Turkey - Power Projects

Docaumt of The World Bank FOR OMCLAL USE ONLY Rqwt No. 12069 PERFORMANCE AUDIT REPORT TuRKEY ELBIST4N PROJECT (LOAN 1023-TU) SECOND TEK TRANSMISSION PROJECT (LOAN 1194-TU) KARARAYA YDROPOER PROJECT (LOAN 1844-TU) ThIRD TEK TRANSMISSION PROJECT (LOAN 2322-TU) AND ELBISTAN OPERATION AND MAINTENANCE ASSISTANCE PROJECT (WAN 2650-TU) JUNE 24, 1993 MICROGRAPHICS Report No: 12069 Type: PPAR Operations Evaluation Department This document has a restricted distribution and may be used by recipien only In the performance of their official duties. Its contents may not otherwise be disclosed withot World Bank authorisation. CURRENCY EQUIVALENTS Currency Unit - Lira (ML) At Appraisal (March 1980) - US$1 - TL 70 TL 1000 - US$14.29 June 1990: - US$1 - TL 2610 TL 1000 - US$0.3831 GLOSSARY OF ABBREVIATIONS DSI - Dev1et Su Isleri (State Hydraulice Agency) BID - European Investment Bank GOT - Government of Turkey KW - Kreditanstalt ffr Wiederaufbau MENR - Ministry of Energy and Natural Resources PCR - Project Completion Report SEE - State Economic Enterprise SPO - State Planning Organisation TEK Turkiye Elektrik Ruruu (Turkish Electricty Authority) TKT - Turkuye Eomur Isletmeleri lurumu (Turkish Coal Authority) Fiscal Year January 1 - December 31 FOR OFFICIAL USE ONLY THE WORLD BANK Washirgton, D.C. 20433 U.S.A. Ottie of Dimector-General Operations fealuation June 24, 1993 MMORANDUM TO THE EXlCUTIVE DIRECTORS AND THE PRESIDENT SUBJECTs Project Audit Report on Turkey Elbistan Project (Loan 1023-TU); Second TEK Transmission Project (Loan 1194-TU); Karakays Hydropower Project (Loan 1844-TU); Third TEK Transmission Project (Loan 2322-TU) and ZIbistan Operation and Maintenance Assistance Prolect (Loan 2650-TU) Attached is the "Performance Audit eport on Turkey - Elbistan Project (Loan 1023-TU), Second TEK Transmission Project (Loan 1194-TU), Karakaya Hydropower Project (Loan 1844-TU), Third TEK Transmission Project (Loan 2322-TU), and Elbistan Operation and Maintenance Assistance Project (Loan 2650-TU)" prepared by OED. The physical elements of the projects represent some 151 of Turkey's installed capacity as well as moat of the associated transmission facilities. They were successfully completed except for the Elbistan power plant, which, though commissioned in 1984, is not yet able to provide the full planned output as long as the associated lignite mine does not supply the fuel in accordance with the specifications. This is likely to take place only once the Government agrees, inter alia, to an adequate pricing of the lignite. The projects also a atmed at strengthening Turkey's main electricity utility, TEK, a goal that was achieved to a very limited degree, the main progress being achieved during the . early 1980s in connection with the implementation of several Bank-supported structural adjustment programs. This success was largely undone by poor fiscal . discipline in the second half of the 1980s. The Bank's hopc that its support of the larakaya Project would help Turkey, Syria, and Iraq reach an agreement on the use ,of Euphrates water was not fulfilled. The outcome of the Karakays project is rated satisfactory, whereas that . of the four other operations qualifies as unsatisfactory. Sustainability of the benefits from the projects is nevertheless considered to be likely except for the two 91bistan operations where it is uncertain. This dcuwmt has a restrfcted dfatrfbution and my be used by recpients only in the perforsence of their officilat duties. Its contents amy not otherwise be disctosed without Wortd Bank authorization. FOR OfTICIAL USE ONLY PERFORMANCE AUDIT REPORT TURREY ELISTAN LIGNITE MINE AND POWER STATION PROJECT (LOA 1023-TU) SECOND TEK TRANSMISSION PROJECT (LOAN 1194-TU) KARAAYA HYDROPOER PROJECT (LOAN 1844-TU) TEIMD TEK TEANSMISSION PROJECT (LOAN 2322-TU) ELISTAN OPERATION AND MAINTENANCE ASSISTANCE PROJECT <LOAN 2650-TU) TABLE OF CONTENTS Palt. No. Preface .....................................****.**........ 0...*. I Basic Data Sheets............... . . .... ................... iii Evaluation..................** * * * * .......... ***.. .................. xii I. BACKGROUND........................... 1 A. Development of the Economy.................... ............ 1 5. The Bank's Early Involvement in the Turkich Pouer Sector. 2 C. The Efergence of TSK.*................................... 2 II. T PE ROJECTS AND THEIR IMPLEMENTATION .......................... 3 A. The Projectos......... **........................... 3 5. The OrigIn of the Elbiatan and Xarakaya Projecta........ 4 C. Elbistan and TEK Tranmission 11....................... 4 D. Karakgya ................................................. 6 E. Third TEK Tranmission Project........................... 7 III. DROAD COUNTRY AND PROJECT FACTORS APECTING TaE OUTCOME OF THE PRJECTS..*................................ 8 A. Political and Economic Environment...........44*449 .... 8 5. The Approval Culture and the Pressure to Lend............ 9 C. Broad Agrement Betveen Bank and Turkey, Project Ownerehip 10 IV. PERORMANCE OF TRE ELBISTAN CMLEX.. ........................... 12 A. Power Plant Operation.................................... 12 B. Pover Plant Adjustments................................... 12 C* The Lignite Quality Is6ue................................. 13 D. The Elbistan Nin......................... 13 E. Bucket Wheel Exeavatore and Hard Layera in the Mine....... 14 7. The Groundwater Issue..................................... 15 0. The MinIng Operation...................................... 15 1. Hoot Urgent Improvement.................................. 17 1. Bank Involvement....................................... 19 This docunent has a restricted distribution and may be used by reciplents only In the performance of their offcial duties. Its contents may not otherwise be discosed without World Bank authoriation. TABLE OF CONTENTS (Cont'd) Palte 5o. V. PERFOmNANCE OFXÉRAYA......................................... 19 VI. ENVIRONMENTAL ISSUES............................................ 20 A. Ilbistan....................................... ..... 20 I . Karakaya .................................................. 22 VII. PROJECT COSTS ATD TEK INVESTMeT................................ 22 A. Introduction .............................................. 22 5. Bag coIts.*............................................. 22 C. Balancng TKMö Investment ................................ 25 VIII. PROJECT TIANA .............................................. 25 A. Elbistan .................................................. 25 B. TE I1 and T11......................................... 27 C. Karskaya ........... . ............................... 27 D. 2Cocuions............................. ...... 37 IX. TEK' PRTOS.X....................... 28 A. Power and E8ne.......................................... 31 1. Pm nel................... ....................... 29 C. Revenues and CTa pef................................... 29 D. Fiacs.........................29 x. CORCLUSIONS AMD LESSONS U[~ND........................... 30 A. Acco=pliohmete........................... 31 B. Drawbacks.................................. ......... 31 C. Bank Proace.....................31 D. Letinos................................................... 32 1. Bank Operations In the Power eco............34 2. Rwaenue and Costa per km8l..............35 I TURREY LBISTAN ILOAN 1023-TU) SECOND TEX TRANSMISSION PROJECT (LOAN 1194-TU) ZARAAXA ,DROPOWER PROJECT (LOAN 1844-TU) THIRD TER TRANSMISSION PROJECT (LOAN 2322-TU) AND ELUISTAN OPERATION & MAINTENANCE ASSISTANCE PROJECT WLOAN 2650-TU) PREACE 1. This report sets forth the results of a performance audit of five loans, three to TRK, the Government-cwned utility responsible for some 901 of the public power supply in Turkey, one to TEK and TKI, the State Coal Authority, and one to the Government of Turkey with D81, the State Hydraulic Works, as the Beneficiarys DSI is the Government-owned entity in charge, inter alia, of planning, design, and execution of hydroelectric developments which, after completion, it transfers to TEX for operation. The loans characterised in the table below were made and disbursed during the 16-year period 1974-1989. Amont Amount Sonth month Vanth loan Wabar Iggec Approved Disbursed Loan Loan Lam B Il l UUSu$ Aweroed ZffeStAn $19AAA 1023-J mibtstan 148.0 148.0 06174 06/76 06/83 1194-U II 56.0 55.4 11175 04178 12181 1844-W arakays 120.0 120.0 05/80 10/80 12/87 2322-W Wa III 163.0 97.8 06/83 11/83 09189 ista 12188 2650-U 0.& N. 10.0 9.9 02/86 05/86 Assistance 2. Except for Loan 1844-TU for larakaya, which was closed on the date anticipated in the loan documents, the actual closing dates of the operations were one to two years late with respect to those originally foreseen. As, in the case of 1bistan, large cost overruns led to the drawdown of the loan much in advance of construction progress, closing was only one year late though plant commissioning too': place seven years later than anticipated at appraisal. 3. The present Performance Audit Report (PAR) is based on four Project Completion Reports (PCRs) prepared by Country Department 1 of the Europe and Central Asia Regional Office and its predecessor. They have already been distributed to the Executive Directors. The PAR is based on a review of This column shows the abbreviated titles of the projects used throughout the present report. it pertinent Bank documents, especially the operational Ziles, and on interviews with Bank staff presently and formerly associated with the Bank's activities in the electricity sector of Turkey. A consultant contributed a desk analysis which is available in ORD's files and which the PAR uses extensively to support the findings concerning the broad issues such as those raised in the recent report of the Bank's Portfolio Management Task F3rce2. In the fall of 1992, another OD consultant visited Ankara and Elbistan. He thus had an opportunity to discuss power sector and project issues related to the audited projects vich Government, TZK, 7T's officials and the Bank's Resident Mission. 