61155 The World Bank Asia Sustainable and Alternative Energy Program Kingdom of Tonga Electric Supply System Load Forecast March 2010 Photo copyrights: ASTAE on the inside back cover © pintxomoruno's photostream The findings, interpretations, and conclusions expressed in this report are entirely those of the authors and should not be attributed in any manner to the World Bank, or its affiliated organizations, or to members of its board of executive directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. The boundaries, colors, denom- inations, or other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. Contents Acknowledgments ......................................................................................................................vi Abbreviations and Acronyms ....................................................................................................vi Currency ......................................................................................................................................vi 1. Introduction ............................................................................................................................1 2. An Overview of Tonga ...........................................................................................................3 3. Data Records ..........................................................................................................................5 4. The Electricity Supply System ..............................................................................................7 4.1 The Tongatapu Grid System...............................................................................................................................8 4.2 The Vava'u Grid................................................................................................................................................10 4.3 The Ha'apai Grid .............................................................................................................................................. 11 4.4 The `Eua Grid...................................................................................................................................................13 4.5 Individual Household Systems ........................................................................................................................14 5. Electricity Rates ....................................................................................................................15 6. The Tongatapu Electricity Consumer Database ................................................................17 6.1 Meter Reading and Customer Database .........................................................................................................17 6.2 Observations on Tongatapu Customer Consumption ......................................................................................18 7. A Forecast of Electricity Demand Growth .........................................................................21 7.1 A Review of Factors Driving Demand Growth ................................................................................................21 7.2 The Recent TPL Forecast ................................................................................................................................23 7.3 System Losses ................................................................................................................................................23 7.4 Demand-Side Management Programs ............................................................................................................24 7.5 Assumptions Used in the Forecast Scenarios ................................................................................................25 7.6 Forecasts of Energy and Peak Demand ..........................................................................................................25 8. Capacity Planning Analyses ................................................................................................33 8.1 Tongatapu ........................................................................................................................................................33 8.2 Vava'u ..............................................................................................................................................................35 8.3 Ha'apai ............................................................................................................................................................35 8.4 `Eua .................................................................................................................................................................37 9. Base Case Energy Analysis .................................................................................................41 Appendices 1 Interim Report Load Forecasts, April 2009 ......................................................................................................45 2 Enterprise Capacity Enhancement for TPL ......................................................................................................55 3 Details of the Load Forecasts..........................................................................................................................59 4 Oil Price Projections ........................................................................................................................................77 iii iv Contents Figures 1 Energy Sent Out, Tongatapu..............................................................................................................................9 2 A Sample Load Shape on Tongatapu .................................................................................................................9 3 Energy Sent Out, Vava'u..................................................................................................................................10 4 A Sample Load Shape on Vava'u ..................................................................................................................... 11 5 Energy Sent Out, Ha'apai ................................................................................................................................12 6 A Sample Load Shape on Ha'apai ...................................................................................................................12 7 Energy Sent Out, `Eua.....................................................................................................................................13 8 A Sample Load Shape on `Eua ........................................................................................................................14 9 Electricity Rates History ..................................................................................................................................15 10 Average Monthly Use of the Largest 500 Meter Records, Tongatapu ............................................................19 11 Average Consumption by Large Consumption Groups, Tongatapu .................................................................19 12 Quarterly Remittances ....................................................................................................................................22 13 Quarterly Energy versus Tariff Rates, Tongatapu .............................................................................................23 14 High, Median, and Low Forecasts for Tongatapu ............................................................................................31 15 High, Median, and Low Forecasts for Vava'u ..................................................................................................31 16 High, Median, and Low Forecasts for Ha'apai.................................................................................................32 17 High, Median, and Low Forecasts for `Eua .....................................................................................................32 A.1 Peak Demand Forecasts for Tongatapu, Interim Report ..................................................................................45 A1.2 Peak Demand Forecasts for Vava'u, Interim Report ........................................................................................47 A1.3 Peak Demand Forecasts for Ha'apai, Interim Report ......................................................................................47 A1.4 Peak Demand Forecasts for `Eua, Interim Report ...........................................................................................49 A3.1: Median Energy Forecasts for Tongatapu .........................................................................................................59 A3.2: Median Energy Forecasts for Vava'u ...............................................................................................................60 A3.3: Median Energy Forecasts for Ha'apai .............................................................................................................60 A3.4: Median Energy Forecasts for `Eua ..................................................................................................................60 A4.1: Ratio of Singapore Gasoil Price to Average Price of World Crude ...................................................................78 A4.2: Implied Shipping and Handling Charges, Gasoil ..............................................................................................78 A4.3: Crude Oil Forecast Scenarios ..........................................................................................................................79 Tables 1 Available Historical Data Records ......................................................................................................................6 2 The Popua Power Station, Tongatapu ................................................................................................................8 3 The Taumu'aloto Power Station, Vava'u ...........................................................................................................10 4 Description of Electricity Tariff Changes, $T/kWh ...........................................................................................16 5: Statistical Findings for Selected Meter Records, Tongatapu ...........................................................................18 6: Summary of Key Load Forecast Assumptions ................................................................................................26 7 Median Forecast, Loss Reduction Plus DSM Implemented on Tongatapu .....................................................27 8 Median Forecast, Loss Reduction Plus DSM Implemented on Vava'u............................................................28 9 Median Forecast, Loss Reduction Plus DSM Implemented on Ha'apai ..........................................................29 10 Median Forecast, Loss Reduction Plus DSM Implemented on `Eua ..............................................................30 11 Low Forecast Capacity Changes, Tongatapu ...................................................................................................34 12 Median Forecast Capacity Changes, Tongatapu .............................................................................................34 13 High Forecast Capacity Changes, Tongatapu ..................................................................................................35 14 Low Forecast Capacity Changes, Vava'u .........................................................................................................36 15 Median Forecast Capacity Changes, Vava'u ....................................................................................................36 16 High Forecast Capacity Changes, Vava'u ........................................................................................................37 17 Low Forecast Capacity Changes, Ha'apai .......................................................................................................37 18 Median Forecast Capacity Changes, Ha'apai ..................................................................................................38 Contents v 19 High Forecast Capacity Changes, Ha'apai .......................................................................................................38 20 Low Forecast Capacity Changes, `Eua ............................................................................................................39 21 Median Forecast Capacity Changes, `Eua .......................................................................................................39 22 High Forecast Capacity Changes, `Eua ...........................................................................................................40 23 Assumed Unit Capacity and Fuel Efficiencies .................................................................................................41 24 Energy and Fuel Use Analysis, Median Forecast, Tongatapu ..........................................................................42 25 Energy and Fuel Use Analysis, Median Forecast, Vava'u.................................................................................43 26 Energy and Fuel Use Analysis, Median Forecast, Ha'apai ...............................................................................43 27 Energy and Fuel Use Analysis, Median Forecast, `Eua ...................................................................................44 A1.1 Tongatapu Low Growth Load Forecast Scenario, Interim Report....................................................................46 A1.2 Tongatapu Median Growth Load Forecast Scenario, Interim Report...............................................................46 A1.3 Tongatapu High Growth Load Forecast Scenario, Interim Report ...................................................................48 A1.4 Vava'u Low Load Growth Forecast Scenario, Interim Report ..........................................................................48 A1.5 Vava'u Median Load Growth Forecast Scenario, Interim Report .....................................................................50 A1.6 Vava'u High Load Growth Forecast Scenario, Interim Report .........................................................................50 A1.7 Ha'apai Low Load Growth Forecast Scenario, Interim Report ........................................................................51 A1.8 Ha'apai Median Load Growth Forecast Scenario, Interim Report ...................................................................51 A1.9 Ha'apai High Load Growth Forecast Scenario, Interim Report ........................................................................52 A1.10 `Eua Low Load Growth Forecast Scenario, Interim Report .............................................................................52 A1.11 `Eua Median Load Growth Scenario, Interim Report.......................................................................................53 A1.12 `Eua High Load Growth Scenario, Interim Report ...........................................................................................53 A2.1 Summary of Budgeted Capital Investments, T$..............................................................................................56 A2.2 Summary of Budgeted Capital Investments, US$ ..........................................................................................56 A3.1 "Do Nothing" Median Forecast, Tongatapu ....................................................................................................61 A3.2 " "Efficient TPL Median Forecast, Tongatapu....................................................................................................62 A3.3 "Efficient TPL Plus DSM" Low Forecast, Tongatapu .......................................................................................63 A3.4 "Efficient TPL Plus DSM" High Forecast, Tongatapu ......................................................................................64 A3.5: "Do Nothing" Median Forecast, Vava'u ..........................................................................................................65 A3.6 " "Efficient TPL Median Forecast, Vava'u ..........................................................................................................66 A3.7 "Efficient TPL Plus DSM" Low Forecast, Vava'u .............................................................................................67 A3.8 "Efficient TPL Plus DSM" High Forecast, Vava'u ............................................................................................68 A3.9 "Do Nothing" Median Forecast, Ha'apai.........................................................................................................69 A3.10 " "Efficient TPL Median Forecast, Ha'apai ........................................................................................................70 A3.11 "Efficient TPL Plus DSM" Low Forecast, Ha'apai ...........................................................................................71 A3.12 "Efficient TPL Plus DSM" High Forecast, Ha'apai ..........................................................................................72 A3.13 "Do Nothing" Median Forecast, `Eua .............................................................................................................73 A3.14 " "Efficient TPL Median Forecast, `Eua............................................................................................................. 74 A3.15 "Efficient TPL Plus DSM" Low Forecast, `Eua ................................................................................................75 A3.16 "Efficient TPL Plus DSM" High Forecast, `Eua ...............................................................................................76 A4.1 Reference Oil Price Projection.........................................................................................................................79 A4.2 Low Oil Price Projection ..................................................................................................................................80 A4.3 High Oil Price Projection..................................................................................................................................80 Abbreviations and Acronyms DSM Demand-side management MW Megawatt GoT Government of the Kingdom of Tonga MWh Megawatt hour (1,000 kWh) GWh Gigawatt hour (1,000,000 kWh) NRBT National Reserve Bank of Tonga kV kilovolts (thousands of volts) T$ Tonga Pa'anga kW kilowatt US$ United States Dollar kWh kilowatt-hour Currency T$1 = US$0.55 T$1 = 100 seniti Acknowledgments This report was prepared under the guidance of Ms. Wendy Hughes and Mr. Frédéric Asseline, World BankTask Team Leaders for the ASTAE assistance, and authored by Mr. Martin Swales, consultant. The author gratefully acknowledges the considerable assistance provided by Mr. David Wright, Chairman; Mr. Peter McGill, CEO; Mr. Ian Skelton; and all the staff of Tonga Power Limited. Without their help, this report would not have been possible. The author also gratefully acknowledges the assistance of Chief `Akau'ola, Coordinator, Renewable Energy Committee; and Mr. Ofa Sefana, Energy Specialist, Energy Planning Unit, Ministry of Lands, Survey, Natural Resources and Environment; among many others. vi 1 Introduction This report has been prepared based on information and data supplied by the government of the Kingdom of Tonga (GoT) and Tonga Power Limited (TPL) during a visit to Tonga in April and May 2009. The report forms a part of the Tonga Energy Road Map, a project identified during discussions between the GoT and representatives of the Asian Development Bank, the European Commission, the International Union for Conservation of Nature, and the World Bank during April 2009. In recognition of the vulnerability of the Tonga electricity sector to the price of oil, the GoT has set a target of 50 percent renewable energy for the grid-based electricity supply in three years. The study is to identify the options available, evaluate their potential role in grid electricity supplies, and determine the best combination of renewable energy projects to meet the target. The work is to be completed by early 2010 by mak- ing maximum use of work that has been completed on energy issues in Tonga and other Pacific Island countries. 1 2 An Overview of Tonga Tonga is an island kingdom made up of four island was reported to be 1.6 percent in the third quarter of groups, Tongatapu, Ha'apai, Vava'u, and Niuas, consist- 2009.2 Tonga's larger export commodities include fish, ing of some 176 islands. Approximately 36 islands have squash, taro, yams, sweet potatoes, vanilla, cassava, permanent settlements. The population is estimated and kava. The volumes of these exports have varied to be 101,000, more than 75 percent of whom live on during the past decade, averaging about T$29.3 million Tongatapu, the main island and location of the capital, (US$16.12 million) annually between 2003 and 2007 but , Nuku'alofa. the total export value has declined significantly to about .7 T$14 million (US$7 million) in 2007/08 and 2008/09.3 The country is a constitutional monarchy. The current King The market for squash peaked in 2005 but has since vir- was crowned in mid-2008. The King appoints the major- tually disappeared. The value of exports covered almost ity of government ministers. The legislature is unicam- 17 percent of imports in 2003/04, but covered less than 9 eral. The Prime Minister and his Cabinet are members percent of imports in 2007/08. The greatest cause for this who are appointed by the King, nine seats are selected decline in relative values relates to the rapid rise in total by the national nobility, and nine seats are filled through import value, increasing by 55 percent from T$169 mil- popular election. The political system is set for change lion (US$93 million) in 2003/04 to T$262 million (US$144 early in 2010 because the King has agreed to divest him- million) in 2007/08. Tonga saw an important increase in self of most of his powers, which will be taken up by tourist receipts recently, increasing from about T$ 25 mil- a democratically elected government. A Constitutional lion (US$13.75 million) in 2006/07 to over T$ 40 million Commission is currently consulting on and developing a (US$22 million) in 2007/08. new Constitution. A very important element driving the Tongan economy The Tongan economy grew at an average annual rate of is the remittances being sent from Tongans working , about 0.7 percent between 2001/02 and 2006/07 the lat- overseas. The largest numbers of overseas Tongans are est fiscal year for which an estimate is available from the working in the United States, followed by those working GoT Statistics Department.1 The economy was estimated in New Zealand and Australia. The statistical relationship to have shrunk by 3.2 percent in 2006/07 but to have between remittances and the consumption of electricity grown at 2.7 percent in 2007/08. Annual inflation reached is explored in section 7.1. more than 10 percent in 2004 but has since declined and 2. National Reserve Bank of Tonga, Quarterly Bulletin, September 2009. 