Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy. Final Report Executive Summary. Submitted to: World Bank

The Power of Experience Submitted to: World Bank Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report Executive Summary Submitted By: March 2010

Contents Section Page Section 1 Executive Summary 1-1 1.1 Introduction... 1-1 1.2 Approach... 1-6 1.3 Load Forecast... 1-6 1.4 Fuel Supply... 1-9 1.5 Project and Technology Analysis... 1-10 1.6 Regional Strategies... 1-25 1.7 Country Summaries... 1-26 1.8 Recommendations... 1-42 Table Page Acronyms and Abbreviations... ii Table 1-1 Net Peak Demand Load Forecast (MW)... 1-7 Table 1-2 Net Generation Forecast (GWh)... 1-8 Table 1-3 Fuel Prices Based on Yearly Demand 2014-2028... 1-9 Table 1-4 Scenario NPV Cost Differences - Base Case Minus Other Scenario Costs (million US$)... 1-26 Figure Page Figure 1-1 Caribbean Regional Map... 1-3 Figure 1-2 Countries Included in the Study... 1-4 Figure 1-3 Other Relevant Countries Addressed in the Study... 1-5 Figure 1-4 Fossil LCL for Dominican Republic... 1-11 Figure 1-5 Other Options for Dominican Republic... 1-12 Figure 1-6 Eastern Caribbean Gas Pipeline (ECGP) Proposed Route... 1-14 Figure 1-7 Dominica Interconnections... 1-16 Figure 1-8 Nevis Puerto Rico and Nevis US Virgin Islands Interconnections... 1-17 Figure 1-9 Saba St. Maarten Interconnection... 1-18 Figure 1-10 Haiti Dominican Republic Interconnection... 1-19 Figure 1-11 United States (Florida) Cuba Interconnection... 1-20 Figure 1-12 Northern Ring Set of Interconnections... 1-21 Figure 1-13 Northern Ring Interconnections Alternative... 1-22 Figure 1-14 Distillate LCL vs. Renewable Energy Options... 1-24 Figure 1-15 Barbados LCL vs. Renewable Energy Options... 1-24 Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report i

Contents Acronyms and Abbreviations APC APUA BL&P CC, CCGT CFB CNG CO2 CSP CT DOMLEC DR ECGP EDH, EdH GJ GRENLEC GT GWH, GWh HFO IBRD IDA JPS km kv kw kwh LCL LNG LSD LUCELEC m MEM MSD MVA MW NPV O&M PV USA US$ V VINLEC yr Antigua Power Company Antigua Public Utility Authority Barbados Light and Power Combined cycle gas turbine Circulating fluidized bed Compressed natural gas Carbon dioxide Concentrating solar power Combustion turbine Dominica Electricity Services Limited Dominican Republic Eastern Caribbean Gas Pipeline Electricite d'haiti Gigajoule Grenada Energy Services Ltd. Gas turbine Gigawatt-hour Heavy fuel oil International Bank for Reconstruction and Development International Development Association Jamaica Public Service Kilometer Kilovolt Kilowatt Kilowatt-hour Least cost line Liquefied natural gas Low speed diesel St. Lucia Electricity Services Ltd. Meter Ministry of Energy and Mining (Jamaica) Medium speed diesel Megavolt-ampere Megawatt Net present value Operations and maintenance Photovoltaic United States of America United States dollar Volt St. Vincent Electricity Service Ltd. Year Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report ii

Section 1 Executive Summary 1.1 INTRODUCTION The objective of this Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy (the Study) is to analyze the availability of technically and financially sound regional and sub-regional energy solutions for power generation rather than specific energy solutions for each Caribbean country. The energy solutions involve new fuels or fuel transport modes (pipeline gas, compressed natural gas [CNG], liquefied natural gas [LNG], coal), new energy resources for power generation (primarily wind and geothermal), and new electrical interconnections among islands, none of which are presently interconnected. The immediate goal of studying these areas was to reduce the Caribbean islands dependence on high price imported distillate and heavy fuel oil (HFO). A related goal was that solutions would emerge that reduced costs, reduced environmental impacts, and increased the integration of the Caribbean islands. The entire Caribbean region is presented on Figure 1-1. We note at the outset that we have identified no truly regional energy solutions, not even one covering the nine countries of primary emphasis mentioned in the second paragraph below. We have identified and analyzed, to varying degrees of detail, 11 submarine cable electrical interconnections between two countries and one land-based interconnection. Some of these twocountry sub-regional interconnections are part of larger schemes involving three or more countries. The only sub-regional fuel project the Study evaluated was the five-country Eastern Caribbean Gas Pipeline (ECGP). Schemes involving LNG or CNG implicitly or explicitly rely on some common facilities when more than one country is a user, but in that sense they are not different from the current delivery modes for distillate and HFO and were analyzed on a countryby-country basis. It is interesting that the goal of reducing dependence on high price imported oil products and the goal of reducing environmental impacts and increasing the integration of the region turned out to be complementary. The most direct benefit of an interconnection comes when one country has a source of low cost power and its neighbor does not. The three lowest cost resources for operation at capacity factors above about 30% are renewables: geothermal, wind (including the cost of backup generation), and small hydro. This assumes that high quality sites can be identified and acquired. Geothermal is the source of generation and drives the benefits for many of the interconnections. Thus geothermal on a local and sub-regional basis, and wind on a local basis, provide a path toward a less oil-dependent, lower cost, lower environmental impact, more sustainable future. The primary emphasis of the Study is on the nine countries in the Caribbean eligible for support from the International Development Association (IDA) and/or the International Bank for Reconstruction and Development (IBRD). Those countries are presented, together with their relative electricity market share, on Figure 1-2. These nine include: Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-1