4. The audit found that, in general, the PCR candidly and adequately described the salient features of the projects and their execution as well as the main issues faced by the borrowers and the Bank during implementation. The PAR snalyses the background against which these operations, of which only that for the Karakay& bydroplant had a reasonably successful outcome, were prepared and carried out. It highlights the fact that most of the shortcomings of the audited projects had at their root the broad issues discussed in the Task Force report mentioned above. It also sets forth some findings, none of them new, concerning the measures needed in order to maimize the benefits to be reaptd from the Elbistan complex. 5. Although the audited projects touched some of the wider sector issues, in particular personnel policies, as well as commercial and financial management, only more recent operations such as the Energy Sector Adjustment Loan and the TEK Restructuring Project, which OED will review in the years to come, had these broad issues at their core. 6. Following standard OED procedures, copies of the draft PAR were sent to the Government and the Borrowers for comments. However, no comments were received. a Effective Implementations key to Development Impact, October 1992. narP0a CE AUDiT EPR TURMT ELISTAN LGUITE MINE AND POER STATION PROJECT (LOAN 1023-TU) ELBISTAN OPERATION & NAäINTEANCE AS$ISTANCE PROJECT (LOAN 2650-TU) B 0PAT S EEY PROJCT DATA Appraal Actuel or Actual as I lim Exetarlion Currenat E4t1~ate of AM. et. Total Project Coet (US$ mlian> 963.4 2485.6 2582 Loa = ount (US$ (Mulln 158 158 1002 Date Phycal Componte Compleed: 03/80 11/88 EcoaM Ite of eturf (M) 17.62 3.42 CURUAIM ESTMNTED A= ACTA DIEBURUEffNT Loam 1023 Z ZZ W. ZZZ ZZ Z2 E Zyfl FY83 Z Y84 Appra~eul Estimate <Us$ m11u~n> 13.6 81.9 111.3 128.6 139.1 148.0 148.0 148.0 148.0 A*tual (US$ mluo=) - - 55.3 81.8 90.4 112.8 127.3 145.4 148.0 Actual a % of Eaimte 50 64 65 76 86 98 100 Date of Final Diaburoemnts 09/83 lam-2650 Appraal E*timate (U$ 9±11~on> 2.0 6.2 10.0 Actual (Us$ mflin) 0 2.8 8.1 9.9 Actual ag 9 of Estimate 0 45 81 99 Date of 11na Disburaement: 03/89 ROETDAT93 nji Oriaaellan Agtga1 1981 ZiR 1981 2650 Identification 03/71 12/84 Appraleal 10/73 04/85 Neottation 05/74 11/85 Board Approval 06/74 02/86 Loa Signature 06/74 02/86 Losn Effectiveness 11/74 05/76 06/76 05/86 Loan Closng 06/82 06/88 06/83 12/88 Project Caupletion 12/81 12/87 11/88 11/88 STAT m PT Måtff-VMeka) mai w£ra ma Bli nu nu nu nu nu må nu må =i ro zou .3 5.2 1.0 19.3 5.4 31.0 h8@ttfane 3.7 Senewtiton .3 4.9 15.8 8.8 11.4 9.7 10.6 6.8 2 * 22.4 .3 93.8 1.2 .1 2.0 4.6 Total .5 '6.4 1.1 21.3 4'.0 4.9 15.8 8.8 11.4 9.7 10.6 6.8 2.7 22.4 .3 166.8 I M 2 6 5 0 z u M = m i = 2 expual Appra.al 2.3 2.3 Ne~gtiationa 3.4 3.4 3.7 4.7 9.4 10.2 13.0 41.0 5.3 5.3 Total 14.7 4.7 9.4 10.2 13.0 52.1 Mission DATA Staga of month No. of Day* in per.$ P--o1!§t C7CIG Inar- Perxa Fil aulna Identification 09173 4 22 Preparation 03/74 3 6 SUpwvison 1 09176 3 20 3 Supcvion II 10/77 1 9 SupGvison III 05178 3 10 3 Suporvision IV 05/79 5 10 3 Supervloton V 05/80 3 10 3 Spervision VI 09180 2 10 3 Superviion VII 02/81 2 11 3 Supovision VIII 06/82 4 11 3 Supevision ix 05/83 6 18 3 Supervision 2 07/83 4 12 3 Loe 2650 Identifiestion 10/83 2 10 Preparation 03/85 3 2 2 Supervilio I 06/86 2 23 1 Supe~vtsion II 03/87 2 26 1 SuperviLon III 06188 2 23 1 Supervision IV 12/88 2 20 1 * fectiv Ioplementations Key to Development lpct, October 1992. V OTHE PROJCT Borrover: Turkish Electricity Authority (TEK) Executing Agencys Turkish Electricity Authority (TER) Follow-up Projects TER Transmission IV (Loan 2856-TU) vi PERFORMANCE AUDIT REPORT TURKEY SECOND TEK TRANSMISSION PROJECT (LOAN 1194-TU) BASIC DATA SHET KEY PROJECT DATA Appraisal Actual or Actual as no Expectationg Currant Estimateg of Agyr. Est. Total Project Cost (US$ million) 145.7 175.0 '205 Loan Amounat (US$ million) 56.0 56.0 1002 Date Physical Components Completedt 06179 06/84 Economics Rate of Return (M) 16% 5% CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS Appraisal Estimate (US$ million) 4.8 20.4 41.7 56.0 56.0 56.0 56.0 56.0 Actual (US$ million) - - 2.0 21.1 39.6 48.6 55.4 55.4 Actual as 2 of Eatisate 0 0 5 38 71 87 99 99 Date of Final Disbursements 02/09/82 ligg OriyIal Plan eiged Actual. Identification 09/74 Appraisal 12/74 Negotiatione 05/75 Board Approval 11/75 loan Signature 06/76 Loan Effectiveness 09176 11/76 04/78 Loan Closing 12/79 12/80 12/81 Project Completion vit rna EDi må nuzxz Mi m m£ mg n zmn lma må nu aa PMaPPZ~aa 4.8 .0 .5 .15. Appr 4.8 .030.8 30.8 megouations 7.9 8.7 16.6 luperstatom 6.3 18.1 21.9 13.6 2.8 1.4 1.9 1.8 6.9 .2 76.9 Other 24.3 3.9 28.1 totaI 4.8 .0 63.3 19.0 18.1 21.9 13.6 2.8 1.4 1.9 1.8 4.9 .2 157.9 WISSION DATA Staga of Month No. of Day& ta Proict Cvelm Year Piald ideantificeation* Preparation Supeirvion 1 03176 2 3 Supevwison II 10176 3 9 Supeirvion III 07/77 2 3 Superwision IV 10/77 1 1 Superviaon V 06/79 3 1 Supervision VI 06/82 1 2 OTER PROJECT DATA Borroer: Turkish Electricity Authority (TEU) Executing Agency: Turkish Electricity Authority (T=K) FolU<>r-up Project: TEK Trandmission Project IV (Loa 2586-TU) L lo ideutification or preparation missions occured. ooi~toSfT2ro uivo 08/80 ----a--- *Swl03 uso 08/50 ---- ---~aey uso 6L1TT tYWaddy oovva1 ZOEMt T6m 30g«alm 00*001 00*501 00*911 00*CT 00'19 00*6 00° --- 3 30 z W mn o0*ogt oo*ozi 09*911 o9*Z o£*0 ov*c oiloz --- (U*~~ $M)a e~~VV ooo1 00*911 00*zol 00*69 00*VL 00'Ls 00,6t oo*t (u@Ttw $sa) 0~o33 T~WTe~dd wu Tmu Tm nua Tm tm Tm T 1TWERmisla Tiev mv cuv =2 UK ~ ~ ~ ~ ~9 *naW 3 2~ uu~ wo00 00*0*1 00*01 (U0lT $En) 0WnOu U, Ze$ "*«IT coglot <UMTW~ z~l)~ 200~* ~o T5~Tu¶-55 ssuu -55 55555 155TN555U!I WiT Z SW tnay zo zmnzy twr~dy (U-vv8 om> w'amlar0 IMOdU0a vivnvxVI ~ff~iU5r£= H IX (Staff-Weeks) m n2 m n m u n£ E. nu nu nu nu i nu Ma.z na nut ffi l0s1 Prappraial 1.9 2.1 25.6 7.0 2.3 Appisal 54.9 .6 .8 27.9 84.2 x*otlations 5.* 16.2 9.3 29.3 Supeistoa . .9 17.4 7.7 3.3 .5 3.4 2.5 $.7 6.9 5.5 5.3 70.1 14.6 7.1 .2 14.7 .0 .4 44.0 Total 1.9 3.4 31.4 12.6 23.9 2.7 55.2 17.4 7.7 3.3 8.5 3.4 2.5 8.7 7.3 5.5 5.3 2- 4 MISSIONA Stage of Month No. of Days in Performance Proiect C~0cle Xer Per*ne Fild Ratinh Tdentlfication 10/75 1 13 2 Preparation 11/76 3 34 Supervision 1 09/80 3 8 2 Supervision II 02/81 1 12 3 Supervision III 12/81 1 6 3 Supervision IV 06/82 1 10 3 Supervision V 09/82 2 30 3 Supervision VI 06/83 2 17 2 Supervision VII 10/83 2 7 2 Superv"Gion VIII 03/84 2 13 2 Supervision IX 10/84 2 22 1 Supervision X 04/85 2 20 2 Supervision XI 10/85 2 24 2 Supervision XII 03/86 2 18 2 Supervision XIII 06/86 2 23 1 Supervision XIV 10/86 4 31 1 Supervision XV 03/87 2 25 1 Supervision XVI 06/88 3 23 1 Supervision XVII 12/88 2 29 1 OTMER PROJECT DATA Borroer: Turkish Electricity Authority (TEK) Executing ASency: Turk:Ls Blectricity Authority (TEK) Follow-up Project: TEK Transmisaion IV (Lo~n 2856-TU) 1. Problem Free; 2. inor Problem; 3. Major Problems. x PERPOMAMNCB AUDIT REPORT TURm THIRD TEK TRANSMISSION PROJECT (LOAN 2322-TU) BASIC DATA SHEST REY PROJECT DATA Appraital Actual or Actual as 2 rx vOcitons Cret setimatee of Appr. Est. Total Project Cost (WS$ million) 259.1 156.9 602 Loan Amount (US$ million) 163.0 97.8 542 Date Physical Components Completeds 1986 09/1989 Economice Rate of Return (2) 8.4 3.5 cOMuLATIVa ESTIMTED AND ACTUAL DISBURSEMENTS VY33 M4FY8 FY86 FYS7 PYS8 FY89 FY90 Appraisal Estiate (WUS$ million) 1.0 24.5 105.7 158.1 163.0 Actual (US$ million) 0.4 17.4 20.2 32.6 73.3 81.4 92.4 97.8 Actual as 2 of Estimate 402 712 202 212 452 502 572 602 Date of Fisal Diabursemeuts 01/30/90 PROJECT DATES JIM es"I a Plan AsMsal Identification 1981 Appraisal 01-31/02-18/82 Negotiationas 05183 Board Approval 06/83 06-23/83 Loan Signature. 06-27/83 Loan Effectiveness 09/83 12-01/83 Loan Closing 12/87 09-30/89 Project Completion 12/86 09-30/89 xi STAFF INPUTS (Staff-Weeks) fI11 Y$2 I3 MII4 IMA II UlZ FY 8 I=FY89 IIt XMI, Preappraisal .6 11.5 12.1 Appraisal 41.3 41.3 Negotiations 8.8 8.8 Supervision .2 25.9 15.0 30.9 19.8 12.3 7.2 8.7 120.0 Other .1 1.9 8.0 Total 0.7 1.9 69.8 25.9 19.0 30.9 19.8 12.3 7.2 8.7 192.2 MISSION DATA Stage of Month No. of Days in Performance Project Cycle Year Persons Field Rating' Identification 01/83 9 19 Preparation 07/83 3 14 Supervision I 03/84 3 10 1 Supervision II 09/84 1 1 Supervision III 10/84 3 10 Supervision IV 10/85 2 11 2 Supervision V 03/86 4 3 2 Supervision VI 06/86 2 2 Supervision VII 10/86 4 33 Supervision VIII 03/87 2 27 Supervision II 12/87 4 13 Supervision X 06/88 3 25 Supervision II 11/88 3 20 Supervision XII 09/89 3 16 OTHER PROJECT DATA Borrowert Turkish Electricity Authority (TEK) Executing Agency: Turkish Electricity Authority Follow-up Project: TER Transmission IV (Loan 2586-TU) 6 1. Problem Free; 2. Minor Problems; 3. Major Problems. xit PERPORMANCE AUDIT REPORT TURKEY ELBISTAN (LOAN 1023-TU) SECOND TER TRANSMISSION PROJECT (LOAN 1194-TU) KARAKAYA HYDROPOWER PROJECT (LOAN 1844-TU) THIRD TEK TRANSMISSION PROJECT (LOAN 2322-TU) AND ELBISTAN OPERATION & NAINTENANCE ASSISTANCE PROJECT (LOAN 2650-TV) EVALUATION SWARY Introducton - te Karakava &Arower Project approved iii 1980 and supported by 1. The operations audited are: Loan 1844-TU of US$120 millions closed