1. National Reserve Bank of Tonga, Quarterly Bulletin, September 3. National Reserve Bank of Tonga, Quarterly Bulletin, September 2009. 2009. 3 3 Data Records All data are subject to the quality of the measuring equip- to the network after the station metering system so the ment and to the attitude of management to recordkeep- estimates of station service on Ha'apai are artificially low. ing. The persons collecting and maintaining a database Metering problems are also reported on `Eua, and these may be the best qualified, but if managers express lim- problems are believed to be the primary cause for widely ited interest in the data, if faulty meters are not replaced, varying estimates for station service on `Eua in the past. or if there seems to be no use made by the utility of the data, then the quality of the data collected can suffer. The customer meter-reading database includes only about The electric supply system in Tonga may have suffered 85 percent of the energy metered and billed each month. some or all of these problems under its previous owners, The remaining 15 percent of billed energy is taken from Shoreline Power, which was reluctant to share informa- a set of processed readings from current transformer tion about the performance of the utility. meters that are used to capture energy use on feeders where some customer metering is incomplete or where TPL management has generously shared whatever his- meters are known to be faulty. torical performance data are available, and this report is based on that dataset. The basic records provided include The meter database was obtained to provide insights into monthly generation, fuel used, and energy sent out to the how different types and sizes of customers use energy network from as long ago as April 1998 on Tongatapu and seasonally and over the long term. However, the purpose monthly billed energy beginning with July 2006. Records of the database is to maintain monthly records in support for the other three grids are available for shorter periods. of the billing cycle, not in the customer-oriented format Table 1 lists the key database records provided. needed to facilitate customer energy use analyses. Some records were found to be clearly wrong. For example, the As noted above, the quality of the data are not known, consumption on one monthly record was several orders particularly given that the majority of it was collected by of magnitude greater than the monthly consumption for the previous owners. It is known, however, that the com- the same account in all other months. Further discussion puter system to automatically record the performance of these analyses made can be found in section 6. of the generating units at the Popua power station on Tongatapu, the supervisory control and data acquisition Notwithstanding the foregoing caveats about data qual- (SCADA) system, failed sometime in the past and that ity, the available databases were used as the founda- all recent data have been collected manually. There is tion for the current load forecasts. Use of the forecasts no SCADA at the power stations on Vava'u, Ha'apai, and should be tempered by the noted data limitations. `Eua. On Ha'apai, a rented generating unit is connected 5 6 Kingdom of Tonga: Electric Supply System Load Forecast Table 1: available HisTorical DaTa recorDs Data record start end comment Tongatapu, gross generation, net energy sent April 1998 October 2009 Continuous, monthly out, fuel used Tongatapu, peak and minimum hourly demand April 1998 October 2009 Broken monthly records to April 2007 Tongatapu, billed energy July 2006 November 2009 Continuous, monthly Vava'u, gross generation, net energy sent out, January 2004 October 2009 Continuous, monthly fuel used, peak and minimum hourly demand Vava'u, billed energy July 2006 August 2009 Continuous, monthly Ha'apai, gross generation, net energy sent out, January 2004 October 2009 Continuous, monthly fuel used, peak and minimum hourly demand Ha'apai, billed energy July 2006 August 2009 Continuous, monthly `Eua, gross generation, net energy sent out, January 2006 October 2009 Continuous, monthly fuel used `Eua, peak and minimum hourly demand January 2004 October 2009 Broken, monthly 'Eua, billed energy July 2006 August 2009 Continuous, monthly Tongatapu, monthly meter readings from cus- January 2004 September 2009 Approximately 85 percent of the billed tomer meter records database energy, format difficult to analyze Ha'apai, monthly meter readings September 2008 September 2009 All months except October 2008 Fuel oil price paid by Tonga electric utility March 1998 April 2009 Average monthly price paid for automotive diesel oil on each grid Electric tariffs March 2002 Present Rates for each grid system with dates for all changes Source: Compiled by author. 4 The Electricity Supply System There are four grid systems owned and operated by recognize that investors need appropriate incentives to TPL, the major electric utility of Tonga, which generates, invest in maintaining and upgrading the network. distributes, and retails electricity on three of the four major island groups of Tongatapu, Vava'u, and Ha'apai. A legislative program culminated in the passing of the Elec- The fourth grid is on `Eua, an island close to Tongatapu. , tricity Act 2007 which provided for the establishment of The ownership of electricity assets has changed twice in the Electricity Commission to regulate tariffs, consumer recent years, first in 1998 when the national utility was service standards, and electrical safety. The regulatory bought by Shoreline Power4 and more recently in mid- framework employed is a "concession contract" model. 2008 when the GoT bought the assets from Shoreline to Tariffs, tariff adjustment formulas, operational efficiency create TPL as a state-owned enterprise. This latest own- benchmarks, consumer service standards, and penalties ership change was already being planned in 20065 before for nonachievement are specified in a contract between rioting in Nuku'alofa resulted in considerable damage the Electricity Commission, on behalf of the Kingdom, to buildings in the downtown area. The riots destroyed and the electricity provider, TPL. international confidence in investment in Tonga and the GoT was not able to proceed with its preferred option to The electricity provider has legal rights under the con- sell the assets to private interests. Given the desire of cession contract to pass through fuel costs to consum- the Crown Prince to assume a constitutional monarchy ers and charge an inflation-indexed nonfuel tariff. The soon after his coronation and withdraw from commercial nonfuel tariff covers agreed-on operational costs, busi- activities within Tonga, the GoT concluded the agreement ness overhead, interest payments, depreciation, and with the Shoreline Group to purchase the utility in 2008. an allowed return on investment that provides retained earnings for future capital expenditure, shareholder divi- Along with the change in ownership in 2008, the Electric dend payments, or both. Power Board was dissolved. The new regulator, the Elec- tricity Commission, was introduced to reflect the best Consumer service standards are regulated and cover international practice and facilitate a possible future sale both the rights and responsibilities of TPL and its cus- of the GoT's interest. The reforms balance the interests tomers regarding of consumers with regard to tariff and service standards against the need for investors to have certainty about · new connections, reconnections, and disconnec- earning a fair return on their investment. The reforms also tions; · limits on interruptions of supply for planned mainte- nance and unexpected events; 4. The Crown Prince of Tonga held a major share in Shoreline · voltage stability; Power. 5. A Conditional Agreement was in place to sell the assets and · customer service; business to Northpower Limited, a New Zealand company in 2006, · billing; and but the Conditions Precedent of the Sale and Purchase Agreement · the frequency of meter testing. could not be met because of the destruction of Shoreline's corporate office, including all of its records. 7 8 Kingdom of Tonga: Electric Supply System Load Forecast The concession contract provides for a review of the tar- iff at seven-year intervals to reset the nonfuel component Table 2: THe PoPua Power sTaTion, TongaTaPu and divide the benefits of efficiency gains between the operator and consumers. net capacity efficiencya Diesel designation number (kw) (kwh/liter) The Electricity Act provides for public and private sector Caterpillar 3516B 6 1,400 3.80 investment in the electricity operator. The GoT does not Caterpillar PM3516B 1 1,400 3.80 see itself as a long-term owner and the sale of shares to (containerized) one or more of the Tongan public, institutions, and corpo- MaK6CM32C 1 2,880 4.35 rate investors will be considered in light of GoT policy to sell its interests in TPL and other public enterprises. Total 8 12,680 Source: Data from TPL. TPL was established with an independent Board of Note: a. Estimated from operating records. Directors drawn from the commercial sector of Tonga 's and from overseas. TPL objectives are to · provide safe, reliable, and affordable electricity sup- Demand, as measured by the energy sent out from the ply throughout Tonga; and power station, grew steadily through the last 10 years at · maximize shareholder value while maintaining pru- an average rate of 5.5 percent annually until August 2008, dent levels of exposure to operational and financial when demand growth faltered and has since declined. risk. The gross peak demand7 reached almost 8,400 kW during October 2007 through April 2008 and has since declined to between 7 ,400 and 8,200 kW. The 12-month average 4.1 The Tongatapu Grid System annual energy delivered by the Popua power station, net of station service, reached just over 48.2 GWh and The system serving all of Tongatapu is the largest of the remained approximately constant during April through four grid systems operated by TPL and distributes about September 2008. By October 2009, the rolling average 85 percent of the grid-supplied energy in Tonga. All of the had declined to about 42.2 GWh. This pattern may be energy is generated at the Popua power station, which seen in figure 1, which shows the monthly energy sent contains eight diesel generating units with a combined out and the 12-month rolling average monthly energy capacity of 12.68 MW. The numbers, size, and key char- sent out for the period 1998 through October 2009. The acteristics of the diesel units in the power station are annual system load factor, calculated as the ratio of the listed in table 2. The output is transmitted throughout the average energy generated (in kWh) during a year to the island on an 11 kV system and distributed at 415 volts. maximum peak demand (kW) during that same year, has As of early 2009, TPL had more than 14,650 customer generally increased from about 62 percent in 2003 to accounts, up from 14,150 in early 2004, or an increase of over 67 percent in 2008/09. The long-term average has less than 3 percent during the five-year period of records been close to 63 percent. provided (an average annual growth of less than 0.6 per- cent). However, throughout the period 2004­08, some Station service8 at Popua has averaged 2.6 percent since 700­750 of these accounts were inactive or had faulty January 2006. Since 2006, the first date for which energy meters as evidenced by the record showing zero con- invoiced by the utility is available, combined technical and sumption. Furthermore, for another 700­800 accounts, nontechnical losses through the transmission and distri- the meter readings were less than 10 kWh during the bution system have averaged 15.6 percent.9 For a system month.6 Based on these data, it is reasonable to con- the size of the Tongatapu network, and for which recent clude that the primary source of growth of energy use is higher use by customers. . 7 The gross peak demand is the maximum power generated by the diesel sets during a calendar month and includes the station service at the time. 8. Station service is the energy used at the power plant to operate 6. TPL management advises that in 2009 some 85 percent of the the ancillary equipment. It is the difference between the gross gen- zero-reading meters were attached to vacant properties with the bal- eration measured at each generator and the energy sent out onto ance being faulty or tampered with. TPL has an ongoing program to the network. remove or replace all these meters. TPL notes that when meters 9. This figure is calculated as the difference between the energy that previously recorded very low consumption were replaced, the sent out from the power station and the energy billed divided by the customers usually begin to record higher electricity use. energy sent out. The Electricity Supply System 9 Figure 1: energy senT ouT, TongaTaPu Monthly generation 12-month rolling average 4,500 4,000 Monthly energy (MWh) 3,500 3,000 2,500 2,000 Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul 1998 1998 1999 1999 2000 2000 2001 2001 2002 2002 2003 2003 2004 2004 2005 2005 2006 2006 2007 2007 2008 2008 2009 2009 Source: Data from TPL. Figure 2: a samPle loaD sHaPe on TongaTaPu Thursday May 28 to Wednesday June 3, 2009 8,000 7,000 6,000 Kilowatt-hours 5,000 4,000 3,000 2,000 1,000 0 0 6 12 18 24 30 36 42 48 54 60 66 72 78 84 90 96 102 109 114 120 126 132 138 144 150 156 162 168 Hour Source: Data from TPL. investment in the network has been limited,10 technical tional statistics, but there are insufficient data to confirm losses in the lines and transformers might be 8­12 per- the breakdown between these two loss categories. cent. Nontechnical losses, that is, amounts lost through bad meters, errors in meter reading, and unmetered con- The load demand during each day is illustrated in figure 2, sumption would then be about 3­7 percent. Nontechnical which shows the hour-by-hour demand during a week in losses of this magnitude are modest relative to interna- 2009. As the figure shows, the minimum demand occurs in the middle of the night, rises in the daytime hours, and exhibits a sharp peak between 7 and 8 pm. TPL advises 10. According to current management, Shoreline did not invest sig- nificantly in the networks. 10 Kingdom of Tonga: Electric Supply System Load Forecast that the time of the daily peak is later during summer months and earlier in the winter. With reference to the Table 3: THe Taumu'aloTo Power sTaTion, vava'u Saturday load shape (hours 49 to 72), the daytime load is lower than during the preceding weekdays, implying net capacity efficiencya that an important part of daily demand is from govern- Diesel designation number (kw) (kwh/liter) ment and business offices that do not operate on the Cummins KTA19G2 3 300 3.7 weekends. The very low Sunday demand (hours 73 to Cummins LTA10G3 2 186 3.3 96) reflects the strong Christian beliefs in Tonga and the Total 5 1,272 prohibition of virtually all business activities on Sunday. Source: Data from TPL. Note: a. Derived from operating records. 4.2 The Vava'u Grid The next largest grid system is on the main island of 2009 to assume that demand has started to grow again. Vava'u in the Vava'u group. The grid is served from the The gross peak demand reached just over 1,080 kW in Taumu'aloto power station, which contains five diesel early 2008 and it reached 1,010 kW in August 2009. The generator sets. Table 3 lists the key characteristics of the system load factor has varied from about 57 percent in diesel units. The power station is located in the principal 2004 to over 60 percent in 2008. The long-term average town of Neiafu and serves all of the communities on the .5 has been 57 percent. island. Transmission is at 6.6 kV and distribution at 415 volts. The number of customers served as of early 2009 Station service at Taumu'aloto averaged 4.9 percent over was 3,201. the period of record. This rate is relatively high and a long- term rate closer to 3 percent would be expected. Meter- Between January 2004, the first date for which records ing errors may be partially to blame for the high rate. are available, and January 2008, the net energy sent out Since 2004, the first date for which data on the energy from the Taumu'aloto power station grew at an average invoiced by the utility is available, technical and nontech- annual rate of 4.0 percent to 5.26 GWh. Energy con- nical losses together through the transmission and dis- sumption has since declined to an annual rate of 4.73 tribution system have averaged 14.5 percent. As noted GWh. Figure 3 illustrates this pattern by showing the for Tongatapu, technical losses might be 8­12 percent, monthly energy consumption and 12-month rolling aver- implying that nontechnical losses are about 2­6 percent. age monthly consumption. There is too little evidence There are insufficient data to confirm the breakdown from the monthly energy figures during the latter half of between the two loss types. Figure 3: energy senT ouT, vava'u 480 Monthly energy 12-month rolling average 460 440 Monthly energy (MWh) 420 400 380 360 340 320 300 Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul 2004 2004 2005 2005 2006 2006 2007 2007 2008 2008 2009 2009 Source: Data from TPL. The Electricity Supply System 11 Figure 4 shows a sample week from May 2009 of hourly 2009, the annual rate had fallen to 1.36 GWh, but the loads on the Vava'u system as measured by the output energy records of late 2009 indicate that demand may of the Taumu'aloto power station. The record begins on a have begun to grow again. Figure 5 illustrates the energy Friday so the third day shows the Sunday demand profile. record. The gross peak demand reached more than 355 The system peak demand occurs at the same time as in kW in late 2007 and it, too, declined to just less than Tongatapu, that is, shortly after sunset each day. Given 300 kW in March 2009. The system load factor has varied that the time of year might be described as the shoulder widely between 50 percent and 60 percent but has aver- season, it is not clear if there is any significant air condi- age about 57 percent since 2004. tioning load during the weekdays. Station service on Ha'apai averaged about 4 percent over the period to mid-2008, but as noted earlier, this statis- 4.3 The Ha'apai Grid tic may be misleading. Since mid-2008, average station service has declined to about 2.6 percent. However, The Ha'apai power station is on the island of Lifuka. This since 2006, when data on the energy invoiced by the station also serves the adjacent island of Lotofoa via a 6.6 utility first became available, technical and nontechnical kV line. Two Cummins LTA10G3 diesel electric sets with a losses through the transmission and distribution system combined capacity of 372 kW are in the power station and have averaged 11.3 percent. As noted for Tongatapu, the are operated continuously except when an engine must technical losses are believed to be the largest portion of be shut down for maintenance. The Cummins engine this figure, but there are insufficient data to confirm the characteristics are the same as shown in table 2. TPL has breakdown between the two loss types. rented a single Cummins KTA19G4 set of 400 kW and located it adjacent to the power station so supply can Figure 6 shows a sample week from May 2009 of hourly be maintained during outages of the other engine sets. loads on the Ha'apai system as measured by the out- Transmission is at 6.6 kV and distribution at 415 volts. The put of the power station. The record begins on Monday number of customers as of late 2009 was 998. May 4. The last day shows the Sunday demand profile although the demands are not significantly different from Between January 2004, the first date for which records most weekday loads. The system peak demand occurs are available, and March 2007 the net energy sent out , at the same time as at Tongatapu, that is, shortly after from the power station grew rapidly, at an annual rate sunset each day. There appears to be little difference of 10.3 percent to 1.56 GWh. The 12-month rolling aver- between the weekday and weekend demand profiles, age monthly energy remained at about that level until implying that there are few business offices to add load March 2008 when demand began to decline. By August during the typical working weekday on the island. Figure 4: a samPle loaD sHaPe on vava'u Friday May 1 to Thursday May 7, 2009 1,000 900 800 700 Kilowatt-hours 600 500 400 300 200 100 0 0 6 12 18 24 30 36 42 48 54 60 66 72 78 84 90 96 102 108 114 120 