Six small countries in the Lesser Antilles: St. Lucia, St. Vincent and the Grenadines, Grenada, Antigua and Barbuda, St. Kitts and Nevis, and Dominica, total combined population about 600,000 Three countries located on two of the four islands in the Greater Antilles: Haiti and the Dominican Republic, both on the island Hispaniola, and Jamaica, total population about 22,000,000 The Study also considered other relevant countries, presented on Figure 1-3, that might be part of a regional energy solution. In addition to the nine countries mentioned above, we visited or obtained significant data on Barbados 1, Trinidad and Tobago, and Martinique; somewhat less on Guadeloupe; and cursory information on Puerto Rico, Sint Maarten, and Cuba. We also obtained cursory information on power generation in Florida. 1 Barbados was addressed in more details as par of the Eastern Caribbean Gas Pipeline project analysis. Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-2

Figure 1-1 Caribbean Regional Map Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-3

Figure 1-2 Countries Included in the Study Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-4

Figure 1-3 Other Relevant Countries Addressed in the Study Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-5

1.2 APPROACH Our approach included the following main steps: Collect the foundation data needed to conduct the work Prepare a peak and energy demand forecast for each country and the Study countries as a whole Forecast fuel costs for all fuels used, including pipeline gas, LNG, CNG, and coal Estimate fuel transportation costs for each fuel to each country and determine effective fuel price Determine the performance and cost parameters of all existing power generation units Determine the performance and cost parameters of power generation units suitable for meeting future demand Evaluate the cost and performance parameters for power generation from renewable energy, and estimate the availability of renewable energy resources Evaluate submarine cable technology Identify and evaluate submarine cable and land-based transmission interconnections Develop scenarios that include a range of approaches to regional power generation, and combine the most attractive components in an proposed scenario Report on and present the results (such as in this report) 1.3 LOAD FORECAST Tables 1-1 and 1-2 provides peak and energy demand forecasts for each country / island and for the region as a whole. Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-6

Table 1-1 Net Peak Demand Load Forecast (MW) Year Antigua and Barbuda Barbados Dominica Dominican Republic Grenada Haiti Jamaica St. Kitts Nevis St. Lucia St. Vincent and Martinique Guadeloupe Grenadines 2009 54 170 15 2,353 31 226 680 29 10 56 27 242 250 4,142 2010 57 176 15 2,447 33 237 707 30 10 58 28 247 256 4,302 2011 60 182 16 2,544 34 249 736 31 11 61 30 255 263 4,472 2012 63 188 16 2,640 36 261 767 32 11 63 32 263 269 4,643 2013 65 195 17 2,727 38 274 799 33 12 65 35 272 276 4,808 2014 67 201 17 2,803 40 288 832 35 13 68 37 281 284 4,965 2015 69 208 18 2,896 42 303 867 36 13 70 40 290 291 5,143 2016 71 216 18 2,992 45 318 904 37 14 73 42 297 298 5,324 2017 73 223 19 3,091 47 334 943 38 15 76 45 303 305 5,512 2018 75 231 19 3,194 50 350 983 40 16 79 48 310 313 5,708 2019 77 239 20 3,300 52 368 1,026 41 17 82 52 317 321 5,911 2020 80 247 20 3,409 55 386 1,071 43 18 85 55 324 329 6,121 2021 82 256 21 3,522 58 405 1,116 44 19 88 59 331 337 6,339 2022 85 265 21 3,638 61 426 1,165 46 20 91 63 339 346 6,565 2023 87 274 22 3,758 64 447 1,214 47 21 95 68 346 354 6,798 2024 90 284 22 3,882 68 469 1,267 49 23 98 72 354 363 7,041 2025 92 294 23 4,010 72 493 1,322 51 24 102 77 362 372 7,293 2026 95 304 24 4,143 75 517 1,379 52 25 106 83 370 381 7,555 2027 98 314 24 4,280 80 543 1,439 54 27 110 88 378 391 7,827 2028 101 325 25 4,421 84 570 1,502 56 29 114 94 387 400 8,109 Growth Rate 3.3% 3.5% 2.7% 3.4% 5.4% 5.0% 4.3% 3.5% 5.9% 3.8% 6.9% 2.5% 2.5% 3.6% Total Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Interim Report 1-7

Table 1-2 Net Generation Forecast (GWh) Year Antigua and Barbuda Barbados Dominica Dominican Republic Grenada Haiti Jamaica St. Kitts Nevis St. Lucia St. Vincent and Martinique Guadeloupe Grenadines 2009 318 1,039 87 12,638 198 660 4,490 161 60 345 156 1,575 1,663 23,391 2010 315 1,073 89 13,142 209 726 4,674 166 67 356 167 1,620 1,720 24,322 2011 312 1,107 91 13,663 220 799 4,865 171 74 367 178 1,672 1,775 25,295 2012 410 1,143 94 14,179 232 878 5,066 175 82 378 191 1,727 1,832 26,387 2013 422 1,180 96 14,646 244 966 5,277 180 86 390 204 1,783 1,888 27,363 2014 434 1,218 99 15,054 257 1,063 5,494 186 90 402 218 1,840 1,947 28,303 2015 447 1,258 101 15,554 270 1,169 5,726 191 94 415 233 1,900 2,003 29,362 2016 461 1,298 104 16,070 285 1,286 5,974 196 99 428 249 1,938 2,060 30,448 2017 475 1,340 106 16,601 300 1,415 6,232 202 103 442 266 1,978 2,117 31,576 2018 489 1,384 109 17,154 316 1,556 6,497 208 107 455 284 2,018 2,174 32,751 2019 503 1,428 112 17,724 333 1,712 6,777 214 111 470 304 2,058 2,233 33,979 2020 519 1,475 114 18,309 350 1,883 7,073 220 115 484 325 2,100 2,284 35,252 2021 534 1,522 117 18,914 369 1,977 7,376 226 119 500 348 2,142 2,337 36,483 2022 550 1,572 120 19,539 389 2,076 7,696 233 124 515 372 2,186 2,390 37,761 2023 567 1,622 123 20,184 409 2,180 8,024 239 129 531 397 2,230 2,445 39,082 2024 583 1,675 126 20,851 431 2,289 8,370 246 134 548 425 2,275 2,501 40,454 2025 600 1,729 129 21,539 454 2,403 8,734 253 139 565 454 2,321 2,559 41,881 2026 618 1,785 133 22,251 478 2,523 9,114 261 145 583 485 2,368 2,618 43,361 2027 636 1,843 136 22,986 504 2,650 9,510 268 150 601 519 2,416 2,679 44,897 2028 654 1,902 139 23,745 530 2,782 9,924 276 156 620 555 2,465 2,741 46,490 Growth Rate 3.9% 3.2% 2.5% 3.4% 5.3% 7.9% 4.3% 2.9% 5.2% 3.1% 6.9% 2.4% 2.7% 3.7% Total Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Interim Report 1-8