on schedule in December - the Elbltan Project approved in 1987 after full disbursements the 1974 and supported by Loan 1023- Government vas the Borrowerg the TU of US$148 million, closed in state Hydraulic Works (DSI) the June 1983 (i.e. 12 months late) executing agency, and TEK the after full disbursements the ultimate owner and operator; the Turkish Electricity Authority project included the 1200 MV (TEK) and the Turkish Coal Karakaya hydropower plant using Authority (TRI) were the the head created on the Euphrates borrowers and the executiving River by a 173m high concrete agencies for the power plant and arch dam 160 Um downstream from the mine, respectively; the the Keban reservoir in Eastern project includeds (i) a 4 x 300 Turkey; MW (later uprated to 4 x 344 MW) minemouth, lignite fired steam - the Third Transmission power plant in Central Turkey; ftoiect approved in 1983 and (ii) 340 km of 380 kV supported by Loan 2322-TU of transmission lines; and (iii) an US$163 million to TK9 closed in opencast lignite mine supplying September 1989 (i.e. 22 months up to 18 million tons of lignite late) after cancellation of per year to the power plant. US$65.2 million; the project included 1500 kou of 380 kV - the Second TEK Transmison transmission lines and a large Project approved in 1975 and technical assistance component, supported by Loan 1194-TU of inter aliag to support UK when US$56 million to TEK, closed in it incorporated the large December 1981 (i.e. 24 months municipal distribution entities; late), atter cancellation of and US$0.6 millions the project included: (1) 755 km of 380 kV the 21bistan Overation and transmission lines and associated Maintenance Assistance Proiect switching facilities, and Cii) approved in 1986 and supported by various studies and training; Loan 2602-TU of US$10 million to UK, closed in December 1988 (i.e. 6 months late) after xift cancellation of US$0.1 million; Rarakaya, and TE I11). the project included the completion of units 3 and 4 of Inlawntation Ezverience the Elbistan power plant and improvements in TEK's operation 4. Commissioning of Elbist vae and maintenance capability at the seven years late and the US$2.5 plant. billion project cost resulted from overruns in real terms of 1002, 752, Obiectives and Settina and 100%, respectively, for the power plant, the mine, and the transmission 2. Inteneive dialogue about the Bank components. The disappointing projects for Elbistan and Karakaya outcome was due to repeated shortages started in the early 1970s in the of funds and poor project management, wake of a major Bank-sponsored the latter, inter alias a consequence restructuring of the power sector of Goverment personnel policies including the creation of TEK. The (especially on salaries) impeding the two oil price crises in 1973 and hiring of adequately qualified staff 1979, and the politically unsettled in sufficient numbers. The plant periods in the mid-1970. and at the will achieve its full capacity in start of the 1980s, as well as 1994 when all the turbo generators associated shifts in economic will have been modified to avoid the policies affected the projects in a recurrence of technical difficulties major way. In particular, the experienced for the first time liquidity crisis of the second half shortly after commissioning (PAR, of the 1970s slowed down the para. 40). The Un of the project Implementation of the first three was estimated at 32 after completion, projects reviewed. In the first half down from 182 at appraisal. of the 1980s, the Bank/Fund - supported Structural Adjustment 5. The TEK 11 project wa Programs, while still restricting completed at about the cost (in real availability of Government funding, terms) expected at appraisal, allowed the sector to increase sub- although poor project management and stantially internal cash generation. shortages of funds also plagued it However, the politically motivated and resulted in a delays of some 36 relaxation of fiscal discipline in months for the main facilities. (PAR, the second half of the 1980s pushed paras. 71 and 72; PCR TEK 119 Annex the sector (especially TK) into 2). The IE resulted at 52 after technical insolvency from which it completion, down from 162 at appears to emerge. (PAR para. 1). appraisal (PAR, para. 16). 3. The five audited projects aimed 6. The Rarakava Project was at: (i) increasing TEK's generating successfully completed, albeit 26 capacity using non-exportable montas late and at a cost 15Z higher resources (lignite and water power) in real terms than estimated at and strengthening the utility's appraisal. These deviations were transmission capability, and (ii) mainly due to the renegotiation of institutional Improvement in the the civil works contract, temporary sector, especially TEK, in terms of lack of funds, heavy rainfall. in personnel policies, financial 1983, and unexpectedly large performance and management, project stabilization work at the dam management capability, training, abutments. (PAR paras. 71 and 72, etc. (pares. 3.01 of PCR9 E4bistan, PCoI, Karakaya, para. 5.03). xtv Reflecting the relative success of heavy ice loads that appeared in some the project, the IERR estimated at stretches particularly exposed to 15% at appraisal, was still at about cold and humid winds (PCR TEK 11 122 after completion. para. 4.1). 7. The Third TER Project also 10. In connection with the suffered from poor project management Implementation of Karakayaq the Bank which, together with protracted had agreed with Turkey that TEK vould procurement and shortages of funds, operate Keban and Karakaya in such a delayed completion by 34 months. way as to guarantee a monthly average However, costs were 40% lower in real flow in the Euphrates exceeding 500 terms than estimated, mostly because me at the border with Syria. With intense competition reduced bid minimal exceptions, Turkey complied prices, TEK used in part material with this rule as long as only Keban already in store and the original and Rarakeya were affecting the above estimate erroneously included flow, i.e. until 1991, when TEK financial costs not eliminated from filled the reservoir of the AtatUrk earlier bid prices utilised for complex downstream from Karakaya. appraisal (PAR, para. 74). The The Bank had expected that by ti estimated IER fell from 8% at time Turkey, Syria, and Iraq would appraisal to 4% after completion. have concluded an agreement on the use of the Euphrates flow replacing Proiect Results the above rule and taking into account that AtatOrks in contrast to 8. The facilities built under the Karakayap involves abstractions from TEK II, Karakaya, and TEK III the river, making major infringements Projects are operating on the above rule far more likely. satisfactorily. However, the No such agreement materialized, Elbistan complex, besides not yet leaving the Bank in an uncomfortable providing the planned capacity, is position QCR Karakaya, para. 6.04, not operating properly, mainly PAR, paras. 60 and 61). because the mine supplies fuel of a quality, though adequate in average, 11. 1niroMental issues appeared is subject to variations far in only in the context of the two power excess of those admitted by the plant projects. At Karakaya, the specifications. It is likely that enviroamental measures agreed with the problem can be solved by better the Bank were properly implemented mixing the material in the mine. and the resettlement and compensation However, TKI, the present operator of of 17,000 people successfully the mine, has little incentive to do completed with the settlement, in so, as the required adjustments would 1991, of the last claims for increase the already large losses it compensation (PCR Karakaya, para. incurs in its present operation of 6.03 and PAR, para. 69). in the mine (PAR, paras. 40 to 42 and 51 Elbistang construction proceeded in to 56). accordance with the agreements with the Bank. The stipulated pollution 9. Operation of Karakaya and of control equipment was installed and the transmission facilities is likely to operate in accordance constructed under the TEK 11 and III with the specifications soon. Under projects is satisfactory after one normal circumstances, plant operation line along the Black Sea was seems to comply with Turkey s and the reinforced to resist the unexpected Bank's environmental standards at the XV time the project started. However, 1985. But when, in the second half when the fuel does not meet the of the 1980., these revenus specifications, in particular when declined, TEX again va. In default. the ash contains too much calcium Indeed, In 1989 it voa technically oxide, the risk that the fly-ash insolvent. (PAR, par. 88 and 89). imissions are too high in the neighborhood of the plant io 14. In the 1970., accounts substantial. In 1992, monitoring of receivable were consistently far in the environmental impact in the plant exceos of the three month.' billia area at large was insufficient as (to bulk client, at the time) sot as part of the equipment foreseen and the lower limit of an acceptable available for a time after performance. Though they improved in commissioning was removed. There is the 1980., when the distribution a need for the Bank to follow up on utilities were integrated Into TEK* the actual levels of pollution and they remained between 4 and 5 month.t the availability of the pollution billings (now to ultimate consumers) monitoring equipment (PCR Elbistan, In the late 1980., I.e. still In para. 6.05 and PAR paras. 65)o excess of the threshold which remained at 3 month. (PAR, pars.. 89). 12. The institutional objectives of the projects were not achieved. 