126 132 138 144 150 156 162 Hour Source: Data from TPL. 12 Kingdom of Tonga: Electric Supply System Load Forecast Figure 5: energy senT ouT, Ha'aPai 160 Monthly energy 150 12-month rolling average 140 Monthly energy (MWh) 130 120 110 100 90 80 Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul 2004 2004 2005 2005 2006 2006 2007 2007 2008 2008 2009 2009 Source: Data from TPL. Figure 6: a samPle loaD sHaPe on Ha'aPai Monday May 4 to Sunday May 10, 2009 350 300 250 Kilowatt-hours 200 150 100 50 0 0 6 12 18 24 30 36 42 48 54 60 66 72 78 84 90 96 102 108 114 120 126 132 138 144 150 156 162 Hour Source: Data from TPL. The Electricity Supply System 13 Figure 7: energy senT ouT, `eua 110 Monthly energy 12-month rolling average 105 100 Monthly energy (MWh) 95 90 85 80 75 70 65 60 Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul 2004 2004 2005 2005 2006 2006 2007 2007 2008 2008 2009 2009 Source: Data from TPL. 4.4 The `Eua Grid Station service on 'Eua has varied considerably since early 2004. In 2004, station service amounted to about `Eua is an island southeast of Tongatapu and a power 3.5 percent of generation but then began to rise to about station with two diesel units serves the network. Two 6.5 percent during 2006 and again to more than 11 per- Cummins LTA10G3 diesel electric sets with a combined cent during the first half of 2008. Some of this change capacity of 372 kW are operated continuously except can be attributed to errors in metering and the major when an engine must be shut down for maintenance. overhaul that was conducted on the diesel sets during (See table 2 for unit characteristics.) TPL has no other 2008. Since mid-2008, station service has declined but generating equipment on the island so the evening peak remains above 6.5 percent on a 12-month rolling aver- demand is not met whenever there is an outage of age basis as of March 2009. The month-to-month data either engine. The number of customers as of early 2009 indicate that station service declined to about 2.5 per- was 1,062. cent in the first quarter of 2009. With respect to network losses, the records are unreliable because of metering Between January 2004, the date of the earliest available problems. During the first quarter of 2009, rates of about records, and February 2008, net energy generated at 15­16 percent were observed. the power station grew at an annual rate of 6.1 percent to about 1.15 GWh. Since then, the energy generated Figure 8 shows a sample week of hourly loads on the `Eua declined to an annual rate of 1.06 GWh in March 2009 system as measured by the output of the power station. before stabilizing. Figure 7 illustrates the historical energy The record begins on Monday May 4 so the last two days record showing the monthly consumption and 12-month show the weekend demand profile, which shows little rolling average. Because the decline in the rolling average difference from most weekday loads. This implies that stabilized as of March 2009, demand on `Eua may have there is little business activity on the island. The system begun to grow again. The gross peak demand reached peak demand occurs shortly after sunset each day. 310 kW in late 2007 but declined to about 280 kW in March 2009. The system load factor has varied widely .5 but has averaged about 47 percent since 2004. 14 Kingdom of Tonga: Electric Supply System Load Forecast Figure 8: a samPle loaD sHaPe on `eua Monday May 4 to Sunday May 10, 2009 300 250 200 Kilowatt-hours 150 100 50 0 0 6 12 18 24 30 36 42 48 54 60 66 72 78 84 90 96 102 108 114 120 126 132 138 144 150 156 162 Hour Source: Data from TPL. 4.5 Individual Household Systems active in Tonga. These systems are typically large enough to provide low-power lighting and power for a small radio Although there are grid systems on only five of the inhab- or equivalent. The GoT has a program to rehabilitate the ited islands,11 electricity is made available to individual solar systems and extend their availability to all Tongans homes on most islands via small, privately owned diesel who do not have access to grid-supplied energy. Further generator sets and by solar panels with battery back-up discussion may be found in "Tonga OFF-Grid Electrifica- systems. In most instances, home-based solar systems " tion Initiative. 12 were installed with funding from various aid agencies 12. Heinz Boehnke and Lano Fonua, "Tonga OFF-Grid Electrification 11. Two islands are supplied by the Ha'apai grid. " Initiative, International Renewable Energy Agency, 2009. 5 Electricity Rates All electricity on the TPL grid is supplied by diesel electric 102.67 seniti/kWh (US$0.56513) and 104.67 seniti/kWh generation and the price of diesel fuel is a major compo- (US$0.576) on Vava'u. Since mid-2008, international oil nent of the electricity tariff. Figure 9 illustrates how the prices have fallen significantly and the correspondingly cost per unit of electricity on Tongatapu has varied since lower diesel fuel costs were passed on to customers March 2002. through tariff changes in February and May 2009. In addi- tion, the tariff structure changed in May 2009 to make The rate structure has varied by both customer and grid the tariff uniform across all grids at 62.79 seniti/kWh system during the period of record. Table 4 describes the changes from January 2002 through late 2009. 13. The official exchange rate floated before April 2006, when the rate was fixed at T$$2.075/US$ until late July 2009. The average The cost of diesel fuel drove the tariffs to record highs monthly exchange rate of the T$$ has since appreciated to T$1.82/ during 2007 and 2008. The tariff for Tongatapu reached US$. Figure 9: elecTriciTy raTes HisTory 1.2 T$ US$ 1.0 Cost per kWh (T$ and US$) 0.8 0.6 0.4 0.2 0 005 03 02 04 05 02 04 05 03 03 04 05 02 002 004 003 07 06 08 09 06 08 09 07 07 08 09 06 009 006 008 007 Jul-20 Jul-20 Jul-20 Jul-20 Oct-20 Oct-20 Oct-20 Oct-20 Apr-20 Apr-20 Apr-20 Apr-20 Jul-20 Jul-20 Jul-20 Jul-20 Oct-20 Oct-20 Oct-20 Oct-20 Apr-20 Apr-20 Apr-20 Apr-20 Jan-2 Jan-2 Jan-2 Jan-2 Jan-2 Jan-2 Jan-2 Jan-2 Source: Compiled by author from TPL records. 15 16 Kingdom of Tonga: Electric Supply System Load Forecast Table 4: DescriPTion oF elecTriciTy TariFF cHanges, $T/kwH Date of rate change Tongatapu Vava'u Ha'apai `Eua Comment March 2002, start of record 0.360 0.380 n.a. n.a. February 2003 0.420 0.440 n.a. n.a. October 2003 0.455 0.475 n.a. n.a. January 2004 0.455 0.475 0.515 0.515 November 2004 0.565 0.585 0.515 0.515 June 2005 0.565 0.585 0.515 0.515 11 seniti subsidy for first 200 kWh used by domestic consumers on Tongatapu and Vava'u only October 2006 0.6171 0.6371 0.5671 0.5671 September 2007 0.6851 0.7051 0.6351 0.6351 February 2008 0.8315 0.8515 0.7815 0.7815 September 2008 1.0267 1.0467 0.9767 0.9767 March 2009 0.7276 0.7476 0.6776 0.6776 11 seniti subsidy discontinued May 2009 0.6279 0.6279 0.6279 0.6279 All rates set equal on all grids October 2009 0.8314 0.8314 0.8314 0.8314 Source: Data from TPL. Note: n.a. = No data available. (US$0.345).14 By late 2009, tariffs once again had to be is currently set at the interim rate of 38.69 seniti/kWh increased solely to cover higher fuel oil costs. (US$0.213) pending a final adjudication between this fig- ure and 40 seniti/kWh, the amount originally claimed by The regulated tariff is made up of two parts, a fuel com- TPL at the rate hearing. The fuel component has been ponent that covers the cost of diesel fuel and a nonfuel reset at approximately quarterly intervals since late 2007 component that covers allowed operational costs, busi- and is based on the expected cost of fuel in the sub- ness overhead, interest payments, depreciation, and an sequent quarter plus an adjustment for the difference allowed return on investment. The return can be used as between the actual and projected fuel costs during the retained earnings for future capital expenditure, share- preceding quarter. The fuel component of the May 2009 holder dividend payments, or a mixture of both. tariff was 24.1 seniti/kWh (US$0.133), but this rate had to be increased to 44.45 seniti/kWh (US$0.244) in October The nonfuel component was the subject of detailed rate 2009 to cover higher fuel costs than had been expected hearings with the Tonga regulator. This nonfuel component in May. It is probable that the high tariffs during 2008 affected the demand for electricity. 14. The fuel costs in Tonga are based on Singapore prices at least two months earlier. 6 The Tongatapu Electricity Consumer Database In April 2009, TPL made available a sample of its Tonga- A preliminary effort to reorganize the roughly 15,000 tapu customer database in Microsoft Excel format. The meter readings each month from the monthly system data covered the period January 2004 through Decem- to a system that tracks each customer account over the ber 2008. This sample was supplemented in November period of record was begun but it was determined that 2009 by meter readings for July 2008 through Septem- the time required to complete the reorganization was too ber 2009, that is, the period since TPL took over as the great for this study. A part of the difficulties related to operator. A less complete record from July 2008 through the addition and removal of some accounts each month September 2009 was also provided for customer meter so that all records had to be inspected to ensure each readings on Ha'apai. account record was matched from month to month. Limited analyses were made using the original data organization. 6.1 Meter-Reading and Customer Database The interpretation that was possible has serious limita- tions because the monthly datasets are a mix of meter TPL advised that the meter-reading database represents readings taken at various times during each calendar about 85 percent of the energy billed each month, with month and sometimes more than once in a month. Not the balance coming from current transformer meters that all meters were read on the same date each month, so report the energy used on individual feeder lines. This the energy used per period varies with the number of energy record also includes manual adjustments made days between meter readings as much as for any other each month for meters that have been tampered with. seasonal factor. In some months, some meters were read twice, requiring a laborious manual adjustment to The customer meter-reading database was obtained to the data that also proved too time-consuming to com- attempt to identify how different customers and groups plete. Data from these months had to be discarded. of customers have used electricity in the past. Past con- sumption is useful in any projection of future use. How- Notwithstanding the foregoing caveats, some useful ever, the database is organized on a monthly basis as insights can be obtained from the records. The data in the needed by the utility to facilitate monthly meter read- original files provided by TPL was reorganized into sepa- ings and billing. A format that includes the full series of rate files for each month and the resulting data sorted in meter readings by account would be more helpful to an different ways to try to gain an understanding of which investigation of how customers have used electricity customers used the most, the average, and the least over time. energy each month. By this method, the same meters do not always end up in the same group each month, so the statistics must be used with care. 17 18 Kingdom of Tonga: Electric Supply System Load Forecast Data from January 2004 through September 2009 was 6.2 Observations on Tongatapu provided and both the total number of accounts as well Customer Consumption as the average consumption of all accounts grew until early 2008. Thereafter, both the number of accounts The next best thing to examining specific customer and the average consumption per account leveled off records is to group customers (meter accounts16) by before declining in late 2008 and through 2009. In early some characteristic. In this case, grouping was based on 2004, there were approximately 14,200 meters being metered consumption. The 500 accounts with the larg- read each month. This increased to between 15,000 est metered consumption each month were grouped and and 15,200 in late 2008. The number of meters read in reviewed. Also, the 1,000, 1,500, and 2,000 accounts late 2009 declined marginally to about 14,900. There with the largest metered consumption were identified are some 800­1,000 meters each month that record and assessed. In addition, the characteristics of the zero energy use, so that the number of consuming cus- remaining meter records, approximately 13,000 records, tomers was about 13,400 in 2004, increasing to almost were grouped and analyzed. Table 5 summarizes statis- 14,200 by mid-2008. The number of consuming custom- tics for these groups. ers has since declined slightly to fewer than 14,000 in late 2009.15 The users in the first group of 500 are by far the most significant electricity consumers; they number less than As noted above, an effort was made to reorganize some 3.5 percent of the total customer accounts and repre- of the records to facilitate analysis of customers' energy sent a group that TPL is likely to be actively monitoring. use. Records for a few of the largest consumers were Only this group shows a continuing tendency for growth extracted and a time series of energy consumption by into mid-2009, at an annual rate of about 3.5 percent. It calendar month was created. No clear pattern in seasonal would be useful for future load forecasting efforts to bet- use emerged, presumably because some customers ter understand what plans these customers may have use more energy during the November­January period for further development. Figure 10 shows the average while others use more during May through August each use per meter for the period of record as well as the year. With respect to the significance of adjusting for 12-month rolling average and a trend line for the monthly the meter-reading date, it was found that the variation record. in energy use between customers overshadowed the effect of the calendar month adjustments. Overall, it is The spike in use in December 2007 and the gap in the believed that efforts to adjust for the date of each meter records during 2008 are irreconcilable meter-reading reading is unlikely to significantly improve the consis- data. tency of analyses of customer use. The single largest meter record is for a facility operated by the Church of Jesus Christ of Latter-day Saints. This facility began operations in mid-2007 and by August of 15. TPL advised that in a survey during 2009, some 85 percent of zero-reading meters were found to be at unoccupied buildings. TPL's 16. Many large customers have several meters, but for the pur- experience has been that when many low and zero-reading meters poses of this analysis each meter is considered to be an individual are replaced, the customers do begin to record higher electricity "customer." use. Table 5: sTaTisTical FinDings For selecTeD meTer recorDs, TongaTaPu Percentage of total average use per meter Highest average use late 2009 average use meter account group monthly energy demand in 2004 (kwh) (kwh) per meter (kwh) 1st 500 45.80 2,230 2,800 2,645 2nd 500 8.31 415 470 455 3rd 500 5.31 267 295 290 4th 500 4.21 212 238 229 Lowest 13,000 36.37 75 85 82 Source: Compiled by author from TPL records. Knowledge about Air Pollution Caused by Stoves and Fuels 19 Figure 10: average monTHly use oF THe largesT 500 meTer recorDs, TongaTaPu Top 500 average 4,500 12-month rolling average Linear (top 500 average) Average kWh per month per customer 4,000 3,500 3,000 2,500 2,000 1,500 1,000 500 0 004 04 04 04 005 05 05 05 006 06 06 06 007 07 07 07 008 08 08 08 009 09 09 09 Apr-20 Jul-20 Oct-20 Apr-20 Jul-20 Oct-20 Apr-20 Jul-20 Oct-20 Apr-20 Jul-20 Oct-20 Apr-20 Jul-20 Oct-20 Apr-20 Jul-20 Oct-20 Jan-2 Jan-2 Jan-2 Jan-2 Jan-2 Jan-2 Source: Compiled by author from TPL records. that year began using an average 116,000 kWh a month, next largest meter reading is on the order of 25,000 kWh varying between 100,000 kWh and 140,000 kWh. This per month. represents 3.0­3.5 percent of the total billed consump- tion on Tongatapu. These figures may overstate the net With reference to table 5, each of the next three groups 's impact of this facility on TPL load because there is of 500 shows lower average use relative to the next evidence that demand at some of the other facilities larger group. However, all of these groups had growing operated by the same organization declined about the average use per meter until about late 2006 (see figure time the new facility began to consume electricity. The 11) when average use stabilized or began to decline. It is Figure 11: average consumPTion by large consumPTion grouPs, TongaTaPu 550 Average kWh per month per customer 500 450 400 350 300 250 200 150 004 04 04 04 005 05 05 05 006 06 06 06 007 07 07 07 008 08 08 08 009 09 09 09 Apr-20 Jul-20 Oct-20 Apr-20 Jul-20 Oct-20 Apr-20 Jul-20 Oct-20 Apr-20 Jul-20 Oct-20 Apr-20 Jul-20 Oct-20 Apr-20 Jul-20 Oct-20 Jan-2 Jan-2 Jan-2 Jan-2 Jan-2 Jan-2 2nd 500 3rd 500 4th 500 2nd 500 12-month rolling average 3rd 500 12-month rolling average 4th 500 12-month rolling average Source: Compiled by author from TPL records. 20 Kingdom of Tonga: Electric Supply System Load Forecast noteworthy that average use of the second largest group Figure 11 illustrates the average energy consumption by of 500 declined significantly from a high of 470 kWh in late the second, third, and fourth groups of large consumers. . 2006 to about 450 kWh in late 2007 The start of the decline approximately coincides with the civil unrest There are about 13,000 meters in the lowest group. in 2006 so perhaps this group was most affected by Although the total consumption for this group is signifi- those events. Since late 2007 average use by this group , cant at 36 percent of total metered demand, the average increased until early 2008 when it stabilized at about 460 use per meter per month is only 80­85 kWh. This group kWh per month. includes an average of about 940 meters with monthly readings of zero. TPL has advised that some 85 percent The third group of 500 shows the most stable average of these meters are at unused buildings and that the bal- use per month of all the groups. Since mid-2008 when ance are faulty or tampered with. By November 2009, that group's average use was the highest, use has TPL advised that more than 1,000 meters had been declined slightly and there seems to be no indication of replaced. However, more than 900 zero-reading meters any change in the declining trend. The characteristics of 's remain in TPL October 2009 meter database. the fourth group of 500 are similar to the third except that the decline from highest average use in mid-2008 has been greater than in the higher-use groups. 7 A Forecast of Electricity Demand Growth 7.1 A Review of Factors Driving the monthly demand for electricity (see figure 1). Appar- Demand Growth ently, most businesses simply relocated into the sur- rounding suburbs and recommenced operations. A major Any forecast of conditions in the future is inevitably sub- new international resort is reportedly being planned for ject to considerable uncertainty. Nevertheless, because the Vava'u island group, but is to be located on one of consumers can respond much more quickly to economic the remote islands and therefore would not be supplied conditions than can an electricity supply utility, forecasts with electricity from the TPL grid. Some economic spin- of future electricity demand are critical to ensure suf- offs may occur, depending on the staffing and servicing ficient generation and transmission capacity to meet policies of this resort complex. However, these impacts consumer expectations. An earlier forecast was issued are not expected to build until the construction period in June 2009 and these data were used in the "Interim and subsequently during the first years of the resort's Report of the Tonga Road Map Project" issued in Novem- business development. ber 2009. These out-of-date forecasts are summarized in appendix 1 for reference. With reference to figure 1, long-term demand growth on Tongatapu has averaged about 6.4 percent annually There are two sources of demand growth for electric- between 1998 and late 2006, when growth faltered. ity: the addition of new customers and increased use , Thereafter, energy use declined through mid-2007 then by existing customers. Both are economically driven grew again until mid-2008. From mid-2008 until late factors that may be expected to respond to economic 2009, energy use continued to decline. The reasons for change. No major new developments are coming up in the growth between 1998 and late 2006 are as difficult Tongatapu that would have a significant impact on the to identify as are the reasons for the changes in growth number of customers or would add a single large load since 2007 . to the system. Redevelopment of the downtown area of Nuku'alofa, required following the 2006 riots, com- The growth in number of customer accounts since 2004 menced in earnest in 2008 and, although TPL requires has been very small, averaging about 0.6 percent annu- customers to register their new electricity requirements ally. Customer growth may have been higher earlier this in advance, this requirement had not been followed as of decade, but supporting data are not available. Some new mid-2009.17 Hence, it is not known what new electricity customers could have been much larger than the aver- demands the downtown redevelopment will create. It is age; for example, the new facility constructed by the worth noting that the loss of the buildings burned during Church of Latter-Day Saints, which broke ground in 2007 , the 2006 riots did not appear to have a material impact on is an exceptionally large new load that was easy to iden- tify. However, the database format has frustrated efforts to find other new customers that might support such a 17 TPL advised that this requirement is not being followed despite . theory. One can point to the general finding that use by TPL having helped the Energy Ministry to develop a questionnaire the largest 500 customers has continued to increase, landowners could use to forecast demand based on the intended at about 3.5 percent annually, even into mid-2009. This use of the building. 