1.4 FUEL SUPPLY The forecast levelized price of distillate over 2014-2028, the assumed study period for the new projects being considered, is US$22.45/GJ. Each country except Dominica has at least one lower cost fuel option, and many countries have more than one. Table 1-3 provides the comparative prices. Distillate, LNG, and pipeline gas can be compared directly because they can fuel the same generators. Coal fuels generators with higher capital costs and higher heat rates, which must be taken into account in comparing fuel options. The prices of all fuels except distillate vary from country to country because they include transportation costs that vary. Table 1-3 Fuel Prices Based on Yearly Demand 2014-2028 Fuels Selected in Addition to Coal and Distillate Levelized Fuel Price, US$/GJ Fuel Selected Coal Distillate Country Antigua and Barbuda None N/A 12.31 22.45 Barbados Pipeline Gas 7.39 7.77 22.45 Dominica Distillate only N/A N/A 22.45 Dominican Republic LNG 8.73 4.19 22.45 Grenada None N/A 12.31 22.45 Guadeloupe Pipeline Gas 10.88 7.77 22.45 Haiti LNG 12.73 7.77 22.45 Jamaica LNG 10.16 4.85 22.45 Jamaica North LNG 10.90 4.85 22.45 Martinique Pipeline Gas 8.99 7.77 22.45 St. Kitts and Nevis None N/A 12.31 22.45 St. Lucia Pipeline Gas 10.49 9.04 22.45 St. Vincent and Grenadines None N/A 12.31 22.45 Coal is an optional fuel for every country except Dominica, where preliminary analysis showed it to be more costly than distillate on a US$/GJ basis. Table 1-3 shows the following: Every country except Dominica has at least one fuel option lower in price than distillate Pipeline gas is the lowest cost natural gas option for every country reached by the ECGP: Barbados, Martinique, St. Lucia, and Guadeloupe Coal is the only optional fuel for Antigua and Barbuda, Grenada, St. Kitts and Nevis, and St. Vincent and Grenadines LNG is the lowest cost natural gas option for Dominican Republic, Haiti, Jamaica, and Jamaica North Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-9

CNG was considered and was the lowest cost gas option for several countries, but for those countries was always higher in cost than distillate and therefore does not appear in Table 15-1. It was considerably lower than distillate for some countries, but was more costly than LNG in those countries. Though not studied in the same detail as the other fuel options, mid-scale LNG may provide an economically attractive option for some countries. 1.5 PROJECT AND TECHNOLOGY ANALYSIS Screening analysis is an approach to comparing the costs of different technologies to determine the least cost technology across the range of annual capacity factors: Uses simplified representations of generation costs to help identify least cost generating technologies Plots annual cost in $/kw-year vs. capacity factor for a set of power plant and/or fuel options The cost in $/kw-yr can also be easily expressed in cents/kwh, which are also of interest but are curved and somewhat harder to interpret than the straight lines in $/kw-yr Annual cost is sum of: Annualized investment-related costs based on initial capital investment, discount rate, and plant lifetime Fixed annual operation and maintenance (O&M) Variable cost (includes fuel cost and variable O&M costs) per kwh times capacity factor times hours per year Selects lowest cost resources at each capacity factor, producing the least-cost line for that set of resources 1.5.1 Isolated Countries / Islands Figure 1-4 illustrates the screening analysis approach. It presents the Fossil Least Cost Line (Fossil LCL) that applies for the Dominican Republic. The word Fossil means that only fossil fueled generation is included in determining the LCL. The scale in $/kw-yr for the solid lines is on the left, the scale in cents/kwh for the dotted lines is on the right. The Fossil LCL comprises 50 MW GT on LNG for capacity factors of zero through 20%, the 300 MW CC on LNG for capacity factors from 25% through 40%, and the conventional coal plant for capacity factors from 45% through 90%. In other words, the generation expansion plan based on this analysis would include gas turbines for peaking duty, combined cycles for mid-range duty, and conventional coal for base load duty. In order to achieve the Fossil LCL the Dominican Republic would have to undertake large capital investments for expansion related to coal and LNG transportation, and for coal plants themselves. This may pose a challenge, even if the desire to do so exists. LNG is preferable for application at lower capacity factors, coal for application at higher capacity factors. The scenario analysis provides more information on which is preferable overall, if doing both is not feasible. Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-10