15. Sinc the requirement of a Thus, in the early 1990s, TEK was current ratio in excess of 1.0 ia. still highly centralized, its projoct introduced in the context of the UK management capability inadequate, III project In 1984, the utility was internal coordination poor, only able to mt it for a few month. bureaucratic procedures (including In 1984 (PAR par". 89 and 90). procurement) cumbersome, autonomy in personnel and financial matters SuiaMiiy lacking, and power system planning unsatisfactory. All these issues 16. The sustainability of benefits were addressed by one or the other of from the Karskaya, Second and Third the projects reviewed and often taken UK projects seems likely, though not up again by follow-on projects in optimal due to TZK. Inatitutional particular the 1991 TEK Restructuring veakne..e.. it should be Project (see e.g. PCR Karakaya, para. substantially enhanced, if the 6.01, PCR TEK II, Chapter VI). ongoing UK Restructuring Project is reasonably successful. In contrast, TEK's Financial Performance sustainability of the benefits from the 9lbistan project remaine 13. In the period 1975-80 TEK never uncertain as long a the facilities achieved the 61 (for 1976) and 82 do not operate In accordance with the rate of return agreed with the Bank, specifications, which require a fuel as the actual figures varied between of more regular quality. This in 2% and 4% (PCR TEK III, Annex 7). In turns is achievable, but cals inter the 1980s, when a contribution to alias for proper pricing of the investment covenant with a 202 lignite. threshold replaced the rate of return covenant, TE was able to meet the ConclusIo and Lesson requirement between 1982 and 1986, when revenues per kWh were increasing 17. Elbistan and Karokays Increaed fast in real terms and reached some Turkey's generation capacity by some 80o of long run marginal cost in 2300 W (plannd to increase to 2500 xvi MW in 1994), 1.e. about 152 of the Portfolio Management Task Force under country's 1990 installed capacity. the need for (i) *project ownership* The projects further included abeut by Government, Borrowers, and one third of the 8500 km of 380 kV Beneficiaries, and (U) realism and transmission lines operating in 1990. candidness in the Banks evaluations The generation facilities achieved and reportings one of the Government's main objectives, i.e. to increase the use - "Ownership" by the country of domestic non-exportable resources authorities has to cover (i) the for power generation. physical component, its execution and operation and (Ui) the 18. Once the Elbistan reaches its Institutional objectives and the full capacity and the fuel supply way to reach them. allows the facilities to work in accordance with the specifications, - Deeply held conviction. such as the plant should become one of the those reflected in Turkeyos main pillars of power generation in extreme centralism and the Turkey, the more so as it Is situated refusal to relinquish, for the In an environmentally less sensitive sake of efficiency, a reasonable area than other lignite-fired plants veasure of authority to executing located further vest; whose future is organizations (the Government to questionable (PAR, para. 93). the utility, the utility to site managemont and consultants, etc.) 19. The Institutional Improvements cannot be eliminated by a in the power sector at large and in contract between local TK, in particular, which the Bank authorities and the Bank. Thus, (though not Turkey) considered a the Bank has to take such major part of the projects have been convictions into account by achieved only to a minimal degree, stepping around them. At the though they ,were taken over in an same tine it should foresee small adjusted form from project to steps providing pressure towards project, even after those audited a change in the efficiency here. inhibiting attitudes. 20. The PCRs and the Audit concur Institutional reform cannot be In rating the institutional progress divorced from the political and achieved negligible. The PCRs do not economic environment. Therefores explicitly rate the overall outcome the Bank, In an environment of the projects but the judgements hostile to some of its broad made in these documents suggest no sector policies should either not contradiction with the conclusions of lend, or# continuing the the Audit, i.e satisfactory for dialogue, frame the project in Karakaya and unsatisfactory for such a way that whatever Elbistan, TEK II, and TER III. (PAR, Institutional progress it par. 97). envisages is realistically achievable and represents a Lessons significant step towards its long term goals (see PCR TEK 119 para. 21. The main lessons emerging from 8.4). the operations discussed here are closely related to the issues covered - In a project like Elbistan, where In the recent report of the Bank's sustainability of benefits is xvil achievable but still in doubt at project completion, the Bank could consider a close follow-up of the project, in particular when the Bank's and ita colenders performance Is under criticism. A reasonably detailed plan for project management to which the Borrower is fully committed should be an integral part of a Bank project involving major physical facilities. Supervision should always include interviews with people directly involved in the projects. Frequent site visits of complex and/or difficult projects with the corresponding interviews of site management, consultants, and operating staff are a must, as they often provide a valuable perspective, entirely different from that set forth at the higher levels of Government and the utility. PERfORMANCE AUDIT REPORT TURKEY ELBISTAN LIGNITE MINE AND POWER STATION PROJECT (LOAN 1023-TU) SECOND TEX TRANSMISSION PROJECT (LOAN 1194-TU) KARAKAYA HYDROPOWER PROJECT (LOAN 1844-TU) THIRD TEK TRANSMISSION PROJECT (LOAN 2322-TU) AND ELBISTAN OPERATION & MAINTENANCE ASSISTANCE PROJECT (LOAN 2650-TU) 1. BACKGROUND A. Development of the Economy 1. Over the period 1963-72, i.e just before the main projects reviewed here (Elbistan and Karakaya)' started, the Turkish economy grew at a high 6.6% per year. Over the next five years growth reached an average 7.7% per annum despite the first dramatic increase of world market oil prices in 1973. However, the country achieved this fast growth by incurring a large amount of debt. In the politically unstable mid-1970s, high interest rates aggravated by weakening exports and lower remittances of Turkish citizens abroad as well as Government's failure to adjust to the changed economic environment led the country into a severe foreign exchange crisis. Trom 1977, the Government had to implement, in cooperation with the International Monetary Fund (IM) and the Bank, a series of stabilization programs focussing on import liberalization, investment rationalization, public enterprise reform, and an enhanced role for the private sector. In 1980, a change in political regime took place and in the period 1980- 84, the Bank supported the implementation and continuation of the above programs with five Structural Adjustment Loans. The operations were at least in part successful.2' Inflation which in the second half of the 1970s had grown from 20% to over 100% per year, was reduced to 20 - 50% per annum in the period 1980-87. However, in the late 1980s, a relaxation of fiscal discipline, coincident with successive elections at local, regional, and national level, led to a new eruption of inflation with annual rates of 60 to 752 over the period 1988-91. 2. In .e energy sector, an aggressive pricing reform dominated the first half of the 1980s for the Turkish Electricity Authority (TEK), the Government-owned utility providing some 902 of the public power supply to the ultimate consumer in Turkey. Indeed, TEK's tariffs increased By more than 60% in real terms. In the second half of the decade this commitment to improved electricity tariffs weakened. Over the period 1986-89, power sector revenues and in particular those of TEK again were inadequate as the adjustments Government permitted did not even compensate for inflation let alone allow TEK's tariffs to catch up with the The Table in para. 1 of the preface sets forth the names used in this report to designate the projects reviewed. 2 See Evaluation of Structural Adjustment Lending in Turkey, Program Performance Audit Report of the Fourth and Fifth Structural Adjustment Loans (Loans 2321-TUR and 2441-TUR), and Overview of SALe I - V, OED Report No. 7502 dated April 26, 1988. 2 economic cost of supply. In 1990 there was a turn for the better as tariffs stopped to fall in real terms. Most recently they even started to increase again (see Graph in Attachment 2). B. The Bank's Early Involvement in the Turkish Power Sector 3. The Bank's involvement in the Turkish power sector dates back to the early 1950s, when Loan 73-TU3 helped finance the Seyban multipurpose project in the southeastern part of the country. In 1953, the gukurova Elektrik A.S. (gEAS), a mixed-economy corporation with minority Government participation, formed to operate the power component (36MW hydro) of the Seyhan project, was granted a concession for supplying bulk electricity to distributors in the Adana area of South-East Turkey. The next two Bank Group operations (Credit 36-TU in 1963 and Credit 69-TU in 1968) were with 9EAS as the Beneficiary and helped fund extensions of the company's hydro facilities. Two further loans, in 1969 and 1971, again supported 9EAS, financing further extensions of generating capacity, associated transmission facilities, and cost overruns from earlier projects. 