21 22 Kingdom of Tonga: Electric Supply System Load Forecast growth represents about 1.6 percent of total growth up Another factor that may affect energy use is the tariff, to 2006. The record of general economic activity shows particularly given that tariff rates spiked during 2008 to little new industry and points to agricultural exports as three times the level of 2002. Figure 13 illustrates tar- the country's largest economic activity. iff changes and corresponding energy use. The values have been scaled to show both data traces in figure 13. One possible reason for the growth in energy consump- The tariff rates do not appear to have an obvious impact tion would be an increase in remittances by Tongans on electricity use; demand only began to fall during the working abroad. Quarterly remittances18 from 2001 are latter half of 2008, when the tariff rates had already shown in figure 12 together with demand for electricity. approached the record high. In a test of correlation Remittances increased steadily through late 2004, when between the quarterly values, the result is positive, an they appear to have begun a slow decline to early 2007 . intuitively contradictory finding. The correlation should be Throughout this period, electricity demand grew almost negative, indicating that demand is reduced by higher tar- continuously. Remittance levels recovered somewhat iff rates. As noted earlier in the examination of the meter during 2007 before declining again in late 2007 through database, the largest 500 customers' use appeared to late 2009. It is difficult to see a corresponding response in be unaffected by changes in tariff rates through 2007 and the demand for electricity. Correlation between quarterly into early 2009--and these customers represent almost remittances and quarterly electricity demand is very poor, 50 percent of electricity consumption. at 0.53, suggesting other more important factors are influ- encing electricity demand growth. To the extent there is The analysis of remittances and tariff rates suggests that any link between remittances and electricity demand, the other more important influences drive electricity demand. Governor of the National Reserve Bank of Tonga (NRBT) However, to the extent that world economic conditions expects remittance levels to remain below long-term affect electricity demand in Tonga, few would argue that averages into at least the first quarter of 2010.19 world economic activity will recover quickly from the cur- rent situation. Likewise, the current effort to reduce Ton- gan dependence on imported fuel oil may be expected to stabilize tariffs but, if the findings of the "Interim Tonga 18. NRBT Quarterly Bulletins. 19. September 2009 Statement of Monetary Policy, November 3, Energy Road Map Study" remain unchanged, Tonga 2009, NRBT. Figure 12: QuarTerly remiTTances Total remittances 140 12-month rolling average remittances Average monthly electricity use 120 Electricity use (kWh x 105) 100 Remittances (T$ million) 80 60 40 20 0 Q2 Q3 Q4 2008 Q2 Q3 Q4 2009 Q2 Q3 2007 2005 Q2 Q3 Q4 2006 Q2 Q3 Q4 2001 Q2 Q3 Q4 2002 Q2 Q3 Q4 2003 Q2 Q3 Q4 2004 Q2 Q3 Q4 Quarter Source: Compiled by author from TPL and NRBT records. A Forecast of Electricity Demand Growth 23 Figure 13: QuarTerly energy versus TariFF raTes, TongaTaPu 140 Average quarterly tariff Average quarterly electricity use 120 Electricity use (kWh x 105) 100 Tariff rates (T$001/kWh) 80 60 40 20 0 2002 Q2 Q3 Q4 2003 Q2 Q3 Q4 2004 Q2 Q3 Q4 2005 Q2 Q3 Q4 2006 Q2 Q3 Q4 2007 Q2 Q3 Q4 2008 Q2 Q3 Q4 2009 Q2 Q3 Quarter Source: Compiled by author from TPL and NRBT records. cannot expect any significant decline in tariff rates for This analysis assumes that TPL will ensure that both the many years. Popua and Taumu'aloto power stations quickly achieve and maintain station service losses of 2 percent of gross generation. At the two power stations serving Ha'apai 7.2 The Recent TPL Forecast and `Eua, station service was assumed to be higher at 3 percent, in recognition of the smaller installed capacities TPL prepared a single set of demand growth forecasts at those stations. for each of the grid systems in late 2008 and used these as the basis for its 10-year development plan for the util- Network or technical losses are the losses in lines and ity. These forecasts assume 3 percent annual growth in transformers; these losses are related to the capacity of demand for Tongatapu until 2012, when growth is pro- these facilities. Each component is designed to conduct jected to increase to 5 percent annually through 2019. On a defined level of power (voltage times current); when other grids, growth is assumed to be zero until 2010, and that level is exceeded, the loss of energy through that 2 percent annually thereafter. component increases very quickly. Typically, because the voltage is kept as close to the design level as possible, power flow can only increase by increasing the current. 7.3 System Losses Resistive losses, the largest component of loss, increase as the square of the current flow, so there is significant There are three types of losses, or electricity output, that benefit to ensuring equipment is not overloaded. do not create revenue, that are under the control of the utility to varying degrees: station service, network tech- TPL has taken over a system from an owner that under- nical losses, and nontechnical losses. Station service is invested in the networks. Many of the networks in the the energy needed to keep the generating units running villages on Tongatapu are significantly below the stan- by driving pumps, fans, and so forth, and to manage the dards adopted by other countries in the region. TPL has power station. This energy can be managed by the util- obtained the support of the New Zealand Agency for ity and, by and large, the statistics suggest that station International Development to rehabilitate and upgrade service on Tonga is 2­3 percent of gross generation. these village networks and, based on a preliminary 24 Kingdom of Tonga: Electric Supply System Load Forecast survey by the consultants,20 TPL is expecting to reduce percent. TPL has committed to reducing these losses as both technical and nontechnical losses at that end of the a first priority and has taken a more aggressive position system by up to 10 percent. TPL staff have also identi- on payment arrears and reducing the grace period for fied numerous sections of the transmission network for payments before disconnecting delinquent customers. which line capacities are too low for the power being car- In addition, TPL has changed out more than 1,000 meters ried and the insulators are too small for the rated voltage. that were known or suspected to be malfunctioning and By late 2009, TPL had completed a budget for network has suggested that a part of the decline in electricity use upgrading across all grids that calls for T$8.6 million observed during 2009 can be attributed to this new policy. (US$4.73 million) to be invested in system upgrades in The TPL meter records show that overdue bills reached 2009 through 2014, with another T$1.7 million (US$0.93 almost T$1.9 million (US$1.045 million) in late 2008 and million) thereafter to 2018. TPL has not had the capacity that these have since been reduced to about T$600,000 to survey the actual state of its entire network so is not (US$330,000). The project to upgrade village networks is able to explicitly link loss reductions to investments. expected to reduce meter tampering. The forecasts are based on TPL continuing its current aggressive stance on The cost of electricity on Tonga is high and the value of nontechnical loss reduction. reducing the losses is correspondingly high. If only the fuel component of the tariff is considered, the rate at late 2009 was T$0.4445/kWh, so saving 1 kWh per day for the 7.4 Demand-Side Management next 10 years has a present value of T$1,191 (US$655) at Programs 8 percent per year.21 If both components of the tariff are assumed to be saved, then the present value of 1 kWh per In a regional demand-side management (DSM) study,23 day would beT$2,229 (US$1,226). A reduction of 1 percent- the consultants investigated the potential to improve the age point in losses across all four grids would save about conversion efficiency of electricity to useful energy ser- 530,000 kWh and be worth T$1,730,400 (US$951,720) vices on several Pacific Island countries, including Tonga. in fuel savings or T$3,236,500 (US$1,285,075) of the full The consultants made two field visits to Tonga and their 's tariff over 10 years. A review of TPL network investment findings are summarized in this section. plans can be found in appendix 2. The range of potential DSM projects that were identified This analysis assumes that TPL will invest sufficient funds during the study include the following: to reduce the technical losses on the Tongatapu grid to 6 percent by 2020. This reduction halves the rate from its · The upgrading of streetlights to use either light-emit- current estimated level of about 12 percent. The analysis ting diodes (LEDs) or to incorporate dimming (volt- also assumes that TPL will concentrate its investment in age reduction) technology; network improvements over the next two to three years. · The reduction of demand in large government build- To achieve the reduction in network loss, TPL could ings through modifications to the cooling and venti- justify spending between T$8 million and T$16 million lation systems, changes in lighting, and upgrades to (US$4.4 million to US$8.8 million) on Tongatapu, depend- other equipment; and ing on how much of the total tariff TPL determines can be · Reducing lighting demand in the residential sector saved.22 The actual investments that can be justified will by replacing incandescent bulbs with compact fluo- depend on the borrowing rates and what proportion of rescent lights (CFLs). these investments must be funded out of equity. Currently, fewer than 60 percent of the approximately The third loss category is referred to as nontechnical 3,200 streetlights connected to the grids are working, losses and includes a wide range of losses that are vari- but the working ones are estimated to have consumed ously affected by the electricity consumer. Based on total 837 ,700 kWh or some 1.4 percent of electricity gener- 's losses and TPL estimate that technical losses are about ated in 2009. When all lights are working, their combined 12 percent on Tongatapu, nontechnical losses are about 3 demand will represent a little over 2 percent of electric- ity generated. The installation of LEDs is estimated to 20. Electronet Preliminary Report, November 2009. cost US$1.12 million and would reduce energy used by 21. The NRBT lists the prime interest rate as 9.7 percent during streetlights by 55 percent. Over 200,000 liters of fuel oil the third quarter of 2009. Multilateral funding agencies will provide would be saved each year and the investment would be investment funds at rates closer to 4 percent. 22. TPL is subject to target performance objectives set by the Elec- tricity Commission. If the loss reduction exceeds those objectives, 23. Promoting Energy Efficiency in the Pacific, by Econoler Interna- TPL will benefit by saving all of the tariff value of a kWh. tional for the ADB, 2009. A Forecast of Electricity Demand Growth 25 recovered in about three years. The contribution of street consumption by the end of 2012 on both the Tongatapu lighting to the daily peak demand would also be reduced and Vava'u grids by 10 percent, and by 5 percent on by 190 kW from some 346 kW to 156 kW. The light Ha'apai and `Eua. dimmer proposal would be less costly; it could reduce energy use during the off-peak period after midnight by about 30 percent but it would not affect the early eve- 7.5 Assumptions Used in ning peak demand. The cost is estimated to be less than the Forecast Scenarios US$220,000 and the investment would be recovered in less than two years. The assumptions used to prepare the load forecasts are listed in table 6. In all instances, growth was assumed The energy reduction activity in larger government to be slow from 2010 through 2012 and to accelerate buildings will need a more detailed evaluation for both between 2013 and 2015. The forecasts assume that planning and cost estimation because of current data lim- 's TPL investments will be increased during the next three itations. Based on experience elsewhere, the consultant years to ensure a rapid decline in technical losses. TPL estimates that energy use can be cut by up to 16 per- is expected to continue its efforts to limit nontechnical cent. With respect to replacing incandescent light bulbs losses but, based on the experience in other utilities with CFLs, additional surveys will be needed to confirm around the world, it is unrealistic to expect this type of the estimates of the number of incandescent bulbs in loss to fall below 1.5­2.0 percent of generated energy. use and then a well-planned public relations program The DSM program impacts are assumed to be linear over would need to be prepared with incentives to improve a three-year period beginning in 2010. the responses by householders. The total annual electricity savings if the three DSM proj- ects are implemented is estimated to be between 2.5 7.6 Forecasts of Energy and GWh and 2.8 GWh, which represents 4.0­4.5 percent of Peak Demand the generation forecast for all grids together in 2012. The median forecasts for each of the grid systems are The consultant identified at least three additional areas , presented in tables 7 8, 9, and 10. Figures 14, 15, 16, for which DSM programs could be effective but advised and 17 illustrate the energy forecasts for each system that data shortages precluded analysis at this time. The together with the high and low scenarios. The forecasts additional areas were energy use in the public water sup- assume that TPL implements the loss-reduction pro- ply systems; the improvement of load factors at large .3 grams described in section 7 and the DSM projects consumers; and broad efficiency improvements in the described in section 7.4. industrial, commercial, and religious sectors. A more detailed investigation program is advocated together Appendix 3 presents more detail of the forecasts, includ- with programs to increase public awareness. ing scenarios that provide estimates of demand if some or all of the loss-reduction and DSM programs are not Based on the foregoing, the base forecast includes an implemented. In total, four scenarios for each of the four aggressive DSM program that would reduce energy grid systems are summarized in appendix 3. 26 Kingdom of Tonga: Electric Supply System Load Forecast Table 6: summary oF key loaD ForecasT assumPTions Forecast factors 2009 2010 2011 2012 2013 2014 2015 2020 Tongatapu Technical loss rate (%) 12.00 11.00 10.00 9.00 8.00 7.50 7.00 6.00 Nontechnical loss rate (%) 3.95 3.00 2.50 2.25 2.15 2.05 1.95 1.50 Station service (%) 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 DSM program impact (%) 0 -1.67 -5.00 -8.33 -10.00 -10.00 -10.00 -10.00 Billed energy growth rate, low forecast (%) -7.82 0 1.00 2.00 3.00 3.00 3.00 3.50 Billed energy growth rate, median forecast (%) -7.82 1.00 2.00 3.00 4.00 4.00 5.00 5.00 Billed energy growth rate, high forecast (%) -7.82 2.00 3.00 4.00 4.50 5.00 5.50 5.50 Vava'u Technical loss rate (%) 9.50 9.00 8.50 8.00 7.50 7.25 7.00 6.50 Nontechnical loss rate (%) 3.21 3.00 2.75 2.50 2.40 2.20 2.00 1.75 Station service (%) 4.75 4.30 3.50 3.00 2.75 2.50 2.50 2.50 DSM program impact (%) 0 -1.67 -5.00 -8.33 -10.00 -10.00 -10.00 -10.00 Billed energy growth rate, low forecast (%) -4.50 0 1.00 2.00 3.00 3.00 3.00 3.50 Billed energy growth rate, median forecast (%) -4.50 1.00 2.00 3.00 4.00 4.00 5.00 5.00 Billed energy growth rate, high forecast (%) -4.50 2.00 3.00 4.00 4.50 5.00 5.50 5.50 Ha'apai Technical loss rate (%) 6.50 6.50 6.50 6.40 6.30 6.30 6.30 6.20 Nontechnical loss rate (%) 2.94 2.75 2.50 2.10 2.00 2.00 1.90 1.75 Station service (%) 3.13 3.00 3.00 3.00 3.00 3.00 3.00 3.00 DSM program impact (%) 0 -0.83 -2.50 -4.17 -5.00 -5.00 -5.00 -5.00 Billed energy growth rate, low forecast (%) -2.47 0 1.00 2.00 3.00 3.00 3.00 3.50 Billed energy growth rate, median forecast (%) -2.47 1.00 2.00 3.00 4.00 4.00 5.00 5.00 Billed energy growth rate, high forecast (%) -2.47 2.00 3.00 4.00 4.50 5.00 5.50 5.50 `Eua Technical loss late (%) 11.00 10.50 10.25 9.00 8.00 7.50 7.00 6.50 Nontechnical loss rate (%) 5.99 5.00 4.25 3.75 3.00 2.75 2.50 2.00 Station service (%) 2.69 3.00 3.00 3.00 3.00 3.00 3.00 3.00 DSM program impact (%) 0 -0.83 -2.50 -4.17 -5.00 -5.00 -5.00 -5.00 Billed energy growth rate, low forecast (%) -1.51 0 1.00 2.00 3.00 3.00 3.00 3.50 Billed energy growth rate, median forecast (%) -1.51 1.00 2.00 3.00 4.00 4.00 5.00 5.00 Billed energy growth rate, high forecast (%) -1.51 2.00 3.00 4.00 4.50 5.00 5.50 5.50 Source: Author. Table 7: meDian ForecasT, loss reDucTion Plus Dsm imPlemenTeD on TongaTaPu end of 2009 (actual) end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Median growth scenario Annual growth rate (%), billed energy -7.82 1.00 2.00 3.00 4.00 4.00 5.00 5.00 Billed energy (kWh) 37,012,005 37,382,125 38,129,768 39,273,661 40,844,607 42,478,391 44,602,311 56,925,107 DSM effect, % reduction of billed energy n.a. 1.67 5.00 8.33 10.00 10.00 10.00 10.00 Impact of DSM on billed energy (kWh) 0 -623,160 -1,906,488 -3,272,674 -4,084,461 -4,247,839 -4,460,231 -5,692,511 Net billed energy (kWh) 37,012,005 36,758,965 36,223,279 36,000,986 36,760,146 38,230,552 40,142,080 51,232,596 Technical losses (estimated kWh) 5,284,297 4,701,728 4,139,803 3,650,804 3,273,024 3,170,029 3,086,157 3,323,195 Technical loss as % of energy sent out 12.00 11.00 10.00 9.00 8.00 7.50 7.00 6.00 Nontechnical losses (estimated kWh) 1,739,510 1,282,289 1,034,951 912,701 879,625 866,475 859,715 830,799 Nontechnical loss as % of energy sent out 3.95 3.00 2.50 2.25 2.15 2.05 1.95 1.50 Total network losses (kWh) 7,023,807 5,984,018 5,174,754 4,563,505 4,152,649 4,036,504 3,945,872 4,153,994 Total network loss as % of energy sent out 15.95 14.00 12.50 11.25 10.15 9.55 8.95 7.50 Energy sent out from Popua (kWh) 44,035,812 42,742,983 41,398,033 40,564,492 40,912,795 42,267,056 44,087,951 55,386,591 Station service (kWh) 898,690 872,306 844,858 827,847 834,955 862,593 899,754 1,130,339 Station service as % of energy generated 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 Total energy generated (kWh) 44,934,502 43,615,288 42,242,891 41,392,339 41,747,750 43,129,649 44,987,706 56,516,929 Annual peak demand (kW) 8,142 7,903 7,654 7,500 7,565 7,815 8,152 10,241 Source: Author. Note: n.a. = Not applicable. A Forecast of Electricity Demand Growth 27 28 Table 8: meDian ForecasT, loss reDucTion Plus Dsm imPlemenTeD on vava'u end of 2009 (actual) end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Median growth scenario Annual growth rate (%), billed energy -4.50 1.00 2.00 3.00 4.00 4.00 5.00 5.00 Billed energy 4,064,748 4,105,395 4,187,503 4,313,128 4,485,653 4,665,079 4,898,333 6,251,652 DSM effect, % reduction of billed energy n.a. 1.667 5.000 8.333 10.000 10.000 10.000 10.000 Impact of DSM on billed energy (kWh) 0 -68,437 -209,375 -359,413 -448,565 -466,508 -489,833 -625,165 Net billed energy 4,064,748 4,036,958 3,978,128 3,953,715 4,037,088 4,198,571 4,408,500 5,626,487 Kingdom of Tonga: Electric Supply System Load Forecast Technical losses (estimated kWh) 442,361 412,871 381,004 353,405 336,051 336,164 339,115 398,607 Technical loss as % of energy sent out 9.50 9.00 8.50 8.00 7.50 7.25 7.00 6.50 Nontechnical losses (estimated kWh) 149,325 137,624 123,266 110,439 107,536 102,008 96,890 107,317 Nontechnical loss as % of energy sent out 3.21 3.00 2.75 2.50 2.40 2.20 2.00 1.75 Total network losses (kWh) 591,687 550,494 504,270 463,844 443,587 438,172 436,005 505,924 Total network loss as % of energy sent out 12.71 12.00 11.25 10.50 9.90 9.45 9.00 8.25 Energy sent out from Taumu'aloto (kWh) 4,656,434 4,587,452 4,482,398 4,417,559 4,480,675 4,636,744 4,844,505 6,132,411 Station service (kWh) 232,046 206,124 162,574 136,626 126,703 118,891 124,218 157,241 Station service as % of energy generated 4.75 4.30 3.50 3.00 2.75 2.50 2.50 2.50 Total energy generated (kWh) 4,888,480 4,793,576 4,644,972 4,554,184 4,607,378 4,755,635 4,968,724 6,289,652 Annual peak demand (kW) 1,041 1,021 989 970 981 1,013 1,058 1,340 Source: Author. Note: n.a. = Not applicable. Table 9: meDian ForecasT, loss reDucTion Plus Dsm imPlemenTeD on Ha'aPai end of 2009 (actual) end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Median growth scenario Annual growth rate (%), billed energy -2.47 1.00 2.00 3.00 4.00 4.00 5.00 5.00 Billed energy (kWh) 1,234,505 1,246,850 1,271,787 1,309,941 1,362,338 1,416,832 1,487,673 1,898,690 DSM effect,% reduction of billed energy n.a. 0.833 2.500 4.167 5.000 5.000 5.000 5.000 Impact of DSM on billed energy (kWh) 0 -10,390 -31,795 -54,581 -68,117 -70,842 -74,384 -94,935 Net billed energy (kWh) 1,234,505 1,236,460 1,239,992 1,255,359 1,294,221 1,345,990 1,413,290 1,803,756 Technical losses (estimated kWh) 88,606 88,687 88,882 88,541 89,796 93,326 97,928 122,088 Technical loss as % of energy sent out 6.50 6.50 6.50 6.40 6.30 6.30 6.30 6.20 Nontechnical losses (estimated kWh) 40,057 37,522 34,185 29,052 28,507 29,627 29,534 34,460 Nontechnical loss as % of energy sent out 2.94 2.75 2.50 2.10 2.00 2.00 1.90 1.75 Total network losses (kWh) 128,663 126,209 123,067 117,593 118,302 122,953 127,461 156,549 Total network loss as % of energy sent out 9.44 9.38 9.32 9.26 9.20 9.14 9.08 8.40 Energy sent out from Ha'apai (kWh) 1,363,168 1,364,423 1,367,415 1,383,446 1,425,331 1,481,365 1,554,407 1,969,166 Station service (kWh) 44,116 42,199 42,291 42,787 44,082 45,815 48,074 60,902 Station service as % of energy generated 3.13 3.00 3.00 3.00 3.00 3.00 3.00 3.00 Total energy generated (kWh) 1,407,284 1,406,621 1,409,706 1,426,233 1,469,413 1,527,180 1,602,481 2,030,068 Annual peak demand (kW) 312 300 300 304 313 325 341 432 Source: Author. Note: n.a. = Not applicable. A Forecast of Electricity Demand Growth 29 30 Table 10: meDian ForecasT, loss reDucTion Plus Dsm imPlemenTeD on `eua end of 2009 (actual) end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Median growth scenario Annual growth rate (%), billed energy -1.51 1.00 2.00 3.00 4.00 4.00 5.00 5.00 Billed energy (kWh) 880,692 889,499 907,289 934,508 971,888 1,010,764 1,061,302 1,354,520 DSM effect, % reduction of billed energy n.a. 0.833 2.500 4.167 5.000 5.000 5.000 5.000 Impact of DSM on billed energy (kWh) 0 -7,412 -22,682 -38,938 -48,594 -50,538 -53,065 -67,726 Net billed energy (kWh) 880,692 882,087 884,607 895,570 923,294 960,226 1,008,237 1,286,794 Kingdom of Tonga: Electric Supply System Load Forecast Technical losses (estimated kWh) 116,706 109,608 106,049 92,380 82,993 80,242 77,985 91,412 Technical loss as % of energy sent out 11.00 10.50 10.25 9.00 8.00 7.50 7.00 6.50 Nontechnical losses (estimated kWh) 63,564 52,195 43,972 38,492 31,122 29,422 27,852 28,127 Nontechnical loss as % of energy sent out 5.99 5.00 4.25 3.75 3.00 2.75 2.50 2.00 Total network losses (kWh) 180,270 161,803 150,021 130,871 114,115 109,664 105,837 119,538 Total network loss as % of energy sent out 16.99 15.50 14.50 12.75 11.00 10.25 9.50 8.50 Energy sent out from `Eua (kWh) 1,060,963 1,043,890 1,034,628 1,026,441 1,037,409 1,069,889 1,114,074 1,406,332 Station service (kWh) 29,282 32,285 31,999 31,746 32,085 33,089 34,456 43,495 Station service as % of energy generated 2.69 3.00 3.00 3.00 3.00 3.00 3.00 3.00 Total energy generated (kWh) 1,090,244 1,076,175 1,066,627 1,058,187 1,069,494 1,102,979 1,148,530 1,449,827 Annual peak demand (kW) 312 229 227 225 228 235 245 309 Source: Author. Note: n.a. = Not applicable. A Forecast of Electricity Demand Growth 31 Figure 14: HigH, meDian, anD low ForecasTs For TongaTaPu 65 Low scenario Annual electricity generated (GWh) 60 Median scenario High scenario 55 50 45 40 35 30 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source: Author. Figure 15: HigH, meDian, anD low ForecasTs For vava'u 7.0 Low scenario Annual electricity generated (GWh) 6.5 Median scenario High scenario 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source: Author. 