Figure 1-5 expands upon Figure 1-4 by adding wind, small hydro, and fossil options to the graph of Figure 1-1. The small hydro line coincidentally overlaps with the wind with backup line at capacity factors from 30% to 40%. Wind with backup, which is typically a better comparison than wind without backup, is now marginally economic at the capacity factors where it might operate at a good site. Wind with backup simply adds the full cost of operation of a 50 MW gas turbine at 5% capacity factor to the costs of wind without backup, which also adds 5% to the capacity factor. Figure 1-5 also illustrates what might occur if neither coal nor LNG is available for future generation for the Dominican Republic. The periwinkle line represents the cost of a 300 MW HFO-fueled steam plant. Without expanded supplies of LNG or coal, costs will more than triple at high capacity factors. Dominican Republic has under construction or planned considerable new small hydro and wind generation. Figure 1-5 illustrates the desirability of such an approach where good sites can be identified. Annual Cost, $/kw-year 2,500 2,000 1,500 1,000 500 Fossil Least Cost Line, US/kW-yr 50 MW GT LNG US$/kW-yr 300 MW CC LNG US$/kW-yr 300 MW Conv Coal, US$/kW-yr Fossil LCL, US cents/kwh 80 70 60 50 40 30 20 10 Cost, US cents/kwh 0 0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% Capacity Factor, % Figure 1-4 Fossil LCL for Dominican Republic Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-11

2,500 80 Annual Cost, $/kw-year 2,000 1,500 1,000 500 Fossil Least Cost Line, US/kW-yr 1.5 MW Wind US$/kW-yr 1.5 MW Wind w/backup US$/kW-yr 300 MW ST HFO DR US$/kW-yr Small Hydro US$/kW-yr Fossil LCL, US cents/kwh 70 60 50 40 30 20 10 Cost, US cents/kwh 0 0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% Capacity Factor, % Figure 1-5 Other Options for Dominican Republic The bullets below summarize the least cost technology/fossil fuel combination by country as determined by screening analysis. This considers the countries and islands as isolated systems. For some countries, imports via submarine cable (to be discussed later) provide a lower cost solution. We eliminated the technology/fuel combinations that were least cost at only one annual capacity factor, such as zero or 90%. Scenario analysis generally supports these conclusions, though multiple fuels were not used as much. Individual Countries Antigua and Barbuda, Grenada, and St. Vincent and Grenadines: 10 MW MSD on distillate for peaking and mid-range duty, and the coal-fueled CFB for base load duty Coal-fueled CFB is only marginally more economic than distillate fueled medium speed diesels (MSD) plants; CO2 costs of US$50/tonne would make the distillatefueled units more economic than the coal-fueled units Dominica, St. Kitts, and Nevis: 5 MW MSD on distillate for peaking, mid-range, and base load duty St. Kitts and Nevis are fortunate that a geothermal resource sufficient to serve all their demand has been confirmed and is in the process of development. For Dominica it seems highly probable that a geothermal resource sufficient to serve at least local demand will be confirmed and developed. None of these islands may not need to install any new distillate-fueled generation. Dominican Republic: 50 MW GT on LNG for peaking duty, 300 MW CC on LNG for mid-range duty, and 300 MW conventional coal plant for base load duty Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-12

The Dominican Republic already has an LNG terminal and coal-fueled power plants. Scenario analysis shows that coal is preferred if only one fuel can be selected for future additions. However, incorporating CO2 costs in the analysis would compromise coal s advantage. With the Dominican Republic s large demand, expanding the use of both fuels may be feasible, even if new facilities are needed. Haiti: 20 MW LSD on LNG for peaking, mid-range, and base load duty. LNG provides very large benefits but requires significant up-front capital expenditures Jamaica and Jamaica North: 50 MW GT on LNG for peaking duty, 20 MW LSD on LNG for mid-range duty, and 50 MW coal-fueled CFB base load duty Today Jamaica and Jamaica North have neither fuel. It seems unlikely that they would want to develop both fuels. If only one is to be developed, LNG is preferred, and its advantage would increase if CO2 costs are incorporated in the analysis. We emphasize that for some countries, imports via submarine cable provide a lower cost solution. This is addressed in the next subsection. Sub-regional Gas Market The ECGP links the markets of the four countries and provides the benefits of economies of scale compared to individual development. Barbados, Guadeloupe, Martinique, and St. Lucia: 20 MW GT on pipeline gas for peaking duty and 20 MW LSD on pipeline gas for mid-range and base load duty For all four countries the pipeline gas is less than half as costly as distillate. For all but St. Lucia, LNG is more costly than pipeline gas but significantly less costly than distillate. The low gas price reduces the benefits of renewables and for Martinique and Guadeloupe makes importing geothermal power from Dominica via submarine cable marginal. Figure 1-6 illustrates the ECGP gas connections among the countries. Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-13

Figure 1-6 Eastern Caribbean Gas Pipeline (ECGP) Proposed Route 1.5.2 Sub-regional Electricity Markets The first three bullets below show the interconnections studied with greatest emphasis. All the interconnections were submarine cables except the Dominican Republic Haiti link noted in the bottom bullet. The interconnections are presented in Figures 1-7 to 1-13. For each interconnection we note its capacity in MW, length in km, cost per kw for interconnection and related facilities only, source of export power, and economic attractiveness. Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-14

