Securing Energy for Development in West Bank and Gaza Annexes June 30, 2017 Annex 1: The Palestinian Electricity Sector Figure A1.1 - Electricity supply system in West Bank and Gaza 2 Figure A1.2 – Power supply to Gaza Source: OCHA, 2016 3 Figure A1.3 – PERC Flat Tariff Structure Flat Tariff Applied to residential, commercial and industrial customers consuming < 60MWh/mnth SEGMENTS 2011 2012 2014 Feb-15 Sep-15 Residential (post-paid) - All West bank Except Jericho & Jordan Valley [NIS/kWh] 1- 160 kWh/month 1 - 100kWh/mnth: 0.4085 0.465 0.490 0.441 0.437 161- 250 kWh/month 101 - 200kWh/mnth: 0.4546 0.510 0.528 0.475 0.471 251-400 kWh/month 0.590 0.635 0.572 0.543 401- 600 KWh/month Abov 200kWh/mnth: 0.4795 0.620 0.665 0.600 0.581 Above 600 KWh 0.690 0.735 0.662 0.642 Fixed fees/day 0.333 0.333 0.333 0.333 0.333 Residential (pre-paid) - All West bank except Jericho & Jordan Valley [NIS/kWh] Flat Rate (no segments) 0.467 0.520 0.565 0.500 0.475 Fixed fees/day 0 0 0 0 0 Residential (post-paid) - Only Jericho & Jordan Valley [NIS/kWh] 1- 500 kWh/month NA 0.480 NA 0.450 0.428 Above 500 kWh/month NA 0.520 NA 0.490 0.466 Residential (pre-paid) -Only Jericho & Jordan Valley [NIS/kWh] Flat Rate (no segments) NA 0.520 NA 0.475 0.451 Commercial (post-paid) - Single & 3-phase [NIS/kWh] Flat Rate (no segments) 0.518 0.630 0.667 0.614 0.596 Fixed fees/day 0.667 0.667 0.667 0.667 0.667 Commercial (pre-paid) - Single & 3-phase [NIS/kWh] Flat Rate (no segments) 0.508 0.600 0.637 0.586 0.568 Fixed fees/day 0.34 0 0 0 0 Industrial (Low Voltage) - less than 60MWh consumption / month [NIS/kWh] Flat Rate (no segments) 0.428 0.500 0.537 0.500 0.485 Fixed fees/day 1 1 1 1 1 Industrial (Med Voltage - 6.6, 11, 33kV) [NIS/kWh] Flat Rate (no segments) 0.399 0.450 0.487 0.440 0.414 Fixed fees/day 4 4 4 4 4 Industrial 2 (Marble and Stone) [NIS/kWh] Flat Rate (no segments) NA NA NA 0.54 0.5238 Water pump [NIS/kWh] Flat Rate (no segments) 0.467 0.500 0.537 0.485 0.460 Fixed fees/day 1 1 1 1 1 Agriculture [NIS/kWh] Flat Rate (no segments) 0.409 0.460 0.497 0.448 0.440 Fixed fees/day 0.333 0.333 0.333 0.333 0.333 Street Lights [NIS/kWh] Flat Rate 0.407 0.466 0.503 0.453 0.450 Fixed fees/day 0.333 0.333 0.333 0.333 0.333 Temporary Service (post-paid) [NIS/kWh] Flat Rate 0.683 0.800 0.837 0.754 0.754 Fixed fees/day 0.667 0.667 0.667 0.667 0.667 Temporary Service (pre-paid) [NIS/kWh] Flat Rate 0.683 0.800 0.837 0.754 0.754 Fixed fees/day 0.34 0.34 0.34 0 0 4 Figure A1.4 – PERC Time of Use (TOU) Tariff Structure TOU tariff Applied to Industrial customers consuming > 60MWh/mnth Low Voltage [NIS/kWh] 2015 Israel 2011 PA 2012 PA Feb 2015 Sep 2015 Season TOU tariff Category Tariff Tariff Tariff PA Tariff PA Tariff Rate-A (Off- Peak) 0.3802 0.3055 0.4364 0.4283 0.4010 Winter Rate-B (Mid - Peak) 0.6035 0.5648 0.7541 0.6798 0.6263 Rate-C (Peak) 1.0052 0.9774 1.2859 1.1323 1.0283 Rate-A (Off- Peak) 0.3304 0.2697 0.3826 0.3722 0.3602 Spring & Rate-B (Mid - Peak) 0.409 0.3408 0.4776 0.4607 0.4400 Autumn Rate-C (Peak) 0.503 0.4259 0.5914 0.5666 0.5303 Rate-A (Off- Peak) 0.3441 0.2756 0.3955 0.3876 0.3767 Summer Rate-B (Mid - Peak) 0.4964 0.4453 0.6026 0.5591 0.5408 Rate-C (Peak) 1.1466 1.0792 1.4105 1.2915 1.1894 Medium Voltage [NIS/kWh] Feb 2015 Sep 2015 Tariff ISL 2011 PA 2012 PA Season TOU tariff Category PA Tariff PA Tariff 2015 [6] Tariff [4] Tariff [3] [2] [1] Rate-A (Off- Peak) 0.3014 0.2684 0.3642 0.3395 0.3149 Winter Rate-B (Mid - Peak) 0.513 0.5165 0.6667 0.5778 0.5285 Rate-C (Peak) 0.8796 0.8963 1.1583 0.9908 0.8939 Rate-A (Off- Peak) 0.2556 0.2355 0.3147 0.2879 0.2780 Spring & Rate-B (Mid - Peak) 0.3259 0.2996 0.4007 0.3671 0.3491 Autumn Rate-C (Peak) 0.4141 0.3795 0.5078 0.4664 0.4336 Rate-A (Off- Peak) 0.2639 0.2367 0.3220 0.2973 0.2886 Summer Rate-B (Mid - Peak) 0.3997 0.3905 0.5108 0.4502 0.4349 Rate-C (Peak) 0.9993 0.9772 1.2627 1.1256 1.0305 5 Figure A1.5 - JDECO BALANCE SHEETS (figures in NIS excluding VAT) 2011 2012 2013 2014 2015 Current Assets Accounts receivable 511,543,102 592,638,906 759,513,337 912,051,741 994,313,285 Cash & Cash Equivalents 66,731,355 78,507,761 89,072,998 81,999,095 76,013,661 Asset inventory in warehouse 50,328,765 32,279,908 41,059,610 46,348,215 45,437,813 Work under implementation 147,115,720 236,841,628 NA NA NA Other current assets 3,493,216 7,661,401 18,406,544 15,083,321 10,010,723 Total Current Assets 779,212,158 947,929,604 908,052,489 1,055,482,372 1,125,775,482 Non-Current Assets Property Plant & Equipment 327,291,487 361,827,870 415,044,614 627,740,662 715,900,801 Projects under construction NA NA 257,407,020 108,497,147 128,373,720 Intangible Assets 50,000 50,000 50,000 50,000 50,000 Other non-current assets 7,130,789 22,717,324 32,248,657 43,470,784 46,554,248 Total non-current assets 334,472,276 384,595,194 704,750,291 779,758,593 890,878,769 Total Assets 1,113,684,434 1,332,524,798 1,612,802,780 1,835,240,965 2,016,654,251 Current liabilities Accounts Payable 285,831,913 441,908,286 881,953,033 1,027,225,379 1,255,331,424 Other current liabilities 115,593,048 116,281,908 132,567,633 139,247,227 143,390,312 Total Current Liabilities 401,424,961 558,190,194 1,014,520,666 1,166,472,606 1,398,721,736 Non-current Liabilities Long term loans 192,511,116 152,496,471 117,087,630 92,051,231 68,657,070 Provision for end of service 67,999,128 68,395,500 86,197,324 81,978,304 89,250,091 Deferred Revenue 127,381,058 179,957,470 114,982,955 118,558,214 130,182,770 Other allocation reserves 3,586,600 3,586,600 3,586,600 3,586,600 5,086,600 Total non-current liabilities 391,477,902 404,436,041 321,854,509 296,174,349 293,176,531 Equity Paid up capital 178,875,000 178,875,000 178,875,000 178,875,000 178,875,000 Treasury shares -3,879,311 -7,666,691 -1,486,709 -3,622,230 -3,622,230 Statutory reserve 9,187,500 9,187,500 9,187,500 9,187,500 9,187,500 Revaluation reserve 86,962,931 69,570,345 53,716,168 33,576,949 67,683,536 Retained earnings 49,635,451 119,932,409 36,135,646 154,576,791 72,632,178 Total equity 320,781,571 369,898,563 276,427,605 372,594,010 324,755,984 Total liabilities & equity 1,113,684,434 1,332,524,798 1,612,802,780 1,835,240,965 2,016,654,251 *Source: 2011-2015 Annual Reports (all years audited by PWC) 6 Figure A1.6 - JDECO INCOME STATEMENTS (figures in NIS excluding VAT) 2011 2012 2013 2014 2015 Operating Income Electricity Sales (billed) 694,862,965 875,140,233 888,860,424 950,714,795 949,052,263 Purchased Electricity -562,555,632 -800,261,437 -831,806,133 -886,356,917 -871,483,182 Gross Profit from sales 132,307,333 74,878,796 57,054,291 64,357,878 77,569,081 Subscriber's contribution to extension of services 35,149,206 22,703,391 68,183,242 54,921,120 55,149,003 Revenue from services 9,828,978 7,166,019 9,646,416 11,571,609 10,182,591 Total Operating Income 177,285,517 104,748,206 134,883,949 130,850,607 142,900,675 Operating Expenses General & Admin expenses -145,790,757 -148,425,865 -162,865,171 -171,517,383 -187,635,103 Depreciation Expenses -23,871,283 -21,160,451 -19,752,101 -29,677,585 -36,690,360 Provision for doubtful receivables NA NA -2,378,492 -2,245,586 -4,000,000 Provision for obsolete & damaged goods NA NA -1,508,245 -1,508,245 -1,815,124 Total Operating Expenses -169,662,040 -169,586,316 -186,504,009 -204,948,799 -230,140,587 Net Income/Losses before other income & expenses 7,623,477 -64,838,110 -51,620,060 -74,098,192 -87,239,912 Financing Expenses -21,769,450 -33,838,017 -28,076,247 15,225,873 10,827,836 Other income 8,993,026 27,270,860 4,725,648 5,217,910 2,191,500 Annual income/loss before income tax -5,152,947 -71,405,267 -74,970,659 -53,654,409 -74,220,576 Income Tax expense -2,589,440 0 0 -2,791,446 -7,658,068 Annual Income/Loss -7,742,387 -71,405,267 -74,970,659 -56,445,855 -81,878,644 *Source: 2011-2015 Annual Reports (all years audited by PWC) 7 Figure A1.7 - SELCO BALANCE SHEETS (figures in NIS excluding VAT) 2011 2012 2013 2014 2015 Current assets Cash and cash equivalents 3,183,063 9,114,211 5,435,813 16,670,961 11,389,588 Checks under collection 4,381,660 6,607,877 5,379,639 2,308,503 4,388,066 Stakeholders net receivables 128,539,580 142,283,171 158,385,189 205,881,442 218,715,066 Inventories 42,268,623 38,112,171 30,555,028 33,265,776 27,559,593 Prepaid payments and debit balances 4,403,698 5,548,385 9,511,609 13,917,083 23,312,365 Total current assets 182,776,624 201,665,815 209,267,278 272,043,765 285,364,678 Non-Current Assets Beit Ummar Municipality 6,892,635 6,892,635 6,892,635 6,892,635 6,892,635 Net fixed assets 86,856,645 107,515,454 107,259,124 108,602,044 138,986,851 Work-in-progress 13,109,462 526,611 4,385,403 9,405,428 0 Other 0 0 0 1,602,317 3,382,006 Total non-current assets 106,858,742 114,934,700 118,537,162 126,502,424 149,261,492 Total assets 289,635,366 316,600,515 327,804,440 398,546,189 434,626,170 Current liabilities Accounts payable 11,928,801 5,054,300 15,598,036 16,196,681 37,843,068 Other current liabilities 7,671,467 10,503,416 11,557,674 17,974,835 18,323,477 Total current liabilities 19,600,268 15,557,716 27,155,710 34,171,516 56,166,545 Long-term liabilities Long-term loans 78,142,027 84,069,134 82,815,894 83,084,752 83,015,967 Severance allowances 2,817,115 3,334,956 4,695,549 4,646,758 5,563,625 Ministry of Finance 179,409,815 223,048,447 257,816,215 325,136,253 333,309,386 Total long-term liabilities 260,368,957 310,452,537 345,327,658 412,867,763 421,888,978 Total liabilities 279,969,225 326,010,253 372,483,368 447,039,279 478,055,523 Equities Paid-in capital 44,250 44,250 44,250 44,250 44,250 Statutory reserve 44,250 44,250 44,250 44,250 44,250 Voluntary reserve 1,869,495 1,869,495 1,869,495 1,869,495 1,869,495 Stakeholders receivables -31,065,858 -40,474,211 -57,391,364 -46,594,927 -61,433,602 Shareholders current account 41,522,376 41,522,376 41,522,376 41,522,376 41,522,376 Accumulative (loss) – Statement B -2,748,372 -12,416,014 -30,767,935 -45,378,534 -25,476,122 Net equities 9,666,141 -9,409,854 -44,678,928 -48,493,090 -43,429,353 Total liabilities and equities 289,635,366 316,600,399 327,804,440 398,546,189 434,626,170 *Source: Financial statements (2011-2013 audited by Talal Abu Gazaleh but 2014-2015 draft/unaudited form) 8 Figure A1.8 - SELCO INCOME STATEMENTS (Figures in NIS excluding VAT) 2011 2012 2013 2014 2015 Revenues Electricity Sales + Discount 48,333,776 55,906,513 53,766,667 76,048,100 67,230,123 Electricity Purchase -42,969,409 -54,065,475 -52,664,853 -70,714,130 -48,869,845 Operating expenses (Wages, rents, -4,024,643 -5,046,172 -8,078,247 -8,251,986 -11,083,814 salaries, maintenance) Installation services revenues 1,537,144 2,743,024 2,442,701 2,178,763 867,314 Other operating revenues 3,424,981 1,205,492 1,293,851 1,171,357 4,758,443 Total profit (loss) 6,301,849 743,382 -3,239,881 432,104 12,902,221 Contributions in kind 131,826 300,269 - 623,738 4,021,210 Currency differential -3,588,404 862,298 2,585,184 -39,483 -249,711 Total profit (loss) before administrative and 2,845,271 1,905,949 -654,697 1,016,359 16,673,720 general expenses Expenses Administrative, general and operating -4,009,067 -4,216,934 -10,058,292 -6,555,902 -7,412,130 expenses Other expenses 2,180,286 1,566,260 1,914,811 347,531 6,909,898 Allowance -4,928,673 -5,493,227 -6,801,094 -7,145,739 -7,115,438 Financing costs -1,574,845 -2,233,066 -2,387,574 -1,933,661 -3,185,965 The provision for doubtful debts -502,135 -1,196,624 -340,034 -339,187 -222,797 Total expenses -8,834,434 -11,573,591 -17,672,183 -15,626,958 -11,026,432 Net income/loss of the year -5,989,163 -9,667,642 -18,326,880 -14,610,599 5,647,288 Accumulative (loss) at the beginning of 2,225,724 -2,748,372 -12,416,014 -30,767,935 -45,378,534 the year Prior-years’ adjustments -19,203 - -25,041 0 14,255,124 Net accumulative (loss) at the end of the -3,782,642 -12,416,014 -30,767,935 -45,378,534 -25,476,122 year – Statement A *Source: Financial statements (2011-2013 audited by Talal Abu Gazaleh but 2014-2015 draft/unaudited form) 9 Figure A1.9 - HEPCO BALANCE SHEETS (figures in NIS excluding VAT) 2011 2,012 2013 2014 2015 Current Assets Cash & Cash Equivalent 10,542,073 13,424,893 13,090,423 2,947,436 5,782,708 Checks Under Collection - Short Term 5,510,679 6,152,219 7,967,833 9,876,538 9,387,145 Accounts Receivables - Net 294,580,488 339,491,140 401,093,998 389,259,352 382,332,268 Inventory 28,680,295 39,868,393 30,765,489 30,762,287 29,994,661 Other Current Assets 3,073,656 357618 1,321,385 8,942,093 5,902,513 Hebron Municipality Current Account 133,858,383 153,102,568 187,741,096 225,874,663 270,649,990 Total Current Assets 476,245,574 552,396,831 641,980,224 667,662,369 704,049,285 Long Term Assets Checks Under Collection - Long Term 1,423,064 1,364,601 3,007,089 6,761,154 11,181,157 Work in Process 0 10537049 19,480,450 2,776,993 7,290,115 Properties, Fixed Assets NBV 128,520,429 127,103,039 126,096,754 142,789,491 137,973,496 Concession Rights 30,444,000 30,444,000 30,444,000 30,444,000 30,444,000 Total Long Term Assets 160,387,493 169,448,689 179,028,293 182,771,638 186,888,768 Total Assets 636,633,067 721,845,520 821,008,517 850,434,007 890,938,053 Liabilities & Owner's Equity Current Liabilities World Bank Loan - Short Term 661,915 686,135 1,029,203 1,805,633 3,419,999 Accounts payable + outstanding 466,273,583 555,898,298 650,710,732 626,763,044 651,371,638 Unearned Revenue 4,526,352 5,960,690 1,705,579 10,422,151 11,022,622 Other Current Liabilities 1,559,105 3,785,656 3,467,256 2,035,616 11898610 Total Current Liabilities 473,020,955 566,330,779 656,912,770 641,026,444 677,712,869 Long Term Liabilities Employees End of Service Benefit – 3,596,685 4,364,141 5,248,868 5,163,790 7,014,518 Provision World Bank Loan - Long Term 9,381,319 8,640,322 8,299,574 7,181,005 6,450,622 Deferred Revenues - grants & in-kind 6,822,433 10,334,784 20,864,471 23,974,775 25,285,153 donations Total Long Term Liabilities 19,800,437 23,339,247 34,412,913 36,319,570 38,750,293 Total Liabilities 492,821,392 589,670,026 691,325,683 677,346,014 716,463,162 Owner's Equity Hebron Municipality Paid in Capital 152,745,000 152,745,000 152,745,000 152,745,000 152,745,000 Prior Period Adjustments - VAT -4,303,468 NA Reconciliation Prior Period Adjustments -8,339,560 -5,448,532 -8,933,325 -20,569,506 -23,062,166 Prior Period Adjustments - MoF 41,222,720 41,222,720 Reconciliation Accumulated Losses -8,236,760 -14,044,297 Total Owner's Equity 143,811,675 132,175,494 129,682,834 173,087,932 174,474,891 Total Liabilities & Owner's Equity 636,633,067 721,845,520 821,008,517 850,433,946 890,938,053 *Source: Annual Reports (NOTE: above financial statements are not audited) 10 Figure A1.10 - HEPCO INCOME STATEMENTS (figures in NIS excluding VAT) 2011 2,012 2013 2,014 2015 Revenues Electricity Sales 144,250,785 159,362,877 171,194,239 179,775,466 183,560,826 Add: Tariff Differences 9,352,890 21,979,208 9,700,234 9,854,794 9,612,348 Add: Fixed Charges NA NA NA 2,991,710 NA Deduct: Cost of Electricity Purchased -136,354,132 -159,809,793 -170,222,657 -175,900,386 -163,700,004 Gross Profit 17,249,543 21,532,292 10,671,816 16,721,584 29,473,170 Other Income Customer Participations 4,247,980 1,596,640 2,165,075 2,640,750 6,896,943 Other Operating Revenues 8,704,419 10,570,952 13,004,102 11,546,126 7,922,121 Accrued of Deferred Revenues 758,048 583,833 600,000 795,173 800,000 Total Other Income 13,710,447 12,751,425 15,769,177 14,982,049 15,619,064 Total Operating Income 30,959,990 34,283,717 26,440,993 31,703,633 45,092,234 Expenses Operating Expenses -1,476,263 -3,067,033 -2,841,731 -1,829,166 -2,722,492 General & Admin Expenses -1,366,052 -1,878,389 -2,706,498 -1,469,510 -1,346,796 Payroll Expenses -10,037,633 -12,013,416 -12,983,957 -11,912,764 -12,358,333 Depreciation -9,002,162 -8,797,369 -9,207,810 -10,251,450 -9,762,652 Community Municipality of Hebron NA -231,614 -178,414 NA -853,424 Contributions Loan Interest Expense -105,696 NA NA NA -195,000 World Bank Loan NA NA NA -170,000 NA Currency Differential Loss -539,704 -120,016 -15,243 -100,000 -100,000 Bad Debt Expenses / Doubtful -1,000,000 -11,472,400 -1,000,000 -3,000,000 -6,000,000 Receivables Net Book Value of Assets Disposed NA NA NA NA -1,000,000 Other NA NA NA -927,688 NA Total Operating Expenses -23,527,510 -37,580,237 -28,933,653 -29,660,578 -34,338,697 Net Income 7,432,480 -3,296,520 -2,492,660 2,043,055 10,753,537 Source: Annual Reports (NOTE: above financial statements are not audited) 11 Figure A1.11 - GEDCO BALANCE SHEET (figures in NIS excluding VAT) 2014 2015 Assets Current Assets Cash and Cash Equivalents 6,389,609 2,197,965 Customers' Receivables 3,545,123,306 3,743,707,146 Materials and Supplies in Warehouses 14,635,146 24,182,036 Partners Current Accounts (Municipalities) 370,615,454 399,610,375 Receivables and Other Current Assets 14,697,288 35,880,987 Total Current Assets 3,951,460,803 4,205,578,509 Non-Current Assets Financial Assets at Fair Value 413,478 484,398 Property, Plant and Equipment, Net 116,354,190 122,062,881 Projects in Progress 4,374,270 10,707,531 Total Non-Current Assets 121,141,938 133,254,810 Total Assets 4,072,602,741 4,338,833,319 Liabilities and Shareholders' Equity Current Liabilities Payables and Other Liabilities 103,218,121 128,883,852 Banks Overdraft 0 13,195,769 Total Current Liabilities 103,218,121 142,079,621 Non-Current Liabilities Palestinian National Authority (PNA) 3,978,060,454 4,208,767,055 Canal Company for Electricity Distribution (Egypt) 100,122,920 151,475,280 Deferred Revenues 80,043,837 97,539,206 Sundry Provisions 59,569,735 61,649,421 Total Non-Current Liabilities 4,217,796,946 4,519,430,962 Total Liabilities 4,321,015,067 4,661,510,583 Shareholders' Equity In-Kind Capital (Electricity Distribution Network) 149,280,948 149,280,948 Revaluation Reserve - Electricity Network 50,011,980 50,0 11,980 Cumulative Change in Fair Value 6,030 76,950 Deferred Losses -1,156,470,721 -447,711,284 This Year loss Prof- Exhibit (B) 708,759,437 -74,335,858 Net Shareholders' Equity - -248,412,326 -322,677,264 Total Liabilities and Shareholders' Equity 4,072 1 602,741 4,338,833,319 Source: GEDCO financial statements (unaudited) 12 Figure A1.12 - GEDCO INCOME STATEMENT (figures in NIS excluding VAT) 2014 2015 Operating Revenues from billed sales 509,181,596 517,553,841 Cost of Sale Cost of Energy Sold -383,614,843 -406,186,153 Energy Lost (Not Billed) -127,853,756 -126,671,379 Operating Expenses -3,299,263 -5,540,937 Total Cost of Sale -514,767,862 -538,398,469 Gross Profit -5,586,266 -20,844,628 Deduct Depreciation of Electricity Network -12,500,915 -13,152,821 Staff Costs -40,490,490 -44,020,905 General and Administrative Expenses -13,817,427 -13,678,708 Losses aggression -35,188,472 -18,726 -101,997,304 -70,871,160 Add Realized Grants and Cash Donations 1,507,955 1,191,189 Realized Grants and In-Kind Donations 17,707,847 5,535,508 Other Revenues 2,882,112 3,823,384 22,097,914 10,550,081 Loss for the Year from Activities -85,485,656 -81,165,707 Other Items: Prior Years Adjustments 794,245,093 6,829,849 Total Other Items 794,245,093 6,829,849 This Year loss Prof -Exhibit (A) 708,759,437 -74,335,858 Source: GEDCO financial statements (unaudited) 13 Figure A1.13 - NEDCO BALANCE SHEETS (figures in NIS excluding VAT) 2011 2012 2013 2014 Current Assets Accounts receivable 48,714,506 97,809,318 142,280,736 128,961,956 Cash on hand at banks 11,223,360 24,619,157 23,092,885 23,137,898 Dues from municipal & village councils NA 56,036,239 80,236,793 70,964,603 Other current assets 52,200,611 22,194,384 37,580,642 19,153,316 Total Current Assets 112,138,477 200,659,098 283,191,056 242,217,773 Non-current assets Property & Equipment 230,296,617 253,445,831 258,628,054 261,864,378 Projects uner construction NA 1,037,694 691,256 3,331,636 Stock items NA 22,683,775 28,566,280 25,916,981 Total Non-current Assets 230,296,617 277,167,300 287,885,590 291,112,995 Total Assets 342,435,094 477,826,398 571,076,646 533,330,768 Current liabilities Accounts Payable 31,620,495 18,117,173 64,573,004 70,652,834 Other current liabilities 76,439,237 185,617,359 219,583,341 170,255,539 Total Current liabilities 108,059,732 203,734,532 284,156,345 240,908,373 Non-curent liabilities Provision for end of service 1,230,496 2,641,153 4,188,232 6,109,462 Deferred earnings NA 33,786,138 37,921,059 39,528,711 Other non-current liabilities 300,700 NA NA NA Total non-current liabilities 1,531,196 36,427,291 42,109,291 45,638,173 Equity Paid-up capital 15,251,594 17,231,440 17,231,440 17,231,440 Shareholder accounts 204,698,552 208,832,878 208,932,490 209,415,550 Statutory reserve 1,289,402 2,191,547 2,896,229 3,766,961 Optional reserve 1,289,402 2,191,547 2,896,229 3,766,961 Retained earnings 10,315,216 7,217,163 12,854,622 12,603,310 Total Equity 232,844,166 237,664,575 244,811,010 246,784,222 Total Liabilities & Equity 342,435,094 477,826,398 571,076,646 533,330,768 Source: Financial Statements audited by Ernst and Young (2015 financial statements not available) 14 Figure A1.14 - NEDCO INCOME STATEMENTS (figures in NIS excluding VAT) 2011 2012 2013 2014 Operating Income Electricity Sales (billed) + subscriptions + services etc 188,881,771 225,555,641 