32 Kingdom of Tonga: Electric Supply System Load Forecast Figure 16: HigH, meDian, anD low ForecasTs For Ha'aPai 2.4 Low scenario Annual electricity generated (GWh) 2.2 Median scenario High scenario 2.0 1.8 1.6 1.4 1.2 1.0 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source: Author. Figure 17: HigH, meDian, anD low ForecasTs For `eua 1.6 Low scenario Annual electricity generated (GWh) 1.5 Median scenario High scenario 1.4 1.3 1.2 1.1 1.0 0.9 0.8 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source: Author. 8 Capacity Planning Analyses This section details the schedule of new generating 8.1 Tongatapu capacity and the retirement of older equipment in accor- 's dance with TPL planning criteria and TPL recently devel- The 10-year capital budget prepared by TPL during late oped 10-year capital investment program. These findings 2008 includes an expectation that demand on Tonga- should be taken as a base case in which TPL continues to tapu will grow enough to require a second 2,880 kW use diesel generating units exclusively. MaK6CM32C medium-speed diesel unit by 2011 or 2012. The current forecasts indicate that additional capacity to TPL has provided guidance about when their existing, meet demand growth is unlikely to be needed before the older generating units will be retired, although TPL will latter half of the decade, about 2017­18 in the median undoubtedly take into account several factors as it makes forecast scenario. Based strictly on demand growth, the final decision about exactly when each unit is to be the second MaK unit would not likely be commissioned decommissioned. These factors will include how well the before 2019. In 2014, three of the existing 1,400kW Cat- unit is functioning, the pattern of maintenance require- erpillar 3516 units are expected to be retired and TPL ments in recent years, and the actual load growth that proposes to replace them with two 2,000k W Caterpillar is occurring. TPL will also take account of the need to C175 units. meet the fuel-efficiency regulations for each generating station. TPL has identified the likely types and capacities The schedules of capacity changes during the decade of new diesel units and these are used in this analysis. to 2020 for each of the forecast scenarios are listed in tables 11, 12, and 13. The tables list key dates during the Given the relatively slow demand growth expected during decade when existing generating units would need to the first few years of the upcoming decade, the sched- be retired and new units commissioned. The "Period" in ule of emplacement of new generating units tends to be each table is the year in which a new unit will have been driven as much by the need to replace units that must be in operation, or an old unit will have been retired, for a full retired as by the need to meet increasing demand. calendar year; thus, the actual commission or retirement occurs at some point in the previous calendar year. 33 34 Kingdom of Tonga: Electric Supply System Load Forecast Table 11: low ForecasT caPaciTy cHanges, TongaTaPu n-1 capacitya excess capacityb Period actions installed capacity (kw) (kw) (kw) 2014 No change from current (2009) 12,680 9,800 2,358 configuration 2015 3 Caterpillar 3516 retired; 12,680 ­ 4,200 + 4,000 = 12,480 9,600 1,985 2 Caterpillar C175 commissioned 2019 2 Caterpillar 3516 retired; 12,480 ­ 2,800 + 2,880 = 12,560 9,780 1,051 1 MaK6CM32C commissioned 2020 2 Caterpillar 3516 retired; 12,560 ­ 2,800 + 4,000 = 13,760 10,880 1,978 2 Caterpillar C175 commissioned Source: Author. Notes: a. The remaining capacity if the unit with the largest capacity is out of service. b. Capacity in excess of peak forecasted demand, with all units in service. Table 12: meDian ForecasT caPaciTy cHanges, TongaTaPu n-1 capacitya excess capacityb Period actions installed capacity (kw) (kw) (kw) 2014 No change from current (2009) 12,680 9,800 1,985 configuration 2015 3 Caterpillar 3516 retired; 2 Caterpillar 12,680 ­ 4,200 + 4,000 = 12,480 9,600 1,448 C175 commissioned 2017 1 Caterpillar C175 commissioned 12,480 + 2,000 = 14,480 11,600 2,669 2019 2 Caterpillar 3516 retired; 14,480 ­ 2,800 + 2,880 = 14,560 11,680 1,896 1 MaK6CM32C commissioned 2020 2 Caterpillar 3516 retired; 12,560 ­ 2,800 + 2,880 = 14,640 11,760 1,519 1 MaK6CM32C commissioned Source: Author. Notes: a. The remaining capacity if the unit with the largest capacity is out of service. b. Capacity in excess of peak forecasted demand, with all units in service. Capacity Planning Analyses 35 Table 13: HigH ForecasT caPaciTy cHanges, TongaTaPu n-1 capacitya excess capacityb Period actions installed capacity (kw) (kw) (kw) 2014 No change from current (2009) 12,680 9,800 1,636 configuration 2015 3 Caterpillar 3516 retired; 2 Caterpillar 12,680 ­ 4,200 + 4,000 = 12,480 9,600 1,044 C175 commissioned 2016 1 Caterpillar C175 commissioned 12,480 + 2,000 = 14,480 11,600 2,602 2019 2 Caterpillar 3516 retired; 14,480 ­ 2,800 + 2,880 = 14,560 11,680 1,214 1 MaK6CM32C commissioned 2020 2 Caterpillar 3516 retired; 12,560 ­ 2,800 + 2,880 = 14,640 11,760 753 1 MaK6CM32C commissioned Source: Author. Notes: a. The remaining capacity if the unit with the largest capacity is out of service. b. Capacity in excess of peak forecasted demand, with all units in service. 8.2 Vava'u Tables 14, 15, and 16 summarize the changes to the installed capacity at the Taumu'aloto power station on Two new units have been ordered for Vava'u and will be Vava'u for the low, median, and high forecast scenarios, installed as soon as they arrive. The units are 600 kW respectively. Cummins QSK23 generators and are expected to be commissioned during 2010. The two 186 kW Cummins LTA10G3 units currently at the Taumu'aloto power station 8.3 Ha'apai will be moved to the Ha'apai power station in 2010. A 400 kW Caterpillar unit on Ha'apai has been rented so TPL has not been able to achieve the N­1 reliability crite- that demand can be met with an adequate N­1 reliability rion on Vava'u since taking over in 2008. As soon as the criterion. This unit is expected to be retired or returned new Cummins units are commissioned, the criterion is to the rental company as soon as the two rehabilitated comfortably achievable. 186 kW Cummins LTA10G3 units now at Vava'u can be installed, expected to be during 2010. One of the existing 300 kW Cummins KTA19G2 units is expected to be retired by early 2013, and the two remain- Based on the forecasts, the schedule for installation of ing similar units will be retired late in 2016. Under the low new generating capacity will be dictated by the need to forecast scenario, a third 600kW Cummins QSK23 will replace the Cummins LTA10G3 units. Two sets are to be be needed for 2016, with a fourth unit added during 2016 retired in late 2014 and the remaining two in late 2017 . to replace the older Cummins units being retired. Under It is expected that as each pair of units is retired, they the median forecast, the third 600 kW Cummins QSK23 will be replaced with two 200 kW Cummins QSL9 units. would be commissioned late in 2013, but this unit would , Tables 17 18, and 19 list the capacity changes. be needed by late 2012. 36 Kingdom of Tonga: Electric Supply System Load Forecast Table 14: low ForecasT caPaciTy cHanges, vava'u n-1 capacitya excess capacityb Period actions installed capacity (kw) (kw) (kw) 2010 No change from current (2009) 1,272 972 ­39 configuration 2011 2 Cummins QSK23 commissioned; 1,272 ­ 372 + 1,200 = 2,100 1,500 530 2 Cummins LTA10G3 moved to Ha'apai 2013 1 Cummins KTA19G2 retired 2,100 ­ 300 = 1,800 1,200 256 2016 1 Cummins QSK23 commissioned 2,400 1,800 779 2017 2 Cummins KTA19G2 retired; 1,800 ­ 600 + 1,200 = 2,400 1,800 745 1 Cummins QSK23 commissioned 2020 No changes 2,400 1,800 636 Source: Author. Notes: a. The remaining capacity if the unit with the largest capacity is out of service. b. Capacity in excess of peak forecasted demand, with all units in service. Table 15: meDian ForecasT caPaciTy cHanges, vava'u n-1 capacitya excess capacityb Period actions installed capacity (kw) (kw) (kw) 2010 No change from current (2009) 1,272 972 ­49 configuration 2011 2 Cummins QSK23 commissioned; 1,272 ­ 372 + 1,200 = 2,100 1,500 511 2 Cummins LTA10G3 moved to Ha'apai 2013 1 Cummins KTA19G2 retired 2,100 ­ 300 = 1,800 1,200 219 2014 1 Cummins QSK23 commissioned 1,800 + 600 = 2,400 1,800 787 2017 2 Cummins KTA19G2 retired; 1,800 ­ 600 + 600 = 2,400 1,800 637 1 Cummins QSK23 commissioned 2020 No changes 2,400 1,800 460 Source: Author. Notes: a. The remaining capacity if the unit with the largest capacity is out of service. b. Capacity in excess of peak forecasted demand, with all units in service. Capacity Planning Analyses 37 8.4 `Eua 's demand. An important factor in TPL decision was the shape of the load curve, illustrated in figure 8, showing Only two 186 kW Cummins LTA10G3 units were on `Eua that the duration of the daily peak is short, typically less in 2009. TPL has made the conscious decision not to add than two hours. another unit until late 2011. Accordingly, the forecasted demand cannot be met if one unit becomes unavailable. By late 2012, TPL expects to be able to install a third If the high forecast comes to pass, the shortfall may Cummins LTA10G3 unit, but at least one of the existing reach 46 kW or some 20 percent of the estimated peak units is expected to be retired by late 2013. A 200kW Table 16: HigH ForecasT caPaciTy cHanges, vava'u n-1 capacitya excess capacityb Period actions installed capacity (kw) (kw) (kw) 2010 No change from current (2009) 1,272 972 ­59 configuration 2011 2 Cummins QSK23 commissioned; 1272 ­ 372 + 1,200 = 2,100 1,500 491 2 Cummins LTA10G3 moved to Ha'apai 2013 1 Cummins KTA19G2 retired; 2,100 ­ 300 +600 = 2,400 1,800 785 1 Cummins QSK23 commissioned 2017 2 Cummins KTA19G2 retired; 1,800 ­ 600 + 600 = 2,400 1,800 568 1 Cummins QSK23 commissioned 2020 No changes 2,400 1,800 360 Source: Author. Notes: a. The remaining capacity if the unit with the largest capacity is out of service. b. Capacity in excess of peak forecasted demand, with all units in service. Table 17: low ForecasT caPaciTy cHanges, Ha'aPai n-1 capacitya excess capacityb Period actions installed capacity (kw) (kw) (kw) 2010 No change from current (2009) 772 372 75 configuration 2011 2 Cummins LTA10G3 moved from 772 ­ 400 + 372 = 744 558 264 Vava'u; Rental unit retired 2015 2 Cummins LTA10G3 retired; 744 ­ 372 + 400 = 772 572 253 2 Cummins QSL9 commissioned 2018 2 Cummins LTA10G3 retired; 772 ­ 372 + 400 = 800 600 248 2 Cummins QSL9 commissioned 2020 No changes 800 600 224 Source: Author. Notes: a. The remaining capacity if the unit with the largest capacity is out of service. b. Capacity in excess of peak forecasted demand, with all units in service. 38 Kingdom of Tonga: Electric Supply System Load Forecast Cummins QSL9 set is to be commissioned by late 2013 N­1 reliability criterion by 2012 and maintain it under all to maintain the N­1 reliability criterion. The second and of the forecast scenarios through 2020. third LTA10G3 units are to be retired in late 2014 and 2015, respectively, and each will be replaced by a 200 Tables 20, 21, and 22 summarize the capacity changes kW QSL9 unit. The system is expected to achieve the for the low, median, and high forecast scenarios, respectively. Table 18: meDian ForecasT caPaciTy cHanges, Ha'aPai n-1 capacitya excess capacityb Period actions installed capacity (kw) (kw) (kw) 2010 No change from current (2009) 772 372 72 configuration 2011 2 Cummins LTA10G3 moved from 772 ­ 400 + 372 = 744 558 258 Vava'u; Rental unit retired 2015 2 Cummins LTA10G3 retired; 744 ­ 372 + 400 = 772 572 231 2 Cummins QSL9 commissioned 2018 2 Cummins LTA10G3 retired; 772 ­ 372 + 400 = 800 600 207 2 Cummins QSL9 commissioned 2020 No changes 800 600 168 Source: Author. Notes: a. The remaining capacity if the unit with the largest capacity is out of service. b. Capacity in excess of peak forecasted demand, with all units in service. Table 19: HigH ForecasT caPaciTy cHanges, Ha'aPai n-1 capacitya excess capacityb Period actions installed capacity (kw) (kw) (kw) 2010 No change from current (2009) 772 372 69 configuration 2011 2 Cummins LTA10G3 moved from 772 ­ 400 + 372 = 744 558 252 Vava'u; Rental unit retired 2015 2 Cummins LTA10G3 retired; 744 ­ 372 + 400 = 772 572 214 2 Cummins QSL9 commissioned 2018 2 Cummins LTA10G3 retired; 772 ­ 372 + 400 = 800 600 181 2 Cummins QSL9 commissioned 2020 No changes 800 600 135 Source: Author. Notes: a. The remaining capacity if the unit with the largest capacity is out of service. b. Capacity in excess of peak forecasted demand, with all units in service. Capacity Planning Analyses 39 Table 20: low ForecasT caPaciTy cHanges, `eua n-1 capacitya excess capacityb Period actions installed capacity (kw) (kw) (kw) 2011 No change from current (2009) 372 186 ­37 configuration 2012 1 Cummins LTA10G3 commissioned 372 + 186 = 558 372 153 2014 1 Cummins LTA10G3 retired; 558 ­ 186 + 200 = 572 372 148 1 Cummins QSL9 commissioned 2015 1 Cummins LTA10G3 retired; 572 ­ 186 + 200 = 586 386 157 1 Cummins QSL9 commissioned 2016 1 Cummins LTA10G3 retired; 586 ­ 186 + 200 = 600 400 164 1 Cummins QSL9 commissioned 2020 No changes 600 400 132 Source: Author. Notes: a. The remaining capacity if the unit with the largest capacity is out of service. b. Capacity in excess of peak forecasted demand, with all units in service. Table 21: meDian ForecasT caPaciTy cHanges, `eua n-1 capacitya excess capacityb Period actions installed capacity (kw) (kw) (kw) 2011 No change from current (2009) 372 186 ­41 configuration 2012 1 Cummins LTA10G3 commissioned 372 + 186 = 558 372 147 2014 1 Cummins LTA10G3 retired; 558 ­ 186 + 200 = 572 372 137 1 Cummins QSL9 commissioned 2015 1 Cummins LTA10G3 retired; 572 ­ 186 + 200 = 586 386 141 1 Cummins QSL9 commissioned 2016 1 Cummins LTA10G3 retired; 586 ­ 186 + 200 = 600 400 144 1 Cummins QSL9 commissioned 2020 No changes 600 400 91 Source: Author. Notes: a. The remaining capacity if the unit with the largest capacity is out of service. b. Capacity in excess of peak forecasted demand, with all units in service. 40 Kingdom of Tonga: Electric Supply System Load Forecast Table 22: HigH ForecasT caPaciTy cHanges, `eua n-1 capacitya excess capacityb Period actions installed capacity (kw) (kw) (kw) 2011 No change from current (2009) 372 186 -­46 configuration 2012 1 Cummins LTA10G3 commissioned 372 + 186 = 558 372 140 2014 1 Cummins LTA10G3 retired; 558 ­ 186 + 200 = 572 372 127 1 Cummins QSL9 commissioned 2015 1 Cummins LTA10G3 retired; 572 ­ 186 + 200 = 586 386 129 1 Cummins QSL9 commissioned 2016 1 Cummins LTA10G3 retired; 586 ­ 186 + 200 = 600 400 130 1 Cummins QSL9 commissioned 2020 No changes 600 400 68 Source: Author. Notes: a. The remaining capacity if the unit with the largest capacity is out of service. b. Capacity in excess of peak forecasted demand, with all units in service. 9 Base Case Energy Analysis This section summarizes the analyses conducted to during 2011 and thereafter. The energy analyses reflect assess the quantities of fuel oil that will be needed to this constraint in 2010 but assume that the constraint is supply the four grid systems, assuming that the electric lifted for all future years. load continues to be supplied entirely with diesel genera- tors fired with imported fuel oil. A simplified probability analysis was used to estimate the energy production of each unit at each of the power No consideration has been given to advancing new diesel stations operated by TPL. In the absence of historical units to improve the energy efficiency of the total gener- records of generator performance, the assurance of TPL ating system. Such an analysis would best be undertaken management that the units have operated well and have when the program of renewable resource development been well maintained has been used as a guide. Each for TPL has been finalized. TPL has also expressed inter- unit is assumed to have unplanned unavailability of one est in converting the medium-speed MaK unit to be able day per month and a planned outage of one week each to burn heavy fuel oil, and in the possibility of using com- year. This results in an annual unplanned failure rate of pressed natural gas in place of automotive gas oil. Nei- 3.35 percent and annual availability of 94.8 percent. ther of these options is considered in this analysis. Table 23 lists the assumptions made regarding the fuel Table 23: assumeD uniT caPaciTy anD efficiencies of the existing and new units. The MaK unit Fuel eFFiciencies in the Popua power station is being operated at about 2,400 kW because the air cooling system is unable to Fuel keep the unit from overheating. While this is appropriate capacity efficiency practice as long as the constraint remains, it does require unit (kw) (kwh/liter) more fuel than necessary to be used at Popua. Because Caterpillar--MaK6CM32C 2,400; 2,880 4.35 these MaK units enjoy a very flat efficiency curve over Caterpillar--Cat 3516B 1,400 3.80 the range of 74 percent to 100 percent of their rated out- Caterpillar--CATC175 2,000 4.20 put, this operating practice has little impact on the fuel efficiency of that unit. However, to the extent that the Cummins --KTA19G2 300 3.70 Caterpillar 3516B units are less efficient and are needed Cummins--LTA10G3 186 3.30 to produce more energy to make up for the MaK's Cummins--QSK23 600 4.00 deficiency, fuel use is higher than necessary at Popua. This analysis assumes that the cooling system will be Cummins--QSL9 200 3.90 upgraded during 2010 so that the best overall station effi- Source: Compiled by author from TPL records. ciency with the existing generating sets can be achieved 41 42 Kingdom of Tonga: Electric Supply System Load Forecast Tables 24, 25, 26, and 27 summarize the estimated TPL has been paying a premium to transport its fuel oil 's energy generation for each system for two of TPL action between Tongatapu and the other grids, these premiums plans--"Do Nothing" and "Efficient TPL Plus DSM"24-- have not been included in these tables. to illustrate the impact that TPL can have on its fuel use through the reduction of losses and the implementation As illustrated in table 24, the fuel used on Tongatapu cur- of a comprehensive DSM project. rently could be reduced by up to 18.6 percent through loss reduction and DSM programs. Fuel use can be Appendix 4 contains alternative projections of world oil reduced comparably for Vava'u but less for Ha'apai and prices and the associated landed price of fuel oil in Tonga. `Eua, in part because DSM is assumed to have a smaller The findings with regard to the expected cost of fuel oil impact on those two grids. must be considered tentative pending the results of the current Tonga Petroleum Supply Chain study. Also, while While the fuel efficiency for the Popua generating sta- tion (table 24) is consistently less than the target rate set by the regulator, TPL is expected to implement a new 24. There is no expectation that TPL will do nothing to improve its program that will include some renewable energy gen- operating efficiency, but this scenario can be used as the reference against which programs to reduce fuel use can be measured. The eration. This action will ensure that TPL meets the fuel "Do Nothing" scenario is believed to be close to the state of the efficiency targets. utility when it was taken over in 2008. Table 24: energy anD Fuel use analysis, meDian ForecasT, TongaTaPu median forecast Fuel Fuel Fuel cost Fuel cost Fuel savings vs. for electricity consumption efficiency reference oil reference oil "Do nothing" year production (kwh) (liters) (kwh/liter) scenario (T$) scenario (us$) scenario (%) 2010 43,615,288 10,701,850 4.075 10,701,850 5,886,018 4.10 2011 42,242,891 10,233,576 4.128 10,233,576 5,628,467 10.18 2012 41,392,339 10,013,794 4.134 10,013,794 5,507,587 14.79 2013 41,747,750 10,105,632 4.131 10,105,632 5,558,098 17.47 2014 43,129,649 10,462,712 4.122 10,462,712 5,754,492 17.99 2015 44,987,706 10,545,700 4.266 10,545,700 5,800,135 17.33 2016 47,088,059 11,022,007 4.272 11,022,007 6,062,104 17.82 2017 49,286,472 11,547,657 4.268 11,547,657 6,351,211 18.13 2018 51,587,523 12,099,179 4.264 12,099,179 6,654,548 18.44 2019 53,996,003 12,481,158 4.326 12,481,158 6,864,637 18.68 2020 56,516,929 13,007,642 4.345 13,007,642 7,154,203 18.60 Source: Author. Base Case Energy Analysis 43 Table 25: energy anD Fuel use analysis, meDian ForecasT, vava'u median forecast Fuel Fuel Fuel cost Fuel cost Fuel savings vs. for electricity consumption efficiency reference oil reference oil "Do nothing" year production (kwh) (liters) (kwh/liter) scenario (T$) scenario (us$) scenario (%) 2010 4,793,576 1,299,197 3.690 1,299,197 714,558 2.97 2011 4,644,972 1,162,302 3.996 1,162,302 639,266 7.80 2012 4,554,184 1,139,525 3.997 1,139,525 626,739 12.24 2013 4,607,378 1,151,920 4.000 1,151,920 633,556 14.71 2014 4,755,635 1,188,995 4.000 1,188,995 653,948 15.25 2015 4,968,724 1,242,285 4.000 1,242,285 683,257 15.67 2016 5,208,602 1,302,277 4.000 1,302,277 716,252 15.81 2017 5,460,062 1,365,015 4.000 1,365,015 750,759 15.94 2018 5,723,661 1,430,915 4.000 1,430,915 787,003 16.08 2019 5,999,987 1,499,997 4.000 1,499,997 824,998 16.21 2020 6,289,652 1,572,413 4.000 1,572,413 864,827 16.35 Source: Author. Table 26: energy anD Fuel use analysis, meDian ForecasT, Ha'aPai median forecast Fuel Fuel Fuel cost Fuel cost Fuel savings vs. for electricity consumption efficiency reference oil reference oil "Do nothing" year production (kwh) (liters) (kwh/liter) scenario (T$) scenario (us$) scenario (%) 2010 1,406,621 426,249 3.300 426,249 234,437 1.05 2011 1,409,706 427,184 3.300 427,184 234,951 2.78 2012 1,426,233 432,192 3.300 432,192 237,705 4.50 2013 1,469,413 445,277 3.300 445,277 244,902 5.40 2014 1,527,180 462,782 3.300 462,782 254,530 5.46 2015 1,602,481 411,763 3.892 411,763 226,470 5.55 2016 1,680,105 431,823 3.891 431,823 237,503 5.70 2017 1,761,489 452,886 3.889 452,886 249,087 5.85 2018 1,846,815 473,542 3.900 473,542 260,448 5.94 2019 1,936,275 496,481 3.900 496,481 273,064 6.08 2020 2,030,068 520,530 3.900 520,530 286,292 6.22 Source: Author. 