Nevis St. Kitts, 50 MW submarine cable capacity, 5 km submarine cable length, US$328/kW (interconnection and related facilities only), geothermal power export, highly economic Dominica Martinique, 100 MW, 70 km, US$588/kW (interconnection and related facilities only), geothermal power export, marginally economic if displaced fuel is gas from ECGP, more economic if displaced fuel is higher cost Dominica Guadeloupe, 100 MW, 70 km, US$588/kW (interconnection and related facilities only), geothermal power export, moderately economic if displaced fuel is gas from ECGP, more economic if displaced fuel is higher cost Nevis Puerto Rico, 400 MW, 400 km, US$1,791/kW (interconnection and related facilities only), geothermal power export, highly economic if displaced fuel is HFO, not economic if displaced fuel is LNG Nevis US Virgin Islands, 80 MVA, 320 km, US$3,541/kW (interconnection and related facilities only), geothermal power export, only marginally economic even though the displaced fuel is distillate Saba St. Maarten, 100 MW, 60 km, US$528/kW (interconnection and related facilities only), geothermal power export, highly economic if displaced fuel is distillate and St. Maarten can accept 100 MW United States (Florida) Cuba, 400 MW, 400 km, US$1,791/kW (interconnection and related facilities only), export from coal-fueled steam plant or gas-fueled combined cycle, highly economic if displaced fuel is HFO Dominican Republic Haiti, 250 MW, 563 km, US$1,899/kW (interconnection and related facilities only), land interconnection, export from HFO fueled steam plant, not economic unless export is from lower cost unit/fuel combination We also developed basic data and cost estimates for four potential interconnections that might form part of a Northern Ring, a conceptual set of interconnections in the northern Caribbean, potentially linking Florida Cuba Haiti Dominican Republic Puerto Rico Nevis, or some subset of those areas. The Northern Ring interconnections not covered above include: Puerto Rico Dominican Republic, 400 MW, 150 km, US$705/kW (interconnection and related facilities only) Haiti Cuba, 400 MW, 200 km, US$705/kW (interconnection and related facilities only) Haiti Jamaica, 400 MW, 250 km, US$998/kW (interconnection and related facilities only) Florida Haiti, 400 MW, 1,100 km, US$3,488/kW (interconnection and related facilities only We did not conduct economic analysis on these four interconnections. The cost per kw for the three shorter interconnections is in what might be an economically viable range if the sending country had low power costs and the importing country s displaced fuel was distillate, HFO, or crude. The Florida Haiti interconnection appears to be outside that range. Costs for the middle Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-15

islands would involve their sharing some of the costs of interconnections closer to the low-cost source, making favorable economics more difficult to achieve. Figure 1-7 Dominica Interconnections Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-16

Figure 1-8 Nevis Puerto Rico and Nevis US Virgin Islands Interconnections Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-17

Figure 1-9 Saba St. Maarten Interconnection Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-18

Figure 1-10 Haiti Dominican Republic Interconnection Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-19

Figure 1-11 United States (Florida) Cuba Interconnection Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-20

Figure 1-12 Northern Ring Set of Interconnections Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-21

Figure 1-13 Northern Ring Interconnections Alternative Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-22

1.5.3 Renewable Energy Wind, geothermal, small hydro, and biomass technology/fuel combinations have the potential, at a good site, to be considerably less costly than distillate fueled power generation. The three lowest cost resources for operation at capacity factors above about 30% are renewables: geothermal, wind (including the cost of backup generation), and small hydro. This assumes that high quality sites can be identified and acquired. Solar PV and solar trough CSP are not competitive for bulk power generation. There are many small solar PV installations in Martinique due to subsidies, and solar PV is competitive for off-grid locations. If a lower cost fuel such as pipeline gas were the competitive fuel, the advantage of the renewable technology would be less. Figure 1-14 compares renewable technologies to the Distillate LCL that results when distillate is the only fuel available. The scale in $/kw-yr for the solid lines is on the left, the scale in cents/kwh for the dotted lines is on the right. The Distillate LCL, in blue, represents the benefit of a renewable energy option. Its generation would displace generation at a cost along that line. Where a renewable energy option s line is below the blue line, there is a net benefit. It reduces costs elsewhere that are more than its own costs. Where it is above the blue line, it represents a net cost. Most of the renewable technologies are shown at a range of capacity factors they might reasonably achieve at a good site. Geothermal is also based on a good site, and is shown over the entire capacity factor range because it is not limited by resource availability once the resource has been defined. Wind with backup simply adds the full cost of operation of a 20 MW LSD at 5% capacity factor to the costs of wind without backup, which also adds 5% to the capacity factor. Biomass costs assume that biomass costs the same as export coal in the US. Figure 1-14 shows that all but two of the renewable energy technologies have the potential, at a good site, to be considerably less costly than distillate fueled power generation. Solar PV and solar trough with six hour storage are above the Distillate LCL. If a lower cost fuel such as pipeline gas were the competitive fuel, the advantage of the renewable technology would be less and might disappear. Figure 1-15 compares renewable technologies to the Fossil LCL for Barbados, which has the lowest cost gas fuel of any of the countries studied. The renewable technologies offer much smaller net benefits, small hydro and wind with storage are marginally economic, biomass is not economic, and the technologies that were not economic before are less competitive. Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-23

Annual Cost, $/kw-year 2,500 2,000 1,500 1,000 500 Distillate LCL, US$/kW-yr 1.5 MW Wind Turbine 1.5 MW Wind w/backup Commercial PV 500 kw 20 MW Geothermal Small Hydro Biomass Solar Trough 6 hr Storage Distillate LCL, US cents/kwh 80 70 60 50 40 30 20 Cost, US cents/kwh 10 0 0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% Capacity Factor, % Figure 1-14 Distillate LCL vs. Renewable Energy Options Annual Cost, $/kw-year 2,000 1,500 1,000 500 Fossil Least Cost Line, US/kW-yr 1.5 MW Wind Turbine 1.5 MW Wind w/backup Solar Trough 6 hr Storage Series6 Commercial PV 500 kw 20 MW Geothermal Small Hydro Biomass Fossil LCL, US cents/kwh 60 50 40 30 20 10 Cost, US cents/kwh 0 0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% Capacity Factor, % Figure 1-15 Barbados LCL vs. Renewable Energy Options Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-24