254,389,364 260,922,894 Electricity purchases + salaries & wages + depreciation -178,567,444 -199,879,476 -229,675,461 -249,814,659 Gross profit / Operating Income 10,314,327 25,676,165 24,713,903 11,108,235 Operating Expenses General & Admin expenses -10,119,348 -17,066,283 -12,359,573 -12,741,808 Depreciation -723,176 -1,889,580 -1,697,289 -1,508,697 Provision for doubtful receivables -795,182 1,269,174 -792,774 -5,422,052 Other expenses -300,700 NA NA NA Total Operating Expenses -11,938,406 -17,686,689 -14,849,636 -19,672,557 Net Income/Losses before other income & expenses -1,624,079 7,989,476 9,864,267 -8,564,322 Revenue settlement with MoF 0 0 0 24,865,770 Grant from PENRA 1,804,535 NA NA NA Other income 310,568 2,916,473 1,878,699 -841,050 Annual profit before income taxes 491,024 10,905,949 11,742,966 15,460,398 Income tax expenses -399,893 -1,884,496 -4,696,143 -6,753,083 Annual Profit after income tax 91,131 9,021,453 7,046,823 8,707,315 Source: Financial Statements audited by Ernst and Young (2015 financial statements not available) 15 Figure A1.15 - TEDCO BALANCE SHEETS (figures in NIS excluding VAT) 2011 2012 2013 2014 2015 Current Assets Cash in bank 4,569,288 5,633,391 6,586,378 35,731,934 3,670,399 Checks 0 1,016,042 1,152,079 1,842,673 3,021,252 Accounts Receivable 18,974,046 22,267,521 25,261,915 33,723,803 40,106,657 other receivables 3,980,653 7,003,961 16,088,166 27,497,055 14,732,989 Accessories & spare parts in warehouse 809,865 1,535,601 1,488,859 1,669,365 1,959,394 Prepaid Expenses 141,785 164,227 166,001 225,038 186,853 Total Current Assets 28,475,637 37,620,743 50,743,398 100,689,868 63,677,544 Fixed Assets 20,849,458 22,975,233 24,146,557 25,764,776 28,768,289 Fixed Asset Consumption -6,723,629 -7,846,018 -9,030,484 -10,132,413 -11,451,747 Net Fixed Assets 14,125,829 15,129,215 15,116,073 15,632,363 17,316,542 Total Assets 42,601,466 52,749,958 65,859,471 116,322,231 80,994,086 Liabilities and Equity Accounts Payable 11,608,119 30,917,332 43,164,091 92,984,000 55,848,851 Other Payables 9,498,750 0 1,712,992 1,931,121 0 Due Payments 7,133 1,202,011 41,373 5,500 9,500 Income Tax provision 65,103 65,103 65,103 65,103 572,664 Other Provisions 524,622 755,152 956,586 1,237,181 1,611,351 Total Liabilities 21,703,727 32,939,598 45,940,145 96,222,905 58,042,366 Capital 15,361,808 15,361,808 15,361,808 15,361,808 15,361,808 Capital Reserve 5,374,958 5,374,958 5,374,958 5,374,958 5,374,958 Legal/statutory reserve 481,660 481,660 481,660 510,557 795,796 Earning from previous years 663,091 0 0 0 0 Losses 0 -320,687 -1,408,066 -1,309,997 -1,147,997 Net Loss for the Year -983,778 -1,087,379 108,966 162,000 2,567,155 Total Equity 20,897,739 19,810,360 19,919,326 20,099,326 22,951,720 Total Liabilities & Equity 42,601,466 52,749,958 65,859,471 116,322,231 80,994,086 Source: Financial statements audited by Jamal Abu Farha 16 Figure A1.16 - TEDCO INCOME STATEMENTS (figures in NIS excluding VAT) 2011 2012 2013 2014 2015 Revenues Electricity Sales - Pre paid 17,539,822 20,048,480 21,065,333 Electricity Sales - mechanical counters 7,916,514 11,280,770 10,736,612 29,751,149 34,608,924 Electricity Sales - medium voltage 12,124,651 13,012,333 13,374,891 Electricity Sales - Street Lighting 1,135,804 1,228,366 1,045,340 Electricity Purchases -26,771,696 -33,147,964 -38,208,786 -44,740,672 -42,342,607 Gross Profit Electricity Sales 2,979,453 1,460,960 508,005 829,277 3,879,569 Other Revenues Revenue from misc services 0 2,163,512 1,696,691 3,118,616 3,610,845 Government Support for Elec Production (Subsidy) 0 0 2,976,902 2,472,444 2,821,416 Income from transformer maint center 346,075 825,730 801,657 773,759 1,206,323 Total Other Revenues 346,075 2,989,242 5,475,250 6,364,819 7,638,584 Expenses Operating Expenses -2,781,780 -3,159,127 -3,210,182 -3,985,514 -4,801,092 General & admin expenses -887,262 -1,768,156 -1,906,374 -2,305,516 -2,561,213 Transformer main center expenses -640,264 -610,298 -757,733 -723,066 -730,790 Total Expenses -4,309,306 -5,537,581 -5,874,289 -7,014,096 -8,093,095 Net profit from transformer maint center -294,189 215,432 43,924 50,693 475,533 Total Profit from elec sales (not including transf maint center) -689,589 -1,302,811 65,042 129,307 2,949,525 Total Net Profit (including tranf maint center) -983,778 -1,087,379 108,966 180,000 3,425,058 Income Tax 0 0 16,345 38,000 572,664 Statutory Reserve - 10% 0 0 10,897 18,000 285,239 Net Profit after Taxes & Reserves 0 0 81,724 124,000 2,567,155 Source: Financial statements audited by Jamal Abu Farha 17 Annex 2: Electricity Demand Figure A2.1: Shifting patterns of energy usage for cooking and baking (percentage of households) 100% 90% 80% 70% Not available 60% 50% Other 40% Kerosene 30% Wood 20% Gas 10% 0% Electricity Jul Jan Jul Jan Jul Apr Apr Jan Jul Jul Jul Jan Jul Jan Jul Apr Apr Jan Jul Jul 2003 2004 2005 20062008 2009 20132003 2004 2005 20062008 2009 2013 Cooking (stove top) Baking (oven) Source: PCBS Household Energy Surveys, 2003-2013 Figure A2.2: Shifting patterns of energy usage for water heating (percentage of households) 100% 90% 80% Not available 70% 60% Other 50% Kerosene 40% Solar 30% Wood 20% Gas 10% Electricity 0% 2003 2004 2005 2009 2006 2008 2001 2003 2004 2005 2009 2013 January April July Source: PCBS Household Energy Surveys, 2001-2013 18 Figure A2.1: Electricity consumption regression model results for the summer season Independent variable Observations R-square Household grid electricity consumption, 14,001 0.16 July Variable Coefficient Standard Error t-Statistic Gaza region dummy 191.9 28.8 6.7 North West-Bank region dummy 183.8 28.1 6.5 Mid West-Bank region dummy 344.3 28.8 12.0 South West-Bank region dummy 211.1 28.6 7.4 Ownership of electric air conditioner 108.5 9.8 11.1 Ownership of electric fan 53.6 14.8 3.6 Ownership of solar heater 27.7 4.1 6.7 Main cooking fuel is electricity 10.9 38.5 0.3 Main baking fuel is electricity -2.3 3.9 -0.6 Main water heating fuel is electricity 41.7 6.0 7.0 Ownership of electric generator -20.0 13.5 -1.5 Source: Own elaboration Figure A2.4: Electricity consumption regression model results for the winter season Independent variable Observations R-square Household grid electricity consumption, 6,733 0.20 Jan. Variable Coefficient Standard Error t-Statistic Gaza region dummy 239.4 22.5 10.6 North West-Bank region dummy 217.7 23.5 9.3 Mid West-Bank region dummy 329.8 24.9 13.2 South West-Bank region dummy 268.7 23.5 11.4 Ownership of electrical heater 51.3 4.9 10.5 Ownership of solar water heater 26.4 3.5 7.6 Main cooking fuel is electricity 9.8 21.6 0.5 Main baking fuel is electricity 4.8 5.4 0.9 Main water heating fuel is electricity 67.5 5.2 13.1 Ownership of electric generator -45.9 10.9 -4.2 Source: Own elaboration 19 Figure A2.5: Existing and forecasted electricity needs of the water and wastewater sector in Gaza Water/Wastewater Facility 2014 2017 2018 2020 2025 2030 2035 North Gaza WWTP (NGEST) Component: Terminal Pumping Station TPS 2 1 1 2 2 3 3 Waste Water Treatment Plant WWTP 2 2 2 3 4 5 5 Recovery and Reuse Scheme Phase 1 2 2 3 3 4 5 Recovery and Reuse Scheme Phase 2 3 3 4 5 5 6 NGEST Total 4 8 9 11 14 17 19 Planned central Gaza WWTP(KFW) 7 7 8 11 13 Khanyounis WWTP 2 2 3 5 6 Rafah Existing WWTP 1 2 2 2 2 3 Gaza existing WWTP(shikh Ejleen) 5 3 3 3 3 Central Desalination Plant 35 35 35 55 55 Deiralbalah Desalination Plant 1 1 1 2 2 2 Gaza Deslination Plant 3 3 3 3 3 3 Existing W and WW Facilties 25 33 30 30 30 30 30 Total Gaza Governorates Energy Required 34 46 87 93 99 127 134 for W & WW Facilties 20 Annex 3: Importing Electricity from Israel Figure A3.1: Electricity Generation in Israel by Type of Fuel and Producer, 2014-2015 Coal Natural Gas Gasoil HFO Renewables Total GWh % GWh % GWh % GWh % GWh % GWh % 2014 IEC 30 58% 22 42% 0.05 0% 0.01 0% 0 0% 52 84% IPPs 0 0% 9 91% 0 0% 0.01 0% 0.87 9% 10 16% Total 30 49 31 49.5 0.05 0.1 0.02 0 0.87 1.4 62 100% 2015 IEC 29 58% 21 42% 0.37 1% 0.06 0% 0 0% 50 78% IPPs 0 0% 13 90% 0.12 1% 0.02 0% 1.28 9% 14 22% Total 29 44.6 34 52.6 0.49 0.7 0.08 0.1 1.28 2 65 100% Source: PUA, June 2016 Figure A3.2: Israeli Generation Capacity, 2007-2015 (MW) 2007 2008 2009 2010 2011 2012 2013 2014 2015 Installed 11,297 11,649 11,664 12,769 12,759 13,248 13,483 13,617 13,617 Capacity Figure A3.3: IEC electricity sales by type of consumers, 2014–2015 Total electricity 2014 Total electricity 2015 consumption (%) (MWh) consumption (%) (MWh) Domestic 34.8 17,604 32.1 15,981 Industrial 18.0 9,108 17.9 8,951 Public and Commercial 30.3 15,342 32.0 15,953 Water pumping 4.0 2,018 4.8 2,404 Agriculture 2.6 1,332 3.5 1,769 East Jerusalem Electricity company 4.2 2,128 3.9 1,945 Palestinian Authority 6.1 3,069 5.8 2,899 Total 100 50,601 100 49,902 Source: IEC Financial Statement of 2015 (Published March 31, 2016) 21 Figure A3.4: Electricity demand forecast for Israel, 2016-2030 Year Generation Consumption Peak demand (GWh) (GWh) (MW) 2016 67.8 63.4 13,191 2017 70.1 65.5 13,670 2018 72.4 67.7 14,126 2019 74.8 69.9 14,577 2020 76.9 71.9 14,960 2021 79.2 74.0 15,446 2022 81.5 76.2 15,895 2023 83.9 78.4 16,348 2024 86.2 80.6 16,767 2025 88.5 82.7 17,265 2026 91.0 85.0 17,734 2027 93.6 87.5 18,236 2028 96.1 89.8 18,688 2029 98.7 92.2 19,237 2030 101.1 94.5 19,711 Note: The forecast is based on data from IEC and is based on an annual growth of 1.9% in GDP per capita and extreme heat stress conditions. Source: Ministry of National Infrastructure, Energy and Water resources, http://energy.gov.il/Subjects/Electricity/Pages/GxmsMniAboutElectricity.aspx Figure A3.5: Overview of Israeli transmission and distribution service tariffs (NIS Agorot per KWh, as of 13.09.2015) ToU Transmission Transmission & Distribution Season block tariffs * distribution tariffs** tariffs*** Off peak 0.89 3.46 2.55 Winter Shoulder 1.10 3.89 2.78 Peak 2.80 7.22 4.38 Off peak 0.81 3.22 2.41 Transition Shoulder 1.36 4.17 2.80 Peak 1.79 4.82 3.01 Off peak 1.42 4.20 2.77 Summer Shoulder 2.60 6.32 3.68 Peak 6.12 12.13 5.90 * Ultra high voltage producer selling to ultra high voltage consumer ** Ultra high voltage or high voltage producer selling to "far away" high voltage consumer *** Ultra high voltage producer selling to "close by" high voltage consumer Exchange rate: US$ 1 = NIS 3.846 (June 30, 2016) Ultra high voltage = 400 Kilo Volt & 161 KV, high voltage = 22 KV & 33 KV Source: Israel PUA 22 Figure A3.6: Efficiency Coefficients and Return on Equity Used in Israeli Tariff-Setting (%) Share of Efficiency Weighted Annual Weighted assets Coefficient Efficiency Return On Return On Coefficient Equity Equity Generation 50.1 2.0 1.00 7.62 3.82 Transmission 19.5 1.3 0.25 5.50 1.07 Distribution 30.3 3.7 1.12 6.20 1.88 100.0 2.38 6.78 Source: Israel PUA and IEC Financial Statements, 2015. Figure A3.7: Period Definitions for Israeli Time-of-Use Tariff Rates Time of Use period definitions Fridays and Season Time of day Saturdays and days before Weekdays holidays holidays Peak 1017 Summer Shoulder 0710 , 1721 (July – August)  24  Off-peak 00 00 24 0007 , 2124 1622 Winter Peak 1719 1620 0608 (December – Shoulder 1921 ,0816, 2224 February) Off-peak 0017 , 2124 0016 , 2024 0006 Transition Peak 0620 (Remaining Shoulder 1721 0620 2022 months) Off-peak 0017 , 2124 0006 , 2024 0006 , 2224 Source: Israel PUA. Last update 15.02.2010. 23 Figure A3.8: Israeli Time-of-Use Tariffs (NIS Agorot* per KWh) Season ToU block Low High Ultra- High Voltage ** Voltage** Voltage** Winter Off-peak 35.60 27.96 25.18 Shoulder 55.60 46.92 43.62 Peak 91.29 79.36 73.93 Transition Off-peak 31.98 24.68 22.10 Shoulder 39.06 30.99 27.89 Peak 47.08 38.49 35.06 Summer Off-peak 33.44 25.62 22.63 Shoulder 48.01 38.61 34.47 Peak 105.59 91.49 84.18 Source: Israel PUA as of 13.09.2015. * Exchange rate: US$ 1 = NIS 3.846 (June 30, 2016) ** Ultra-high voltage = 400 KV & 161 KV, high = 22 KV & 33 KV, low = 400 Volt Figure A3.9: Israeli Bulk Supply Tariffs (NIS Agorot per KWh) Low voltage High Voltage 44.35 35.92 Source: Israel PUA as of 13.09.2015 Figure A3.10: Israeli System Management Services Tariffs (NIS Agorot/kWh) Season Time-of- Administrative System Backup Other system Total Use costs balance services services Off-peak 0.27 0.58 0.40 4.01 5.26 Winter Shoulder 0.27 0.58 0.77 4.01 5.63 On-peak 0.27 0.58 1.35 4.01 6.21 Off-peak 0.27 0.58 0.34 4.01 5.20 Transition Shoulder 0.27 0.58 0.43 4.01 5.30 On-peak 0.27 0.58 0.56 4.01 5.42 Off-peak 0.27 0.58 0.34 4.01 5.20 Summer Shoulder 0.27 0.58 0.54 4.01 5.41 On-peak 0.27 0.58 1.41 4.01 6.27 Average tariff 0.27 0.58 0.54 4.01 5.40 Source: Israel PUA as of 13.09.2015 Exchange rate: US$ 1 = NIS 3.846 (June 30, 2016) 24 Annex 4: Importing Natural Gas for Domestic Power Generation Figure A4.1: Prospective industrial consumers of natural gas in the West Bank Diesel LPG Natural Gas Name of No. City Type of Factory Consumption Consumption Demand Company (L/year) (kg/year) 1000 CM/Year* 1 BPC company Ramallah Pharmaceutical 134,785 0 126 2 Star Factory Ramallah Chemical 59,512 31,055 94 Al-Juneidi 3 Hebron Food 1,020,000 0 954 Factory 4 Aziza Factory Tulkarem Food 170,138 0 159 5 NBC Factory Ramallah Food 134,611 0 126 Siniora 6 Aziza Food 202,042 0 189 Factory Sinokrot 7 Ramallah Food 166,307 116,798 301 Factory AL- Jebrini 8 Hebron Food 92,028 121,575 238 Factory Al-Arz 9 Nablus Food 0 112,794 141 Company Al-Safa 10 Nablus Food 0 116,575 146 Factory Al-Betra 11 Hebron Food 0 67,192 84 Company NAPCO 12 Nablus Aluminium 0 379,464 474 Company Total Consumption/year 1,979,423 945,453 3,033 Source: Palestinian Federation of Industry - Eco Energy's Calculation of NG Demand * Natural gas conversion factors: 1000 liters of Diesel = 935 cubic meters of gas 1 ton LPG=1250 cubic meters Figure A4.2: Forecast demand for natural gas based on power generation in the West Bank Installed Electricity Total electricity % of domestic Natural gas Capacity 1 Generation Demand 2 Production 3 demand4 Year MW GWh GWh % BCM 2022 200 1226 6417 19 0.24 2023 200 1226 6802 18 0.24 2024 400 2453 7210 34 0.47 2025 400 2453 7643 32 0.47 2026 400 2453 8101 30 0.47 2027 400 2453 8587 29 0.47 2028 600 3679 9103 40 0.71 2029 600 3679 9649 38 0.71 2030 600 3679 10228 36 0.71 1 Jenin IPP: 200 MW at 2022, 400 MW at 2024; Tarkumiye IPP: 200 MW at 2028 2 Based on 2015 demand in the WB of 4286 GW and assumed growth rate of 6% per annum 3 Share of domestic gas based generation of total electricity demand in the West Bank 4 CCGTs have 57% efficiency and operated at 70% capacity 25 Figure A4.3: Forecast demand for natural gas based on power generation in Gaza Total Natural Gas based power Capacity & Generation % domestic Natural gas electric Converted New Total Generation Demand 2 production 3 demand 4 GPP1 CCGT1 capacity Year MW MW MW GWh GWh % BCM 2022 70 70 429 1462 29 0.11 2023 70 70 429 1550 28 0.11 2024 140 140 858 1643 52 0.21 2025 140 140 858 1741 49 0.21 2026 140 100 240 1472 1846 80 0.33 2027 140 100 240 1472 1956 75 0.33 2028 140 100 240 1472 2074 71 0.33 2029 140 100 240 1472 2198 67 0.33 2030 140 100 240 1472 2330 63 0.33 1 GPP: 70 MW conversion from gasoil to gas at 2022, additional 70 MW at 2024; New 100 MW CCGT at 2026 2 Based on 2015 demand in Gaza of 972 GWh, and assumed average growth rate of 6% per annum 3 Share of domestic gas based generation of total electricity demand in Gaza 4 Converted GPP works at 45% efficiency; new CCGT works at 57% efficiency, all plants work at 70% capacity Figure A4.4: Natural Gas Prices in Israel 2016 ($/mmbtu) consumer Initial price Indexation IEC 5.7 U.S. CPI +- 1%/yr * Major IPPs 4.7-5.0 IEC generation tariff with ceiling Major Industries 4.7-5.5 basket of fuels with cap Marketing companies 5.2-5.8 heavy fuel oil with cap final price for small industries 6.0-7.0 heavy fuel oil with cap * IEC's price indexation formula: U.S. CPI+1% /year until 2020 and then U.S. CPI - 1%/year for 7 years 26 Annex 5: Importing Electricity from Jordan & Egypt Figure A5.1: Projected fossil fuel supply situation in the Egyptian power market Average capacity utilization Unit 2015 2016 2017 2018 2019 2020 2021 Average (fossil fuel) % 54% 55% 52% 44% 40% 41% 41% Specific generation cost Unit 2015 2016 2017 2018 2019 2020 2021 (marginal cash cost for EEHC) Coal USD/kWh 0.000 0.000 0.000 0.000 0.000 0.030 0.033 Heavy Fuel Oil USD/kWh 0.042 0.044 0.045 0.047 0.048 0.050 0.052 Light Fuel Oil USD/kWh 0.059 0.059 0.060 0.062 0.064 0.065 0.068 Natural Gas USD/kWh 0.022 0.024 0.025 0.027 0.029 0.032 0.034 Average (fossil fuel) USD/kWh 0.027 0.029 0.030 0.032 0.033 0.035 0.038 Specific generation cost (marginal Unit 2015 2016 2017 2018 2019 2020 2021 economic cost) Coal USD/kWh 0.000 0.000 0.000 0.000 0.000 0.030 0.033 Heavy Fuel Oil USD/kWh 0.042 0.035 0.045 0.051 0.057 0.063 0.070 Light Fuel Oil USD/kWh 0.096 0.078 0.106 0.119 0.133 0.149 0.167 Natural Gas USD/kWh 0.039 0.033 0.042 0.046 0.050 0.056 0.063 Average (fossil fuel) USD/kWh 0.040 0.034 0.043 0.047 0.052 0.057 0.062 Generation Unit 2015 2016 2017 2018 2019 2020 2021 Coal GWh - - - - - 5,995 19,376 Heavy Fuel Oil GWh 37,489 39,701 39,726 41,118 43,935 45,067 44,987 Light Fuel Oil GWh 451 450 427 364 328 336 336 Natural Gas GWh 124,005 131,543 138,433 147,148 154,814 158,803 158,520 Total (fossil fuel) GWh 161,946 171,694 178,586 188,629 199,076 210,202 223,218 Capacity Unit 2015 2016 2017 2018 2019 2020 2021 Combined-Cycle Gas Turbine MW 11,730 12,480 17,230 23,730 29,730 29,730 29,730 Gas Turbine MW 6,794 7,020 5,820 7,030 7,030 7,030 7,030 Steam turbine (oil & gas boiler) MW 2,800 2,800 2,800 2,832 2,832 2,832 2,832 Steam turbine (coal boiler) MW - - - - - 1,600 5,180 Total (fossil fuel) MW 21,324 22,300 25,850 33,592 39,592 41,192 44,772 27 Annex 6: Developing Domestic Renewable Power Generation Table A6.1: Estimation of disaggregated potential for residential rooftop PV Number of Residential Rooftops % # % Household Number Governorate Population* individual rooftop population (HH) size* of HHs houses* for PV West Bank 2,790,331 Region 57.4% 4.9 569,455 326,867 Jenin 303,565 WB-N 1,094,815 24.1% 223,432 128,250 Tubas 62,627 WB-N Tulkam 178,774 WB-N Nablus 372,621 WB-N Qualqilya 108,049 WB-N Salfit 69,179 WB-N Ramallah 338,383 WB-C 1,011,269 22.2% 206,381 118,463 Jericho 50,762 WB-C Jerusalem 411,640 WB-C Bethlehem 210,484 WB-C Hebron 684,247 WB-S 684,247 15.0% 139,642 80,155 Gaza Strip 1,760,037 Gaza 1,760,037 38.7% 29.3% 5.7 308,778 90,472 North 348,808 Gaza 606,749 Dier al Balah 255,705 Khan Yunis 331,017 Rafah 217,758 TOTAL 4,550,368 100.0% 878,234 417,339 *Source: PCBS Table A6.2: Estimation of number of rooftops for non-residential rooftop PV Public Administration 200 Schools 2200 Commercial 5000 Source: PEC Table A6.3: Assumptions regarding land requirements for solar power generation in Palestine (m2/KWp) Rooftop solar 8-12 Utility scale PV 24-32 CSP 31-40 Wind 210-330 Source: NREL – Land-use requirements for solar power plants in the US (2013), Land-Use Requirements of Modern Wind Power Plants in the US (2009) 28 Table A6.4: Estimation of overall potential for rooftop solar PV in Palestine Maximum Potential Capacity – Rooftop Solar West Bank Area per Potential # rooftop % rooftop % well Available rooftop Capacity for PV available oriented Surface (m2) (m2) (MW) Residential 326,867 150 30% 30% 4,412,709 490 Public 123 200 40% 100% 9,811 1 Schools 1,349 160 50% 100% 107,925 12 Commercial 3,066 300 30% 100% 275,944 31 Gaza Area per Potential # rooftop % rooftop % well Available rooftop Capacity for PV available oriented Surface (m2) (m2) (MW) Residential 90,472 150 30% 30% 1,221,373 136 Public 77 200 40% 100% 6,189 1 Schools 851 160 50% 100% 68,075 8 Commercial 1,934 300 30% 100% 174,056 19 TOTAL West Bank & Gaza 697 Table A6.5: Estimation of potential for utility scale solar power generation in Palestine Maximum Potential Capacity – Utility Scale Solar Available Available total surface Potential Capacity according to PETL according to PETL (km2) (%) (%) (km2) (MWp) Area A&B 2488 40% 0.12% 3 103 Area C 3732 60% 2.64% 98.5 3374 TOTAL 6220 100% 2.76% 101.5 3476 29 Figure A6.6: Currently installed and on-going solar projects at Gaza hospitals Unit benefiting of the Capacity Budget MOH facility Donor Status project (W) (USD) Italian workers Shifa hospital Cardiac care 50,000 Completed 1 syndicate 4,500 2 Shifa hospital ICU JICA 30,000 150,000 Completed Nassr pediatric NCU (Nursery) Sawaed Society 90,000 Completed 3 hospital 20,000 Harazen maternity OT, Lab, lighting UNDP 60,000 Completed 4 hospital 12,000 Emirati RC maternity OT lights UNDP 40,000 Completed 5 hospital 8,000 6 EGH ICU ICRC 30,000 140,000 Completed Refrigerators for 32 PHC clinics ICRC 190,000 Completed 7 vaccines 750 Tahreer maternity OT, delivery wards, Human Appeal 217,180 On-going 8 hospital NCU, ED Int. 50,000 9 Al-Aqsa hospital OT, NCU, Cardiac care UNDP 60,000 225,000 On-going 10 Indonesian hospital OT, ED UNDP 60,000 225,000 On-going Rantissi Specialized Welfare NCU, part of Lab 150,000 On-going 11 hospital Association 30,720 TOTAL 305,970 1,537,180 Source: World Health Organization (WHO) 30 Figure A6.7: Critical units in Gaza MOH hospitals in need of solar energy Hours of Capacity Budget