4. In parallel with the two operations just mentioned, which for fifteen years were to remain the last ones with gEAS, the Bank started its support of the public enterprises in the sector with ETIBANK's Keban Transmission Project in 1968, followed by the Power Transmission Project in 1971 for which the then newly created TEK, that succeeded ETIBANK, was the Borrower. A 1973 loan for which the Istanbul municipal utility was the Borrower, and which focussed on distribution, completed the set of Bank Group operations which preceded the projects audited here. 5. The Bank's shift towards lending to the Government-owned utilities reflected the desire to get involved more centrally in Turkey's power sector development, which would have been difficult had the Bank confined itself to supporting qEAS as it had done earlier. C. The Emergence of TER 6. Until 1970, four state organizations (inter alia ETIBANK), some six private and mixed-economy utilities (e.g. 99AS), a large number of industrial firms and more than 600 municipalities and villages were involved in the power sector. Therefore, as a prelude to the above shift of emphasis in its lending to the sector, the Bank, in 1967, provided Government with a US$1.95 million grant to help finance: (i) studies of sector reorganization, and (ii) implementation of the relevant recommendations. In 1970, as a result of these studies, Government established TEK as a State Economic Enterprise (SEE) mainly responsible for planning, designing, constructing, and operating generation, transmission and distribution facilities except for the installations still to be owned by the private and mixed-economy utilities such as 9EAS and for the distribution systems of major municipalities. Further, planning, design, and construction of hydroplants was to remain with DSI (State Hydraulic Works), which would transfer the completed facilities to TEK for their operation. 3 Attachment 1 sets forth a full list of Bank Group operations in the power sector with the amounts and critical dates. 3 II. THE PROJECTS AND THEIR IMPLEMENTATION A. The roiects 7. The five audited operations covered: - Loan 1023-TU to TKI and TER for the Elbistan mine and 4 x 344 MW (originally 4 x 300 MW) mine-mouth, lignite-fired steam electric station implemented and later operated by TKI and TEK, respectively; and 540 km of 380 kV transmission lines and associated switching facilities; - Loan 1194-TU to TEK for the TER II project including 755 km of 380kV transmission line and associated switching facilities; - Loan 1844-TU to Government for the 6 x 300 MW Karakaya hydroelectric facility to be carried out by DSI and, after completion, to be taken over by TEK for operations - Loan 2322-TU to TEK for the TEK III project including 1500 km of 380 kV power transmission facilities permitting inter alia the evacuation of the energy produced at Karakaya; and - Loan 2650-TU to TEK for assistance in project completion, early operation, and the setting up of maintenance at the Elbistan station. 8. Bilateral contributions to the Elistan project which ultimately cost some US$2.5 billion (about 2.6 times the amount originally estimated) came from Germany (mostly Kreditanstalt fUr Wiedersufbau - KfW - and the European Investment Bank - EIB-). In the US$ 1.1 billion Karakaya project, a group of Swiss banks, RIB, and Italy were the main foreign financiers besides the Bank. 9. Whereas the TER II project was the subject of a Project Completion Report (PCR) that ORD processed in 1984 (OED Report No. 5304), the other four operations were reviewed in three PCRe written in 1990 in the Energy Operations Division of Country Department I in the then Europe, Middle East, and North Africa Regional Office. The PCRe were prepared by current and retired Bank staff who, for the amet part, had participated actively in the supervision of the loans discussed here and in the appraisal and supervision of several subsequent operations in the energy and power sector (see Attachment 1). However, no one among the original appraisal team for the Elbistan and Karakays loans is on the Bank staff today, nor is anyone who participated in the earliest supervision activities. Further, the Bank's files do not contain any useful material dated before 1975, even though the earliest activities on these projects go back to the late 1960s. As set forth in the PCRs, except for Elbistan, the physical project objectives of these operations were achieved, albeit at some considerable delay and, in the case of Elbitan, with huge cost overruns. The PCRs conclude that, in contrast, the other purpose of the reviewed operations - to foster institutional improvements - was not only not achieved but that progress towards such goals was extremely disappointing, indeed. The following chapters will confirm this performance rating and elaborate on some of its aspects. 4 10. The Elbistan project and, toa lesser degree, the other operations reviewed here, exemplify several of the shortcomings the recent report by the Bank's Portfolio Management Task Force' identified as major factors in the unsatisfactory performance of many Bank projects. Further chapters of the present report analyse some of these factors. 11. The audit finds it regrettable that the PCR for the Elbistan project became available only seven years after the Bank closed the loan and some five years after substantial completion of the project. Successful and unsuccessful projects are equally valuable when it comes to learning from experience. In the case of an operation with an outcome as poor as that in Elbistan exposing the Bank and its co-financiers to widespread criticism, the audit considers it highly desirable that the facts be set out in a PCR as early as possible after the closing date even if, as in this case, a closely related operation was still ongoing. B. The Origin of the Elbstan ad X#rakaya Projects 12. In 1960, a systematic geological survey identified a large deposit of minable, low-grade lignite in the Elbistan-Afein area of Central Turkey. In subsequent years, with the assistance of German financing and expertise, a project for a mine fuelling a mine-mouth thermal power plant was developed. At about the same time, DSI advanced the Xarakaya project for a large hydro power plant which was to use the water of the Euphrates River, as regulated by the Keban reservoir, over the head created by a 173m high concrete arch dam permitting to store 5.6 billion m3. Turkey asked the Bank to help finance both projects, which, in 1973, after the first steep increace of oil prices, had acquire- a high profile as they both were to use non-exportable resources. Preparation of the two projects thus proceeded at the same time in an environment unsettled by a major restructuring of the power sector and the tentativeness associated with the initial operation in a new setup. The Bank documents at the time did not identify this as the major risk it proved to be. C. Elbistan and TEK Transmission II 13. It appears that the Bank, though it followed with interest the development of the Elbistan project, was not actively involved in its genesis. It assumed a high profile only when financing for all contracts except that for the civil works had already been arranged. Therefore, there was pressure to proceed quickly. Though Turkey was in default of several covenants agreed under previous Bank loans to the power sector, the Board approved the Elbistan project in June 1974, before Turkey had eliminated the defaults; this elimination was made a condition of effectiveness of the loan. Not only did the Bank not insist on compliance before it completed the processing of the operation, but it was preparing another project, larakaya, and, when this operation ran into major obstacles (see para. 17) it quickly introduced a new operation (TEK II) into the lending program and two months later, in September 1974, appraised it. The elimination of the above default also became a condition of presentation to the Board of the transmission loan. 4 "Effective Implementation: Key to Development Impact", October 1992. 5 14. A new government took a particularly hard stand towards the Bank and refused to take any measure, even if agreed with the Bank, that the public might see as contributing to inflation. The Bank, in turn, refused to act on the three power loans that were either in the pipeline or waiting to become effective. This led to an acrimonious press campaign against the Banks it also taxed the patience of the Bank*s co-lenders for the Elbietan project as they could not start disbursing although substantial costs that they were supposed to finance were accruings. Finally, Turkey made some concessions, though, the main Bank condition, a tariff increase that would have permitted TEK to earn an 8% return in 1976 was never entirely fulfilled. Following this, the Bank approved the transmission loan in November 1975 and made the Elbistan loan effective in June 1976, two years after project approval. It then took 28 months from project approval to make the transmission loan effective, although the loan documents did not set any special condition of effectiveness. Indeed, declaring the loan effective earlier would have resulted in immediate default because the downturn in the Turkish economy that took place in the mid-1970s (see para. 1) had led TEK into financial straights. Only when the Bank agreed to weakening TEK's revenue covenants was it in a position to make the TEK II loan effective without the utility immediately defaulting. 