1.5.4 CO2 Costs If a tax or similar levy were attributed to each tonne of CO2 emissions, the cost of using fuels would increase. This would open wider the economic window for technologies that produce lower or no CO2 emissions. However, all the countries today primary fuel is distillate and/or HFO, so the window is already quite wide. We investigated the impact if a cost of US$50/tonne were attributed to CO2 emissions. At US$50/tonne, the effective price of fuels would increase in a range from US$2.52 for distillate to US$4.41 for coal, representing increases ranging from 15% for distillate to 91% for the lowest cost coal for the Study islands. In the bullets below we measure the impact of CO2 costs by how technology choices change when it is applied. Countries with small demand: The fuel prices are high even when coal fuels some of the least-cost generation. For Antigua and Barbuda, Grenada, and St. Vincent and Grenadines, the preferred fuel would switch from coal to distillate. The renewable energy resources that were economic before are now somewhat more economic, and those that were not economic edge closer to being competitive. Countries with medium or high demand. The fuels are much less expensive than distillate and therefore the displaced generation is lower in cost, narrowing the economic window for alternatives. For the Dominican Republic, Jamaica, and Jamaica North, incorporating CO2 costs in the analysis would probably eliminate coal s advantage over LNG or increase LNG s advantage over coal. With no CO2 cost the renewables that were economic for the islands with small demand are still economic, though in some cases only marginally so. Incorporating CO2 costs makes renewables more competitive. 1.6 REGIONAL STRATEGIES For all Study countries combined, costs including fuel savings from exports and interconnection costs were: US$31,985 million for the Base Case Scenario US$29,424 million for the Fuel Scenario US$29,415 million for the Interconnection/Renewable Scenario US$27,619 million for the Integrated Scenario Table 1-4 presents cost differences among Scenarios by system as well as differences in Scenario total costs. The Fuel Scenario and Interconnection/Renewable Scenarios both reduce costs by about US$2.5 billion compared to the Base Case. The Integrated Scenario reduces costs by Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-25

about US$ 4.3 billion, showing that the Integrated Scenario captures most of the individual benefits of each of the other two Scenarios. Table 1-4 Scenario NPV Cost Differences - Base Case Minus Other Scenario Costs (million US$) Fuel Scenario Interconnection/ Renewable Scenario Integrated Scenario Antigua and Barbuda 12 20 31 Barbados 906 39 912 Dominica 0 604 10 Dominican Republic 444 350 721 Grenada 32 17 45 Haiti 433 76 476 Jamaica 500 138 628 St. Kitts 0 159 159 Nevis 0 1,135 1,135 St. Lucia 216 18 221 St. Vincent and Grenadines 18 14 29 Total 2,561 2,570 4,365 The costs of the interconnections and the fuel savings from the exports of geothermal power are attributed to Dominica and Nevis. All numbers in Table 15-2 have positive values (except for zeros for Dominica, St. Kitts, and Nevis for the Fuel Scenario), meaning that each Scenario and each country in each Scenario provides cost savings compared to the Base Case. 1.7 COUNTRY SUMMARIES Each country summary below presents paragraphs on: Overview Current and Forecast Load Fossil Fuel Options Renewable Generation Potential Development Scenarios (development plans for the Base Case Scenario, the Fuel Scenario, the Interconnection/Renewable Scenario, and the Integrations Scenario) Discussion of Country Results 1.7.1 Antigua and Barbuda Overview: Antigua Public Utility Authority (APUA) is responsible for the power generation, transmission, and distribution of electricity in Antigua and Barbuda. APUA purchases most of the power from Antigua Power Company (APC), a private company. Antigua and Barbuda Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-26

currently rely exclusively on diesel for power generation. Efforts are underway to convert some of the diesel engines to HFO as an alternate fuel. Current and Forecast Load: The country s 2009 peak demand is just over 50 MW, with net generation of over 300 GWh. By 2028 peak demand is projected to increase to around 100 MW, with net generation increasing to around 650 GWh (increase rate of 3.9% per year). Losses in the transmission and distribution system are projected to decrease from over 30% in 2009 to around 10% by 2028. Fossil Fuel Options: Imported coal was considered as an alternative fuel. Due to the location and electricity demand on the island, the Study did not find natural gas to be an economically viable fuel option. Renewable Generation Potential: Wind is the most promising renewable resource for Antigua and Barbuda. A 2008 Energy Engineering Corp. report indicated that up to 400 MW of wind power can be developed on the islands, primarily on Barbuda. Solar PV potential is estimated at 27 MW of installed capacity, but bulk power development would not be economic based on current estimates. Development Scenarios: All four Study Scenarios assumed that the committed system additions of the Casada Gardens units will be installed during 2011-2013. With those unit additions, system reserve margin requirements would be satisfied until 2019. During 2020-2028 the system demand growth will require building additional generation units. For the Base Case Scenario, new unit additions are assumed to be 10 MW medium speed diesel units using distillate oil. By 2028 the system will need another 30 MW (3 x 10 MW units) to meet the required capacity. For the Fuel Scenario, coal-fueled circulating fluidized bed (CFB) plants are marginally more economic than distillate fueled medium speed diesels (MSD) plants; new unit additions are assumed to be 10 MW CFB units using imported coal. CO2 costs of US$50/tonne would make the distillate-fueled units more economic than the coal-fueled units. Conventional (large-scale) LNG is more costly than either (distillate or coal) option. Though not studied in the same detail as the other fuel options, mid-scale LNG may provide an economically attractive option. By 2028 the system will need another 30 MW (3 x 10 MW units) to meet the required capacity. The Fuel Scenario results show that the introduction of coal provides net present worth savings of US$12 million compared to the Base Case Scenario. The Interconnection/Renewable Scenario assumed development of new diesel units as in the Base Case Scenario, with the addition of 14 MW of new wind units. This assumes that sites with good winds and low development costs can be identified and acquired. There is no electrical interconnection. The Interconnection/Renewable Scenario results show that the introduction of wind generation provides net present worth savings of US$20 million compared to the Base Case Scenario. The Integrated Scenario assumed new generation units are 10 MW CFB units, as in the Fuel Scenario, and the addition of 14 MW of new wind units, as in the Interconnection/Renewable Scenario. The Integrated Scenario results show that including both coal as a fuel and wind Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-27