MOH facility Targeted Unit power supply (KWp) USD Hemodialysis (38 HD unit + desalination plant) 12 100 500,000 NCU for premature babies (35 incubator) 24 30 120,000 1 Shifa hospital Cardiac Care 24 30 120,000 Laboratory 24 20 80,000 Sterilization unit (to operate one OT sterilizer) 12 40 160,000 OT rooms (eight rooms) 6 30 120,000 NCU (14 beds) 24 30 120,000 2 EGH Neurology care (12 beds) 24 30 120,000 Laboratory 24 20 80,000 Sterilization unit (to operate one OT sterilizer) 12 40 160,000 OT rooms (three rooms) 6 20 80,000 ICU (16 beds) 24 30 120,000 Nasser Hemodialysis (18 HD unit + desalination plant) 12 50 200,000 3 hospital NCU for premature babies (20 incubator) 24 30 120,000 (Khanyounis) Laboratory 24 20 80,000 Sterilization unit (to operate one OT sterilizer) 12 40 160,000 Rantissi ICU (4 beds) 24 10 40,000 4 Specialized Hemodialysis (5 HD units) 12 10 40,000 hospital Dorra ICU (6 beds) 24 20 80,000 5 pediatric Laboratory 24 20 80,000 hospital OT rooms ( three rooms) 6 15 60,000 6 Eye hospital Sterilization unit (to operate one OT sterilizer) 12 40 160,000 OT rooms ( two rooms) 6 15 60,000 Beit Hanoun 7 Laboratory 24 20 80,000 hospital Sterilization unit (to operate one OT sterilizer) 12 40 160,000 Hemodialysis (18 HD units) 12 50 200,000 Al-Aqsa 8 Laboratory 24 25 100,000 hospital Sterilization unit (to operate one OT sterilizer) 12 40 160,000 ICU (6 beds) 24 15 60,000 9 Najjar hospital Laboratory 24 20 80,000 Sterilization unit (to operate one OT sterilizer) 12 40 160,000 Emirati RC NCU 24 10 40,000 10 maternity Laboratory 24 20 80,000 hospital Sterilization unit (to operate one OT sterilizer) 12 40 160,000 Grand Total in USD 1,010 4,140,000 Source: World Health Organization (WHO) 31 Annex 7: Developing Transmission Infrastructure Figure A7.1: Map of existing transmission infrastructure in Palestine Power Plant Existing IEC substation Existing 22kV distribution lines Existing 33kV distribution lines Existing IEC 161kV transmission lines Northern WB is fed from: - 2x33kV feeders from IEC Beisan s/s - 12x33kV feeders from IEC Arail s/s - 2x33kV feeders from IEC Afraym s/s - 2x33kV feeders from IEC Immanuel s/s - 5x22kV feeders from IEC Southern WB is fed from: - 10x33kV feeders from IEC Hebron s/s - 4x33kV feeders from portable s/s Gaza is fed from: - 10x33kV feeders from IEC - 3x33kV feeders from Egypt - 220kV feeder from GPP 32 Figure A7.2: Location and service area of new PETL high voltage substations Jenin Al Jalamah s/s Nablus Sarra s/s New PETL substation Jalamah s/s service area Ramallah Qalandia s/s Sarra s/s service area Qalandia s/s service area Beit Ula s/s service area Tarqumiya Beit Ula s/s 33 Figure A7.3: Land in the West Bank is divided into Areas A and B, under Palestinian civil administration, and Area C under Israeli civil administration 34 Annex 8: Robust Planning Methodology and Detailed Technical Results Introduction This Technical Annex describes the methodological considerations, input assumptions and additional results to support messages in the section on power system expansion planning included in the main report of the Securing Energy for Development study. The analysis applies concepts from the Robust Decision Making (RDM) framework to develop a generation capacity expansion plan for both the West Bank and Gaza that takes into consideration endogenous and exogenous uncertainties. It is built around a linear programming (LP) optimization and simulation model that generates scenarios to capture the pervasive uncertainties in a region where geopolitical conditions heavily influence the availability of electricity and fuel supply. The analysis compares the output of classical methods to power system expansion planning with the RDM-related approach to show how uncertainties can affect the choice of technology options. The analysis also presents results that take into account constraints with access to project financing. The geopolitical conditions, and related uncertainties, call for a significant shift from traditional planning methods. To the extent that political, social and economic uncertainties can be abstracted for modelling purposes, they have been incorporated in the analysis largely through varying assumptions related to the availability, timing and cost of infrastructure development. Objective Given the uncertainties described, the objective of the generation expansion planning component of the study is to propose a set of generation options that perform well under various conditions. They are designed to be able to satisfy peak load and energy demand up to 2030 reliably, and at the most efficient cost. The analysis thus seeks to answer the following questions: i) In a deterministic analysis, what does a least cost capacity expansion plan look like, which ensures Palestine is self-reliant and able to meet demand securely (assuming there are no capital constraints)? ii) What are the features of a capacity plan that ensures Palestine can respond to a wide range of uncertainties including contingencies around electricity imports? What are the cost implications of such a plan? How well does this second plan perform in terms of costs compared with a classic least cost plan? iii) How does the average cost of production change by sharing reserve margin requirements with neighboring countries? iv) Given capital constraints, what is a balanced mix that combines (ii) and (iii) to keep average costs of production at a specified annual level? 35 v) How does access to regulated land (known as Area C) affect the generation mix and system costs? vi) If political uncertainties are not resolved over the planning horizon, what is the impact of only implementing projects that are solely within the control of the Palestinian Authority (PA), and what is the impact of delayed action or inaction on unmet demand? Limitations of the Study While efforts have been made to include some uncertainties in the expansion plan, it is not comprehensive in this regard. Importantly, climate risks are not included in the analysis. For example, rising air temperatures reduce the efficiency of plants (including most types of solar PV plants) while increasing demand (for cooling) during the summer months, amongst others. The location of plants and financial structuring of potential projects are other important issues that can affect which specific projects materialize. These issues are beyond the scope of the study, which does not consider locational issues. Delays during construction are also not considered in the analysis. Construction delays can impact project costs, and large projects tend to be more exposed to this risk. As an example, doubling the construction time for a gas turbine from 24 to 48 months could increase project costs by close to 10%1. Delays also expose the project to fluctuations in material prices linked to international commodity prices. Including these issues tends to further strengthen the case for distributed generation options. Methodological Considerations Planning for the expansion of power sectors in developing countries is challenging in part due to the uncertainty associated with demand projections because historical trends are typically different from expected growth patterns. The power sector in Palestine falls in this category. Additionally, the geopolitical situation in Palestine (one of the Territories listed on the bank’s list of fragile situations)2 introduces additional layers of significant uncertainty. Constraints imposed by fragility have been routinely left out of power sector planning in most conflict-prone countries. It manifests through various impacts on technology choices, timing, cost of and access to financing, etc. The lack of financing, delays and damages are all constraints and risks that need to be considered in planning for these to be more effective. Although these issues are well understood in qualitative terms and practiced in the field, it is only recently that consideration is being given to quantitatively formalize this trade-off to produce a power system 1 Based on the following: overnight capex - $750/kW; weighted average cost of capital – 10%; plant life – 40 years; half of capex needed from start of construction. 2 http://siteresources.worldbank.org/EXTLICUS/Resources/511777-1269623894864/FY15FragileSituationList.pdf 36 plan that finds a good balance between cost and risks and characterizes the ever changing dynamics of fragile states. (Bazilian & Chattopadhyay, 2016). For Palestine in particular, uncertainties around demand projections, the level of electricity imports, timing of fuel availability, volumes and costs of fuel, granting of access to expand infrastructure, and the risk of high outage rates mean the classic approach to least cost expansion planning is not adequate. The classic approach to least cost planning typically assumes expected outage rates, fuel and plant availability and a load growth forecast to project a generation mix that satisfies peak load and energy demand under a predefined set of constraints. The robustness of the plan is tested through a carefully selected set of scenarios. Under the current circumstances in Palestine, the sheer number of uncertainties leaves such an approach too vulnerable to failure measured by the inability to meet demand. An approach to counter this risk could be to plan for the worst or close to worst case scenario but this comes at cost. While partly justifiable, this cost in the form of capital expenditure is likely to be unwarranted because the system will be overly designed. The approach adopted for this study therefore seeks to balance the goal of meeting demand at all times, with the risk of stranded assets under multiple scenarios. Bazilian and Chattopadhyay (2016) describe three possible planning techniques for fragile states, namely: a) Least-cost planning tools that include risk premiums as inputs; b) Extension of least- cost planning models with a simulation component reflect some of the uncertainties associated with fragile and conflict states; and c) Stochastic programming and robust decision models that are specifically designed to facilitate decision making under uncertainty. In a case study for the Republic of South Sudan, Bazilian and Chattopadhyay employed the first approach (a) above to look at the impact of differentiating the cost of capital for risky projects (typically large, scale-efficient infrastructure which are cheaper but highly exposed to the risk of destruction and significant delays) from smaller but less risky options. By using a higher weighted average cost of capital (WACC) for riskier projects as a proxy to capture a wide range of financing risks, the analysis results in a shift away from large, centralized technology options to more decentralized choices. For the case of Palestine, there is no evidence to conclude that financing for scale-efficient projects like thermal power plants will be more expensive than financing for more decentralized options or by how much. In another case study, Spyrou and Hobbs (2016) use a two-stage stochastic planning model to analyze the impact of climate risks on the power system expansion plan for Bangladesh - one of the most vulnerable countries to climate change (World Bank, 2013). The analysis concludes that modeling the relationship between climate and power system parameters could save up to 1.6 billion 2015 US$. In a two-stage stochastic programming model, an action is taken in the first stage (or the present) knowing that the future could evolve in many different ways based on a set of random parameters. A set of recourse decisions is then defined to determine a course of action in the second stage that responds to the outcome of the uncertain parameter. The goal is to find a solution that optimizes the expected outcome of a decision. Stochastic programming 37 models are reliant on the fact that probability distributions governing the data are known or can be estimated. (Shapiro & Philpott, 2007) (Shapiro, Dentcheva, & Ruszczynski, 2009) Establishing such probability distributions for Palestine can be challenging. There is little historical data to inform the design of the distribution parameters. The distribution shapes could be attempted through the use of expert judgement, but the political underpinnings in Palestine increase the subjectivity of such an exercise. The approach used to deal with uncertainties in this study involves using Monte Carlo simulations to draw scenarios from a range of parametric uncertainties and then running them through a deterministic linear programming (LP) model. The simulations serve to “recognize” the stochastic nature of the parameters, but this process falls short of a full stochastic model because the final selection of the capacity plan is decided by the modeller rather than the model. The Monte Carlo simulation process considers random variation in uncertain parameters such as demand and generation availability, fuel prices, etc to form a composite sample that represents one realization or “future” of all possible uncertain parameters. The LP dispatch optimization is solved for the sample to obtain one point estimate of system costs, prices, etc. The process is repeated for a large set of samples (e.g., somewhere between 100 and 1000 depending on the number of parameters and their variance) to form a distribution of the outcomes. The process is fundamentally not very different from running a large number of alternative scenarios with the exception that: a) we directly represent the probability distribution of each uncertainty parameter rather than accepting a pre-defined scenario with a specific view on the uncertain parameters; and b) we therefore can evaluate the impact of multiple uncertain parameters on the final output – be it prices or system costs. The process flow for the analysis is shown in Figure A - 1. We start by identifying all the input parameters which are uncertain. These include: the timing and availability of fuel, energy demand, fuel prices, amount of imports, availability of power plants, investment costs, amongst others (see Table A - 1). The ranges for these uncertain parameters were developed by the project team after consultations in Jordan, Israel, Egypt, and Palestine. 38 1. Scenarios: develop multiple future scenarios from predefined ranges of uncertain parameters using Monte Carlo sampling. 2. Multiple least cost plans: For each future scenario, develop a least cost plan. 3. Review options: Evaluate least cost plans to rank technologies and capacities according to frequency of selection across scenarios. 4. Stack and test options: Select technologies and capacities ranked highest – i.e., selected in 100% of scenarios. Test resultant plan across multiple scenarios and observe total system costs. Reduce preference ranking and select associated technologies and capacities until least selected options (i.e. only in 1% of scenarios) are available. 5. Robust plan: Select plan with the lowest average of total system costs across multiple scenarios. Figure A - 1: Process flow for selecting robust options In this type of analysis, it is important to identify the correlations between parameters. Of particular interest was the correlation between fuel and electricity import prices. Since the transition from oil based generation to gas based generation, electricity prices in Israel have been decoupled from international oil prices because most of the production is driven by long term gas purchase agreements. The growing share of renewable energy will further decouple the two prices. This is likely to be the case for Egypt as well. We therefore assume no correlation between fuel prices and imports from Israel and Egypt. Electricity prices in Jordan on the other hand are correlated with fuel prices and this assumption has been used in the study. At present, gas generation is based on LNG and so prices are linked to the international gas market. However there are considerations to import gas from Israel or other countries. Jordan is also considering oil shale and nuclear power as generation options together with a strong renewable energy 39 portfolio. These plans, if implemented will reduce the linkage between international oil and electricity prices. Table A - 1: Input parameters and associated uncertainties Parameter Uncertainty Fuel Prices, volumes and availability of diesel (as a result of attacks and politically imposed constraints). Timing, volumes, prices and availability of gas. Capex Variations in PV, wind and CSP capex Electricity Prices, volumes and availability imports Demand High volatility in projected demand Transmission Uncertainties around the commissioning of West Bank backbone and West Bank-Gaza interconnection Plant Extended outages as a result of damage or difficulty in reaching plant availability locations. Damage due to sabotage incurs a cost to the system. Details of the process flow are: Step 1: Develop multiple scenarios After identifying the uncertain parameters, multiple scenarios were generated using Monte Carlo sampling techniques. Each scenario contains a random draw from within the distribution of the individual parameters for every year in the planning horizon. For example, it will include a certain demand profile for every year, plant availability for every year, fuel prices for every year, etc. Since the draws from the various parameters are completely independent, two valid questions arise at this stage: a) how certain are we that the combination of individual draws adequately cover the worst case scenarios?; and b) how plausible is the combination of all the individual draws? The coverage of scenarios depends on the number of draws for the Monte Carlo analysis. A higher number of draws increases the coverage of scenarios. This needs to be balanced with the computational requirements. The number of scenarios was selected to minimize the variance in total system costs across scenarios; an indication that enough samples across the range of uncertainty have been selected. In this study, 100 draws was used to demonstrate the merits of the study approach. For question (b), the primary issue of concern was the link between outages caused by sabotage and demand. As an example, how plausible is a scenario with high outages and high demand 40 growth? Using the study approach, it is also possible to discard combinations of parameters that are unreasonable. To design the scenarios, the study categorizes four possible states in the West Bank and Gaza: war, siege, economic stagnation, or economic prosperity. These conditions are characterized by different levels of investments in power sector infrastructure and demand characteristics as shown in figure A-4. •Significant disruptions to supply •Disruptions to supply due to due to sabotage sabotage and •Limited imports •Limited imports •WB Transmission backbone may •Limited fuel (diesel) be commissioned •WB Transmission backbone may •Very limited fuel (diesel) be in place •High unmet demand •Low load growth A: War B: Siege •High unmet demand C: D: Stagnant Prosperous •Disruptions due to •Open imports •Gas available •Higher level of imports •WB Transmission backbone •Limited fuel (diesel or gas) commissioned •WB transmission backbone may •Increased uncercainty in load be commissioned growth •Increased uncertainty in load •Low unmet demand growth •Relatively higher unmet demand Figure A - 2: Characterizing four possible economic conditions of Palestine For extended periods in the planning horizon, or the entire duration, either territories of Palestine could be in a state of: siege (as is the case in Gaza presently); stagnation (as is the case in the West Bank); or prosperity where there are no limits to infrastructure development. A state of war is more transient marked by particular years with high outages, limited imports and limited fuel supply. For example, over the entire planning horizon, a territory could be in a state of stagnation with spot disturbances in which supply options are disrupted. History shows destroyed equipment are restored and this is assumed to continue. It was also decided to establish a minimum availability threshold for imports from Israel, the main source of imports since it was unrealistic to anticipate this to be completely unavailable. Step 2: Develop multiple least cost plans 41 For each of the scenarios above, we develop an expansion plan using the core LP model. It is important to note that the model considers the entire planning horizon and optimizes capacity and dispatch to minimize system costs for the horizon. It does not optimize the solution on a year to year basis. For example, if a scenario draws on natural gas being available but also draws low availability for the gas plants in some years, the model consider this availability and may install more capacity within defined constraints to satisfy the supply-demand balance. It will not only use the higher availability of preceding years to determine optimal capacity and timing. The core LP model is explained in a later section. Step 3: Review plans and rank options In this step, we analyze for every year in every scenario, the frequency with which technology options are picked and, when picked, the capacity that is installed in that year. The aim is to identify those technology-capacity options that are robust across multiple scenarios. These are then ranked by the percentage of times the technology-capacity mix is selected in every scenario and for every year. To determine the capacity that is picked up across multiple scenarios, we approximate the plant capacities to the nearest 10MW. This is similar to creating 10MW “bins”3. Step 4: Stack and test options The most preferred technology-capacity options are those that are selected in most scenarios. Figure A - 3 shows variation of the capacity of available generation options across multiple scenarios. Existing capacity is always included. As seen from the figure, 194MW of the plant PVr- WBC is robust in all scenarios, while the first concentrated solar power (CSP) of 20MW is only picked up in 11% of scenarios. Technology and capacity options that appear in 100% of scenarios constitute no-regret options. Moving to the right of the chart, the capacities of technologies increase but are less robust. 