15. Implementation of both projects was exceedingly slow before the loans became effective and remained so later, delaying commissioning of the main facilities by four (TEK II) and seven (Elbistan) years. The main reasons common to both operations were perennial shortages of funds reflecting the country's economic difficulties, and poor project management (discussed further in Chapter VII). Elbistan suffered from further major problems, such as the steep increase in estimated cost, which amounted to 55Z in US$ by the end of 1975 and was followed by other upward adjustments. In 1976, the cumulation of difficulties led to a virtual standstill in project implementation. Therefore, in a meeting of the co-lenders that took place in Luxembourg in January 1977, 8IB, KfW, and the Bank agreed to threaten suspending their respective loans to Turkey's power sector. As a consequence, what had been a complex organizational, managerial, and technical problem became also a political one. The controversy escalated to a point where Government approached the President of the Bank stating that the many meetings (on Elbistan) between the Turkish authorities and the Bank were leading to a loss of time and requested the President "to use his personal influence in discouraging the insistence on details so that there will be no delay in the execution of this project." The "details" to which the Prime Minister referred were inter alia Turkey's compliance with major covenants of the power loans and TEK's commitment to set up an appropriate project management! Despite the drama, construction performance did not improve, and by 1979, i.e. five years after the project had started, it was only about 15% completed. 16. In 1980, the renewed threat by the co-lenders to suspend disbursements from their loans seems to have been instrumental in getting from the Turkey the promise to set up a strong site management under expatriate leadership. However, it is evident that the two sides understood this agreement to mean two quite different things: Though the presence of high level expatriate consultants A particularly sharp rejoinder by the German Ministry of Economic Cooperation is dated August 25, 1975. 6 increased, the Turkish authorities never delegated to site management more than a fraction of the authority that would be considered a normal requirement in most other countries. Progress ultimately improved somewhat, and TEK was able to commission the first unit in 1984 and the last in 1988, 75 and 104 monthe later than estimated at appraisall its utilization remaining highly unsatisfactory. That many of the difficulties experienced in Elbistan were characteristic of Turkey rather than the project is also suggested by the collapse of project management in the case of the rather simple 1EK I project (para. 82). The IEERRs for the Elbistan and TER II projects were estimated after completion at 32 and 52, respectively, well below the 18% and 16% envisaged at appraisal. Unfortunately, as further discussed in Chapter IV, operation of the complex was also ill-starred. The 1986 Elbistan Operation and Maintenance Assistance project tried -- with limited success (see Chapter IV) -- to address the plant's operational pr>blems. It was supported by Loan 2650-TU of US$10 million, a nominal amount in view of the project's financial requirements even at that late time. But it allowed the Bank to continue its supervision of the project, a declared objective of the complementary operation (SAR, para. 4.01). D. Earakayg 17. As mentioned earlier, in the mid-1970s, when it was processing the Elbistan loan, the Bank was moving forward on still another large project, the Karakays Hydroelectric development on the Euphrates for which it had received the 1970 feasibility study some 18 months before the appraisal of the Elbistan project. In the previous decade the Bank had refused to help finance the Keban hydroelectric project upstream from Karakaya in part because of the concern over the management of the water of the Euphrates which after leaving Turkey flows through Syria and Iraq. Yet, in June 1974, the Bank sent an appraisal mission, which was inconclusive as project information proved to be lacking. Subsequently, in parallel with the further preparation of the project proper, discussions of the international issue was taken up at the level of the Bank's President and management on the one hand and the responsible ministers of the three countries involved on the other. Turkey's position was all along thats (i) the project did not imply any water abstractions (except for modest increases in evaporation losses), and (ii) the benefits of the Karakays. project for Syria and Iraq in the form of a better regulation of river flows were substantial, and (III) far from objecting to its construction, the two countries should be willing to help finance itt The Bank was trying to advance the technical basis for the international negotiations by developing at the request of the three countries a computer model for the Euphrates flows6. Nevertheless, it remained wary of getting involved into international politics and suggested (through the President) that, in the absence of a tripartite agreement, Turkey might wish to finance Karakaya from sources other than the Bank and submit other projects for the Bank"s consideration, a solution the Turkish Governmsnt considered unsatisfactory. 6 It appears that whereas Syria and Iraq accepted the validity of the model, Turkey did not wish to use it. 7 18. In late 1976 the Bank appraised the project and, in early 1977, for the first time clearly outlined a solution for the water management issue that it would be ready to accept as a basis for lendings (i) Turkey and the Bank would agree on criteria for operating Karakayal (ii) Turkey would advise Syria and Iraq of the criteria and contract with the Bank to meet them, and (iii) Turkey would undertake to arrange tripartite studies of the water management issue for the long range. The thrust of all this was to lay the foundation for a Bank loan in the absence of a tripartite agreement. 19. In subsequent years, Turkey's economy was in crisis (see para. 1) and while implementation of the Karakays project had started, the approval of the Bank loan was held hostage inter alia to the conclusion of a Standby Agreement between Turkey and the IW. At the same time, the Bank, through its President and through missions to Syria and Iraq, tried -- unsuccessfully -- to convince these countries to negotiate a settlement with Turkey. In November 1979, the Bank reappraised the project, now envisaging a joint operation with RIB to finance the only major project component not yet funded, the civil works contract. This contract had been under implementation since 1976 and the subject of a major dispute between DSI and the foreign consortium that had won the award. At the time of the re-appraisal this issue seemed to be settled, but in fact, it took another two years until physical progress resumed under reasonably normal conditions. Also in connection with the re-appraisal, the Bank accepted Turkey's guaranty that it would manage the reservoirs on the Euphrates river in such a way as to provide a flow at the border with Syria that would exceed 500 m3/s in average over any single month ("rule of 500") . This rule was confirmed by an official policy statement endorsed by the Turkish Parliament. After loan approval in May 1980, both Syria and Iraq protested the Bank's decision on water management, which a 1982 Bank mission re-examined and found still appropriate. 20. On the technical side, until 1981, progress remained unsatisfactory. However, when the dispute with the contractor was finally settled and DSI had a reasonably efficient project management on site (see PAR para. 83 and SAR Karakaya paras 5.01 to 5.03), progress became satisfactory. The facilities were commissioned in 1989, some 27 months later than expected at appraisal. The IERR estimated after completion at 12% (down from 151 at appraisal) reflects the relative success of the project. 21. Although in the mid-1980s, Government and TER were again in default of several major covenants, the Bank was preparing a supplemental loan of some US$100 million to support the completion of larakaya. It then expected the Turkish side to agree to prepare a new simulation study of the river flows in the Euphrates basin to demonstrate that Turkey's planned management of the waters would not be detrimental to the downstream countries. As Turkey failed to provide some crucial hydrological data, the Bank stopped considering the supplementary loan (PCR Karakaya, para. 9.03). . Third TE. Transmission Proiect 22. The Bank appraised this project in early 1983 while TEK was still struggling to complete Elbistan, during the main construction phase of the larakays project. While, as mentioned above, Government and TEK were in default of major covenants, the US$163 million operation vent ahead. Beyond financing 8 major components of TEK's 1982-86 transmission program, it included US$9 million for technical assistance that wae aimed at addressing the weaknesses in training, accounting, financial reporting, electronic data processing, load research, and demand management that had marred the sector's performance during the preceding decade. As in its other projects reviewed here, TEK, while completing the physical project components, incurred major delays (3 years) with respect to the timetable at appraisal and failed to achieve meaningful progress in the institutional components. Accordingly, the IERR fell from over 8% as estimated at appraisal to a disappointing 3.51 after completion, although the project experienced a substantial cost underrun. III. BROAD COUNTRY AND PROJECT FACTORS AFFECTING THE OUTCOME OF THE PROJECTS A. Political and Economic Enironment 23. Progress on the Bank projects was affected in a major way by the political environment and the country's economic performance, which, in turn, was heavily influenced by national and global developments, especially the two massive increases in oil prices in 1973 and 1979. In Turkey, the foreign exchange crisis of the second half of the 1970s affected in a major way the Elbistan and the TEK II projects by limiting the availability of funds. This dearth of resources doubtlessly also contributed to the slow progress of the Karakaya project in the late 1970s, i.e. before the Bank approved its loan for this development. Cost overruns in Elbistan and Karakaya, especially Elbistan, (paras. 