generation provides combined savings of US$31 million over the Base Case Scenario. The Integrated Scenario results show savings close to the sum of the savings of other two Scenarios. Discussion of Country Results Adding coal-fueled CFB technology reduces net present worth costs by US$12 million compared to the Base Case Scenario, with a cost advantage compared to distillate-fueled MSD technology ranging from 2% at 55% capacity factor to 10% at 80% capacity factor. That cost advantage disappears if costs of US$50/tonne are attributed to CO2 emissions. Conventional LNG is more costly than distillate, but mid-scale LNG might be a viable fuel option, justifying a more detailed analysis. Development of wind generation reduces net present worth costs by US$20 million compared to the Base Case Scenario, assuming that sites with good winds and low development costs can be identified and acquired. With that assumption wind is much lower in cost than distillate fueled generation. Small hydro and biomass would also be economic, if good sites can be identified. The benefits are relatively unaffected by the choice of fuel for the country s fossil units. 1.7.2 Barbados Overview: Barbados Light and Power (BL&P), a private company, is responsible for power generation, transmission, and distribution of electricity in Barbados. Existing installed generation of around 240 MW, mostly comprising of low and medium speed diesel units, substantially exceeds peak demand and provides a comfortable reserve margin. BL&P is looking to diversify its fuel mix which is mostly dependent on imported oil products. Current and Forecast Load: The country s 2008 peak demand was 164 MW, with net generation of over 1,000 GWh. By 2028 peak demand is projected to double to around 325 MW, with net generation increasing to around 1,900 GWh (increase rate of 3.5% per year). Fossil Fuel Options: Natural gas, delivered as LNG or through the Eastern Caribbean Gas Pipeline (ECGP), and imported coal were considered as alternative fuel options. Due to the location and electricity demand on the island, the Study found natural gas delivered through ECGP to be the most economically attractive fuel option. Renewable Generation Potential: No studies on country-specific overall wind and solar potential are available. We estimated Barbados wind potential to be at least 10 MW based on an already approved project. Solar PV potential is estimated at 26 MW of installed capacity, but bulk power development would not be economic based on current estimates. Development Scenarios: All four Scenarios assumed that the committed system additions of the nine 16 MW Trent units will be installed. The first six units were added during 2011-2013 while the next three units were added when required to match the load growth. All Trent unit additions would satisfy reserve margin requirements until 2025. During 2026-2028 the Barbados system will require new capacity additions. Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-28

For the Base Case Scenario, new additions are assumed to be 20 MW low speed diesel units using distillate oil. By 2028 the system will need another 40 MW (2 x 20 MW units) to meet the required capacity. For the Fuel Scenario, assumed system additions are the same as for the Base Case Scenario. The difference is that in this scenario most existing and all new units are assumed to use natural gas as a fuel, supplied through the ECGP. The Fuel Scenario shows that the introduction of ECGP natural gas provides net present worth savings of US$906 million compared to the Base Case Scenario.. For the Interconnection/Renewable Scenario, most assumed system additions are the same as for the Base Case Scenario. The difference in this Scenario is the addition of 45 MW of new wind units. This assumes that sites with good winds and low development costs can be identified and acquired. There is no electrical interconnection. The Interconnection/Renewable Scenario shows that the introduction of wind generation provides net present worth savings of US$39 million compared to the Base Case Scenario. For the Integrated Scenario, the availability of natural gas is assumed, as in the Fuel Scenario, combined with the addition of 45 MW of new wind units, as in the Interconnection/Renewable Scenario. The Integrated Scenario results show net present worth savings of US$912 million over the Base Scenario, only slightly more than for the Fuel Scenario. Discussion of Country Results Barbados has four fossil fuel options that offer significant economic benefits compared to continued reliance on oil products: natural gas via the ECGP, LNG, CNG, and coal. By far the most attractive is the ECGP option, which provides net present worth savings of $906 million compared to the Base Case Scenario. Its cost per kwh compared to distillate fueled generation ranges from less than half at 20% capacity factor to less than 40% at 80% capacity factor. If the ECGP does not materialize, the other fuel options should be considered. They would offer significant savings compared to distillate, though not as dramatic as ECGP gas offers. Development of wind generation reduces Interconnection/Renewable Scenario net present worth costs by US$39 million compared to the Base Case Scenario, assuming that sites with good winds and low development costs can be identified and acquired. However, when ECGP gas is available, as is assumed in the Integrated Scenario, adding wind generation increases savings by only US$6 million. Wind is only marginally economic, as would be small hydro if good sites can be identified, but biomass would be marginally uneconomic. This illustrates the high dependence of wind generation savings on the assumed fuel supply option, and the possibility that wind generation penetration might be limited to only a few of the best wind sites. 1.7.3 Dominica Overview: Dominica Electricity Services Limited (DOMLEC) is a sole producer responsible for the power generation, transmission, and distribution of electricity in Dominica. Existing installed generation, comprising high and medium speed diesel units and hydro units, exceeds peak Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-29