3 Plant capacities are modelled as continuous variables to reduce computational time so the capacity plans likely contain different plant capacities for each of the scenarios (even if by a small number). In reality, plant capacities are discrete variables rather than continuous. For example, generation plants are typically commissioned in blocks equal to the size of the units (eg, 48MW could be configured as 12MW x 4 units). In the model, we assume this to be continuous allowing capacity increases that do not necessarily match unit sizes. CCGT capacity of 1.1MW could be added, for example, which is not realistic. However, the objective of this exercise is to develop a sense of the generation mix going forward. The error introduced by this approximation is therefore not important. 42 Cumulative capacity by technology [2020] Israel-WBC 2000 Capacity gets less robust across future scenarios as we move DiesGen-Gaza 1800 towards the right CC-Gaza 1600 CC-WBN Cumulative capacity (MW) 1400 No regret capacity/options DiesGen-WBC 1200 DiesGen-WBN 1000 PVcC-WBC 800 PVr-WBC 600 CSP-WBC 400 200 0 45% 100% 95% 90% 85% 80% 75% 70% 65% 60% 55% 50% 40% 35% 30% 25% 20% 15% 10% 5% Percentage of occurrence across future scenarios Figure A - 3: Cumulative capacity by technology While it would be ideal to only include the most robust technology-capacity options, this is unlikely to meet demand, and so it is necessary to add more capacity. We therefore stack less preferred options to increase installed capacity (ie, select capacities to right in Figure A - 3). At this point, we have little idea about the target capacity that will be adequate for the system because the demand forecast is uncertain. To deal with this problem, we run multiple experiments (multiple scenarios that draw from the uncertain parameters) whenever we include additional options in the stack and observe the performance of the capacity plan. It is important to distinguish this step from step 2 where we develop multiple capacity plans to satisfy multiple future scenarios. In step 2, the question we answer is this: if the future looked like a particular scenario, what types of plants should be built? When should they be built and how should they be dispatched? In this step, whenever we add onto the stack of technology-capacity options, we take this as our capacity plan and ask the question: if this was my capacity plan, how would it perform in multiple scenarios? The stack developed in step 4 can also be used to select projects to prioritize. The more robust options at the bottom of the stack should be given a higher priority. Step 5: Select robust plan The performance of each potential capacity plan is evaluated mainly through the change in total system costs or the objective function (least cost). By observing the total system costs across multiple scenarios, we select the plan with the lowest cost as the “robust” plan. 43 Jiang and Vogt-Schilb (2016) employ a similar approach in a case study for Bangladesh as the quantitative basis for a “robust adaptive strategy that performs acceptably over several dimensions in as many plausible futures as possible”. The main differences are: a) Latin hypercube sampling was used to reduce the need for large number of simulations, while we use Monte Carlo sampling for this study; and b) after generating multiple expansion plans, the technology-capacity options were placed in large bins to reduce the number of plans, each of which was then tested across multiple scenarios and performance assessed independently; in this study, we develop the capacity plan starting with no-regret options and increasingly adding less preferred options as described in Step 4 above. Structure of the model West Bank is modeled as three separate zones – WB_North, WB_Central and WB_South (see Figure A - 4). Electricity imports from Israel is injected through three points in the North, Central and South. Demand and solar resource availability is accordingly distributed amongst the three zones in West Bank. Gaza is modeled as one separate zone.4 IEC_WBNorth IEC injection point in north WB_North IEC_WBCentral IEC injection WB_Central point in center Jordan_WB IEC_Gaza IEC Egypt_WB injection point Imports from in Gaza Jordan and IEC_WBSouth Egypt IEC injection point in south WB_South Egypt_Gaza Injection point for Egypt imports 4 Within the scope of the project, all imports are modeled as generators connected to the relevant zones. This mathematically yields similar results as the primary focus is on the impact of energy imports into Palestine and not necessarily energy exchange between. 44 Figure A - 4: Categorization of zones and power import connections in West Bank and Gaza There is currently no transmission network in West Bank or Gaza, and the two territories are not directly interconnected either. The analysis of transmission requirements is undertaken separately and not included in the model. The model used for for Palestine is a GAMS5-based least-cost planning tool that is in many ways, simpler to populate (through an Excel front end) and easier to customize or build algorithms/procedures around the basic model to deal with uncertainties. Given the sparse data available and the wide range of uncertainties necessary for the analysis, this flexibility is critical. Mathematical description of the LP model This section briefly describes the LP least cost planning model at the core of the analysis. The deterministic least cost planning model takes into consideration the following as input:  Cost: investment costs for generation expansion, fuel prices, fixed and variable operation and maintenance (O&M) costs;  Load: Load forecasts in the form of load duration curves (LDC) for the planning years;  Generation: Operational characteristics of generation plants such as thermal efficiency, maximum utlization factors, amongst others;  Transmission: Transfer capacity limits between zones and associated losses; The main output of the model is a set of generation options and their associated timing, dispatch levels and residual or unmet demand. From this, the average cost of generation per year (or block of the load duration curve), associated emissions, total system costs, capex requirements, reserve margins amongst others can be calculated. The objective function to be minimized is the discounted net cost (at a rate of 10%) for the planning period and is calculated as shown in EQ 1. EQ Set 1: Objective function 5 General Algebraic Modeling System (GAMS) is a fully documented model and has been used for other Bank assignments in Ukraine, Bangladesh, Bulgaria, and South Africa. 45 Where the sets and decision variables are defined as follows: Sets Decision Variables 46 Sum of Cap*CAPEXperkW determines the total annualized investment for all thermal generators in a particular year. The cost recovery factor (CRF) is calculated using a weighted average cost of capital (WACC) of 10%. For renewable energy (RE) plants, a separate term, BuiltRE, adds the annualized capex requirements to the objective function. This has been separated as a new variable because the capex for RE plants changes with time. Any new installation therefore applies the capex requirements for the year of installation calculated in a separate equation. Sum of Cap*FOMperMW adds the fixed operating and maintenance costs for all generators per installed kW every year. VC + VOM together make up the short run marginal cost for each generator, VC being the cost of fuel and VOM being variable operating and maintenance costs. Sum of CumRepair*CostOfRepair add a cost whenever there is damage to a plant. The CostOfRepair is assumed to be a third of the Capex and CumRepair carries the total annualized repair costs through the planning horizon. Repair costs are assumed to be recovered over 12 years (irrespective of the plant). 6 Sum of GenEmis*CO2Price adds the cost of CO2 emissions if required and is set by the flag IncludeCO2Price. CO2 prices are not included in the analysis for Palestine. 6 The cost per kW to repair a plant depends on the extent of damage. To simplify the problem, we assume a single cost amortized over 12 years. This has been estimated from past Bank projects that refurbished/rehabilitated thermal plants mostly to improve efficiency. 47 In addition to regular costs, the objective function includes penalties associated with violation of demand constraint, VoLL - which is set at USD 750 per MWh for this study, and reserve limit, VoLLReserve which is set at USD 5,000 per KW in this study. Mathematically, the unserved energy variable is relaxes the demand balance constraint to avoid infeasibility in time periods with excess demand. In practice, it is an indirect measure of system reliability. Its valuation is an economic concept that indicates the willingness to pay of electricity consumers to avoid supply interruption (Electricity Commission, 2008). The study reports VoLL to be as high as USD 44,500 per MWh in Australia and USD 960 per MWh for Chile. Mathematically, because of the role this plays in balancing demand and supply, the value for lost load needs to be high enough to prevent the model from curtailing load as a means of minimizing system costs. The selected VoLL is set at the cost of self-generation through portable household gasoline generators. The objective function is minimized subject to various constraints described here: EQ Set 2: Capacity Balance and Firstly, the dynamic linkages across the years is captured in the first equality constraint that defines the variable Cap. Capacity may be augmented by building new units (i.e., Build), or it can be mothballed (i.e., Retire). Secondly, the first year capacity is restricted to the existing capacity and the total capacity that can be built for a new station over the entire planning horizon is restricted to the planned capacity addition. Additionally constraints ensure new plants are not mothballed and existing capacity is included. Finally, the capacity addition, annual capacity and power output are subject to a set of conditions as represented by the last two constraints. EQ. Set 3: Capacity Utilization Limits 48 Generation from all units (i.e., existing and new) is limited by the maximum capacity factors on the Cap. Availability is the maximum capacity factor and it is one of the random parameters sampled. Imports (which are modelled as generators running on an “import fuel”), are limited to import caps and the availability of the tie line. This is defined by the second equation. The third equation ensures that sum of imports from Israel to the three zones of West Bank do not exceed Israel-West Bank import cap. EQ. Set 4: Destroyed capacity Major outages due to damage incur a cost and the probability of damage is defined as another uncertain parameter. DestroyedCap is a fraction that represents the installed capacity destroyed. For distributed generation sources, damage has less of an impact and DestroydCap is small. Centralized units are more exposed to the risk of damage which takes the entire plant out of service. These repairs incur a cost to the system which is amortized over 12 years. The first equation calculates the destroyed capacity in a year. The second and third calculated total repairs for all years and the first year respectively. It is this variable that is multiplied by the per kW repair costs in the objective function. EQ. Set 5: Zonal balance 49 By Kirchoff’s First Law (also know as KCL Kirchoff’s current law), the total line flows into and out of a node must equal the difference between the generation flowing into the node and the off- takes. Thus, the nodal balance constraints equate demand, generation, losses, electricity flows to/from the node. Generation deficit violation variables (USE1) are also included in order to deal with those rare situations in which the system may be unable to meet the load at a node, due to a general shortage of generation, or to transmission system failure. Lines have a conventional direction associated with them. A positive Tran variable represents power flowing into the node for some lines, and power flowing out for others. A fraction of the loss (LS) is attributed to the load end of the line. Demand is one of the key random parameters in the model. Given a distribution of peak and energy, the random sampling process draws a demand profile for each of the load block and the dispatch optimization is repeated for each such demand sample (along with other random parameters). The first equation is used in energy efficiency scenarios and the second is used when energy efficiency is not part of the scenario. The third equation limits transfers to the capacity of the transmission corridor (pTransferLimit) which is also randomly sampled. EQ. Set 6: Reserve requirements 50 Reserve is modelled for each territory (b). Although provision of regulation is fundamentally different from provision of contingency reserve and the former is also governed by additional set of constraints in real-time, the nature of constraints that are relevant in a long-term planning framework is largely similar across regulation and reserve. Contingency reserve response is expected to occur automatically when frequency falls, and generators with spinning reserve respond under ‘free governor action”. In the longer timeframe, generators that are not currently synchronised may synchronise and commence generation. ReserveContribution is a factor that determines available capacity that contributes to the reserve requirements. EQ. Set 7: Fuel consumption and constraints The first equation defines the fuel consumption as a function of the generation from that type of fuel across all generating units over all LDC blocks in that year. We have assumed a constant heat rate over the entire generation range of a generator, but this can be changed to represent the detailed heat rate characteristic of the unit using a piecewise linear function. The second constraint is a simple bound on the maximum amount of fuel that is available in a year. It is also possible to represent any take-or-pay fuel contracts for individual generating stations/companies. For this stage of the planning exercise, we do not impose take-or-pay constraints even though this is likely to be the case for gas supply. Our objective at this stage is to establish what volume of gas for the power sector is least cost. This, together with other domestic uses of gas in Palestine will inform the structure of any take-or-pay contract, the details of which will need to be thoroughly analyzed. The third constraint ensures that total generation in every load block does not exceed rated capacity. EQ. Set 8: Variable renewable energy profiles 51 These equations define the limits of generation from variable RE sources. REProfile is the maximum utilization of the RE plant in every time block. All three types of VRE (PV, CSP and wind) have unique profiles. EQ. Set 9: Combined solar and PV capacity cannot exceed total land available Input Parameters In this section, we describe the main inputs required for the model including the RE profiles, load blocks, and generator data. Generator data Generators are defined by the parameters defined in Table A - 2. The Gaza Power Plant (GPP) is the only operating power plant. The installation costs for variable RE (VRE) technologies are modeled to reduce cover the planning horizon. The rate at which VRE prices reduce is sampled and discussed in a later section. Table A - 2: Generator parameters Fuel Installation Fixed Variable Contribution Base Max Heat cost (2018) O&M O&M to reserve* unit repair rate size costs $/kW $/kW- $/MWh % MW $/kW MMBTU yr /MWh 1Rooftop PV (PVr) Solar 2,591 15 0 0 0 864 0 1Utility PV (PVc) Solar 1,646 13 0 0 0.5 549 0 52 1Concentrated Solar Solar 5,552 59 10 0.75 10 1,851 0 Power/Thermal (CSP) 1Wind (Wind) Wind 1,863 51 0 0 1 621 0 4Biogas (Bio) Landfill/ 3,942 107 5 1 2 1,314 214.5 Manure 2Distributed Diesel Diesel 800 15 15 1 2 263 10 Genset (DiesGen) 2Combined Cycle Gas Gas/ 1,300 6.2 3.5 1 140 433 6.7 Turbine (CC) Diesel 2Simple Cycle Gas Gas/ 1,000 25 7.5 1 100 333 9 Turbine (GT) Diesel 3Imports from Jordan 5 0 Scenario 0 0 3Imports from Israel 5 0 Scenario 0 0 3Imports from Egypt 5 0 Scenario 0 0 * A factor that determines available capacity that contributes to the reserve requirements. PV and wind for example are not firm and do not contribute to the reserve margin in the analysis. SOURCES: 1 – Team estimates based on NREL Annual Technology Baseline; 2 – High end of Lazard’s levelized cost of energy analysis (Version 9.0). 3 – Team estimates. 4 - IEA World Energy Outlook Renewable energy sources could be a significant part of the energy mix in Palestine with total potential of approximately between 3,100 and 4,000 MW (depending on the share of CSP and PV. See Table A - 3). There is considerable technical potential for solar PV and CSP (at least 98% of RE potential), but the bulk (at least 76% of potential solar generation) is in Area C of the West Bank and can only be realized if Israel grants access to the land. The technical potential for gas/diesel fired plants depends on the volume of fuel. 53 Table A - 3: Potential generator capacities Gaza Potential Capacity (MW) Rooftop PV (PVr) 163 Biogas (Bio) 2 Distributed Diesel Genset (DiesGen) Unconstrained Combined Cycle Gas Turbine (CC) Unconstrained Simple Cycle Gas Turbine (GT) Unconstrained West Bank North Potential Capacity (MW) Rooftop PV (PVr) 210 Commercial PV (PVcAB) - Area A, B 14 Wind (WindC) - Area C 9 Biogas (Bio) 10 Distributed Diesel Genset (DiesGen) Unconstrained Combined Cycle Gas Turbine (CC) Unconstrained Simple Cycle Gas Turbine (GT) Unconstrained West Bank Central Potential Capacity (MW) Rooftop PV (PVr) 194 Commercial PV (PVcAB) - Area A, B 7 Commercial PV (PVcC) - Area C 3,200 Concentrated Solar Power/Thermal (CSP) - Area C 2,424 Biogas (Bio) 8 Distributed Diesel Genset (Diesel) Unconstrained West Bank South Potential Capacity (MW) Rooftop PV (PVr) 131 Commercial PV (PVcAB) - Area A, B 14 Wind (WindC) - Area C 36 Biogas (Bio) 7 Distributed Diesel Genset (DiesGen) Unconstrained Combined Cycle Gas Turbine (CC) Unconstrained Simple Cycle Gas Turbine (GT) Unconstrained SOURCE: Team estimates While there continues to be several discussions with potential investors and the PA around new sources of generation such as the Jennin power plant, there are no committed generation projects so we do not include specific candidate projects in the plan. We instead use generic generators to determine the capacities of various technologies that are robust. Electricity imports are modelled as generators with no capex requirements. The fuel Fixed O&M costs include the cost of providing ancillary services to Palestine which is priced at 12 $/MW-h. The capacity is therefore optimized to minimize system costs within export limit constraints and gives the minimum transfer capacity when sizing the interconnection. Import sources considered are shown in Table A - 4. 54 Table A - 4: Current capacity and pricing of electricity imports From To Current capacity (MW) 2016 price ($/MWh) Israel West Bank 800 90.0 Israel Gaza 120 90.0 Jordan West Bank 30 95.9 Egypt Gaza 10 50.0 Egypt West Bank (through 0 N/A1 Jordan) 1 There is currently no power import from Egypt to the West Bank but this could include the cost of generation of 81 $/MWh to Egypt and 6.5 $/MWh wheeling charges to Jordan based on current transmission wheeling charges for renewables in Jordan at 4.5 fils/kWh SOURCE: Team estimates Under 2018 cost conditions, solar PV has the lowest levelized cost of energy (LCOE) of all technologies available as shown in Figure A - 5. At low utilization rates, diesel has the lowest costs making them good candidates for providing back-up services. This also means frequent outages which affect the availability of plants increases average system costs. 