72 and 73) which were essentially unrelated to the political and economic development, compounded the funding problem, which Government and the Bank tried to alleviate by earmarking counterpart funds generated in the context of the Bank's 1978 Program Loan (Loan 1627-TU) and the Structural Adjustment Loans for the financing of the local costs of the TEK projects. 24. Political factors also played a major role in project outcome. Thus, the 1975 change of government led to politically motivated defaults with respect to agreements previous governments had reached with the Bank (para. 13). The Bank reacted by postponing the effectiveness of approved loans and interrupting the processing of operations in the pipeline. Ultimately, after much debate within the Bank and some concessions by Turkey on the power tariff front, the Bank resumed advancing its leading operations in the sector. Undoubtedly, the "approval culture" discussed in the next section played a role in this decision as did the politically highly significant fact that Turkey was an economy at the very front of the battle between the concepts supported by the Bank and those then still dominating the Eastern European countries. But, in any event, such It seems worth noting that had the SAR used a more realistic, lower estimate project cost (see discussion in para. 73), the 1ERR at appraisal would have been substantially higher and the shortfall, accordingly, larger. 9 turmoil diverts energy from project implementation and negatively affects the outcome. 25. Symmetrically, in the early 1980s, when Government, the IF, and the Bank were in substantial agreement and closely collaborated in the implementation of successive Structural Adjustment Programs, TEK enjoyed the benefits of the programs inter alia in the form of much improved tariffs. However, in the later 1980s, elections looming large, fiscal discipline weakened leading the sector into the dismal financial situation (in 1990, TEK was technically bankrupt) from which it presently struggles to emerge. B. The Anroval Culture and the Pressure to Lend 26. Chapter I provides several examples illustrating shortcomings which relate to what the recent report of the Bank's Portfolio Management Task force labels as the "Approval Culture" or even worse to the pressure to lend. These deficiencies tend to lead to a project start on a flawed basis and, therefore to affect negatively project outcome. 27. The events surrounding the preparation of the TEK II project (paras. 13 and 14) provide a good example. At the time, the Bank made a loan (for Elbistan), which it could not make effective because Borrower and Guarantor did not fulfill the conditions of effectiveness, including the elimination of defaults under an earlier loan (for the Power Transmission Project). Far from slowing down the processing of further lending, it prepared another operation (for the TEK II project) to persuade the Turks to live up to their obligations under the Elbistan loan to be made effective. To eliminate some possible obstacles on the way to project start, the agreements for the TEK II loan did not include special conditions of effectiveness, which, as it turned out, was of little help (para. 14). 28. The spirit prevailing during this time is manifested by many internal memoranda that in one way or another, some explicitly, propose not to react to defaults by retaliation, but by an increase of the efforts to advance lending in order "to impress on the government the importance the Bank attaches to the resolution of the problems" which were at the root of the defaults. The result of this "no stick, double carrot approach" was a better fulfillment of the lending program. But it also led to a substantial delay in disbursements (the consequences of which appeared only much later in the Bank's performance analysis) and to an accumulation of commitment charges in excess of US$2 million before the start of disbursements, charges that affect the Borrower, not the Bank. 29. These manifestations of the Bank's "approval culture" need some perspective. Indeed, if, at least in the late 1970s, the Bank seems to have consistently given in to the Turkish side, this normally only occurred over a long time starting with the Bank formulating strong conditions that the Turks did not fulfill. Months of negotiations usually brought some concessions or some action that appeared to represent a substantial progress towards the set goals and a basis for a conciliatory move by the Bank. However, with hindsight, the evaluation of this progress often seems to have reflected more the effort invested in it than the significance of the objectives actually achieved. 10 30. The above examples reflect an approach that was manifestly flawed and which, after 1980, does not seem to have been as commonly followed. It was harshly criticised by some of the key staff involved. This is not unusual, but, what in this case was unusual was the nu=ber of protests and the cynicism of some memoranda barely masking doubts of the integrity of the institution. 31. The projects reviewed here illustrate a more insidious manifestation of the "approval culture", in particular certain facets of the Bank's lack of realism in its project presentations. Indeed, the specialists who originally prepare the projects, tend to be biased as they prefer to show that their project is feasible rather than unfeaeible. Bank operational staff often espouse this bias (Elbistan may well have been a case in point, perhape because they perceive that rejecting projects brings less reward than approving operations; they also tend to understate project risks. 32. The Elbistan project and, to a lesser degree, the TEK II and III projects are cases of underestimated risks. The SAR for Elbistan deals with the project risks in one sentence: "Emphasis has been placed on a proven and reliable system." The design chosen was indeed proven reliable in Germany. It had, at the time been used once in Greece, where units of about half the size of those at E1bistan were installed; some technical difficulties appeared but were mastered. In Turkey, the technology was not only entirely new, but the project management setup was improper and known to be so (see paras. 79 to 81), as, in particular it was too centralized and did not allow the sorely needed expertise of the engineering consultants to have the impact it should. In all three above projects, the Bank underrated the risks associated with TEK's institutional weaknesses, especially concerning the availability of qualified staff in adequate numbers. 33. That second, less obvious side of the approval culture is still quite coVmmon. Therefore, evaluations of projects at their preparation or implementation stage should not only be more realistic but also their presentation be more candid. In particular, where risks are substantial, they need to be clearly identified together with the possible measures to hedge against them. A further analysis should define the hedging that is justified. The implementing agencies have to take the remaining risks for which they should carry out a contingency planning (what if...?) to be reflected inter alia in the contingency allowances of the cost estimates and in the approaches envisaged for the case one or the other of the risks materializes. C. Broad Aareement Between Bank and Turkey, Proiect Ownership 34. During most of the period covering preparation and implementation of the projects reviewed here, there was a flagrant lack of a broad consensus between Government and sector on the one hand, and the Bank on the other concerning the main features of the development of the power sector. This discrepancy existed already in their understanding of what a Bank project is and should be. Indeed, for the Turkish side, such an operation is essentially a vehicle for a transfer of resources to be applied to the implementation of a physical project such as Elbistan. This attitude is best illustrated by the declaration of a Turkish minister, who, in 1978 said that he was reluctant to seek Bank financing as the Bank did not limit itself to project financing but took a missionary approach 11 about sectoral policies and institutional aspects, which he believed misplaced and an interference in Government's internal policies. Whereas the audit has no specific indication that presently Government still thinks along these lines, it is evident that the view still prevails in many people. A high executive of one of the sector entities put it very bluntly to the audit: "With the Bank one should only talk hardware"t If, as the audit suspects, this attitude is still widespread, it represents a major obstacle to the success of Bank projects, and the only thing the Bank can do about it, is carefully assess the degree to which it shares the broad outlook about the sector and the project with the Turkish side and ask for action in the areas of broad agreement leaving other areas the subject of a continuing dialogue. However, if the Bank considers the common ground insufficient for acceptable progress, it should not lend, because the experience with Turkey, which only confirms that in many other