44 Kingdom of Tonga: Electric Supply System Load Forecast Table 27: energy anD Fuel use analysis, meDian ForecasT, `eua median forecast Fuel Fuel Fuel cost Fuel cost Fuel savings vs. for electricity consumption efficiency reference oil reference oil "Do nothing" year production (kwh) (liters) (kwh/liter) scenario (T$) scenario (us$) scenario (%) 2010 1,076,175 326,114 3.300 326,114 179,363 4.04 2011 1,066,627 323,220 3.300 323,220 177,771 6.75 2012 1,058,187 320,663 3.300 320,663 176,364 10.19 2013 1,069,494 324,089 3.300 324,089 178,249 12.72 2014 1,102,979 285,880 3.858 285,880 157,234 13.70 2015 1,148,530 294,703 3.897 294,703 162,087 14.20 2016 1,203,309 308,541 3.900 308,541 169,697 14.35 2017 1,260,700 323,256 3.900 323,256 177,791 14.54 2018 1,320,829 338,674 3.900 338,674 186,271 14.73 2019 1,383,826 354,827 3.900 354,827 195,155 14.92 2020 1,449,827 371,751 3.900 371,751 204,463 15.10 Source: Author. APPENDIX 1 Interim Report Load Forecasts, April 2009 A1.1 The Tongatapu Load Forecasts A1.2 The Vava'u Load Forecasts, Interim Report Tables A1.1, A1.2, and A1.3 set out the three scenarios for the Tongatapu grid. Annual forecasts were made through Tables A1.4, A1.5, and A1.6 set out the three scenarios 2015 and a longer-term estimate is shown for 2020. The for the Vava'u grid. Annual forecasts were made through billed energy at the end of 2008 was calibrated such that 2015 and a longer-term estimate is shown for 2020. For the energy generated at the power station equals the the median and high scenarios, the billed energy at the record for December 2008. This adjustment was neces- end of 2008 was calibrated such that the energy gener- sary because the actual losses for 2008 are not the same ated at the power station equals the record for Decem- as have been assumed for the forecast period. ber 2008. Because there is little evidence of resumption in growth during the first quarter of 2009, for the low sce- Evidence from the monthly records indicates that growth nario, the billed energy for 2008 was reduced to reflect may have restarted during the first quarter of 2009. To this continuing decline in demand. reflect this possibility in the high scenario, the assumed billed energy for the end of 2008 was increased relative to the energy levels used in the median and low scenarios. Figure a.1: Peak DemanD ForecasTs For Of the losses of 15.6 percent on the energy sent out TongaTaPu, inTerim rePorT from Popua, 12 percentage points are assumed to be network related and it is assumed that TPL will imple- 13,000 Low ment an upgrade program to reduce these losses to Median about 10 percent in 2015 and by a further 1 percentage 12,000 High point by 2020. The nontechnical losses are interpreted to 's be about 3.2 percent and TPL loss-reduction program Peak demand (kw) 11,000 is expected to reduce these to 2 percent by 2020. As a result, overall losses are projected to decline from 15.6 10,000 percent to just over 11 percent by 2020. 9,000 Station service has averaged 2.6 percent over the histori- cal records and the same rate was assumed throughout 8,000 the forecast. The annual peak demand was calculated using the long-term annual load factor of 63 percent. 7,000 2008 2009 2010 2011 2012 2013 2014 2015 Figure A1.1 shows the forecasted annual peak demands Source: Author. for Tongatapu for each scenario. 45 46 Table a1.1: TongaTaPu low growTH loaD ForecasT scenario, inTerim rePorT indicator end of 2008 end of 2009 end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Low growth scenario Annual growth rate (%), billed energy -1.00 2.00 2.00 4.00 4.00 5.00 5.50 5.50 Billed energy (kWh) 40,243,565 39,841,129 40,637,952 41,450,711 43,108,739 44,833,089 47,074,743 49,663,854 64,908,671 Technical loss rate (%) 12.00 12.00 11.67 11.34 11.01 10.68 10.35 10.02 9.00 Nontechnical loss rate (%) 3.21 3.21 3.05 2.88 2.71 2.55 2.38 2.21 2.00 Total loss rate (%) 15.60 15.60 15.07 14.55 14.02 13.50 12.98 12.45 11.18 Energy sent out from Popua (kWh) 47,681,949 47,205,130 47,850,396 48,506,977 50,139,278 51,829,100 54,093,693 56,728,839 73,078,891 Station service (kWh) 8,711,208 8,624,096 8,489,765 8,351,114 8,368,959 8,379,539 8,462,929 8,579,307 10,120,991 Total energy generated (kWh) 48,954,773 48,465,225 49,127,717 49,801,825 51,477,698 53,212,628 55,537,673 58,243,162 75,029,663 Annual peak demand (kW) 8,871 8,782 8,902 9,024 9,328 9,642 10,063 10,554 13,595 Source: Author. Kingdom of Tonga: Electric Supply System Load Forecast Table a1.2: TongaTaPu meDian growTH loaD ForecasT scenario, inTerim rePorT indicator end of 2008 end of 2009 end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Median growth scenario Annual growth rate (%), billed energy 2.50 3.00 3.50 4.00 4.50 5.00 5.50 5.50 Billed energy (kWh) 40,243,565 41,249,654 42,487,144 43,974,194 45,733,162 47,791,154 50,180,711 52,940,651 69,191,313 Technical loss rate (%) 12.00 12.00 11.67 11.34 11.01 10.68 10.35 10.02 9.00 Nontechnical loss rate (%) 3.21 3.21 3.05 2.88 2.71 2.55 2.38 2.21 2.00 Total loss rate (%) 15.60 15.60 15.07 14.55 14.02 13.50 12.98 12.45 11.18 Energy sent out from Popua (kWh) 47,681,949 48,873,998 50,027,784 51,460,039 53,191,713 55,248,758 57,662,768 60,471,780 77,900,600 Station service (kWh) 8,711,208 8,928,988 8,876,084 8,859,522 8,878,454 8,932,417 9,021,309 9,145,366 10,788,769 Total energy generated (kWh) 48,954,773 50,178,642 51,363,227 52,833,716 54,611,615 56,723,570 59,202,021 62,086,016 79,980,082 Annual peak demand (kW) 8,871 9,092 9,307 9,573 9,896 10,278 10,727 11,250 14,492 Source: Author. Appendix 1. Interim Report Load Forecasts, April 2009 47 Of the losses of 14.8 percent on the energy sent out the energy generated at the power station equals the from Taumu'aloto, 12 percentage points are assumed record for December 2008. Indications are that growth to be network related and it is assumed that TPL will of demand may have restarted during the first quarter implement an upgrade program to reduce these to about of 2009. Hence, for the high scenario, the billed energy 10 percent in 2015 and by a further 1 percentage point for 2008 was increased modestly to reflect these higher by 2020. The nontechnical losses are interpreted to be demands. 's about 2.5 percent, and TPL loss-reduction program is expected to reduce these to 1.5 percent by 2020. As a Average losses on Ha'apai have been about 11 percent result, overall losses are projected to decline from 14.8 over the long term, but in the last 12 months they have percent to just about 10.5 percent by 2020. averaged only 10.5 percent. TPL believes network losses are the larger part of the total loss. It has been assumed Station service has averaged 4.83 percent over the his- that the breakdown between network and nontechni- torical records and this is believed to be higher than nec- cal losses is 8 percent in network losses and some 2.3 essary. Accordingly, it has been assumed that TPL will percent in nontechnical losses. A loss-reduction program work to reduce the station service to 3 percent of gen- should reduce network losses by about 0.5 percentage eration by 2014 and keep it at that level for the future. The 's points by 2015. TPL loss reduction program aimed at annual peak demand was calculated using the long-term nontechnical losses is expected to reduce these to 2.0 annual load factor of 53.6 percent. percent by 2020. As a result, overall losses are projected to decline from 10.5 percent to about 9.5 percent by Figure A1.2 shows the forecast annual peak demands for 2020. Vava'u for each scenario. Station service has averaged 2.85 percent over the his- torical records and this is believed to be about the right A1.3 The Ha'apai Load Forecasts, level. It was assumed that this loss will continue through Interim Report the forecast period. The annual peak demand was cal- culated using the long-term annual load factor of 52.4 Tables A1.7 A1.8, and A1.9 set out the three scenarios for , percent. Figure A1.3 shows the forecast annual peak the Ha'apai grid. For the low and median scenarios, the demands for Ha'apai for each scenario. billed energy at the end of 2008 was calibrated such that Figure a1.2: Peak DemanD ForecasTs For vava'u, Figure a1.3: Peak DemanD ForecasTs For inTerim rePorT Ha'aPai, inTerim rePorT 1,500 500 Low Low Median Median 1,400 High 450 High 1,300 Peak demand (kw) Peak demand (kw) 400 1,200 1,100 350 1,000 300 900 800 250 2008 2009 2010 2011 2012 2013 2014 2015 2008 2009 2010 2011 2012 2013 2014 2015 Source: Author. Source: Author. 48 Table a1.3: TongaTaPu HigH growTH loaD ForecasT scenario, inTerim rePorT indicator end of 2008 end of 2009 end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 High growth scenario Annual growth rate (%), billed energy 4.00 5.00 5.50 5.50 6.00 6.00 6.00 5.50 Billed energy (kWh) 40,700,000 42,328,000 44,444,400 46,888,842 49,467,728 52,435,792 55,581,940 58,916,856 77,001,974 Technical loss rate (%) 12.00 12.00 11.67 11.34 11.01 10.68 10.35 10.02 9.00 Nontechnical loss rate (%) 3.21 3.21 3.05 2.88 2.71 2.55 2.38 2.21 2.00 Total loss rate (%) 15.60 15.60 15.07 14.55 14.02 13.50 12.98 12.45 11.18 Energy sent out from Popua (kWh) 48,222,749 50,151,659 52,332,415 54,870,856 57,535,345 60,618,172 63,869,331 67,298,137 86,694,409 Station service (kWh) 8,810,009 9,162,409 9,284,978 9,446,739 9,603,468 9,800,524 9,992,323 10,177,740 12,006,659 Total energy generated (kWh) 49,510,009 51,490,409 53,729,378 56,335,581 59,071,196 62,236,316 65,574,262 69,094,596 89,008,634 Annual peak demand (kW) 8,971 9,330 9,736 10,208 10,704 11,277 11,882 12,520 16,128 Source: Author. Kingdom of Tonga: Electric Supply System Load Forecast Table a1.4: vava'u low loaD growTH ForecasT scenario, inTerim rePorT indicator end of 2008 end of 2009 end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Low growth scenario Annual growth rate (%), billed energy -2.00 2.00 2.00 4.00 4.00 5.00 5.50 5.50 Billed energy (kWh) 4,140,000 4,057,200 4,138,344 4,221,111 4,389,955 4,565,554 4,793,831 5,057,492 6,609,940 Technical loss rate (%) 12.00 12.00 11.64 11.29 10.95 10.62 10.30 10.00 9.00 Nontechnical loss rate (%) 2.50 2.50 2.33 2.17 2.00 1.84 1.67 1.50 1.50 Total loss rate (%) 14.80 14.80 14.25 13.70 13.17 12.65 12.15 11.65 10.64 Energy sent out from Taumu'aloto (kWh) 4,859,155 4,761,972 4,825,814 4,891,409 5,055,999 5,227,009 5,456,644 5,724,382 7,396,564 Station service (kWh) 965,763 946,448 912,220 877,825 860,303 842,629 831,575 843,933 1,015,384 Total energy generated (kWh) 5,105,763 5,003,648 5,050,564 5,098,936 5,250,259 5,408,183 5,625,406 5,901,425 7,625,324 Annual peak demand (kW) 1,087 1,066 1,076 1,086 1,118 1,152 1,198 1,257 1,624 Source: Author. Appendix 1. Interim Report Load Forecasts, April 2009 49 A1.4 The `Eua Load Forecasts, Figure a1.4: Peak DemanD ForecasTs For Interim Report `eua, inTerim rePorT Tables A1.10, A1.11, and A1.12 set out the three scenar- 430 Low ios for the `Eua grid. For the low and median scenarios, 410 Median the billed energy at the end of 2008 was calibrated such High 390 that the energy generated at the power station equals Peak demand (kw) the record for December 2008. Evidence indicates that 370 growth of demand may have restarted during the first 350 quarter of 2009. Hence, for the high scenario, the billed 330 energy for 2008 was increased modestly to reflect these 310 higher demands. 290 Average losses on `Eua have been about 13 percent over 270 the long term but in the last 12 months, they have appar- 250 ently increased to about 15 percent. TPL advises that the 2008 2009 2010 2011 2012 2013 2014 2015 loss data may be unreliable as a result of metering prob- lems. It was assumed that the breakdown between net- Source: Author. work and nontechnical losses is 12 percent in network losses and some 2.68 percent in nontechnical losses. A loss-reduction program should reduce network losses by about 1.5 percentage points by 2015 and by 2020 to Station service has varied widely at the `Eua power sta- 's 10 percent. TPL nontechnical loss reduction program tion but is now about 3 percent. It was assumed that this is assumed to reduce these to 2 percent by 2020. As loss will continue through the forecast period. The annual a result, overall losses are projected to decline from 15 peak demand was calculated using the long-term annual percent to just over 12 percent by 2020. load factor of 45.3 percent. Figure A1.4 shows the forecast annual peak demands for `Eua for each scenario. 50 Table a1.5: vava'u meDian loaD growTH ForecasT scenario, inTerim rePorT indicator end of 2008 end of 2009 end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Low growth scenario Annual growth rate (%), billed energy 0.00 2.00 3.50 4.00 4.50 5.50 5.50 5.50 Billed energy (kWh) 4,225,000 4,225,000 4,309,500 4,460,333 4,638,746 4,847,489 5,114,101 5,395,377 7,051,542 Technical loss rate (%) 12.00 12.00 11.64 11.29 10.95 10.62 10.30 10.00 9.00 Nontechnical loss rate (%) 2.50 2.50 2.33 2.17 2.00 1.84 1.67 1.50 1.50 Total loss rate (%) 14.80 14.80 14.25 13.70 13.17 12.65 12.15 11.65 10.64 Energy sent out from Taumu'aloto (kWh) 4,958,920 4,958,920 5,025,403 5,168,619 5,342,536 5,549,791 5,821,196 6,106,821 7,890,720 Station service (kWh) 985,592 985,592 949,948 927,574 909,059 894,664 887,132 900,315 1,083,221 Total energy generated (kWh) 5,210,592 5,210,592 5,259,448 5,387,906 5,547,805 5,742,153 6,001,233 6,295,692 8,134,763 Annual peak demand (kW) 1,110 1,110 1,120 1,147 1,182 1,223 1,278 1,341 1,733 Source: Author. Kingdom of Tonga: Electric Supply System Load Forecast Table a1.6: vava'u HigH loaD growTH ForecasT scenario, inTerim rePorT indicator end of 2008 end of 2009 end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 High growth scenario Annual growth rate (%), billed energy 1.00 3.00 4.00 5.00 5.50 6.00 6.00 5.50 Billed energy (kWh) 4,225,000 4,267,250 4,395,268 4,571,078 4,799,632 5,063,612 5,367,429 5,689,474 7,435,915 Technical loss rate (%) 12.00 12.00 11.64 11.29 10.95 10.62 10.30 10.00 9.00 Nontechnical loss rate (%) 2.50 2.50 2.33 2.17 2.00 1.84 1.67 1.50 1.50 Total loss rate (%) 14.80 14.80 14.25 13.70 13.17 12.65 12.15 11.65 10.64 Energy sent out from Taumu'aloto (kWh) 4,958,920 5,008,509 5,125,418 5,296,950 5,527,832 5,797,225 6,109,549 6,439,699 8,320,836 Station service (kWh) 985,592 995,448 968,854 950,605 940,588 934,552 931,076 949,391 1,142,266 Total energy generaated (kWh) 5,210,592 5,262,698 5,364,121 5,521,683 5,740,220 5,998,164 6,298,504 6,638,865 8,578,182 Annual peak demand (kW) 1,110 1,121 1,142 1,176 1,223 1,277 1,341 1,414 1,827 Source: Author. Table a1.7: Ha'aPai low loaD growTH ForecasT scenario, inTerim rePorT indicator end of 2008 end of 2009 end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Low growth scenario Annual growth rate (%), billed energy -1.00 2.00 2.00 4.00 4.00 5.00 5.50 5.50 Billed energy (kWh) 1,276,000 1,263,240 1,288,505 1,314,275 1,366,846 1,421,520 1,492,596 1,574,688 2,058,055 Technical loss rate (%) 8.00 8.00 7.90 7.80 7.70 7.60 7.50 7.40 7.40 Nontechnical loss rate (%) 2.31 2.31 2.26 2.21 2.16 2.11 2.06 2.01 2.00 Total loss rate (%) 10.50 10.50 10.34 10.19 10.03 9.88 9.72 9.56 9.55 Energy sent out from Ha'apai (kWh) 1,409,980 1,395,880 1,421,784 1,448,168 1,503,961 1,561,902 1,637,671 1,725,290 2,254,558 Station service (kWh) 174,164 172,423 173,800 175,165 179,978 184,897 191,749 199,773 260,758 Total energy generated (kWh) 1,450,164 1,435,663 1,462,305 1,489,440 1,546,824 1,606,417 1,684,345 1,774,461 2,318,813 Annual peak demand (kW) 316 313 319 324 337 350 367 387 505 Source: Author. Table a1.8: Ha'aPai meDian loaD growTH ForecasT scenario, inTerim rePorT indicator end of 2008 end of 2009 end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Median growth scenario Annual growth rate (%), billed energy 2.50 3.00 3.50 4.00 4.50 5.00 5.50 5.50 Billed energy (kWh) 1,276,000 1,307,900 1,347,137 1,394,287 1,450,058 1,515,311 1,591,076 1,678,586 2,193,844 Technical loss rate (%) 8.00 8.00 7.90 7.80 7.70 7.60 7.50 7.40 7.40 Nontechnical loss rate (%) 2.31 2.31 2.26 2.21 2.16 2.11 2.06 2.01 2.00 Total loss rate (%) 10.50 10.50 10.34 10.19 10.03 9.88 9.72 9.56 9.55 Energy sent out from Ha'apai (kWh) 1,409,980 1,445,230 1,486,481 1,536,331 1,595,521 1,664,956 1,745,724 1,839,124 2,403,313 Station service (kWh) 174,164 178,519 181,709 185,829 190,935 197,096 204,401 212,953 277,963 Total energy generated (kWh) 1,450,164 1,486,419 1,528,846 1,580,116 1,640,993 1,712,407 1,795,477 1,891,539 2,471,807 Annual peak demand (kW) 316 324 333 344 357 373 391 412 538 Appendix 1. Interim Report Load Forecasts, April 2009 Source: Author. 51 52 Table a1.9: Ha'aPai HigH loaD growTH ForecasT scenario, inTerim rePorT indicator end of 2008 end of 2009 end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 High growth scenario Annual growth rate (%), billed energy 4.00 5.00 5.50 5.50 6.00 6.00 6.00 5.50 Billed energy (kWh) 1,296,000 1,347,840 1,415,232 1,493,070 1,575,189 1,669,700 1,769,882 1,876,075 2,451,955 Technical loss rate (%) 8.00 8.00 8.00 8.00 8.00 8.00 8.00 8.00 8.00 Nontechnical loss rate (%) 2.31 2.31 2.31 2.31 2.31 2.31 2.31 2.31 2.31 Total loss rate (%) 10.50 10.50 10.50 10.50 10.50 10.50 10.50 10.50 10.50 Energy sent out from Ha'apai (kWh) 1,432,080 1,489,363 1,563,831 1,649,842 1,740,583 1,845,018 1,955,720 2,073,063 2,709,410 Station service (kWh) 176,894 183,970 193,169 203,793 215,001 227,902 241,576 256,070 334,673 Total energy generated (kWh) 1,472,894 1,531,810 1,608,401 1,696,863 1,790,190 1,897,601 2,011,458 2,132,145 2,786,628 Annual peak demand (kW) 321 334 350 370 390 413 438 464 607 Source: Author. Kingdom of Tonga: Electric Supply System Load Forecast Table a1.10: `eua low loaD growTH ForecasT scenario, inTerim rePorT indicator end of 2008 end of 2009 end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Low growth scenario Annual growth rate (%), billed energy -1.00 2.00 2.00 4.00 4.00 5.00 5.50 5.50 Billed energy (kWh) 962,000 952,380 971,428 990,856 1,030,490 1,071,710 1,125,296 1,187,187 1,551,606 Technical loss rate (%) 12.00 12.00 11.70 11.40 11.10 10.80 10.50 10.20 10.00 Nontechnical loss rate (%) 2.68 2.68 2.61 2.54 2.48 2.41 2.34 2.28 2.00 Total loss rate (%) 15.00 15.00 14.62 14.23 13.85 13.47 13.09 12.71 12.20 Energy sent out from `Eua (kWh) 1,106,300 1,095,237 1,113,422 1,131,901 1,173,240 1,216,079 1,272,593 1,338,064 1,740,902 Station service (kWh) 177,489 175,714 175,397 175,002 177,947 180,851 185,475 191,019 241,523 Total energy generated (kWh) 1,139,489 1,128,094 1,146,825 1,165,858 1,208,437 1,252,562 1,310,771 1,378,206 1,793,129 Annual peak demand (kW) 287 284 289 294 305 316 330 347 452 Source: Author. Table a1.11: `eua meDian loaD growTH scenario, inTerim rePorT indicator end of 2008 end of 2009 end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Median growth scenario Annual growth rate (%), billed energy 2.50 3.00 3.50 4.00 4.50 5.00 5.50 5.50 Billed energy (kWh) 962,000 986,050 1,015,632 1,051,179 1,093,226 1,142,421 1,199,542 1,265,517 1,653,980 Technical loss rate (%) 12.00 12.00 11.70 11.40 11.10 10.80 10.50 10.20 10.00 Nontechnical loss rate (%) 2.68 2.68 2.61 2.54 2.48 2.41 2.34 2.28 2.00 Total loss rate (%) 15.00 15.00 14.62 14.23 13.85 13.47 13.09 12.71 12.20 Energy sent out from `Eua (kWh) 1,106,300 1,133,958 1,164,088 1,200,810 1,244,666 1,296,315 1,356,558 1,426,349 1,855,765 Station service (kWh) 177,489 181,926 183,379 185,656 188,780 192,784 197,713 203,622 257,458 Total energy generated (kWh) 1,139,489 1,167,976 1,199,010 1,236,835 1,282,006 1,335,205 1,397,254 1,469,139 1,911,438 Annual peak demand (kW) 287 294 302 312 323 336 352 370 482 Source: Author. Table a1.12: `eua HigH loaD growTH scenario, inTerim rePorT indicator end of 2008 end of 2009 end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 High growth scenario Annual growth rate (%), billed energy 4.00 5.00 5.50 5.50 6.00 6.00 6.00 5.50 Billed energy (kWh) 970,000 1,008,800 1,059,240 1,117,498 1,178,961 1,249,698 1,324,680 1,404,161 1,835,182 Technical loss rate (%) 12.00 12.00 11.70 11.40 11.10 10.80 10.50 10.20 10.00 Nontechnical loss rate (%) 2.68 2.68 2.61 2.54 2.48 2.41 2.34 2.28 2.00 Total loss rate (%) 15.00 15.00 14.62 14.23 13.85 13.47 13.09 12.71 12.20 Energy sent out from `Eua (kWh) 1,115,500 1,160,120 1,214,070 1,276,570 1,342,277 1,418,044 1,498,076 1,582,613 2,059,074 Station service (kWh) 178,965 186,124 191,253 197,369 203,585 210,887 218,338 225,930 285,664 Total energy generated (kWh) 1,148,965 1,194,924 1,250,493 1,314,867 1,382,545 1,460,585 1,543,018 1,630,091 2,120,847 Annual peak demand (kW) 290 301 315 331 348 368 389 411 534 Appendix 1. Interim Report Load Forecasts, April 2009 Source: Author. 