2018 2025 Distributed Diesel Distributed Diesel 80.0 CCGT - Diesel 80.0 CCGT - Gas GT - Diesel GT - Gas 70.0 70.0 Utility PV Utility PV CSP CSP 60.0 60.0 IEC IEC Jordan-West Bank Jordan-West Bank 50.0 50.0 Egypt-Gaza Egypt-Gaza 40.0 40.0 30.0 30.0 20.0 20.0 10.0 10.0 - - 2% 6% 10% 14% 18% 22% 26% 30% 34% 38% 2% 6% 10% 14% 18% 22% 26% 30% 34% 38% Utilization Utilization NOTES: Capex and O&M as shown in Table A - 3; Diesel= 21 NOTES: Capex and O&M as shown in Table A - 3; Diesel= 34.2 $/MMBTU; WACC=10% ; 20 year life for PV and 30 for all $/MMBTU; Gas=5.5$/MMBTU; WACC=10% ; 20 year life for PV others and 30 for all others Figure A - 5: Comparison of LCOE in US cents/kWh for technology options 55 As the utilization rate increases, the high cost of diesel makes these far less attractive. Scale efficient units like CCGTs become more cost effective. We also see CSP outperforming CCGT running on diesel due to the high cost of diesel. These comparisons do not take into account other benefits of the various technologies such as: the provision of ancillary services and system support for thermal units, and avoided generation emissions for renewable energy technologies. A different picture emerges in 2025 when the capex for solar PV in particular is expected to be lower and gas is more likely to be available. In this scenario, the LCOE for CCGT at 70% utilization is on par with the LCOE for utility scale PV at 5.4 US cents/kWh at a gas price of 5.5 US$/MMBTU. CSP costs are also lower with a high end LCOE close to 15 US cents/kWh. Demand Data The demand forecast developed shows a wide range of uncertainty in outer years rising to nearly 40% of the low forecast scenario in 2030. Peak demand forecasts was calculated with an assumed load factor of 60% based on historical data from JDECO. The resultant peak load is between 1,800MW and 2,500MW in Palestine (Figure A - 6). a) Demand forecast (GWh) 8,000 Gaza West Bank 8,000 6,000 6,000 4,000 4,000 2,000 2,000 - - 2024 2016 2017 2018 2019 2020 2021 2022 2023 2025 2026 2027 2028 2029 2030 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 Low Case Central Case High Case Low Case Central Case High Case b) Peak load forecast (MW) Gaza West Bank 1,600 2,000 1,500 1,100 1,000 600 500 100 - 2018 2023 2028 2016 2017 2019 2020 2021 2022 2024 2025 2026 2027 2029 2030 2024 2016 2017 2018 2019 2020 2021 2022 2023 2025 2026 2027 2028 2029 2030 (400) Low Case Central Case High Case Low Case Central Case High Case 56 Figure A - 6: Forecast demand and peak load in West Bank and Gaza The peak and energy forecasts were used to develop load blocks for each year in the horizon. (Load blocks are used to reduce the size of the LP and computing resource requirements.) In the absence of system data, load blocks were developed using simulated hourly load data. Ideally, a year of system hourly load data is the least requirement to first generate the load duration curve and then the load blocks. This becomes even more critical if variable renewable resources are included in the model because the coincidence between hourly load data and renewable resource availability is important. For this analysis, we obtained daily load curves for the winter and summer seasons for the Jerusalem area from the distribution company (JDECO). This gives a sense of the daily characteristics of demand. The same daily curve was assumed for West Bank and Gaza. The same profile was assumed for weekends and weekdays but was shifted downwards to simulate lower demand on weekend days. We also obtained monthly energy consumption for West Bank which gives a sense of the seasonality and month-to-month variations in demand. Combining this data, we generated a rudimentary load duration curve which characterizes monthly load and seasonal day load variations. The load block definition was maintained for all forecasted years and every demand growth path (low, medium, high and robust). The demand for West Bank was distributed across the three zones using historical sales data from the distribution companies. 54% of the load was allocated to central West Bank, 26% to northern West Bank and 20% to southern West Bank as shown in Table A - 5. Table A - 5: Distribution of demand in West Bank Zone Share of West Bank demand West Bank North 26% West Bank Central 54% West Bank South 20% Renewable Energy Data Variable renewable energy technologies considered were wind and solar for PV and CSP applications. Unlike a broad resource assessment for a region, hourly energy output data is required to ensure that output is correctly matched to the various load blocks. The System Advisor Model (SAM) from the National Renewable Energy Laboratory (NREL) was used to calculate hourly energy output from resource data for all three technologies. 7 Solar (PV and CSP): To consider year to year variations in solar data, we use typical meteorological year (TMY) data provided by various bodies and entities. TMY data includes monthly data that represent typical conditions and is selected from a multi-year data set8. The 7 The System Advisor Model (SAM) is a performance and financial model for RE planning from NREL (https://sam.nrel.gov/) 8 Weather Data Overview available here (https://www.nrel.gov/analysis/sam/help/html- php/index.html?weather_format.htm) 57 study utilized TMY data from three locations in Israel which are close to West Bank namely: Tel Aviv for northern West Bank, Atarot for central West Bank and Bersheva for southern West Bank. TMY data from Al Arish in Egypt was used for Gaza.9 Wind: Hourly wind speed data is not as readily available as irradiation data for most locations. To obtain hourly data for the study, we scaled wind speeds from weather station data (which is typically measured at approximately 10m) to 80m. 10 Understanding the complementarity between wind and solar resources will require several years of data but the daily profiles used show that wind and solar output coincide with each other as shown in Figure A - 7. CSP could be used to better complement the resources. 1 0.8 0.6 0.4 0.2 0 0 4 8 12 16 20 0 4 8 12 16 20 0 4 8 12 16 20 0 4 8 12 16 20 0 4 8 12 16 20 0 4 8 12 16 20 Jan Mar May Jul Sep Nov PV CSP Wind Demand Figure A - 7: 24 hour profiles from selected months for West Bank Other input assumptions Cost of capital: The weighted average cost of capital (WACC) is assumed at 10%. Discount rate: The discount rate used to determine the system net present value is assumed at 10%. 9 Solar resource data obtained from EnergyPlus (https://energyplus.net/weather). EnergyPlus is a tool funded by the U.S. Department of Energy’s (DOE) Building Technologies Office (BTO), and managed by NREL. 10 Wind speed data from weather stations was obtained from the Iowa Environmental Mesonet program of the Iowa State University of Science and Technology. It collects environmental data from cooperating members with observing networks (http://mesonet.agron.iastate.edu/ASOS/) 58 Characterizing Uncertainties In this section, we discuss the representation of uncertainties in the model. While most of the uncertainties stem from the political conditions, others such as uncertainties around demand forecasts and fuel prices are common to most power systems. Demand Demand is sampled between the low forecast and high forecast. The uncertainty is observed in the rate at which demand will grow. We first select a demand point in the first year and then for each subsequent year, we select a demand point between the demand of the preceding year and the high load forecast trajectory. This ensures demand increases steadily (albeit at an unknown rate) as would be expected in Palestine. Fuel Volumes and Pricing Natural Gas: Gas from various fields is considered as potential fuel for both Gaza and West Bank with variable timing, volumes and pricing (for example from different gas fields in Israel available in the north of West Bank, from Gaza or Egypt in the south of West Bank, etc). The model is passive to the source of gas and the matrix is simplified as three sources of gas – one source for Gaza (GazaGas), one for south of West Bank (WB_SouthGas - for plants in the south such as Hebron) and a third source for gas in the north (WB_NorthGas - for plants such as the Jenin power plant). Each source of gas has a range of dates gas could be expected, an associated range of possible volumes and also a range of prices. The sampled scenarios draw from these ranges to determine the year gas is available for power production, the volume, and its price. Table A - 6: Uncertain parameters around gas supply for power generation: timing, volume and pricing Available for power Annual Vol BCM Price ($/MMBTU) Earliest Latest Min Max Min Max GazaGas 2022 2035 0.2 2.0 4.00 7.50 WB_NorthGas 2021 2035 0.2 2.0 4.00 6.50 WB_SouthGas 2024 2035 0.2 2.0 4.00 7.50 SOURCE: Team estimates The sampled volume is the maximum annual volume of gas used for the planning horizon. For example, if the sampled parameters for Gaza gas are 2025 available year, 1.3 BCM volume and price of 5.2 $/MMBTU, there is no gas available in the model until 2025 and 1.3 BCM available from 2025 at a price of 5.2 $/MMBTU. Diesel: The volume of diesel is unconstrained in the model unless an incident reduces supply. Shortages in the past have largely been due to the inability to pay for fuel. Future diesel prices follow the trajectory for international oil price forecasts. Between 2000 and 2015, oil prices in 59 real 2010 US dollars ranged between $32 and $98 per barrel or 66% and 200% of 2015 prices 11. There is a strong correlation between the price of diesel and crude oil in most markets (EIA, 2015) so, we assume range of 66%-200% of the average cost of fuel for the GPP in 2015 as an uncertainty range for the price of diesel. Therefore price per liter of diesel for every year is sampled between $0.51 and $1.57 for every scenario (see Figure A - 8). 1.80 1.51 1.57 1.57 1.60 1.41 1.32 1.40 1.24 1.15 1.08 $/liter 1.20 1.01 0.94 1.00 0.77 0.82 0.88 0.72 0.80 0.55 0.60 0.40 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 Upper bound Base case Lower bound Figure A - 8: Diesel price range The volume of diesel is unconstrained. There is a risk to the availability of fuel caused either by damage to pipelines (in the case of gas supply) or restrictions to the movement of fuel tankers (for diesel) this is dealt with in a later section. Import Volumes and Pricing Two forms of uncertainties are simulated for power imports namely: a) when import limits could increase due to changes in Palestine’s network to accept higher imports or changes in the generation and network capacity of exporting countries to be able to export more power; and b) import prices following a change in import volumes. We first sample a year when anticipated changes in the networks allow for increased power imports. Capacity before this year is fixed at current levels and then allowed to increase to an upper limit from the sampled year. For some imports, the upper limit is also a range because it is unclear. After the capacity change, the price is sampled between the price of a preceding year and an upper limit. The cost of imports from Jordan is indexed to the cost of fuel but other import sources are independent of fuel prices because they are largely based on national gas reserves and contracted under long term purchase agreements. 11 World Bank Global Economic Monitor Commodities (http://databank.worldbank.org/data/reports.aspx?source=Global-Economic-Monitor-(GEM)-Commodities) 60 Imports from Israel: The increase in power imports from Israel to West Bank is contingent on the commissioning of four transmission substations in West Bank. Electricity imports could increase from 850 MW in 2017 to between 1,400 MW and 1,800 MW from 2020. The current import from Israel to Gaza is approximately 120 MW and could be increased to between 220 MW and 270 MW from 2022. The IEC sells electricity to JDECO at time-of-use tariffs, but to the rest of Palestine at a bulk tariff rate of approximately US$ 90 per MWh (close to the weighted average of the time-of-use tariffs). There is the possibility that the rest of Palestine will be transitioned to time-of-use tariffs. There is also the likelihood that the price of electricity sold to Palestine will increase with the change in import volumes. The team estimates this to be up to 110 $/MWh. Pricing before the change in import limit is increased by 1% per annum. The 1% annual increase then continues from the new price. Consider an example in which 2020 is the year sampled for an increase in Israeli imports into West Bank. By 2020, the cost of imports would be approximately 92.7 $/MWh due to the 1% change in prices. The price beyond 2020 is sampled between 92.7 $/MWh and a defined upper limit. If the new price is sampled as 95 $/MWh for example, it is applied from 2021 and increased at 1% per annum from 2022. Imports from Jordan: Increase in power imports from Jordan to the West Bank is contingent on upgrading the current interconnection and the availability of excess capacity/energy in Jordan. Both parameters are uncertain together with the price at which power will be sold eventually. It is estimated that the earliest the interconnection could be upgraded is 2022 increasing import capacity up to 1,000 MW. The cost of imports from Jordan is indexed to fuel prices. The relationship is simplified using the polynomial function that best approximates the correlation between forecast fuel prices and forecast tariffs from 2016-2025 as shown in Figure A - 9 .. 61 20.00 y = -8.2309x2 + 27.597x - 3.3149 18.00 16.00 cents/kWh 14.00 12.00 10.00 8.00 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 Diesel prices ($/liter) Figure A - 9: Relationship between cost of imports from Jordan and diesel prices Imports from Egypt: Increase in power imports from Egypt to Gaza is contingent on upgrading the current distribution link through the Sinai region, increasing the capacity of the grid in Gaza, the availability of excess generation in Egypt, and power transfer capability of the Egyptian network to wheel power to the point of the interconnection line. The earliest this could be expected is estimated to be in 2021 at a capacity between 70-150 MW12. Because this is uncertain, the volume of imports is also sampled within this range. Jordan is interconnected with the Egyptian network at 500 kV and it is possible for Jordan to wheel power from Egypt through to the West Bank. This is contingent on the completion of the Green Corridor project13 by 2018/2019, and the availability of excess capacity/energy in Egypt. If this is incremental to current exports from Jordan, the Jordan-West Bank interconnection needs to have been commissioned as well. An additional 50-200 MW could be wheeled to Egypt through Jordan. Table A - 7: Changes in electricity import sources and capacities From To Change in capacity Capacity (MW) Price After Change ($/MWh) Earliest Latest Before Max After Min Max Israel West Bank 2020 2030 850 1400-1800 Preceding year 110 Israel Gaza 2022 2035 120 270 Preceding year 110 Jordan West Bank 2022 2035 30 100-200 Based on oil price Egypt Gaza 2021 2035 10 70-150 81 100 Egypt West Bank Same as Jordan-West Bank 0 50-200 87.51 106.51 1Same rate sold to Gaza plus 6.5 $/MWh wheeling charge to Jordan SOURCE: Team estimates 12 Capacity range is based on NEPCo Annual Reports 13 The Greed Corridor Project is a major grid upgrade to the North-South transmission corridor in Jordan. 62 Cost and Land Access for VRE technologies Installation costs for VRE technologies are expected to decline but the speed of decline is unclear (see a comparison of costs by NREL for example14). Investment cost estimates for the region produced by the International Energy Agency (IEA) for the World Energy Outlook15 are considered to be on the high side, especially when compared with the results of tenders from various countries (for example, a recent bid in Zambia yielded 6 cents/kWh). While this is unlikely to be the case in Palestine, there is uncertainty around what could be expected in the future. We therefore include a range of installation costs for VRE technologies. The team ’s capex estimates are based on NREL’s 2016 Annual Technology Baseline and the range of capex variation is shown in Figure A - 10. We ensure that capex for rooftop and utility scale PV are increased or reduced in tandem, but assume that the capex for wind and CSP are independent from other technologies. 3.0 2.9 Rooftop PV (PVr) Commercial PV (PVc) 2.8 2.5 2.6 2.4 3.0 2.3 2.2 2.0 2.0 1.9 2.5 1.8 1.7 1.7 1.6 1.6 2.0 1.5 1.5 1.5 1.8 1.6 1.5 1.4 1.0 1.3 1.2 1.1 1.1 1.0 1.0 0.5 0.5 0.0 0.0 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2026 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2027 2028 2029 2030 14 Annual technology costs (http://www.nrel.gov/docs/fy16osti/66944.pdf) 15 IEA’s World Energy Outlook Model (http://www.worldenergyoutlook.org/weomodel/) 63 7.0 Concentrated Solar (CSP) Wind 6.3 3.0 6.0 5.9 5.6 2.5 5.0 5.2 4.8 4.6 4.5 4.3 2.0 4.0 4.1 4.0 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 1.9 3.8 3.7 3.5 3.4 3.2 1.5 3.0 1.0 2.0 1.0 0.5 0.0 0.0 2018 2025 2016 2017 2019 2020 2021 2022 2023 2024 2026 2027 2028 2029 2030 2030 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 NOTES: Base case scenario capex estimates based on NREL’s 2016 Annual Technology Baseline. Lower bounds and upper bound for wind are estimated by NREL in the 2016 Annual Technology Baseline. For wind, capex is expected to increase due to the need for higher masts and bigger turbines to maximize low wind speeds. Upper bound for PV is from the IEA World Energy Outlook. Upper bound for CSP is calculated by team using IEA trajectory for CSP without storage. Figure A - 10: Capex ($/W) range for VRE technologies CSP and PV have different land requirements, capital costs and generation profiles. CSP requires 30% more land per MW installed, and is more than 3 times more expensive than PV but is dispatchable, while PV is not. The share of CSP and PV is optimized by the model subject to land constraints. The land constraint is defined by EQ. Set 9 which ensures that the total installed capacity of CSP and PV does not exceed available land. As noted, at least 76% of the solar potential is in Area C, and therefore projects are subject to Israel granting access to the site. We assume the earliest Palestine could access Area C is 2018 and the latest date of 2035. Allowing for a two year construction period, the earliest solar projects are allowed from 2020. The study will demonstrate the benefit of utilizing the solar resource in Area C so the model is set up such that from the year access is granted, all potential land is available for use. We do not take the fact that access is granted on a project by project basis into consideration. Fuel Interruptions and Plant Outages There are several risks that could ultimately affect the volume of fuel available for power generation including the risk of vandalism to gas pipelines, reduced fuel volumes due to lack of payment and restrictions to the transportation of diesel. We define a probability of outage on the sources of fuel supply and a range of availability for each year as shown in Table A - 8Table A - 1 The minimum availability defines the longest period of outage. 64 Table A - 8: Annual probability of outage and minimum availability for fuel sources. Probability of Minimum Interruption Availability Gaza Gas 0.12 0.3 WB_NorthGas 0.05 0.6 WB_SouthGas 0.10 0.6 GazaDiesel 0.40 0.3 WestBank Diesel 0.30 0.4 Israel Imports 0.02 0.8 Jordan Imports 0.05 0.7 Egypt Imports 0.20 0.6 SOURCE: Team assumptions The annual probability of interruption and availability of fuel are selected to show the relative risks associated with different sources following discussions with various stakeholders. Details of the risks associated with the various fuel sources are included in a separate section of the main report. The probability of outage on imported electricity together with the availability also captures possible challenges in exporting countries. A similar set of risks apply to power plants and the interconnection lines. Centralized plants are more vulnerable to destruction than decentralized options. In the model, we specify a share of installed capacity that is taken out of service when there are damages. This percentage is assumed to be 50%, 75% or 100% of capacity for centralized units and less than 5% for decentralized units. In effect, this penalizes larger units because the availability can be significantly reduced due to damages which affect the cost of per unit of production. Damages incur costs to the system determined by RepairCosts. Without knowing apriori the extent of damage, it is difficult to assess the costs or length of outage. For examply, replacing the step-up transformers or fuel tanks to a gas plant both result in a complete shut-down of the plant but the cost implications and duration of outages are completely different. In the model, we assume the cost of repairs to be a third of the cost of installation and randomly select a duration of outage between a full year to as little as one month16. The availability of plants is calculated as: (1 − ℎ )×(1 − )) × Plants in Gaza are at a higher risk than those in the West Bank as shown in Table A - 9. 