demand by 35% providing a comfortable reserve margin. Dominica is looking to diversify its fuel mix, which is mostly dependent on imported oil products. Current and Forecast Load: The country s current peak demand is around 15 MW, with net generation of around 90 GWh. By 2028 peak demand is projected to increase to 25 MW, with net generation increasing to around 150 GWh (increase rate of 2.5% per year). Fossil Fuel Options: Due to the low electricity demand on the island, the least-cost fuel is distillate because the fixed costs associated with all other fuels produce higher unit costs in US$/GJ. Renewable Generation Potential: Based on the ongoing assessment of potential at the Watton Waven field in central Dominica, and West Indies Power s exploration in the Soufriere area, geothermal potential is estimated to be adequate to supply 100 MW of geothermal power plants. Drilling of the first three slim (exploratory) wells is scheduled to start in June 2010 in the Soufriere area near the southern coast. Solar PV potential is estimated at 45 MW of installed capacity, but bulk power development would not be economic based on current estimates. Dominica also has small-size hydro and wind potential. Development Scenarios: Starting in 2012 Dominica will require new capacity additions. For the Base Case Scenario, new additions are assumed to be 5 MW medium speed diesel units using distillate oil. By 2028 the system will need another 15 MW (3 x 5 MW units) to meet the required capacity. Dominica does not have a potentially less expensive fossil fuel option. The Interconnection/Renewable Scenario assumes the addition of a 20 MW geothermal unit in 2012 to satisfy local needs. It also assumes submarine cable electrical interconnections with Martinique and Guadeloupe, and the addition of two 92.5 MW units in 2014 to support exports to those two countries. The results show large benefits of geothermal development in this Scenario, with net present worth savings of US$604 million compared with the Base Case Scenario. The Integrated Scenario assumed assumes the same geothermal additions as in the Interconnection/Renewable Scenario. The key is the assumed fuel savings due to energy exports to Martinique and Guadeloupe. The Integrated Scenario assumes construction of the ECGP and natural gas deliveries to those two countries, so fuel savings on Martinique and Guadeloupe are reduced because the imports are replacing lower cost natural gas (rather than distillate) based generation. The Integrated Scenario result shows combined savings of only US$10 million, demonstrating that savings are highly dependent on the assumed fuel supply option for Martinique and Guadeloupe. Savings of US$10 million is considerably less than the savings from the much smaller supply to Dominica alone. Discussion of Country Results: Because its low demand means that no fossil fuel options appear economic compared to distillate, geothermal development is particularly important for Dominica. It seems probable that a geothermal resource at least large enough to serve Dominica s demand will be confirmed. This would be the most important result from the country s point of view, as it would insulate the Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-30

country from the high price of distillate, and the uncertainty associated with variation in that price over time. It would also reduce CO2 emissions. Considering the low cost of power of geothermal power, wind generation is only marginally economic compared to geothermal for domestic consumption on Dominica. Small hydro also would be marginally economic if good sites can be identified, but biomass would be marginally uneconomic. Confirmation of a resource sufficient to serve exports to Martinique and/or Guadeloupe is less certain. The benefits of such development are also less certain. Almost US$600 million in savings when distillate is the displaced fuel disappear when ECGP gas is assumed to be the displaced fuel. Martinique and Guadeloupe have another fossil fuel option, LNG, with lower cost than distillate. In the Fuel Scenario, LNG would provide significant savings compared to the Base Case Scenario, though less than the savings with ECGP gas. It is not clear how much the savings would be in the Integrated Scenario, but they would be higher than US$10 million. It is clear that more detailed analysis of the Martinique and Guadeloupe systems would be required to determine the desirability of developing geothermal power on Dominica for export to those countries. That will depend on the fuel supply (continuing with distillate, ECGP, LNG) they select as well as factors such as costs and the number of units that could be converted to natural gas. 1.7.4 Dominican Republic Overview: Prior to 1997 all the generation, transmission, and distribution assets of the Dominican Republic (DR) were owned by the state owned company CDE. In 1997 a capitalization process divided the three entities and the stocks of the companies were sold t private investors. Now the DR has eleven different private thermal power generating companies and a government owned hydroelectric entity, Empresa de Generación Hidroeléctrica Dominicana (EGEHID). AES Dominica, the largest thermal power generator, is owned by AES an international utility company. Other generation companies are La Empresa Generadora de Electricidad Haina (EGE Haina), Generadora Palamara La Vega (GPLV), La Compañía De Electricidad De San Pedro De Macorís (CESPM) and five smaller companies. There are three private and one public distribution companies and a public owned transmission company. Current and Forecast Load: The country s 2008 peak demand was 2,168 MW, with net generation of over 11,600 GWh, making it by far the largest power market of all studied countries. By 2028 peak demand is projected to double to over 4,400 MW, with net generation increasing to around 23,750 GWh (increase rate of 3.4% per year). Fossil Fuel Options: Today the DR has power plants using coal and natural gas derived from LNG, but most of its existing generation uses HFO. Expanding the use of coal and LNG offers the potential to reduce costs and were considered as alternative fuel options. Renewable Generation Potential: The government enacted a law in 2007 defining goals for future renewable energy development. The goal is to have 25% renewable energy by 2025. About 350 MW of wind projects have already been approved. In addition, there is significant additional wind potential based on provisional studies. There are also estimates of 2,899 MW of Caribbean Regional Electricity Generation, Interconnection, and Fuels Supply Strategy Final Report 1-31