53 APPENDIX 2 Enterprise Capacity Enhancement for TPL A2.1 TPL's Existing Investment During 2008, TPL management directed all of its depart- Program ment managers to prepare 10-year budgets for their areas of responsibility. The primary objective was to sup- TPL became the new operator of the Tonga electricity 's port TPL request to the regulator for an adjustment to grid supply system in July 2008. TPL inherited a system 's the fixed portion of the tariff. However, TPL submission that had a backlog of important investments needed to also provided an opportunity for the utility to identify the reduce operating losses throughout the system. TPL work that would be required to maintain, upgrade, and is regulated by the Electricity Commission, which has grow the capacity of the system. This budget exercise established a set of target efficiencies for each of the was initially completed in 2008 but was revisited follow- grid systems. The profitability of TPL will be reduced if ing the hearings with the regulator, in part to reflect the it is unable to achieve these targets within the set time- 's extent to which TPL planned investments were allowed frame. Key objectives for system losses and fuel use per to be recovered through the tariff. The following review is unit of generated electricity will be tightened over time to based on the revised investment budgets. reflect the desire to improve the performance of the util- ity. The regulator has set the system loss target at 13 per- TPL has advised that notwithstanding the extent to which cent by January 2011, to remain at that level until 2015. the overall budgets may have been accepted during the Fuel efficiency targets are 4.25 kWh/liter for Tongatapu, 's regulatory hearings, TPL Board must approve invest- 3.75 for Vava'u, 3.70 for Ha'apai, and 3.70 for 'Eua. TPL ments for specific projects before their implementation. will have to go before the Electricity Commission before Because revenues declined during late 2008 through 2015 to apply for a new level for the fixed portion of the late 2009, the Board directed TPL management to limit tariff, at which time the commission is expected to set new investment as long as future revenues remain new targets for losses and fuel efficiency. unpredictable. Given that the total losses on Tongatapu in 2009 were The budgets for the network have been separated into almost 17 percent, including station service, TPL will .3 two main categories, maintenance and capital works. need to act quickly to achieve the target loss rate of 13 Over the 10-year period, TPL expects to spend an average percent. Likewise, with the exception of Vava'u, where of T$2.75 million (US$1.51 million) each year on general new units are scheduled to be installed during 2010, the maintenance and to invest T$19 million (US$10.45 mil- system is unlikely to achieve the fuel efficiency targets lion) in capital works. Over the 10-year planning horizon, without new generating equipment at the Popua power maintenance activities are expected to cost some T$27 .5 station and on `Eua. million (US$15.13 million) or about 45 percent more than investments in new works. 55 56 Kingdom of Tonga: Electric Supply System Load Forecast Table a2.1: summary oF buDgeTeD caPiTal invesTmenTs, T$ budget topic 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015­18 Growth-driven investments 566,722 621,534 320,210 345,912 118,144 109,624 109,624 Upgrades to resolve safety 407,926 382,366 187,471 217,269 25,560 25,560 38,340 concerns System improvements 2,547,128 2,582,160 1,327,036 893,991 1,233,152 626,777 1,031,597 Revenue meter replacement 1,359,969 0 0 0 0 0 0 Office equipment, tools, 339,651 385,070 359,642 259,642 283,570 261,142 1,104,854 vehicles, and the like Source: Compiled by author from TPL data. Table a2.2: summary oF buDgeTeD caPiTal invesTmenTs, us$ budget topic 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015­18 Growth-driven investments 19,759 311,697 341,844 176,116 190,252 64,979 60,293 Upgrades to resolve safety 31,631 224,360 210,302 103,109 119,498 14,058 21,087 concerns System improvements 217,903 1,400,920 1,420,188 729,870 491,695 678,233 567,378 Revenue meter replacement 0 747,983 0 0 0 0 0 Office equipment, tools, 241,570 186,808 211,789 197,803 142,803 155,964 607,670 vehicles, and the like Source: Compiled by author from TPL data. TPL has broken the capital investment budget into several A2.2 TPL's Network Investment Plans subtopics that can be grouped into five main categories: investment needed to meet growth, to replace danger- TPL has functioned primarily as an operating entity, with ous installations, to improve the network, to upgrade its staff focused on operating, maintaining, and repairing the revenue meters, and to add equipment for the TPL existing equipment. Because relatively little investment offices and staff. The budgeted items are similar for each has been made in new network facilities, the staff have of the grid systems, although in the category of invest- little experience with the management and implementa- ments to meet demand growth, specific new lines will tion of such projects. To meet the loss targets set by the be needed on Tongatapu whereas the work on the other regulator, new investment is likely required for new trans- grids is more generic. Tables A2.1 and A2.2 summarize mission lines, to upgrade existing lines and transformers, the budgeted annual investments. and to modify the operation of the system. New generat- ing equipment is being procured and existing generat- TPL has not been able to link the budgeted capital invest- ing equipment is being relocated. All these activities will ments to a level of loss reduction. It will be beneficial for involve planning and management on the part of teams future planning to upgrade metering and test equipment of TPL staff, as well as management and monitoring of to measure the performance of the existing network contractors. and to improve the capacity of TPL staff to conduct loss- reduction analyses. Appendix 2. Enterprise Capacity Enhancement for TPL 57 TPL has limited capacity to undertake the expected The most cost-effective training options will be to arrange workload during this period of system improvements. for selected staff to attend short courses at universities, TPL management is encouraged to commit to a fast- colleges, and utilities in the region, likely in New Zea- track program of skills upgrading for key personnel in the land, Australia, or on other Pacific Island countries as company. Such a program should be properly planned, the opportunities arise. Staff expenses, that is, travel, so it is recommended that a specialist be retained who accommodations, and per diem allowances, will prob- can evaluate the skills and capacity of personnel and then ably be the largest expense, but the cost to purchase develop a custom series of short and medium-length and maintain some of the more-sophisticated electricity courses for each involved staff member. The topics to be system planning and analysis software is likely to repre- covered should include the following: sent the largest cost over time. Course attendance fees are the least expensive part of training. · Planning, to encompass training on typical utility planning requirements, planning data requirements, TPL may wish to consider one or more longer courses database management, load-forecasting techniques, for key staff members who have the potential to develop generation system simulation methods, and net- into senior managers of the company. A careful assess- work planning, including appropriate computer soft- ment of corporate objectives, management needs, and ware for all stages of the planning process. Courses the capacity of existing potential candidates should be should include essential engineering economic anal- made with the assistance of a specialist advisor. ysis methods for system upgrades. · Project management requirements and techniques for all in-house planning work. · Project implementation, including training in project management, consultant and contractor manage- ment, contract monitoring, and related software to support these activities. APPENDIX 3 Details of the Load Forecasts In addition to developing a set of forecasts to indicate the High, low, and median forecasts were prepared for each expected range of future demands for electricity when of the efficiency scenarios described above. The detailed loss-reduction and DSM programs are implemented, sep- buildup of the median forecast for Tongatapu is shown arate forecasts were prepared of the potential demands in table A3.1 for the "Do Nothing" and table A3.2 for if the DSM program is not implemented and if there is no " the "Efficient TPL efficiency assumptions. Tables A3.3 effort to reduce losses. The scenarios have been named and A3.4 show the detailed buildup of the low and high "Do Nothing" to represent business continuing as it was scenarios for the "Efficient TPL Plus DSM" efficiency before July 2008, when TPL took over; "Efficient TPL to " assumptions. The details of the median forecast are represent what is expected to happen as TPL upgrades .6 contained in section 7 of the main report. Each of the the system; and "Plus DSM" to represent the final fore- median electricity forecasts for Tongatapu is illustrated in casts that include all reasonable projects to improve sys- figure A3.1. tem efficiency and so minimize fuel use while still relying on diesel generating equipment consuming imported Comparable tables and figures are included for each of fuels. the other three grids on Vava'u, Ha'apai, and `Eua. Figure a3.1: meDian energy ForecasTs For TongaTaPu 75 "Do nothing" scenario Annual electriity generatation (GWh) 70 "Efficient TPL" scenario "Plus DSN" scenario 65 60 55 50 45 40 35 30 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source: Author. 59 60 Kingdom of Tonga: Electric Supply System Load Forecast Figure a3.2: meDian energy ForecasTs For vava'u 8.0 "Do nothing" scenario Annual electriity generatation (GWh) 7.5 "Efficient TPL" scenario 7.0 "Plus DSN" scenario 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source: Author. Figure a3.3: meDian energy ForecasTs For Ha'aPai 2.4 "Do nothing" scenario Annual electriity generatation (GWh) 2.2 "Efficient TPL" scenario "Plus DSN" scenario 2.0 1.8 1.6 1.4 1.2 1.0 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source: Author. Figure a3.4: meDian energy ForecasTs For `eua 1.8 "Do nothing" scenario Annual electriity generatation (GWh) 1.7 "Efficient TPL" scenario 1.6 "Plus DSN" scenario 1.5 1.4 1.3 1.2 1.1 1.0 0.9 0.8 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source: Author. Table a3.1: "Do noTHing" meDian ForecasT, TongaTaPu end of 2009 indicator (actual end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Median growth scenario Annual growth rate (%), billed energy -7.82 1.00 2.00 3.00 4.00 4.00 5.00 5.00 Billed energy (kWh) 37,012,005 37,382,125 38,129,768 39,273,661 40,844,607 42,478,391 44,602,311 56,925,107 Technical losses (estimated kWh) 5,284,297 5,337,140 5,443,883 5,607,200 5,831,488 6,064,747 6,367,985 8,127,341 Technical loss as % of energy sent out 12.00 12.00 12.00 12.00 12.00 12.00 12.00 12.00 Nontechnical losses (estimated kWh) 1,739,510 1,756,905 1,792,043 1,845,804 1,919,637 1,996,422 2,096,243 2,675,396 Nontechnical loss as % of energy sent out 3.95 3.95 3.95 3.95 3.95 3.95 3.95 3.95 Total network losses (kWh) 7,023,807 7,094,045 7,235,926 7,453,004 7,751,124 8,061,169 8,464,228 10,802,738 Total network loss as % of energy sent out 15.95 15.95 15.95 15.95 15.95 15.95 15.95 15.95 Energy sent out from Popua (kWh) 44,035,812 44,476,170 45,365,694 46,726,665 48,595,731 50,539,561 53,066,539 67,727,845 Station service (kWh) 898,690 907,677 925,830 953,605 991,750 1,031,420 1,082,991 1,382,201 Station service as % of energy generated 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 Total energy generated (kWh) 44,934,502 45,383,847 46,291,524 47,680,270 49,587,481 51,570,980 54,149,529 69,110,046 Annual peak demand (kW) 8,142 8,223 8,388 8,640 8,985 9,345 9,812 12,523 Source: Author. Appendix 3. Details of the Load Forecasts 61 62 Table a3.2: "eFFicienT TPl" meDian ForecasT, TongaTaPu end of 2009 indicator (actual end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Median growth scenario Annual growth rate (%), billed energy -7.82 1.00 2.00 3.00 4.00 4.00 5.00 5.00 Billed energy (kWh) 37,012,005 37,382,125 38,129,768 39,273,661 40,844,607 42,478,391 44,602,311 56,925,107 Technical losses (estimated kWh) 5,284,297 4,781,435 4,357,688 3,982,681 3,636,693 3,522,255 3,429,063 3,692,439 Technical loss as % of energy sent out 12.00 11.00 10.00 9.00 8.00 7.50 7.00 6.00 Kingdom of Tonga: Electric Supply System Load Forecast Nontechnical losses (estimated kWh) 1,739,510 1,304,028 1,089,422 995,670 977,361 962,750 955,239 923,110 Nontechnical loss as % of energy sent out 3.95 3.00 2.50 2.25 2.15 2.05 1.95 1.50 Total network losses (kWh) 7,023,807 6,085,462 5,447,110 4,978,351 4,614,054 4,485,004 4,384,302 4,615,549 Total network loss as % of energy sent out 15.95 14.00 12.50 11.25 10.15 9.55 8.95 7.50 Energy sent out from Popua (kWh) 44,035,812 43,467,587 43,576,877 44,252,012 45,458,661 46,963,396 48,986,613 61,540,656 Station service (kWh) 898,690 887,094 889,324 903,102 927,728 958,437 999,727 1,255,932 Station service as % of energy generated 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 Total energy generated (kWh) 44,934,502 44,354,681 44,466,201 45,155,114 46,386,389 47,921,832 49,986,339 62,796,588 Annual peak demand (kW) 8,142 8,037 8,057 8,182 8,405 8,683 9,057 11,379 Source: Author. Table a3.3: "eFFicienT TPl Plus Dsm" low ForecasT, TongaTaPu end of 2009 indicator (actual end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Low growth scenario Annual growth rate (%), billed energy -7.82 0 1.00 2.00 3.00 3.00 3.00 3.50 Billed energy (kWh) 37,012,005 37,012,005 37,382,125 38,129,768 39,273,661 40,451,870 41,665,427 49,485,456 DSM effect, % reduction of billed energy n.a. 1.667 5.000 8.333 10.000 10.000 10.000 10.000 Impact of DSM on billed energy (kWh) 0 -616,990 -1,869,106 -3,177,354 -3,927,366 -4,045,187 -4,166,543 -4,948,546 Net billed energy (kWh) 37,012,005 36,395,015 35,513,019 34,952,414 35,346,295 36,406,683 37,498,884 44,536,911 Technical losses (estimated kWh) 5,284,297 4,655,176 4,058,631 3,544,470 3,147,138 3,018,796 2,882,945 2,888,881 Technical loss as % of energy sent out 12.00 11.00 10.00 9.00 8.00 7.50 7.00 6.00 Nontechnical losses (estimated kWh) 1,739,510 1,269,594 1,014,658 886,118 845,793 825,138 803,106 722,220 Nontechnical loss as % of energy sent out 3.95 3.00 2.50 2.25 2.15 2.05 1.95 1.50 Total network losses (kWh) 7,023,807 5,924,770 5,073,288 4,430,588 3,992,931 3,843,934 3,686,052 3,611,101 Total network loss as % of energy sent out 15.95 14.00 12.50 11.25 10.15 9.55 8.95 7.50 Energy sent out from Popua (kWh) 44,035,812 42,319,785 40,586,307 39,383,002 39,339,226 40,250,617 41,184,936 48,148,012 Station service (kWh) 898,690 863,669 828,292 803,735 802,841 821,441 840,509 982,612 Station service as % of energy generated 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 Total energy generated (kWh) 44,934,502 43,183,454 41,414,599 40,186,736 40,142,067 41,072,058 42,025,444 49,130,624 Annual peak demand (kW) 8,142 7,825 7,504 7,282 7,274 7,442 7,615 8,902 Source: Author. Note: n.a. = Not applicable. Appendix 3. Details of the Load Forecasts 63 64 Table a3.4: "eFFicienT TPl Plus Dsm" HigH ForecasT, TongaTaPu end of 2009 indicator (actual end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 High growth scenario Annual growth rate (%), billed energy -7.82 2.00 3.00 4.00 4.50 5.00 5.50 5.50 Billed energy (kWh) 37,012,005 37,752,245 38,884,812 40,440,205 42,260,014 44,373,015 46,813,531 61,183,412 DSM effect, % reduction of billed energy n.a. 1.667 5.000 8.333 10.000 10.000 10.000 10.000 Impact of DSM on billed energy (kWh) 0 -629,330 -1,944,241 -3,369,882 -4,226,001 -4,437,301 -4,681,353 -6,118,341 Net billed energy (kWh) 37,012,005 37,122,915 36,940,572 37,070,323 38,034,013 39,935,713 42,132,178 55,065,071 Technical losses (estimated kWh) 5,284,297 4,748,280 4,221,780 3,759,244 3,386,445 3,311,419 3,239,157 3,571,788 Kingdom of Tonga: Electric Supply System Load Forecast Techni 3 cal loss as % of energy sent out 12.00 11.00 10.00 9.00 8.00 7.50 7.00 6.00 Nontechnical losses (estimated kWh) 1,739,510 1,294,985 1,055,445 939,811 910,107 905,121 902,337 892,947 Nontechnical loss as % of energy sent out 3.95 3.00 2.50 2.25 2.15 2.05 1.95 1.50 Total network losses (kWh) 7,023,807 6,043,265 5,277,225 4,699,055 4,296,552 4,216,540 4,141,494 4,464,735 Total network loss as % of energy sent out 15.95 14.00 12.50 11.25 10.15 9.55 8.95 7.50 Energy sent out from Popua (kWh) 44,035,812 43,166,180 42,217,796 41,769,378 42,330,565 44,152,254 46,273,671 59,529,807 Station service (kWh) 898,690 880,942 861,588 852,436 863,889 901,066 944,361 1,214,894 Station service as % of energy generated 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 Total energy generated (kWh) 44,934,502 44,047,123 43,079,384 42,621,814 43,194,454 45,053,320 47,218,032 60,744,701 Annual peak demand (kW) 8,142 7,981 7,806 7,723 7,827 8,164 8,556 11,007 Source: Author. Note: n.a. = Not applicable. Table a3.5: "Do noTHing" meDian ForecasT, vava'u end of 2009 indicator (actual end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Median growth scenario Annual growth rate (%), billed energy -4.50 1.00 2.00 3.00 4.00 4.00 5.00 5.00 Billed energy (kWh) 4,064,748 4,105,395 4,187,503 4,313,128 4,485,653 4,665,079 4,898,333 6,251,652 Technical losses (estimated kWh) 442,361 446,801 455,737 469,409 488,185 507,713 533,098 680,384 Technical loss as % of energy sent out 9.50 9.50 9.50 9.50 9.50 9.50 9.50 9.50 Nontechnical losses (estimated kWh) 149,472 150,972 153,991 158,611 164,955 171,553 180,131 229,898 Nontechnical loss as % of energy sent out 3.21 3.21 3.21 3.21 3.21 3.21 3.21 3.21 Total network losses (kWh) 591,833 597,773 609,728 628,020 653,141 679,266 713,230 910,282 Total network loss as % of energy sent out 25.00 12.71 12.71 12.71 12.71 12.71 12.71 12.71 Energy sent out from Taumu'aloto (kWh) 4,656,434 4,703,168 4,797,231 4,941,148 5,138,794 5,344,346 5,611,563 7,161,934 Station service (kWh) 232,046 234,541 239,232 246,409 256,265 266,516 279,842 357,157 Station service as % of energy generated 4.75 4.75 4.75 4.75 4.75 4.75 4.75 4.75 Total energy generated (kWh) 4,888,480 4,937,709 5,036,463 5,187,557 5,395,059 5,610,862 5,891,405 7,519,091 Annual peak demand (kW) 1,041 1,052 1,073 1,105 1,149 1,195 1,255 1,601 Source: Author. Appendix 3. Details of the Load Forecasts 65 66 Table a3.6: "eFFicienT TPl" meDian ForecasT, vava'u end of 2009 indicator (actual end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Median growth scenario Annual growth rate (%), billed energy -4.50 1.00 2.00 3.00 4.00 4.00 5.00 5.00 Billed energy (kWh) 4,064,748 4,105,395 4,187,503 4,313,128 4,485,653 4,665,079 4,898,333 6,251,652 Technical losses (estimated kWh) 442,361 419,870 401,057 385,531 373,390 373,515 376,795 442,896 Technical loss as % of energy sent out 9.50 9.00 8.50 8.00 7.50 7.25 7.00 6.50 Kingdom of Tonga: Electric Supply System Load Forecast Nontechnical losses (estimated kWh) 149,325 139,957 129,754 120,478 119,485 113,343 107,656 119,241 Nontechnical loss as % of energy sent out 3.21 3.00 2.75 2.50 2.40 2.20 2.00 1.75 Total network losses (kWh) 591,687 559,827 530,810 506,009 492,874 486,858 484,451 562,138 Total network loss as % of energy sent out 12.71 12.00 11.25 10.50 9.90 9.45 9.00 8.25 Energy sent out from Taumu'aloto (kWh) 4,656,434 4,665,222 4,718,313 4,819,138 4,978,527 5,151,937 5,382,784 6,813,790 Station service (kWh) 232,046 209,618 171,131 149,045 140,781 132,101 138,020 174,713 Station service as % of energy generated 4.75 4.30 3.50 3.00 2.75 2.50 2.50 2.50 Total energy generated (kWh) 4,888,480 4,874,840 4,889,444 4,968,183 5,119,308 5,284,038 5,520,804 6,988,503 Annual peak demand (kW) 1,041 1,038 1,041 1,058 1,090 1,125 1,176 1,488 Source: Author. Table a3.7: "eFFicienT TPl Plus Dsm" low ForecasT, vava'u end of 2009 indicator (actual end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Low growth scenario Annual growth rate (%), billed energy -4.50 0.00 1.00 2.00 3.00 3.00 3.00 3.50 Billed energy (kWh) 4,064,748 4,064,748 4,105,395 