16 The cost of repairs is based on the rehabilitation of thermal plants carried out by the World Bank. 65 Table A - 9: Annual risks associated with generators Gaza Probability % of Minimum Maximum of damage Capacity Duration of Duration of Subject to Outage (% of Outage Damage year) (% of year) Rooftop PV (PVr) 0.05 1 8 50 Biogas (Bio) 0.05 100 10 80 Distributed Diesel Genset (Diesel) 0.10 1 10 80 Combined Cycle Gas Turbine (CC) 0.15 50-100 10 100 Simple Cycle Gas Turbine (GT) 0.15 50-100 10 100 Israel-Gaza 0.02 100 5 25 Egypt-Gaza 0.20 100 8 50 West Bank Probability % of Minimum Maximum of damage Capacity Duration of Duration of Subject to Outage (% of Outage Damage year) (% of year) Rooftop PV (PVr) 0.05 1 8 50 Commercial PV (PVcAB) 0.05 1 10 70 Concentrated Solar Power/Thermal (CSP) 0.10 50-100 10 100 Wind (WindC) - Area C 0.05 5 10 80 Biogas (Bio) 0.05 100 10 80 Distributed Diesel Genset (Diesel) 0.10 1 10 80 Combined Cycle Gas Turbine (CC) 0.10 50-100 10 100 Simple Cycle Gas Turbine (GT) 0.10 50-100 10 100 Israel-WB 0.02 100 5 25 Jordan-WB 0.10 100 5 70 Egypt-WB 0.15 100 5 70 SOURCE: Own elaboration based on team assumptions Distribution of uncertain parameters Multiple experiments were performed to develop the capacity plan and, before discussing the results of the analysis, it is worth examining the uncertain parameters to understand their distribution. The energy demand for 2030 sampled across multiple scenarios is negatively skewed with a mean of 4,032 GWh and 6,856 GWh in Gaza and West Bank respectively, as seen in 66 Figure A - 11. This is higher than the 2030 median forecast of3,548 GWh and 6,004 GWh. We can also see that 30% of scenarios sample 2030 PV prices at or below $1/w, and the cost of CSP is relatively high in comparison. The price range for gas between $4/MMBTU and $7.5/MMBTU (which translates to $2.7-5 cents/kWh for the fuel cost assuming a CCGT) makes it a competitive option, so the critical parameter related to gas is the timing or availability. 83% of the experiments included the availability of gas in West Bank by 2030 compared with 66% of samples in Gaza. This is because there are two likely sources of gas in West Bank (either north or south) and gas in the north has an earlier likelihood of materializing. Finally, access to area C in the West Bank is critical for large scale deployment of solar technologies and construction of the transmission backbone; 68% of the samples allowed access to Area C by 2030 with 30% allowing access by 2022. 67 Figure A - 11: Distribution of uncertain parameters from the experiments a) Distribution of energy demand in 2030 in Gaza and West Bank (GWh) Gaza 25 100 West Bank 20 80 15 100 Cumulative Frequency 15 60 80 Cumulative Frequency 10 60 10 40 5 40 5 20 20 0 0 0 0 3650 3700 3750 3800 3850 3900 3950 4000 4050 4100 4150 6450 6550 6650 6750 6850 6950 7050 b) Distribution of capex for CSP and Utility-scale PV in 2030 ($/kW) CSP Utility PV 15 100 20 100 80 80 Cumulative 15 Frequency Cumulative Frequency 10 60 60 10 40 40 5 5 20 20 0 0 0 0 1,000 1,100 1,200 1,300 1,400 1,500 1,600 600 700 800 900 3,100 2,900 3,300 3,500 3,700 3,900 4,100 4,300 c) Distribution of the timing of gas for power production in Gaza and West Bank (year) Gaza 100 15 West Bank 100 15 80 80 Cumulative Frequency Cumulative Frequency 60 10 10 60 40 5 40 20 5 20 0 0 0 0 2022 2024 2026 2028 2030 2032 2034 2033 2021 2023 2025 2027 2029 2031 2035 e) Distribution of access to Area C in West Bank 12 100 10 80 Cumulative Frequency 8 60 6 40 4 2 20 0 0 2018 2020 2022 2024 2026 2028 2030 2032 2034 68 SOURCE: Own elaboration Combining these randomly generated parameters yields a vast array of possible future scenarios, each with a different set of implications. For example, a scenario with access to area C 2020 may have a relatively higher PV capex trajectory resulting in little installed PV or vice versa. To keep track of the multiple future scenarios, we adopted a scoring system from which we can assess the underlying conditions for each scenario. Detailed Results This section describes the scenarios analyzed highlighting the differences in results with the aim of providing insights to these differences. A discussion of the policy implications and relevance of the scenarios to the context of Palestine is included in the main report. Nine expansion scenarios were developed for West Bank and six for Gaza to answer the questions raised by the study. The scenarios are not necessarily incremental and some are used to illustrate the impact of modelling and policy choices as described in Table A - 10. Additionally, existing capacity options are tested and a situation where no action taken is also simulated to show the impact of inaction. Table A - 10: Expansion plans developed to answer study questions Case Description WEST BANK WC1: Classic least cost Least cost plan based on a best estimate of the future in Palestine with reserve requirements satisfied domestically (high security) WC2: Domestic reserves Robust expansion plan that considers uncertainties with reserve requirements satisfied domestically (high security) WC3: Shared reserves Robust expansion plan that considers uncertainties with reserve requirements shared with imports (partial reliance on electricity imports for security) WC4: Area C access WC3 with full access to Area C from 2018 WC5: Cost cap WC3 with system average cost capped at IEC import tariffs. WC6: PENRA vision Palestinian Energy And Natural Resources Authority (PENRA) vision to limit generation from any source to under 50% with IEC providing reserves WC7: Planned future Scenario with current generation options under consideration by PENRA WC8: High IEC Scenario with full supply from IEC and minimal investments in near-committed RE projects 69 WC9: Do nothing Continuation of the status quo with limited increase in IEC imports. GAZA GC1: Planned future Scenario with current generation options under consideration by PENRA GC2: PENRA vision PENRA vision to limit generation from any source to under 50% with IEC providing reserves GC3: Full Supply w/ GPP Full supply to Gaza with the Gaza Power Plant (GPP) GC4: High IEC Full supply to Gaza with IEC and GPP shut down and minimal investments in RE GC5: Meet demand with Gas Full supply with gas from Gaza marine gas fields GC6: Do nothing Continuation of the status quo with limited increase in IEC imports. A deterministic least cost plan could be costly because it is tailored for a particular scenario and there is a high chance of regret or failure or under-utilized assets when underlying assumptions change. To improve the resilience of the capacity plan to uncertainties, we employ the methodology described to develop subsequent capacity plans. A plan that performs well under uncertainty may not necessarily be optimal for any one particular future scenario – even the most likely future scenario – but will reduce the risk of over or under-investment. We first look at a robust plan that ensures the West Bank is able to cover all contingencies internally. Given that there is currently very little installed capacity, such a plan will require significant investments over short periods. We therefore look at a scenario where reserve requirements are shared with interconnected systems to benefit from one of the main benefits of interconnections – that is, the distribution of reserve capacity requirements amongst. Given the high potential of solar in Area C, the impact of access to Area C on generation options is evaluated in WC4. In WC5, we examine the impact imposing cost constraints on the model based on the policy of the PA to cap the cost of energy to the costs of imports from IEC. These scenarios are only relevant for West Bank because supply options to Gaza are much more limited and most constraints result in unmet demand. Scenarios WC6 and GC2 evaluate PENRAs long term vision to limit electricity generation from any source to under 50% which diversifies the energy mix. Apart from C1 in which parameters were fixed, most other cases considered uncertainties around demand, fuel pricing and availability as described in previous sections. Features of the scenarios are described in Error! Reference source not found.. The presentation of results follows the sequence of questions raised. Results for West Bank are first presented followed by results for Gaza. West Bank 70 A classic least cost plan In a deterministic analysis, what does a least cost capacity expansion plan look like, which ensures Palestine is self-reliant and able to meet demand securely (assuming there no capital constraints)? A classic least cost plan based on the planners’ best estimate of the future performs extremely well if that future materializes. Under the static conditions described in Error! Reference source not found., power generation switches from IEC imports to gas and meets entire demand. At approximately 6 US cents per kWh, CCGT is the least cost option when gas is available followed by utility scale PV at approximately US$ 1,041 per kW and 7 US cents per kWh. Table A - 11: Underlying assumptions for the deterministic plan Parameter Assumption Demand Central case Diesel prices Base case Gas prices 5.75 $/MMBTU Increase in Israel-WB 2021 Increase in Jordan-WB 2024 Egypt-WB 2024 Increase in Israel-Gaza 2024 Increase in Egypt-Gaza 2023 Israel Import price 90 $/MWh + 1% p.a. Jordan Import price Based on diesel price Egypt Import price 81 $/MWh + 1% p.a. Timing of gas (WB) 2022 Timing of gas (Gaza) 2023 Volume of gas (WB) 1.1 BCM Volume of gas (Gaza) 1.1BCM Reserve margin requirements 15% Access to area C 2020 Financial constraints No Unplanned outages No RE Capex Base case 428MW of distributed diesel capacity is installed from 2018 largely to satisfy reserve margin requirements and this is maintained through to 2030. 71 Energy mix (GWh) 2030 Installed capacity (MW) 7,000.00 PV- Jordan Imports Other, 6,000.00 Diesel , 10 15 Genset, 5,000.00 Israel 428 Imports 4,000.00 , 741 3,000.00 2,000.00 1,000.00 GT, 0 - CCGT, 2026 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2027 2028 2029 2030 886 PV-Other PV-Area C CSP Wind Biogas Diesel Genset CCGT GT Israel Imports Egypt Imports Jordan Imports Figure A - 12: Energy and mix and 2030 capacity share in a deterministic scenario How well does the least cost plan perform under uncertainty? To test the performance of the classic least cost plan, it is subjected to 100 simulations in which various parameters such as demand, fuel availability and pricing, disruptions and import volumes are varied. On average (across the 100 samples), the share of gas in the energy mix drops to 45%, largely substituted by imports from Israel (Figure A - 13). Energy mix (GWh) Unserved Energy 8,000.00 Jordan Imports 7,000.00 Egypt Imports Israel Imports 6,000.00 GT 5,000.00 CCGT 4,000.00 Diesel Genset 3,000.00 Biogas Wind 2,000.00 CSP 1,000.00 PV-Area C - PV-Other 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 Demand Figure A - 13: Energy and mix for the deterministic plan under test conditions In terms of reliability measured by the level of unmet demand, there is little impact as unserved energy is just 334 GWh over the planning horizon. Outages are covered by a combination of diesel 72 generators and imports from Egypt, Jordan and Israel. However, a comparison of costs shows that undiscounted system costs nearly double from US$ 8.7 bn to US$ 16.5 bn (Figure A - 14). 16.46 1.73 3.36 8.72 0.38 1.13 1.88 8.54 1.32 0.46 0.00 1.43 4.95 0.00 CAPEX Fixed + Var O&M Repair Costs CAPEX Fixed + Var O&M Repair Costs Unserved Energy Unserved Energy Fuel Total Fuel Total Reserve Penalties Reserve Penalties Costs Costs Deterministic plan Deterministic plan under shocks Figure A - 14: Comparison of system costs in US$ billion for the deterministic scenario In some scenarios, gas materializes earlier than 2023 so on average, there is some gas in 2022. This helps reduce system costs but the gains are offset by the other scenarios when gas is available beyond 2023. The absence of gas is substituted by imports to the extent permitted. Destructions also add to the cost but the highest change in costs is due to higher fuel costs (US$ 3.6 bn higher) as gas is substituted with more expensive options when delayed or unavailable. Incorporating uncertainties What are the features of a capacity plan that ensures that the West Bank can respond to a wide range of uncertainties including contingencies around electricity imports? What are the cost implications of such a plan? An optimal capacity plan was generated for multiple sampled scenario. Each plan is perfectly suited for the particular scenario but the plans will perform differently across multiple scenarios. 100 future scenarios were developed and for each of these, a capacity plan was generated using the core LP model. 100% 80% 60% 40% 20% 0% 16 49 82 1 4 7 10 13 19 22 25 28 31 34 37 40 43 46 52 55 58 61 64 67 70 73 76 79 85 88 91 94 97 100 Wind Biogas CCGT PV-Area A/B PV-Area C 73 Figure A - 15: Total energy mix (2020-2030) for 100 possible future scenarios As seen from Figure A - 15, IEC imports continue to be a steady source of imports for West Bank across multiple scenarios. Generation from gas (CCGT) is also prominent as well as PV which was not picked up in the deterministic scenario. In terms of capacity, the standard deviations in Figure A - 16 show Israel imports provide a steady source of capacity. Technologies like CSP are also picked up in later years when access to Area C is granted. There is wide range around the solar PV capacity in area C due to the combination of capex and access to land. Diesel generators appear to be a robust option across multiple scenarios but, as seen from Figure A - 15, the utilization of these units is low because of the high cost of fuel. However, they are best suited for providing system reserve because of the low capex requirements. 2500 Average capacity Standard deviation 2000 1500 MW 1000 500 0 Biogas Biogas Biogas Israel Imports Israel Imports Israel Imports PV-Other PV-Other PV-Other CCGT GT PV-Area C GT PV-Area C GT PV-Area C CSP CCGT CSP CCGT CSP Diesel Genset Wind Diesel Genset Wind Diesel Genset Wind Jordan Imports Jordan Imports Jordan Imports 2020 2025 2030 Figure A - 16: Mean capacity by fuel for specific years showing range as error bars and standard deviation as dots From the capacity plans, the most robust was developed. Capacities of a particular technology with high frequency of selection across scenarios are more robust. A plan comprising solely of no-regret options is unlikely to satisfy demand and will result in high system costs. Therefore capacity and technology options that are less preferred across the scenarios need to be included in the expansion plan. Whenever additional capacity is added, the capacity plan is tested across multiple scenarios. As more capacity is added, capex requirements increase but unmet demand reduces and total system costs reduce accordingly. Beyond a certain point, unmet demand is minimized and additional investments increase system costs. The lowest point is selected as the optimum capacity plan. 74 Avg discounted system costs ($US 35 34 33 32 31 Selected bn) 30 capacity plan 29 28 27 26 25 8 9 10 11 12 13 14 Total Capex ($US bn) Figure A - 17: Total capex and system costs for 100 capacity plans tested across multiple scenarios. NOTE: Installed capacity is increased with increasingly less preferred technology capacities across the 100 capacity plans. The resultant capacity plan performs better than the deterministic plan by saving nearly US$ 1.2 billion over the planning horizon. The NPV of the robust plan is US$ 7.1 billion compared with the deterministic plan at US$ 8.6 billion. Capex requirements in the robust capacity plan is US$ 1.3 billion higher than the classic case but unserved energy costs are 47% lower and repair costs are less than half the results from the deterministic scenario. Total capacity is 3,484 MW for average expected peak capacity of 1,300 MW (Figure A - 18). The total capacity is high but required to meet the planning requirements. System reserve requirements is set at 15% above peak demand and must be satisfied internally. Import capacity therefore does not contribute to reserve requirements. Additionally, PV does not provide firm capacity and so does not contribute to the reserve margin limits. While the low capex requirements for distributed diesel plants make them an attractive option to meet reserve margins, energy output shows they are low on the merit order of dispatch because of the relatively higher cost of fuel and utilization is approximately 1%. PV capacity helps reduce fuel and repair costs. a) Capacity (MW) b) Energy (GWh) 4,000 8,000 3,500 7,000 3,000 6,000 2,500 5,000 2,000 4,000 1,500 3,000 1,000 2,000 500 1,000 - - 2021 2016 2017 2018 2019 2020 2022 2023 2024 2025 2026 2027 2028 2029 2030 2019 2023 2016 2017 2018 2020 2021 2022 2024 2025 2026 2027 2028 2029 2030 75 PV-Other PV-Area C CSP Wind Biogas Diesel Genset CCGT GT Israel Imports Egypt Imports Jordan Imports Unserved Energy Demand Figure A - 18: Robust capacity plan and energy mix Transforming a system with no installed capacity to one that is self-reliant in a short period of time as illustrated in this scenario is extreme and impractical but illustrates the merits of considering uncertainties in the planning process. US$ 945 million is required in 2018 alone to avoid the reserve requirement penalty. Capex requirements are annualized and its impact on the average cost of generation is distributed across several years. For a self-reliant system, the average cost of generation increases from 9.2 US cents per kWh to 12.4 US cents per kWh in 2019 and drops in later years to 11.4 US cents per kWh (Figure A - 19). While the impact on the average cost is reasonable, raising the required capital, associated infrastructure and human capital needs will be more challenging. To reduce the financial burden on consumers, a longer timeframe will be required to develop a capacity mix that is self-reliant. During this period, imports will continue to play a significant role in the energy mix. 1,600 14 Average cost (US cents/kWh) USD million Repair Costs 1,400 12 Reserve Penalties 1,200 10 1,000 Unserved Energy Costs 8 800 Fixed + Var O&M 6 600 CAPEX 4 400 200 2 Fuel - 0 Average cost of generation 2023 2016 2017 2018 2019 2020 2021 2022 2024 2025 2026 2027 2028 2029 (US cents /kWh) 2030 Figure A - 19: Associated costs for the self-reliant robust capacity plan How does the average cost of production change by sharing reserve margin requirements with neighboring countries? A power system designed to be operated independently misses the benefits of large interconnected systems. A major benefit of interconnecting power systems is the ability to distribute reserve margin requirements thereby reducing total system costs. While this is technically optimal, other non-technical considerations may constrain the benefits associated with operating an interconnected system. To evaluate the cost of a completely self-reliant system for Palestine, the robust plan is compared with a plan that is not constrained to satisfy reserve margin requirements internally. The same steps outlined in the flowchart (Figure A - 1Error! Reference source not found.) are followed to 76 develop the alternative plan with the main difference being that the need to meet reserve margin requirements internally is removed. Partially relying on imports reduces capex requirements from US$ 2.2 bn to US$ 1.4 bn (Figure A - 20). There is some loss of reliability as unserved energy increases from 0.3% to 1.1%. However, the average cost is more stable and does not exceed 10.6 US cents per kWh with nearly US$ 200 million in fuel savings. 