7 47.6 180.9 187.4 61.0 38.8 520.46 Mae Moh 12 3.? 34.3 54.r 41.? 37.3 11.1 187.88 Ao Phai 2 3.58 44.5 171.81 184.54 65.58 30.39 500.4 Ao Phal 3 3.58 44.5 170.9 175.97 54.45 29.33 478.71 Ao I'i.c