4,187,503 4,313,128 4,442,522 4,575,798 5,434,612 DSM effect, % reduction of billed energy n.a. 1.667 5.000 8.333 10.000 10.000 10.000 10.000 Impact of DSM on billed energy (kWh) 0 -67,759 -205,269 -348,945 -431,313 -444,252 -457,580 -543,461 Net billed energy (kWh) 4,064,748 3,996,989 3,900,126 3,838,558 3,881,815 3,998,270 4,118,218 4,891,151 Technical losses (estimated kWh) 442,361 415,713 393,193 374,302 359,028 355,696 351,984 385,013 Technical loss as % of energy sent out 9.50 9.00 8.50 8.00 7.50 7.25 7.00 6.50 Nontechnical losses (estimated kWh) 149,325 138,571 127,209 116,969 114,889 107,935 100,567 103,657 Nontechnical loss as % of energy sent out 3.21 3.00 2.75 2.50 2.40 2.20 2.00 1.75 Total network losses (kWh) 591,687 554,284 520,402 491,271 473,918 463,631 452,551 488,671 Total network loss as % of energy sent out 12.71 12.00 11.25 10.50 9.90 9.45 9.00 8.25 Energy sent out from Taumu'aloto (kWh) 4,656,434 4,619,031 4,625,797 4,678,774 4,787,046 4,906,153 5,028,349 5,923,283 Station service (kWh) 232,046 207,543 167,775 144,704 135,366 125,799 128,932 151,879 Station service as % of energy generated 4.75 4.30 3.50 3.00 2.75 2.50 2.50 2.50 Total energy generated (kWh) 4,888,480 4,826,574 4,793,572 4,823,479 4,922,412 5,031,952 5,157,281 6,075,162 Annual peak demand (kW) 1,041 1,028 1,021 1,027 1,048 1,072 1,098 1,294 Source: Author. Note: n.a. = Not applicable. Appendix 3. Details of the Load Forecasts 67 68 Table a3.8: "eFFicienT TPl Plus Dsm" HigH ForecasT, vava'u end of 2009 indicator (actual end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 High growth scenario Annual growth rate (%), billed energy -4.50 2.00 3.00 4.00 4.50 5.00 5.50 5.50 Billed energy (kWh) 4,064,748 4,146,043 4,270,424 4,441,241 4,641,097 4,873,151 5,141,175 6,719,310 DSM effect, % reduction of billed energy n.a. 1.667 5.000 8.333 10.000 10.000 10.000 10.000 Impact of DSM on billed energy (kWh) 0 -69,115 -213,521 -370,089 -464,110 -487,315 -514,118 -671,931 Net billed energy (kWh) 4,064,748 4,076,928 4,056,903 4,071,152 4,176,987 4,385,836 4,627,057 6,047,379 Technical losses (estimated kWh) 442,361 424,027 408,998 396,982 386,329 390,175 395,475 476,027 Kingdom of Tonga: Electric Supply System Load Forecast Technical loss as % of energy sent out 9.50 9.00 8.50 8.00 7.50 7.25 7.00 6.50 Nontechnical losses (estimated kWh) 149,325 141,342 132,323 124,057 123,625 118,398 112,993 128,161 Nontechnical loss as % of energy sent out 3.21 3.00 2.75 2.50 2.40 2.20 2.00 1.75 Total network losses (kWh) 591,687 565,369 541,321 521,039 509,954 508,573 508,468 604,189 Total network loss as % of energy sent out 12.71 12.00 11.25 10.50 9.90 9.45 9.00 8.25 Energy sent out from Taumu'aloto (kWh) 4,656,434 4,711,412 4,811,745 4,962,280 5,151,051 5,381,724 5,649,643 7,323,498 Station service (kWh) 232,046 211,694 174,519 153,473 145,660 137,993 144,863 187,782 Station service as % of energy generated 4.75 4.30 3.50 3.00 2.75 2.50 2.50 2.50 Total energy generated (kWh) 4,888,480 4,923,106 4,986,264 5,115,753 5,296,710 5,519,717 5,794,505 7,511,280 Annual peak demand (kW) 1,041 1,049 1,062 1,090 1,128 1,176 1,234 1,600 Source: Author. Note: n.a. = Not applicable. Table a3.9: "Do noTHing" meDian ForecasT, Ha'aPai end of 2009 indicator (actual end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Median growth scenario Annual growth rate (%), billed energy -2.47 1.00 2.00 3.00 4.00 4.00 5.00 5.00 Billed energy (kWh) 1,234,505 1,246,850 1,271,787 1,309,941 1,362,338 1,416,832 1,487,673 1,898,690 Technical losses (estimated kWh) 88,606 89,491 91,281 94,020 97,781 101,692 106,776 136,277 Technical loss as % of energy sent out 6.50 6.50 6.50 6.50 6.50 6.50 6.50 6.50 Nontechnical losses (estimated kWh) 40,050 40,450 41,259 42,497 44,197 45,965 48,263 61,597 Nontechnical loss as % of energy sent out 2.94 2.94 2.94 2.94 2.94 2.94 2.94 2.94 Total network losses (kWh) 128,656 129,942 132,540 136,517 141,977 147,656 155,039 197,874 Total network loss as % of energy sent out 9.44 9.44 9.44 9.44 9.44 9.44 9.44 9.44 Energy sent out from Ha'apai (kWh) 1,363,168 1,376,792 1,404,327 1,446,457 1,504,316 1,564,488 1,642,713 2,096,564 Station service (kWh) 44,116 44,779 45,675 47,045 48,927 50,884 53,428 68,190 Station service as % of energy generated 3.13 3.15 3.15 3.15 3.15 3.15 3.15 3.15 Total energy generated (kWh) 1,407,284 1,421,571 1,450,003 1,493,503 1,553,243 1,615,372 1,696,141 2,164,754 Annual peak demand (kW) 312 303 309 318 331 344 361 461 Source: Author. Appendix 3. Details of the Load Forecasts 69 70 Table a3.10: "eFFicienT TPl" meDian ForecasT, Ha'aPai end of 2009 indicator (actual end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Median growth scenario Annual growth rate (%), billed energy -2.47 1.00 2.00 3.00 4.00 4.00 5.00 5.00 Billed energy (kWh) 1,234,505 1,246,850 1,271,787 1,309,941 1,362,338 1,416,832 1,487,673 1,898,690 Technical losses (estimated kWh) 88,606 89,306 90,842 91,624 93,596 97,340 102,095 127,886 Technical loss as % of energy sent out 6.50 6.50 6.50 6.40 6.30 6.30 6.30 6.20 Kingdom of Tonga: Electric Supply System Load Forecast Nontechnical losses (estimated kWh) 40,057 37,783 34,939 30,064 29,713 30,901 30,791 36,097 Nontechnical loss as % of energy sent out 2.94 2.75 2.50 2.10 2.00 2.00 1.90 1.75 Total network losses (kWh) 128,663 127,089 125,781 121,688 123,309 128,241 132,886 163,982 Total network loss as % of energy sent out 9.44 9.25 9.00 8.50 8.30 8.30 8.20 7.95 Energy sent out from Ha'apai (kWh) 1,363,168 1,373,939 1,397,568 1,431,629 1,485,647 1,545,073 1,620,559 2,062,673 Station service (kWh) 44,116 42,493 43,224 44,277 45,948 47,786 50,120 63,794 Station service as % of energy generated 3.13 3.00 3.00 3.00 3.00 3.00 3.00 3.00 Total energy generated (kWh) 1,407,284 1,416,432 1,440,792 1,475,906 1,531,595 1,592,859 1,670,680 2,126,467 Annual peak demand (kW) 312 302 307 314 326 339 356 453 Source: Author. Table a3.11: "eFFicienT TPl Plus Dsm" low ForecasT, Ha'aPai end of 2009 indicator (actual end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Low growth scenario Annual growth rate (%), billed energy Source: 0.00 1.00 2.00 3.00 3.00 3.00 3.50 Author. Billed energy (kWh) Note: n.a. 1,234,505 1,246,850 1,271,787 1,309,941 1,349,239 1,389,716 1,650,547 = Not applicable. DSM effect, % reduction of billed energy n.a. 0.833 2.500 4.167 5.000 5.000 5.000 5.000 Impact of DSM on billed energy (kWh) 0 -10,287 -31,171 -52,992 -65,497 -67,462 -69,486 -82,527 Net billed energy (kWh) 1,234,505 1,224,218 1,215,679 1,218,796 1,244,444 1,281,777 1,320,230 1,568,019 Technical losses (estimated kWh) 88,606 87,809 87,139 85,962 86,342 88,874 91,479 106,132 Technical loss as % of energy sent out 6.50 6.50 6.50 6.40 6.30 6.30 6.30 6.20 Nontechnical losses (estimated kWh) 40,057 37,150 33,515 28,206 27,410 28,214 27,589 29,957 Nontechnical loss as % of energy sent out 2.94 2.75 2.50 2.10 2.00 2.00 1.90 1.75 Total network losses (kWh) 128,663 124,960 120,654 114,168 113,752 117,088 119,069 136,089 Total network loss as % of energy sent out 9.44 9.38 9.32 9.26 9.20 9.14 9.08 8.40 Energy sent out from Ha'apai (kWh) 1,363,168 1,350,914 1,340,603 1,343,151 1,370,510 1,410,693 1,452,055 1,711,812 Station service (kWh) 44,116 41,781 41,462 41,541 42,387 43,630 44,909 52,943 Station service as % of energy generated 3.13 3.00 3.00 3.00 3.00 3.00 3.00 3.00 Total energy generated (kWh) 1,407,284 1,392,694 1,382,065 1,384,692 1,412,897 1,454,323 1,496,964 1,764,754 Annual peak demand (kW) 312 297 294 295 301 310 319 376 Source: Author. Note: n.a. = Not applicable. Appendix 3. Details of the Load Forecasts 71 72 Table a3.12: "eFFicienT TPl Plus Dsm" HigH ForecasT, Ha'aPai end of 2009 indicator (actual end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 High growth scenario Annual growth rate (%), billed energy -2.47 2.00 3.00 4.00 4.50 5.00 5.50 5.50 Billed energy (kWh) 1,234,505 1,259,195 1,296,971 1,348,850 1,409,548 1,480,025 1,561,427 2,040,722 DSM effect, % reduction of billed energy n.a. 0.833 2.500 4.167 5.000 5.000 5.000 5.000 Impact of DSM on billed energy (kWh) 0 -10,493 -32,424 -56,203 -70,477 -74,001 -78,071 -102,036 Net billed energy (kWh) 1,234,505 1,248,702 1,264,547 1,292,647 1,339,071 1,406,024 1,483,355 1,938,686 Technical losses (estimated kWh) 88,606 89,566 90,642 91,170 92,908 97,489 102,783 131,221 Kingdom of Tonga: Electric Supply System Load Forecast Technical loss as % of energy sent out 6.50 6.50 6.50 6.40 6.30 6.30 6.30 6.20 Nontechnical losses (estimated kWh) 40,057 37,893 34,862 29,915 29,494 30,949 30,998 37,038 Nontechnical loss as % of energy sent out 2.94 2.75 2.50 2.10 2.00 2.00 1.90 1.75 Total network losses (kWh) 128,663 127,459 125,504 121,086 122,402 128,437 133,780 168,259 Total network loss as % of energy sent out 9.44 9.38 9.32 9.26 9.20 9.14 9.08 8.40 Energy sent out from Ha'apai (kWh) 1,363,168 1,377,932 1,394,493 1,424,538 1,474,723 1,547,437 1,631,469 2,116,470 Station service (kWh) 44,116 42,616 43,129 44,058 45,610 47,859 50,458 65,458 Station service as % of energy generated 3.13 3.00 3.00 3.00 3.00 3.00 3.00 3.00 Total energy generated (kWh) 1,407,284 1,420,548 1,437,621 1,468,596 1,520,333 1,595,296 1,681,926 2,181,928 Annual peak demand (kW) 312 303 306 313 324 340 358 465 Source: Author. Note: n.a. = Not applicable. Table a3.13: "Do noTHing" meDian ForecasT, `eua end of 2009 indicator (actual end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Median growth scenario Annual growth rate (%), billed energy -1.51 1.00 2.00 3.00 4.00 4.00 5.00 5.00 Billed energy (kWh) 880,692 889,499 907,289 934,508 971,888 1,010,764 1,061,302 1,354,520 Technical losses (estimated kWh) 116,706 117,871 120,229 123,836 128,789 133,941 140,638 179,493 Technical loss as % of energy sent out 11.00 11.00 11.00 11.00 11.00 11.00 11.00 11.00 Nontechnical losses (estimated kWh) 63,552 64,186 65,470 67,434 70,131 72,937 76,584 97,742 Nontechnical loss as % of energy sent out 5.99 5.99 5.99 5.99 5.99 5.99 5.99 5.99 Total network losses (kWh) 180,258 182,057 185,699 191,270 198,920 206,877 217,221 277,235 Total network loss as % of energy sent out 16.99 16.99 16.99 16.99 16.99 16.99 16.99 16.99 Energy sent out from `Eua(kWh) 1,060,963 1,071,557 1,092,988 1,125,778 1,170,809 1,217,641 1,278,523 1,631,755 Station service (kWh) 29,282 49,905 50,903 52,430 54,527 56,709 59,544 75,995 Station service as % of energy generated 2.69 4.45 4.45 4.45 4.45 4.45 4.45 4.45 Total energy generated (kWh) 1,090,244 1,121,462 1,143,891 1,178,208 1,225,336 1,274,350 1,338,067 1,707,750 Annual peak demand (kW) 312 239 244 251 261 271 285 364 Source: Author. Appendix 3. Details of the Load Forecasts 73 74 Table a3.14: "eFFicienT TPl" meDian ForecasT, `eua end of 2009 indicator (actual end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Median growth scenario Annual growth rate (%), billed energy -1.51 1.00 2.00 3.00 4.00 4.00 5.00 5.00 Billed energy (kWh) 880,692 889,499 907,289 934,508 971,888 1,010,764 1,061,302 1,354,520 Technical losses (estimated kWh) 116,706 110,529 108,769 96,396 87,361 84,465 82,090 96,223 Technical loss as % of energy sent out 11.00 10.50 10.25 9.00 8.00 7.50 7.00 6.50 Kingdom of Tonga: Electric Supply System Load Forecast Nontechnical losses (estimated kWh) 63,564 52,633 45,099 40,165 32,760 30,970 29,318 29,607 Nontechnical loss as % of energy sent out 5.99 5.00 4.25 3.75 3.00 2.75 2.50 2.00 Total network losses (kWh) 180,270 163,163 153,868 136,561 120,121 115,435 111,407 125,830 Total network loss as % of energy sent out 16.99 15.50 14.50 12.75 11.00 10.25 9.50 8.50 Energy sent out from `Eua (kWh) 1,060,963 1,052,662 1,061,157 1,071,069 1,092,009 1,126,199 1,172,709 1,480,350 Station service (kWh) 29,282 32,557 32,819 33,126 33,773 34,831 36,269 45,784 Station service as % of energy generated 2.69 3.00 3.00 3.00 3.00 3.00 3.00 3.00 Total energy generated (kWh) 1,090,244 1,085,218 1,093,976 1,104,195 1,125,783 1,161,030 1,208,979 1,526,134 Annual peak demand (kW) 312 231 233 235 240 247 257 325 Source: Author. Table a3.15: "eFFicienT TPl Plus Dsm" low ForecasT, `eua end of 2009 indicator (actual end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 Low growth scenario Annual growth rate (%), billed energy -1.51 0.00 1.00 2.00 3.00 3.00 3.00 3.50 Billed energy (kWh) 880,692 880,692 889,499 907,289 934,508 962,543 991,419 1,177,495 DSM effect, % reduction of billed energy n.a. 0.833 2.500 4.167 5.000 5.000 5.000 5.000 Impact of DSM on billed energy (kWh) 0 -7,339 -22,237 -37,804 -46,725 -48,127 -49,571 -58,875 Net billed energy (kWh) 880,692 873,354 867,262 869,485 887,783 914,416 941,849 1,118,621 Technical losses (estimated kWh) 116,706 108,523 103,970 89,689 79,801 76,414 72,850 79,465 Technical loss as % of energy sent out 11.00 10.50 10.25 9.00 8.00 7.50 7.00 6.50 Nontechnical losses (estimated kWh) 63,564 51,678 43,110 37,370 29,925 28,018 26,018 24,451 Nontechnical loss as % of energy sent out 5.99 5.00 4.25 3.75 3.00 2.75 2.50 2.00 Total network losses (kWh) 180,270 160,201 147,079 127,059 109,726 104,432 98,868 103,916 Total network loss as % of energy sent out 16.99 15.50 14.50 12.75 11.00 10.25 9.50 8.50 Energy sent out from `Eua (kWh) 1,060,963 1,033,554 1,014,341 996,545 997,508 1,018,848 1,040,717 1,222,536 Station service (kWh) 29,282 31,966 31,371 30,821 30,851 31,511 32,187 37,810 Station service as % of energy generated 2.69 3.00 3.00 3.00 3.00 3.00 3.00 3.00 Total energy generated (kWh) 1,090,244 1,065,520 1,045,713 1,027,366 1,028,359 1,050,359 1,072,904 1,260,347 Annual peak demand (kW) 312 227 223 219 219 224 229 268 Source: Author. Note: n.a. = Not applicable. Appendix 3. Details of the Load Forecasts 75 76 Table a3.16: "eFFicienT TPl Plus Dsm" HigH ForecasT, `eua end of 2009 indicator (actual end of 2010 end of 2011 end of 2012 end of 2013 end of 2014 end of 2015 end of 2020 High growth scenario Annual growth rate (%), billed energy -1.51 2.00 3.00 4.00 4.50 5.00 5.50 5.50 Billed energy (kWh) 880,692 898,306 925,255 962,266 1,005,568 1,055,846 1,113,917 1,455,846 DSM effect, % reduction of billed energy n.a. 0.833 2.500 4.167 5.000 5.000 5.000 5.000 Impact of DSM on billed energy (kWh) 0 -7,486 -23,131 -40,095 -50,278 -52,792 -55,696 -72,792 Net billed energy (kWh) 880,692 890,821 902,124 922,171 955,289 1,003,054 1,058,222 1,383,053 Technical losses (estimated kWh) 116,706 110,694 108,149 95,124 85,869 83,821 81,851 98,250 Kingdom of Tonga: Electric Supply System Load Forecast Technical loss as % of energy sent out 11.00 10.50 10.25 9.00 8.00 7.50 7.00 6.50 Nontechnical losses (estimated kWh) 63,564 52,711 44,842 39,635 32,201 30,734 29,233 30,231 Nontechnical loss as % of energy sent out 5.99 5.00 4.25 3.75 3.00 2.75 2.50 2.00 Total network losses (kWh) 180,270 163,405 152,992 134,758 118,069 114,555 111,084 128,480 Total network loss as % of energy sent out 16.99 15.50 14.50 12.75 11.00 10.25 9.50 8.50 Energy sent out from `Eua (kWh) 1,060,963 1,054,226 1,055,116 1,056,929 1,073,359 1,117,609 1,169,306 1,511,534 Station service (kWh) 29,282 32,605 32,632 32,689 33,197 34,565 36,164 46,748 Station service as % of energy generated 2.69 3.00 3.00 3.00 3.00 3.00 3.00 3.00 Total energy generated (kWh) 1,090,244 1,086,831 1,087,748 1,089,618 1,106,555 1,152,174 1,205,470 1,558,282 Annual peak demand (kW) 312 231 232 232 236 245 257 332 Source: Author. Note: n.a. = Not applicable. APPENDIX 4 Oil Price Projections A4.1 Purpose The relationship between average world crude oil prices and the Singapore price for gasoil is shown in figure A4.1. This appendix describes the analysis conducted to pre- The relationship was erratic over the period of record. In pare scenarios for the possible fuel prices TPL will face a simple interpretation, the ratio has trended to increase during the period to 2020, the time covered by the Tonga . from about 1.20 to 1.27 The average over the period was Energy Road Map Study. 1.236. The forecast assumes that the ratio of the price of gasoil to the price of crude will remain constant at 1.24. The primary reference for the historical data and the pro- jections was the U.S. Department of Energy (www.eia. Figure A4.2 shows the difference between the price doe.gov). TPL pays for fuel at Tongatapu and the price of gasoil at Singapore. The Singapore price of at least two months before the record price at Tongatapu is the most appro- A4.2 Methodology priate basis for comparison, to account for transit time. That difference is what is illustrated by figure A4.2. The Components of the fuel oil price include the price of figure includes a simple trend line showing that the crude, the premium on that price for refined fuel oil (gas- cost of shipping and handling has increased from about oil, the European term for high-speed diesel oil), and the T$0.14 to T$0.34 per liter. However, the figure could be cost of shipping and handling to Tonga. The historical oil interpreted as showing the cost of shipping and handling prices are typically presented in US$ and converted to T$ to have been relatively constant during the period up using actual exchange rates. to 2001, to have had a step increase that applied from early 2001 through 2006, and to have had another step increase for the period since 2006. Errors were made in the fuel price calculations in recent months, but this A4.3 Analysis is expected to be resolved soon. Unless advised to the contrary, the cost of fuel shipping and handling that will TPL provided a list of prices it paid for diesel fuel for each be added to the forecast price of gasoil at Singapore will month between March 1998 and March 2009. Monthly be T$0.35 per liter. average prices for gasoil at Singapore were obtained from the U.S. Department of Energy Web site. Historical The forecast of future crude oil prices was taken from the world crude oil prices were taken from the same source. International Energy Outlook 2009.25 Because the fore- The historic exchange rates between US$ and T$ ware casts in that document provide expected price levels for the interbank rates on the OANDA Web site. 25. Energy Information Administration, International Energy Outlook 2009, (Washington, DC: U.S. Department of Energy) 2009. 77 78 Kingdom of Tonga: Electric Supply System Load Forecast 2015 and 2030, manual adjustments to annual growth in described above. With respect to the future exchange the intervening years were applied to smooth the projec- rate for the Tonga Pa'anga (T$), the currency had been tions. The results are shown in figure A4.3 to 2020, the fixed relative to the US$ since March 2006 but it began time frame of the Tonga Energy Road Map. to float in late July 2009. As of early September 2009, the value had increased from US$0.482 to about US$0.52. Tables A4.1 to A4.3 summarize the development of An exchange rate of US$0.52 per T$ was used in the the prices for fuel oil in Tonga based on the reference, tables. low, and high crude oil forecasts and the relationships Figure a4.1 raTio oF singaPore gasoil Price To average Price oF worlD cruDe 1.50 1.45 Ratio of Singapore gasoil price to 1.40 average price of world crude 1.35 1.30 1.25 1.20 1.15 1.10 1.05 1.0 Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Source: Compiled by author from U.S. Department of Energy database, May 2009. Figure a4.2 imPlieD sHiPPing anD HanDling cHarges, gasoil 0.6 0.5 0.4 T$ per liter 0.3 0.2 0.1 0 Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Source: Compiled by author from U.S. Department of Energy database, May 2009. Appendix 4. Oil Price Projects 79 Figure a4.3 cruDe oil ForecasT scenarios 250 Reference case Low case 200 High case US$ per barrel 150 100 50 0 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source: Compiled by author from U.S. Department of Energy database, May 2009. Table a4.1: reFerence oil Price ProjecTion world crude oil world crude oil singapore gasoil Tongatapu fuel oil Tongatapu fuel oil year (us$/bbl) (us$ per liter) (us$ per liter) (us$ per liter) (T$/liter) 2009 61.00 0.384 0.476 0.658 1.26 2010 70.00 0.440 0.546 0.728 1.40 2011 80.00 0.503 0.624 0.806 1.55 2012 90.00 0.566 0.702 0.884 1.70 2013 98.00 0.616 0.764 0.946 1.82 2014 104.00 0.654 0.811 0.993 1.91 2015 110.00 0.692 0.858 1.040 2.00 2016 111.23 0.700 0.868 1.050 2.02 2017 112.48 0.707 0.877 1.059 2.04 2018 113.74 0.715 0.887 1.069 2.06 2019 115.01 0.723 0.897 1.079 2.08 2020 116.30 0.732 0.907 1.089 2.09 Source: Compiled by author from U.S. Department of Energy database, May 2009. 80 Kingdom of Tonga: Electric Supply System Load Forecast Table a4.2: low oil Price ProjecTion world crude oil world crude oil singapore gasoil Tongatapu fuel oil Tongatapu fuel oil year (us$/bbl) (us$ per liter) (us$ per liter) (us$ per liter) (T$/liter) 2009 61.00 0.384 0.476 0.658 1.26 2010 59.40 0.374 0.463 0.645 1.24 2011 57.84 0.364 0.451 0.633 1.22 2012 56.32 0.354 0.439 0.621 1.19 2013 54.84 0.345 0.428 0.610 1.17 2014 53.40 0.336 0.417 0.599 1.15 2015 52.00 0.327 0.406 0.588 1.13 2016 51.86 0.326 0.405 0.587 1.13 2017 51.73 0.325 0.403 0.585 1.13 2018 51.59 0.325 0.402 0.584 1.12 2019 51.46 0.324 0.401 0.583 1.12 2020 51.32 0.323 0.400 0.582 1.12 Source: Compiled by author from U.S. Department of Energy database, May 2009. Table a4.3: HigH oil Price ProjecTion world crude oil world crude oil singapore gasoil Tongatapu fuel oil Tongatapu fuel oil year (us$/bbl) (us$ per liter) (us$ per liter) (us$ per liter) (T$/liter) 2009 61.00 0.384 0.476 0.658 1.26 2010 78.00 0.491 0.608 0.790 1.52 2011 93.00 0.585 0.725 0.907 1.74 2012 108.00 0.679 0.842 1.024 1.97 2013 122.00 0.767 0.952 1.134 2.18 2014 134.00 0.843 1.045 1.227 2.36 2015 145.00 0.912 1.131 1.313 2.52 2016 148.14 0.932 1.155 1.337 2.57 2017 151.35 0.952 1.180 1.362 2.62 2018 154.63 0.973 1.206 1.388 2.67 2019 157.98 0.994 1.232 1.414 2.72 2020 161.41 1.015 1.259 1.441 2.77 Source: Compiled by author from U.S. Department of Energy database, May 2009. The World Bank The World Bank Group Asia Sustainable and Alternative Energy Program 1818 H Street, NW Washington, DC 20433 USA www.worldbank.org/astae