1,600 14 USD million Average cost (US cents/kWh) 1,400 12 Repair Costs 1,200 10 Reserve Penalties 1,000 8 Unserved Energy Costs 800 6 Fixed + Var O&M 600 4 CAPEX 400 200 2 Fuel - 0 2027 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2028 2029 2030 Average cost of generation (US cents/kWh) Figure A - 20: Associated costs for the robust capacity plan with shared reserves Comparison of least cost and policy-based scenarios Table A - 12 summarizes the results for West Bank. Supply from Israel to West Bank has been stable in the past and if this continues, the cost of inaction on system reliability (WC9) is modest with 4% unmet demand over the planning horizon as demand outgrows the pace of system expansion. By 2030, unmet demand is estimated at 9% of unmet demand. If supply from IEC grows with demand (WC8), unmet demand estimated at 1.3% and the average cost of electricity is approximately 9.8 US cents per kWh. PENRA’s current expansion plan is also robust with unmet demand under 1% but with US$ 927 million in capex requirements is more expensive than reliance on IEC. On the other hand, PENRA’s policy to cap projects below the costs of IEC imports delays investment and results in 4.2% of unmet demand. Across the scenarios, WC7 (the planned future) has the lowest combination of capex requirements and unmet demand. In general, the scenarios with higher local generation have lower unmet demand. Additional tests were run to assess the performance of the plans under stability (labelled peace) and extreme shocks (labelled war). The tests were carried out over the period from 2025-2030 for a select number of scenarios. The more diversified scenarios performed better under sever shocks than the less diversified scenarios. For example, WC4 has a low combination of fuel costs 77 and unmet demand under shocks Figure A - 21. Domestic reserves is more expensive both under stability and during shocks but provides the highest security of supply (least unmet demand). WEST BANK - Share of Unserved Energy (%) WEST BANK - Fuel Costs (US$ bil) - WC WC WC WC WC WC WC 32% High IEC supply 9 7 6 2 3 4 8 WC WC WC WC WC WC WC High IEC supply - 9 7 6 2 3 4 8 3% 20% High Area C 0.74 High Area C 1% 0.23 25% Shared reserves 1.19 Shared reserves 2% 0.63 14% Domestic reserves 4.22 Domestic reserves 0% 1.20 PENRA vision 21% PENRA vision 1.10 2% 0.50 Planned future 26% Planned future 0.72 3% 0.20 Do Nothing 36% Do Nothing - 7% - 0% 20% 40% - 1.00 2.00 3.00 4.00 5.00 War Peace War Peace Figure A - 21: Performance of scenarios under stability and extreme shocks in West Bank 78 Table A - 12: Summary of results for West Bank Classic Domestic Shared High PENRA Planned High Do Price cap LCP reserve Reserve Area C Vision future IEC nothing WC1 WC2 WC3 WC4 WC5 WC6 WC7 WC8 WC9 1. 2030 Total Available Capacity MW 2,066 3,485 2,440 2,792 2,052 2,641 2,127 1,607 987 PV-Other MW 22 109 159 39 39 574 147 147 15 PV-Area C MW 0 554 554 1,094 0 0 0 0 0 CSP MW 0 7 0 0 0 0 0 0 0 Wind MW 9 10 10 10 10 50 0 0 0 Biogas MW 25 30 30 30 20 30 0 0 0 Diesel Genset MW 597 1,190 50 30 0 30 0 0 0 CCGT MW 572 720 720 590 780 730 520 0 0 GT MW 0 0 0 0 0 0 0 0 0 Israel Imports MW 699 761 806 887 1,111 1,119 1,430 1,430 942 Egypt Imports MW 93 82 82 82 78 77 0 0 0 Jordan Imports MW 49 21 29 30 14 31 30 30 30 2. Peak Demand MW 1,304 1,304 1,304 1,304 1,304 1,304 1,304 1,304 1,304 3. Domestic capacity - 2030 MW 1,226 2,621 1,523 1,793 849 1,414 667 147 15 4. Domestic cap. as share of peak - 2030 % 94% 201% 117% 137% 65% 108% 51% 11% 1% 5. Average Cost of Energy US cents/kWh 13.57 11.77 9.94 9.88 9.49 10.16 10.06 9.78 9.79 6. Total Capex USD mill 1,323 2,833 1,982 2,284 1,139 2,133 850 174 0 7. Total Opex USD mill 378 464 241 347 100 280 174 102 71 8. Total Fuel USD mill 3,606 869 673 482 374 748 566 0 0 9. Unserved Energy Costs USD mill 1,135 169 658 782 2,553 547 222 811 2,645 10. Total Penalties (other) USD mill 5,093 2,908 794 657 363 2,697 1,183 391 482 11. Total System Costs USD mill 11,535 7,243 4,348 4,551 4,530 6,406 2,995 1,478 3,197 12. Total Unmet Demand GWh 1,513 225 878 1,043 3,404 730 296 1,081 3,527 13. Total Energy Demand GWh 81,669 81,669 81,669 81,669 81,669 81,669 81,669 81,669 81,669 14. Share of Total Unment Demand (Total) % 1.9% 0.3% 1.1% 1.3% 4.2% 0.9% 0.4% 1.3% 4.3% 15. Share of Energy Imports - 2030 % 58% 38% 37% 33% 52% 44% 62% 96% 90% 16. Diversity factor - 2030 % 0.37 0.28 0.27 0.25 0.39 0.26 0.49 0.91 0.79 79 17. 2030 Share of Energy Mix 6,873 6,904 6,906 6,894 6,900 6,892 6,883 6,860 6,869 PV-Other % 1% 3% 4% 1% 1% 14% 3% 3% 0% PV-Area C % 0% 13% 13% 25% 0% 0% 0% 0% 0% CSP % 0% 0% 0% 0% 0% 0% 0% 0% 0% Wind % 0% 0% 0% 0% 0% 2% 0% 0% 0% Biogas % 3% 3% 3% 3% 2% 3% 0% 0% 0% Diesel Genset % 0% 0% 0% 0% 0% 0% 0% 0% 0% CCGT % 37% 40% 40% 34% 43% 37% 32% 0% 0% GT % 0% 0% 0% 0% 0% 0% 0% 0% 0% Israel Imports % 48% 30% 30% 26% 44% 32% 62% 95% 88% Egypt Imports % 6% 7% 7% 7% 8% 9% 0% 0% 0% Jordan Imports % 3% 0% 0% 0% 1% 3% 0% 1% 2% Unserved Energy % 0.3% 0.0% 0.0% 0.1% 1.1% 0.0% 0.0% 0.6% 8.6% 80 Gaza Given the limited supply options, the scenarios in Gaza focus on various policy options. With the exception of GC2, all the other scenarios are tests of various supply options under uncertainty. Comparison of least cost and policy-based scenarios Table A - 13 summarizes the results for Gaza. Unlike the West Bank, the cost of inaction on system reliability (GC6) is severe with 52% unmet demand over the planning horizon. Scenario GC4 which allows increased imports from IEC offers the best combination of costs and unmet demand. Additional tests were run to assess the performance of the plans under stability (labelled peace) and extreme shocks (labelled war). The tests were carried out over the period from 2025-2030 for a select number of scenarios. As seen for the West Bank, the more diversified scenarios performed better under sever shocks than the less diversified scenarios. GC2 (PENRA vision) has a low combination of fuel costs and unmet demand under shocks Figure A - 22. GAZA - Share of Unserved Energy (%) GAZA - Fuel Costs (US$ bil) GC6 GC1 GC2 GC3 GC4 GC5 31% 1.15 GC6 GC1 GC2 GC3 GC4 GC5 Meet demand with Gas 5% Meet demand with Gas 0.57 High IEC supply 51% - 31% High IEC supply - Full Supply w/ GPP 28% 1.15 6% Full Supply w/ GPP 0.57 PENRA vision 24% 4% PENRA vision 1.10 29% 0.50 Planned future 7% 0.72 Planned future 0.20 Do Nothing 63% 50% 1.40 Do Nothing 1.71 0% 20% 40% 60% 80% War Peace - 1.00 2.00 War Peace Figure A - 22: Performance of scenarios under stability and extreme shocks in Gaza 81 Table A - 13: Summary of results for Gaza Planned future PENRA Vision Full supply High IEC Gaza supply Do nothing w/GPP with gas GC1 GC2 GC3 GC4 GC5 GC6 1. 2030 Total Available Capacity MW 975 1,077 1,395 971 970 190 PV-Other MW 163 163 163 163 163 0 PV-Area C MW 0 0 0 0 0 0 CSP MW 0 0 0 0 0 0 Wind MW 0 0 0 0 0 0 Biogas MW 2 2 2 2 0 0 Diesel Genset MW 0 120 0 0 0 0 CCGT MW 560 460 560 0 677 60 GT MW 0 60 0 0 0 0 Israel Imports MW 240 199 660 796 120 120 Egypt Imports MW 10 73 10 10 10 10 Jordan Imports MW 0 0 0 0 0 0 2. Peak Demand MW 767 767 767 767 767 767 3. Domestic capacity - 2030 MW 725 805 725 165 840 840 4. Domestic capacity as share of peak - 2030 % 95% 105% 95% 22% 110% 110% 5. Average Cost of Energy US cents/kWh 13.39 15.44 11.41 10.37 15.15 14.68 6. Total Capex USD mill 1,035 1,066 1,035 385 1,185 0 7. Total Opex USD mill 236 280 205 76 246 173 8. Total Fuel USD mill 2,264 3,471 718 4 3,987 1,588 9. Unserved Energy Costs USD mill 2,834 1,630 1,333 2,167 2,390 18,108 10. Total Penalties (other) USD mill 1,644 8,473 1,637 220 1,962 763 11. Total System Costs USD mill 8,013 14,920 4,928 2,851 9,769 20,632 12. Total Unmet Demand GWh 3,779 2,173 1,778 2,889 3,186 24,144 13. Total Energy Demand GWh 46,538 46,538 46,538 46,538 46,538 46,538 14. Share of Total Unment Demand % 8% 5% 4% 6% 7% 52% 15. Share of Energy Imports - 2030 % 29% 45% 46% 93% 16% 26% 16. Diversity factor - 2030 % 0.54 0.36 0.46 0.84 0.72 0.47 82 17. 2030 Share of Energy Mix 4,032 4,032 4,032 4,032 4,032 4,032 PV-Other % 6% 6% 6% 6% 6% 0% PV-Area C % 0% 0% 0% 0% 0% 0% CSP % 0% 0% 0% 0% 0% 0% Wind % 0% 0% 0% 0% 0% 0% Biogas % 0% 0% 0% 0% 0% 0% Diesel Genset % 0% 0% 0% 0% 0% 0% CCGT % 68% 47% 51% 0% 83% 11% GT % 0% 1% 0% 0% 0% 0% Israel Imports % 28% 33% 45% 92% 14% 24% Egypt Imports % 2% 12% 1% 2% 1% 2% Jordan Imports % 0% 0% 0% 0% 0% 0% Unserved Energy % 0% 1% 0% 0% 2% 63% 83 Sequencing investments The analysis identifies options that are robust in multiple possible future scenarios but it also clearly shows reliance on external decisions, least of which is technical. Without a change the underlying geo-political conditions, options for power supply that are directly within the control of the PA are limited. However, we see that these options, while not the cheapest, are indeed least-cost. For example, based on the assumed likelihood of gas availability, CCGT is a robust option across all applicable scenarios. However, the construction of thermal plants is only least cost when gas is available. Translating any plan into reality will require a different approach where decisions are taken as uncertainties resolve over time. There are, however, options that are optimal with either high or limited imports and these are obvious targets for immediate action. The approach used for the study shows how technology and capacity that are robust across multiples scenarios can be determined. For example, the analysis shows that solar (both rooftop PV and centralized) are optimal investments in both West Bank and Gaza because they are less dependent on external factors. In the interim, strengthening imports from Israel also helps keep down average system costs. Imports from Egypt will also help diversify supply and increase reliability of supply. Palestine stands at an advantage to benefit from evolutions in power systems because there are no locked-in technologies. The cost of PV has dropped by over 60% since 2010 and costs of storage technologies such as utility scale batteries or fuel cells are on a downward trajectory. As the unit costs of storage reach parity with cheapest source of imports, it will be beneficial to consider battery storage as a means of improving the security of supply (see Box A-1). A list of generation technologies and triggers for taking action are summarized in Table A - 14.. Table A - 14: Triggers for deciding on various technology options Technology Decision trigger Solar pv and other small RE (Area A and Immediate Gaza) Solar pv (Area B and C) When access is granted Increased imports from Jordan When Jordan is able to export power Increased imports from Egypt When Egypt is able to export power Increased imports from IEC Immediate Storage When unit cost of storage is close to cost of reserves from imports Additional thermal plant in Gaza When there is clarity around gas availability Additional thermal plant in West Bank When there is clarity around gas availability West Bank backbone When access is granted and there is clarity around availability of gas for centralized self- generation or higher imports from Jordan especially. 84 West Bank-Gaza interconnection When access is granted or Israel is willing to construct and operate the line and there is clarity around availability of centralized self- generation or higher imports from Jordan especially. Box A-1: Improving the security of supply with renewable energy technologies Palestine’s ability to generate its own electricity offers relatively higher security than importing electricity because, unlike electricity, fuel can be purchased from several markets which reduces dependence on a single source. The most secure system will be one unconstrained by fuel requirements. Renewable energy technologies (RETs), namely solar, wind and battery, offer this potential for Palestine. As the costs of RETs fall, certain capex combinations for solar technologies, wind and batteries yield over all unit costs that are reasonably low to merit closer examinations. A simplified exercise was undertaken to illustrate the concept. The analysis takes 2030 hourly load conditions and meets this demand with RETs through several combinations of investment costs. The chart below shows combinations of costs that yield average system costs of 12, 13, 14 and 15 US cents per kWh. For example, if the cost of storage drops to $263 per kWh, it could be combined with low cost offshore wind between $2,402 and $3,753 per kW and/or PV between $299 and $790 per kW. These ranges can be combined to yield 15 UScents per kWh. In other words, if the cost of PV drops by 60% reaching $790 per kW and storage for example falls to $263 per kWh, it is a combination that could be attractive for large-scale deployment of RE. Capex ($/kW for generators and $/kWh for storage) Storage 263 66 ~ 15 c/kWh Wind 3,753 2,402 CSP-6 2,563 2,402 CSP-10 3,453 2,801 PV 790 299 Storage 263 66 ~ 14 c/kWh Wind 3,304 2,402 CSP-6 2,523 2,402 CSP-10 2,982 2,801 PV 643 339 Storage 109 65 ~ 13 c/kWh Wind 3,291 2,402 CSP-6 2,667 2,402 CSP-10 3,056 2,801 PV 586 329 Storage 86 61 ~ 12 c/kWh Wind 3,132 2,402 CSP-6 2,402 2,402 CSP-10 3,238 3,012 PV 492 345 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percentage reduction in capex relative to 2016 costs 85 Annex 9: Financial Sector Model Methodology Overall Approach The financial model of the Palestinian Authority (PA) power sector will be built on three levels.  Level 1: simple cash flow models of the six Palestinian power distribution utilities DISCOs (JDECO, GEDCO, HEPCO, NEDCO, SELCO & TEDCO).  Level 2: a simple cash flow model of the new Palestinian transmission utility PETL.  Level 3: a simple characterization of the net impact of the power sector on the budget of the Palestinian Authority in the form of subsidies. The financial model use as its historic reference period, the years 2011-2015. The financial model will project forwards for the period 2016-2030. The objective of the financial modeling is to evaluate the tariffs setting and the creditworthiness of the DISCOs and especially of PETL as an off-taker for a series of major new commercial term commitments for the bulk purchase of power into the Palestinian Territories, and to identify a series of measures that could be taken to improve this creditworthiness. In particular, these measures could include the following.  Improvements in the commercial and operational performance of the DISCOs.  Increases in the retail tariff to the end consumer.  Injection of additional public subsidy to the sector. The financial model uses PA electricity physical demand projections from the Robust Planning Model and T&D costs forecast based on the various planning scenarios. The Financial Model was used to explore the financial impacts on the sector based on 3 scenarios from the Robust Planning model that covered the entire range of power production costs from lowest to highest. For the West Bank these included ‘Planned Future’, ‘Maximum Cooperation’ and ‘PENRA Vision’ scenarios and for Gaza, these included ‘Planned Future’, ‘Maximum Independence’ and ‘Maximum Cooperation’. The detailed description of the scenarios is provided in the Planning Model section of the main report under Part II. The Financial Model assumes that:  PETL will act as a single buyer. PETL will import all the electricity that is available from IEC, Jordan and Egypt and will buy all the electricity produced in the PA (CCGT, Solar, wind, biomass etc.).  PETL will act also as a single supplier to the DISCOs using the IEC transmission infrastructure or its own transmission infrastructure (transmission costs are included in PETL tariffs shown in the Financial Model ). 86  The DISCOs invest in their own distribution infrastructure (distribution costs are included in DISCOs tariffs shown in the Financial Model). Level 1: Distribution Utilities Output variable: Electricity average equilibrium cost and retail tariff in each distribution utility as well as aggregate. Based on data projections and the chosen levels for input parameters, the model solves for the retail electricity price level that assures the financial equilibrium of each utility. This should initially be done at the utility level. However, PERC currently has a policy of charging a single uniform tariff throughout Palestine, so the model will also calculate the average cost recovery tariff across the five distribution utilities, as well as computing the transfers that would be needed across utilities to ensure their individual financial sustainability should the uniform tariff be applied. Those whose financial equilibrium tariff is above the sector average would need to receive a net compensatory transfer, and vice versa. Input variables: distribution losses and revenue collection ratio. The financial model will be set-up in such a way as to allow the user to choose target values for distribution losses and revenue collection ratios for the year 2030. These two parameters reflect the overall operational and financial performance of the utility and can be improved over time through management effort. Given that there are measures underway to improve the poor current performance in these areas, the model assume that the full benefit of these measures will be achieved by 2030. The user should be able to input more or less ambitious targets for both of these variables to see what impact this has on the equilibrium tariff. Data projections: characterizing the revenues and expenditures. Basic data on the revenues and expenditures of the utilities is collected by PERC for the purposes of determining the revenue requirement for the cost plus tariff-setting process. On the revenue side, the model takes historic data on billings and collections in both physical and financial terms. The difference between power purchased and power billed will give distribution losses. The difference between power billed and power collected will give the collection losses. The projection of the revenue side will be based on physical demand projections provided by the Robust Planning Model and on return on equity set by the regulator (PERC). The tariff to be applied to the demand projections will be based on the solution of the model as noted above. On the expenditure side, the model uses data from the DISCOs financial annual reports on O&M, taxes, debt service, planned investments, and power purchase costs. O&M are projected based on demand projections and on Efficiency factor to be set by the regulator. Debt service and planned investments are projected based on information about the repayment profile of currently held debts, interest rate on debt and investment plans for the period. The distribution utilities’ most significant expenditure is power purchase. 87 The projection of the wholesale power price over time will be an output of the Level 2 model covering PETL (see below). Affordability check: how power bills weigh on household budgets. As an add-on to the financial analysis of the DISCOs, the model includes a module that will allow to check for affordability and compute the potential value of consumer subsidies. The affordability check is based on data for the average household income across ten deciles of the Palestinian income distribution that is derived from the PCBS Labor Force Survey for 2013. These will need to be rolled forward to reflect anticipated real income growth through 2030. Subsistence electricity consumption can be estimated as the amount of electricity needed to provide a basic package of energy services in the household. Such information was derived from the PCBS Household Energy Surveys. Based on an estimate of subsistence electricity consumption the weight of the power bill associated with the equilibrium retail tariff can be calculated as a share of household income. When this share exceeds 5%, an affordability issue is presumed to arise. On this basis, it is possible to calculate the total amount of government demand-side subsidy that would be needed to keep the cost of subsistence consumption below the 5% threshold. The Model calculates and displays two distinct subsidies. The first is the subsidy requirement to maintain financial equilibrium if retail tariffs are not adjusted as additional power supply options come online. The second is the subsidy requirement to provide targeted subsidies to the poorest who cannot afford increases in tariffs. The subsidies are then compared in scenarios where DISCO efficiencies are, and are not, improved to provide a sense of the impact of DISCO inefficiencies on the PA budget. Level 2: PETL Output variable: average wholesale price of electricity to be charged by PETL to DISCOs Based on data projections and the chosen levels for input parameters, the model solves for the average wholesale power price level that assures the financial equilibrium of the PA power sector, and for Gaza and for the West Bank separately. Input variable: average unit subsidy to the wholesale price of electricity to be applied by PA. The financial model allow the user to choose the percentage of the wholesale electricity price that would be subsidized by the PA. The value of this supply side subsidy is initially set to zero so as to understand the full tariff implications of the proposed investment plan. If the resulting retail tariff proves to be unaffordable (based on the affordability check), then the problem can be addressed either through incorporating a supply side subsidy at the level of PETL, or a demand-side subsidy directly to consumers of the distribution utilities, or a combination of the two. Although in practice, supply-side subsidies are more commonplace, demand-side subsidies are far preferable from an economic standpoint. 88 Data projections: characterizing revenues and expenditures On the expenditure side, PETL’s expenditures can be divided between those associated with wholesale power purchase and those associated with operating the transmission system. In terms of wholesale power purchase, in the future PETL will be the holder of various PPAs signed with different suppliers that may include IEC, Israeli IPPs, gas-fired IPPs in Palestine, solar IPPs in Palestine, power import contracts with Jordan and Egypt. The output of the planning model will give the total amount of power from each source that PETL will need to purchase in any given year. The planning model will also have unit cost information for each of these projects. On the basis of this information, a financial PPA price will need to be estimated bearing in mind the potential financing conditions for power generation infrastructure in Palestine, particular for domestic IPPs. Multiplying each power purchase price by the corresponding volume of power will give the total wholesale power purchase bill in any given year. The overall volume of supply should be compatible with the demand projections used in the planning model and also feeding into the Level 1 DISCOs model. PETL also faces other costs associated with operating and developing the transmission system. These include transmission losses, O&M expenditures, taxes, debt service, and any investments needed to upgrade the transmission network. These data is obtainable from the PETL PWC Business Plan or from PETL itself. On the revenue side, PETL’s future revenue will be the wholesale power tariff multiplied by the total amount of energy demanded by the DISCOs. Level 3: Palestinian Authority Data projections: characterizing fiscal flows to the power sector The model will take stock of all the ways in which the energy sector results in revenues or expenditures to the public budget. On the revenue side, the power sector contributes tax revenues through the application of VAT and corporation taxes (VAT is not calculated in the model at this stage). It is not clear whether there are any other positive fiscal contributions at present, but the future development of Gaza marine would potentially provide an important revenue source, although certainly not earmarked to the power sector. On the expenditure side, the power sector draws a number of implicit and explicit subsidies from the public budget, for which we do not yet have a comprehensive inventory. The ones that we do know about include net lending, subsidy to DISCOs to compensate for higher IEC prices, and potentially pass through of capital grants and concessional finance from donors. 89 Furthermore, the demand and supply-side subsidies calculated in Levels 1 and 2 respectively enter the Level 3 model as a projected subsidy expenditure for the sector. The impact of this subsidy on the overall fiscal balance of the PA would need to be gauged in order to identify a level of public subsidy that is affordable in fiscal terms. Since power is only one of the sectors handled through the budget, it would be important to know the overall revenue and expenditure balance of the PA and how this is projected to evolve over time. By way of summary, the schematic chart below illustrates the flows into and between the different entities of the PA power sector presented in the financial model. 90 Part I: Current Cash Flow JDECO Residential / Commercial / Public (East Jerusalem) Power Import Electricity Consumers Egypt Municipalities HEPCO & other (Hebron) Power Import NEDCO PETL Jordan (Nablus) (TSO & Single buyer) SELCO Power Import IEC (South WB) Israel TEDCO (North WB) Renewable Energy IPP’s GEDCO (Gaza) IPP Part II: Capital Cash Flow Gaza TAMAR LEVIATAN Foreign/ Domestic IPP Investors Jenin IPP Banks & Hebron Gaza Marin (Gas field) Financial Institutions Part III: Government Take PA (Palestinian Authority) Legend Power cash flow(1) (Kwh x Predicted Electricity Tariff) Taxes, Net lending,(2) Capital Subsidies or Guarantees and Royalties Net Capital Investments(3) O&M and Financial Expenses Loans Principal payments Gas purchase 1) Power cash flow includes repayments of debts to IEC. 2) DISCOs (and some municipalities) are currently paying directly to IEC for the purchased power distributed to Palestinian customers. Since DISCOS do not pay 100% their electricity suppliers, namely IEC, the PA is indirectly subsidizing the DISCOs due to the monthly sums taken by the Israeli Ministry of Finance from Palestinian taxes collected on their behalf (“clearance revenues”) to compensate from the Palestinian DISCO’s non-payment for purchased electricity from IEC (“Net lending”). 3) Capital investments minus return on capital 91 Figure A9.1 - Financial Model inputs and outputs – GEDCO 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 GEDCO Purchases and Sales Purchase of electricity from IEC & Jordan/ PETL GWh 1,763 1,642 1,730 1,385 1,432 1,486 1,504 1,724 2,036 2,328 2,612 2,996 3,149 3,297 3,471 3,649 3,749 3,900 4,035 4,170 Total Losses % 30.0% 30.0% 30.0% 26.5% 26.2% 26.0% 25.8% 25.6% 25.4% 25.2% 24.9% 24.7% 24.5% 24.3% 24.1% 23.9% 23.6% 23.4% 23.2% 23.0% Total Power Sales GWh 1,234 1,149 1,211 1,018 1,056 1,099 1,116 1,283 1,519 1,742 1,960 2,255 2,377 2,496 2,635 2,778 2,862 2,986 3,099 3,211 Collection rate % 65.0% 68.0% 71.0% 64.0% 65.0% 66.7% 68.5% 70.2% 71.9% 73.7% 75.4% 77.1% 78.9% 80.6% 82.3% 84.1% 85.8% 87.5% 89.3% 91.0% Total Power Paid by consumers GWh 802 781 860 652 686 734 764 900 1,093 1,283 1,478 1,739 1,874 2,012 2,170 2,336 2,456 2,614 2,766 2,922 Operating Income (Power sales) mil NIS 615 599 632 509 518 592 575 717 832 942 1,039 869 1,391 1,494 1,587 1,618 1,723 1,680 1,788 1,832 Other Income mil NIS NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA Total Income mil NIS 615 599 632 509 518 592 575 717 832 942 1,039 869 1,391 1,494 1,587 1,618 1,723 1,680 1,788 1,832 GEDCO Operating Costs Electricity Purchase from IEC & Jordan/ PETL mil NIS 701 817 911 916 795 808 760 932 1,054 1,133 1,219 951 1,581 1,665 1,731 1,720 1,799 1,704 1,781 1,786 O & M Expenses mil NIS 53 56 54 58 63 66 66 76 90 103 115 132 139 146 153 161 166 172 178 184 Depreciation Expenses mil NIS NA NA NA 13 13 13 13 13 13 22 22 22 22 22 22 22 22 22 22 22 Running cost mil NIS NA NA NA NA NA 0 0 0 0 16 16 16 16 16 16 16 16 16 16 16 Financial cost mil NIS NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA Return On Equity mil NIS NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA Total Electricity Costs mil NIS 755 873 965 986 872 887 840 1,021 1,157 1,273 1,372 1,121 1,758 1,848 1,921 1,919 2,002 1,914 1,997 2,008 GEDCO Income Annual income/loss before income tax mil NIS -139 -274 -334 -477 -354 -295 -265 -304 -325 -331 -333 -252 -367 -354 -335 -301 -280 -234 -210 -176 Income Tax - 15% mil NIS NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA Net Annual income mil NIS -139 -274 -334 -477 -354 -295 -265 -304 -325 -331 -333 -252 -367 -354 -335 -301 -280 -234 -210 -176 GEDCO Purchase, Sale and Equilibrium tariff Average purchase cost / PETL Tariff NIS/KWh 0.398 0.497 0.527 0.661 0.555 0.544 0.505 0.541 0.518 0.487 0.467 0.318 0.502 0.505 0.499 0.471 0.480 0.437 0.441 0.428 Average retail tariff NIS/KWh 0.498 0.521 0.522 0.500 0.490 0.807 0.753 0.796 0.761 0.734 0.703 0.500 0.742 0.743 0.731 0.693 0.701 0.643 0.646 0.627 Electricity average equilibrium cost NIS/KWh 0.941 1.117 1.123 1.513 1.270 1.209 1.099 1.134 1.058 0.992 0.928 0.645 0.938 0.919 0.886 0.822 0.815 0.732 0.722 0.687 *Assuming 'Planned Future' planning scenario from 2016-2030 92 Figure A9.2 - Financial Model inputs and outputs – JDECO 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 JDECO Purchases and Sales Purchase of electricity from IEC & Jordan/ PETL GWh 1,797 1,943 1,902 1,935 2,114 2,084 2,142 2,199 2,186 2,391 2,511 2,446 2,548 2,509 2,563 2,677 2,796 2,918 2,913 3,012 Total Losses % 27.7% 26.4% 26.2% 24.9% 23.9% 23.8% 23.8% 23.7% 23.6% 23.6% 23.5% 23.5% 23.4% 23.3% 23.3% 23.2% 23.2% 23.1% 23.1% 23.0% Total Power Sales GWh 1,299 1,431 1,403 1,454 1,609 1,588 1,633 1,678 1,670 1,827 1,920 1,872 1,952 1,923 1,966 2,055 2,148 2,244 2,241 2,320 Collection rate % 95.9% 96.6% 83.4% 95.0% 90.5% 90.5% 90.6% 90.6% 90.6% 90.7% 90.7% 90.7% 90.8% 90.8% 90.8% 90.9% 90.9% 90.9% 91.0% 91.0% Total Power Paid by consumers GWh 1,245 1,381 1,171 1,381 1,444 1,437 1,479 1,520 1,513 1,656 1,742 1,698 1,772 1,746 1,786 1,867 1,952 2,040 2,039 2,111 Operating Income (Power sales) mil NIS 695 875 889 951 949 946 943 970 975 1,129 1,184 1,223 1,246 1,205 1,214 1,248 1,278 1,065 1,285 1,309 Other Income mil NIS 54 57 83 72 68 67 68 70 70 76 80 78 81 80 82 85 89 93 93 96 Total Income mil NIS 749 932 971 1,022 1,017 1,012 1,012 1,041 1,045 1,206 1,265 1,301 1,328 1,285 1,296 1,334 1,368 1,158 1,378 1,405 JDECO Operating Costs Electricity Purchase from IEC & Jordan/ PETL mil NIS 563 800 832 886 871 741 731 756 762 909 960 1,011 1,030 991 999 1,030 1,056 814 1,058 1,079 O & M Expenses mil NIS 146 148 163 172 188 185 190 195 194 212 223 217 226 223 228 238 248 259 259 267 Depreciation Expenses mil NIS 24 21 20 30 37 36 36 36 35 37 36 36 36 35 35 35 34 34 34 33 Interest rate on debt % 2.75% 3.52% 2.10% -1.04% -0.64% 3.50% 3.75% 4.00% 4.25% 4.50% 4.75% 5.00% 5.25% 5.50% 5.75% 6.00% 6.25% 6.50% 6.75% 7.00% Financing Expenses mil NIS 22 34 28 -15 -11 59 62 63 64 64 64 63 63 62 61 60 59 58 57 56 Running cost mil NIS NA NA NA NA NA 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 Other Expenses mil NIS NA NA 4 4 6 4 5 5 5 5 5 5 5 5 5 6 6 6 6 6 Return On Equity mil NIS NA NA NA NA NA 23 22 21 20 19 17 15 13 11 9 6 4 1 0 -2 Total Electricity Costs mil NIS 754 1,004 1,046 1,076 1,091 1,045 1,041 1,071 1,076 1,242 1,302 1,344 1,369 1,323 1,332 1,370 1,402 1,167 1,408 1,434 JDECO Income Annual income/loss before income tax mil NIS -5 -71 -75 -54 -74 -14 -13 -14 -16 -22 -25 -32 -33 -32 -33 -35 -37 -14 -36 -38 Income Tax - 15% mil NIS 3 0 0 3 8 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Net Annual income mil NIS -8 -71 -75 -56 -82 -14 -13 -14 -16 -22 -25 -32 -33 -32 -33 -35 -37 -14 -36 -38 JDECO Purchase, Sale and Equilibrium tariff Average purchase cost / PETL Tariff NIS/KWh 0.313 0.412 0.437 0.458 0.412 0.356 0.341 0.344 0.349 0.380 0.382 0.413 0.404 0.395 0.390 0.385 0.378 0.279 0.363 0.358 Average retail tariff NIS/KWh 0.535 0.612 0.633 0.654 0.590 0.658 0.638 0.638 0.644 0.682 0.680 0.720 0.703 0.690 0.680 0.669 0.655 0.522 0.630 0.620 Electricity average equilibrium cost NIS/KWh 0.606 0.727 0.894 0.779 0.755 0.727 0.704 0.705 0.711 0.750 0.747 0.791 0.773 0.758 0.746 0.734 0.718 0.572 0.691 0.680 *Assuming 'Planned Future' planning scenario from 2016-2030 93 Figure A9.3 - Financial Model inputs and outputs – NEDCO 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 NEDCO Purchases and Sales Purchase of electricity from IEC & Jordan/ PETL GWh 414 474 480 502 549 576 1,125 1,193 1,232 1,274 1,289 1,470 1,498 1,668 1,763 1,799 1,842 1,882 2,051 2,122 Total Losses % 19% 17% 12% 14% 17% 17% 17% 18% 18% 19% 19% 20% 20% 20% 21% 21% 22% 22% 23% 23% Total Power Sales GWh 337 392 420 435 458 478 928 979 1,007 1,035 1,042 1,182 1,198 1,327 1,395 1,416 1,442 1,465 1,588 1,634 Collection rate % 79% 70% 87% 86% 98% 99% 99% 100% 100% 101% 101% 102% 102% 103% 103% 104% 104% 105% 105% 106% Total Power Paid by consumers GWh 266 274 365 376 450 472 922 978 1,010 1,044 1,056 1,204 1,226 1,364 1,441 1,470 1,504 1,535 1,672 1,728 Operating Income (Power sales) mil NIS 189 223 232 245 242 252 452 493 521 589 606 734 744 811 852 867 878 706 939 976 Other Income mil NIS 2 6 24 40 NA 46 48 51 53 55 55 63 64 72 76 77 79 81 88 91 Total Income mil NIS 191 230 256 285 NA 298 501 544 574 643 661 797 808 883 928 944 957 787 1,027 1,067 NEDCO Operating Costs Electricity Purchase from IEC & Jordan/ PETL mil NIS 179 200 230 250 247 205 384 410 430 484 493 608 606 659 687 692 696 525 745 760 O & M Expenses mil NIS 10 17 12 13 NA 15 29 30 31 32 33 37 38 42 45 46 47 48 52 54 Depreciation Expenses mil NIS 1 2 2 2 NA 1 2 3 4 6 6 6 6 6 6 6 6 6 6 6 Interest rate on debt % NA NA NA NA NA 4% 4% 4% 4% 5% 5% 5% 5% 6% 6% 6% 6% 7% 7% 7% Financing Expenses mil NIS NA NA NA NA NA 10 11 17 19 21 22 22 26 25 26 25 22 19 5 13 Running cost mil NIS NA NA NA NA NA 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 Other Expenses mil NIS 1 NA 1 5 NA 6 12 13 13 14 14 16 16 18 19 19 20 20 22 23 Return On Equity mil NIS NA NA NA NA NA 17 18 20 22 25 27 30 33 36 40 45 49 54 59 65 Total Electricity Costs mil NIS 191 219 245 269 NA 254 455 493 519 583 596 720 726 788 824 834 841 673 890 922 NEDCO Income Annual income/loss before income tax mil NIS 0.5 11 12 15 NA 60 63 70 77 85 92 107 115 131 144 155 166 168 195 210 Income Tax - 15% mil NIS 0.4 2 5 7 NA 9 9 11 12 13 14 16 17 20 22 23 25 25 29 32 Net Annual income mil NIS 0.1 9 7 9 NA 51 54 60 65 73 78 91 98 112 123 132 141 143 166 179 NEDCO Purchase, Sale and Equilibrium tariff Average purchase cost / PETL Tariff NIS/KWh 0.348 0.410 0.444 0.487 0.450 0.356 0.341 0.344 0.349 0.380 0.382 0.413 0.404 0.395 0.390 0.385 0.378 0.279 0.363 0.358 Average retail tariff NIS/KWh 0.561 0.569 0.551 0.563 0.528 0.533 0.490 0.504 0.516 0.564 0.573 0.610 0.607 0.594 0.591 0.590 0.584 0.460 0.561 0.565 Electricity average equilibrium cost NIS/KWh 0.716 0.797 0.671 0.717 NA 0.539 0.493 0.504 0.514 0.558 0.565 0.598 0.592 0.577 0.572 0.568 0.559 0.438 0.533 0.533 *Assuming 'Planned Future' planning scenario from 2016-2030 94 Figure A9.4 - Financial Model inputs and outputs – TEDCO 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 TEDCO Purchases and Sales Purchase of electricity from IEC & Jordan/ PETL GWh 71 81 85 96 104 118 213 226 234 242 245 279 284 317 334 341 350 357 389 403 Total Losses % 3% 16% 14% 15% 16% 17% 17% 17% 18% 18% 19% 19% 20% 20% 21% 21% 22% 22% 23% 23% Total Power Sales GWh 69 68 73 81 87 99 177 187 192 197 198 225 228 252 265 269 274 278 301 310 Collection rate % 97% 105% 97% 85% 76% 77% 78% 79% 80% 81% 82% 83% 84% 85% 86% 87% 88% 89% 90% 91% Total Power Paid by consumers GWh 67 72 71 68 67 76 139 148 154 160 163 187 192 215 228 234 241 248 271 282 Operating Income (Power sales) mil NIS 30 35 39 46 46 42 73 80 84 96 99 121 123 136 144 148 151 125 165 173 Other Income mil NIS 0 3 5 6 8 9 16 17 17 18 18 20 21 23 25 25 26 26 29 30 Total Income mil NIS 30 38 44 52 54 51 89 96 102 113 117 141 144 159 168 173 176 151 194 202 TEDCO Operating Costs Electricity Purchase from IEC & Jordan/ PETL mil NIS 27 33 38 45 42 42 73 78 82 92 94 115 115 125 130 131 132 100 141 144 O & M Expenses mil NIS 3.7 4.9 5.1 6.3 7.4 8.4 15.1 16.0 16.5 17.1 17.3 19.7 20.1 22.4 23.6 24.1 24.7 25.2 27.5 28.5 Depreciation Expenses mil NIS NA NA NA NA NA 0.0 0.0 0.0 0.0 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 Interest rate on debt % NA NA NA NA NA 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Financing Expenses mil NIS NA NA NA NA NA 2.0 2.2 3.7 4.4 5.1 5.9 6.5 7.9 8.5 9.5 10.2 10.5 10.8 9.2 11.0 Running cost mil NIS NA NA NA NA NA 0.0 0.0 0.0 0.0 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 Other Expenses mil NIS 0.6 0.6 0.8 0.7 0.7 0.8 1.5 1.6 1.6 1.7 1.7 2.0 2.0 2.2 2.3 2.4 2.5 2.5 2.7 2.8 Return On Equity mil NIS NA NA NA NA NA 1.6 1.5 1.3 1.1 1.0 0.8 0.6 0.4 0.3 0.3 0.4 0.7 1.0 1.8 2.5 Total Electricity Costs mil NIS 31.1 38.7 44.1 51.8 50.4 55.0 93.1 100.4 105.3 117.5 120.0 144.7 146.0 159.1 166.8 169.2 171.0 139.8 183.2 189.7 TEDCO Income Annual income/loss before income tax mil NIS -1.0 -1.1 0.1 0.2 3.4 -2.2 -3.1 -2.9 -2.5 -3.2 -2.5 -3.1 -1.7 0.0 1.8 3.8 6.1 12.1 12.2 15.1 Income Tax - 15% mil NIS 0.0 0.0 0.0 0.0 0.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.9 1.8 1.8 2.3 Net Annual income mil NIS -1.0 -1.1 0.1 0.1 2.9 -2.2 -3.1 -2.9 -2.5 -3.2 -2.5 -3.1 -1.7 0.0 1.8 3.8 5.2 10.3 10.4 12.9 TEDCO Purchase, Sale and Equilibrium tariff Average purchase cost / PETL Tariff NIS/KWh 0.378 0.407 0.447 0.468 0.407 0.356 0.341 0.344 0.349 0.380 0.382 0.413 0.404 0.395 0.390 0.385 0.378 0.279 0.363 0.358 Average retail tariff NIS/KWh 0.432 0.506 0.527 0.563 0.529 0.556 0.526 0.538 0.549 0.597 0.606 0.644 0.641 0.631 0.630 0.629 0.625 0.503 0.608 0.612 Electricity average equilibrium cost NIS/KWh 0.465 0.539 0.619 0.756 0.757 0.720 0.672 0.678 0.684 0.733 0.736 0.773 0.761 0.740 0.730 0.722 0.709 0.564 0.675 0.672 *Assuming 'Planned Future' planning scenario from 2016-2030 95 Figure A9.5 - Financial Model inputs and outputs – HEDCO 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 HEPCO Purchases and Sales Purchase of electricity from IEC & Jordan/ PETL GWh 362 369 373 379 411 419 650 673 695 720 744 729 719 738 733 732 792 810 850 901 Total Losses % 22% 19% 20% 19% 20% 21% 21% 21% 21% 21% 21% 22% 22% 22% 22% 22% 22% 23% 23% 23% Total Power Sales GWh 282 300 299 306 328 333 516 532 549 567 585 571 563 576 571 569 614 627 656 694 Collection rate % 74% 74% 70% 82% 81% 82% 83% 83% 84% 85% 85% 86% 87% 87% 88% 88% 89% 90% 90% 91% Total Power Paid by consumers GWh 209 222 209 251 267 273 427 443 461 480 499 491 487 502 501 503 547 562 593 632 Operating Income (Power sales) mil NIS 154 181 181 193 193 175 246 260 274 307 322 339 334 339 338 338 359 300 376 399 Other Income mil NIS 14 13 16 15 16 16 25 26 26 27 28 28 27 28 28 28 30 31 32 34 Total Income mil NIS 167 194 197 208 209 191 270 285 300 335 350 367 362 367 365 366 389 330 408 434 HEPCO Operating Costs Electricity Purchase from IEC & Jordan/ PETL mil NIS 136 160 170 176 164 149 222 231 242 274 285 301 291 292 286 282 299 226 309 323 O & M Expenses mil NIS 13 17 19 15 16 17 26 27 28 29 30 29 29 29 29 29 32 32 34 36 Depreciation Expenses mil NIS 9 9 9 10 10 10 10 9 9 11 11 11 10 10 10 10 10 10 10 10 Interest rate on debt % NA NA NA NA NA 4% 4% 4% 4% 5% 5% 5% 5% 6% 6% 6% 6% 7% 7% 7% Financing Expenses mil NIS 2 12 1 3 7 25 26 31 34 37 41 44 47 50 53 55 57 61 59 66 Running cost mil NIS NA NA NA NA NA 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 Other Expenses mil NIS NA 0 0 1 1 1 1 1 1 1 2 2 1 2 2 2 2 2 2 2 Return On Equity mil NIS NA NA NA NA NA 12 12 12 11 11 10 8 7 6 5 4 3 2 2 1 Total Electricity Costs mil NIS 160 197 199 206 198 214 297 311 326 362 377 394 386 388 384 381 403 333 415 438 HEPCO Income Annual income/loss before income tax mil NIS 7 -3 -2 2 11 -10 -15 -15 -15 -18 -18 -20 -18 -16 -14 -13 -11 -2 -6 -4 Income Tax - 15% mil NIS 1 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Net Annual income mil NIS 6 -3 -2 2 9 -10 -15 -15 -15 -18 -18 -20 -18 -16 -14 -13 -11 -2 -6 -4 HEPCO Purchase, Sale and Equilibrium tariff Average purchase cost / PETL Tariff NIS/KWh 0.377 0.433 0.456 0.464 0.398 0.356 0.341 0.344 0.349 0.380 0.382 0.413 0.404 0.395 0.390 0.385 0.378 0.279 0.363 0.358 Average retail tariff NIS/KWh 0.545 0.604 0.605 0.630 0.590 0.641 0.576 0.585 0.594 0.640 0.645 0.692 0.687 0.675 0.674 0.671 0.657 0.533 0.634 0.632 Electricity average equilibrium cost NIS/KWh 0.766 0.889 0.952 0.821 0.743 0.782 0.696 0.702 0.707 0.755 0.756 0.804 0.792 0.773 0.766 0.758 0.736 0.592 0.701 0.694 *Assuming 'Planned Future' planning scenario from 2016-2030 96 Figure A9.6 - Financial Model inputs and outputs – SELCO 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 SELCO Purchases and Sales Purchase of electricity from IEC & Jordan/ PETL GWh 121 125 124 124 131 178 207 214 221 229 237 232 229 235 233 233 252 258 271 287 Total Losses % 37% 30% 29% 28% 27% 27% 26% 26% 26% 26% 25% 25% 25% 25% 24% 24% 24% 24% 23% 23% Total Power Sales GWh 76 88 88 89 96 131 152 158 164 171 177 174 172 177 177 177 192 197 208 221 Collection rate % 54% 59% 58% 71% 79% 80% 81% 82% 82% 83% 84% 85% 85% 86% 87% 88% 89% 89% 90% 91% Total Power Paid by consumers GWh 41 52 51 63 76 104 123 129 135 142 148 147 147 153 154 156 170 176 187 201 Operating Income (Power sales) mil NIS 48 56 54 76 67 73 93 98 102 112 116 120 118 119 118 117 124 104 130 136 Other Income mil NIS 2 5 6 4 9 9 11 11 11 12 12 12 12 12 12 12 13 13 14 15 Total Income mil NIS 50 61 60 80 77 82 103 108 113 123 128 132 129 131 129 129 137 117 143 151 SELCO Operating Costs Electricity Purchase from IEC & Jordan/ PETL mil NIS 43 54 53 71 49 63 71 74 77 87 91 96 93 93 91 90 95 72 98 103 O & M Expenses mil NIS 8 9 18 15 19 20 23 24 25 25 26 26 25 26 26 26 28 29 30 32 Depreciation Expenses mil NIS 5 5 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 Interest rate on debt % NA NA NA NA NA 4% 4% 4% 4% 5% 5% 5% 5% 6% 6% 6% 6% 7% 7% 7% Financing Expenses mil NIS 2 2 2 2 3 15 14 15 15 14 14 13 12 11 11 10 9 9 8 8 Running cost mil NIS NA NA NA NA NA 0 0 0 0 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23 Other Expenses mil NIS -2 0 -2 0 -7 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Return On Equity mil NIS NA NA NA NA NA 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Total Electricity Costs mil NIS 56 71 78 95 71 105 115 120 123 134 138 142 137 138 135 133 140 116 144 149 SELCO Income Annual income/loss before income tax mil NIS -6 -10 -18 -15 5 -22 -12 -11 -10 -11 -10 -10 -8 -7 -5 -4 -3 1 0 1 Income Tax - 15% mil NIS 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Net Annual income mil NIS -6 -10 -18 -15 5 -22 -12 -11 -10 -11 -10 -10 -8 -7 -5 -4 -3 1 0 1 SELCO Purchase, Sale and Equilibrium tariff Average purchase cost / PETL Tariff NIS/KWh 0.356 0.433 0.426 0.570 0.373 0.356 0.341 0.344 0.349 0.380 0.382 0.413 0.404 0.395 0.390 0.385 0.378 0.279 0.363 0.358 Average retail tariff NIS/KWh 0.639 0.636 0.610 0.851 0.703 0.804 0.754 0.756 0.752 0.789 0.781 0.818 0.800 0.778 0.764 0.751 0.728 0.591 0.692 0.676 Electricity average equilibrium cost NIS/KWh 1.375 1.368 1.544 1.495 0.940 1.005 0.934 0.927 0.913 0.947 0.929 0.964 0.934 0.900 0.876 0.854 0.820 0.660 0.766 0.742 *Assuming 'Planned Future' planning scenario from 2016-2030 97