countries, has shown that conditionality cannot replace such broad agreement. It can nevertheless support the commonality of broad views by formalizing it. 35. In terms of the objectives of the audited projects, there was agreement on the need to further the use in the sector of lignite and water power. There often was also substantial agreement on broad institutional reform of the sector, although the sense of urgency was quite different in Turkey and in the Bank. In the area of cost recovery, short term political considerations time and again prevailed over long term objectives endorsed by the Bank. This was especially the case in the second half of the 1970s and, after a period of heightened fiscal discipline, again in the late 1980s. 36. In the context of the organizational and managerial strengthening of the sector institutions, in particular TEK, another hampering factor was the exceedingly slow change of the institutionsl culture in the State Economic Enterprises (SEEs), which parallels the missing sense of urgency mentioned before. The structural adjustment programs of the early 1980s called for SEE reform, which for TEK, after it had incorporated the public distribution entities in 1982, was officially implemented in 1984-85 giving the utility more autonomy, even allowing it to set its prices (with a proviso that the Council of Ministers can require an adjustment of TEK's rates against an appropriate compensation).a However, in the relations between Government and TER, as well as within TEK itself, this reform was slow to change the way things were dealt with; indeed, so slow that in 1991, the Bank felt that it had to support TEK through the TEK Restructuring Project, which to an extent tries to implement the measures already decided seven years earlier. 37. In discussions with many Turkish officials, the audit detected a strong reluctance to the delegation of power, which is likely to be a major factor in limiting the pace of institutional development. At the top, it manifests itself in Government's refusal to grant TEK the reasonable measure of autonomy it needs to operate efficiently, and this despite the measures officially taken years ago. This reluctance is also pervasive within TEK itself, where the units that most need such an area of autonomy, in particular those far away from the center such as the Ebistan organization, are reined in to an unreasonable degree. This explains in part: (i) the feeling pervading the staff of these units, that they SAR, TEK Restructuring Project, para. 3.01. 12 work in a world entirely controlled but not understood by headquarters, and (ii) the dificulties TEK has to find qualified personnel to work in such outposts both during construction and after the start of operation (another factor, of course, is inadequate remuneration). Therefore, the Bank should be aware that projects that rely on such a change of mentality risk to progress at a slow pace, as sophisticated as they may be in their approach to foster a more reasonable institutional setup. IV. PfiHANCE OF THE ELBISTAN COMPLEX A. Power Plant Oeration 38. Until now, the performance of the Elbistan plant was less than satisfactory inter alia because its availability was low. Only in about one case out of eight did the annual availability of a unit exceed 80%, a reasonable value; the average was only 632. However, the utilization of the units, when available, was high, in average 785, and in excess of 90% in nearly half the cases; only in 1987-88, two relatively vet years during early operation of the Karakaya hydroplant, did the utilization of the units at Elbistan fall substantially though the availability at the time was not particularly low. 39. The above suggests that the low output obtained from this expensive facility is, at the present stage, less an issue of inadequate dispatching than of low plant availability, in part related to lack of adequately trained staff and unsatisfactory operating and maintenance practices (PCR Elbistan, para. 10.01). This said, the audit, on the basis of interviews with the responsible managers of TEK, confirms the PCR's conclusion that there is an urgent need for a more rational dispatching taking into account, in particular, economic cost and environmental impact (para. 66). The basis for such dispatching is to be created in the framework of the Operational Improvement Program (0MIP) which is part of the 1991 TEK Restructuring Project'. However, the hiring of an experienced utility to asist TEK in implementing the prc,ram was, in August 1992, some eight months behind schedule. B. Power Plant Adjustmento 40. After early operation of the units had shown shortfalls with respect to the guaranteed performance of the boilers, the supplier increased the heating surface in the economizer and the superheater of the units. Further, since commissiontg, the design of the last stage of the low pressure turbine, was the source of operational difficulties. In 1991, major damage in Unit 2, required the return of the broken components to the manufacturer for rebuilding according to a modified design. To avoid similar accidents in the other units, TER To an extent, the 1985 Power System Operations Assistance Project supported by Loan 2602-TUR already addressed this issue, though indirectly (see para. 403 (c) of the corresponding SAR). 13 dismounted the critical parts in the respective turbogenerators thus temporarily reducing their maximum output from 344 MW to some 290 NW. When, the re-built turbo-generator of unit 2 will be available, the other machines will need similar adjustments, and, therefore, the full nominal capacity of the plant will be available to the system in 1994 at beat. C. The Lianite Quality Issue 41. The most critical problem for the plant is the inadequate quality of the lignite supplied by TKI. Table 1 below sets forth the fuel specifications derived from the analysis of the lignite to be mined and the values experienced with the fuel actually supplied in 1992. 42. The main problem i not so much the discrepancy between specified and actual average quality but with the variations in quality. The two most critical variations are those of the heat rate and of the calcium oxide content of the ash. When the heat rate drops below the 950 kcal/kg specified as the minimum admissible, TEK needs to inject oil into the furnaca to assure proper combustion and an adequate energy supply to the boiler. When the calcium oxide content of the ash is higher than the 501 maximum specified, the silicate content is low. In this situation, practically no slag falls through the grate of the furnace though the design assumes that about one third of the combustion residues would be evacuated through the slag extractors. This occurs because most of the residue is volatile and therefore exits the furnace together with the flue gases through the electrostatic precipitators, which should eliminate most of the solid components. However as these facilities are not laid out for the higher concentration of flyash, they get overloaded and fail to eliminate a substantial part of the particles, which thus escape through the stacks, possibly polluting the environment at times beyond acceptable levels. When the calcium oxide content of the ash is below the 25% minimum specified, this indicates that the silicate content of the lignite is high and the raw fuel is more abrasive than foreseen in the plant design, causing fast deterioration of the mills and ducts. The plant has been plagued by both phenomena outlined above.'o D. The 81bistan Mine 43. The Elbistan mine is laid out to produce 20 million t of lignite per year of which 18 million t are to feed the power plant. This mining output is based on the assumptions that the installed capacity of the power plant would be used at a rate of 6000 hours per annun1 and that the average heat content of the lignite would be some 1050 kcal/kg which means that the plant would require about s 'When the OED mission visited the plant site in late August 1992, the heat rate of the fuel was 715 kcal/kg, i.e. far below the specified acceptable minimm. The ash content of the lignite, in turn, was at 22%, i.e. still within the acceptable range. However, the calcium oxide content of the ash, at 73.81, substantially exceeded the specified maximum tolerable. it The Audit understands that, whereas the 6000 hours utilization corresponds to the assumption at project inception, TEK's Planning Department is calculating with a 6500 hours average utilization of the plant, i.e. a fiure even further from the present reality. 14 2.4 kg of lignite for a gross generation of 1 kWh. Operation over the past 8 years confirmed this value, vh1ch is not likely to change substantially. Indeed, improvements in operational efficiency are likely to reduce mostly the oil consumption which has remained high and, at best, keep the lignite consumption at about its present level. Table Chracteriltices of the Elbistan Lianite as Specified and as Supplied Specification Actual Neat Rate: average 1050 kcal/kg 1100 kcallkg range 950 - 1600 kcallkg 700 - 1650 kcal/kg Ash Content: average 15.3% 18.5% range 8Z - 232 10% - 35% Humiditys average 571 501 range 501 - 64% 461 - 522 Calcium Oxide average 301 302 Content of Ash range 25% - 50% 72 - 78% 44. Rheinbraun of Germany planned the mine on the basis of 1200 test borings in the area foreseen to be mined to fuel the four units of the first stage of the power plant. This means that the average distance between boreholes was between 100 and 150sm, which seems reasonable. Moreover, the exploitation of the seam has by and large confirmed the findings of the preliminary investigations. The stripping ratio, i.e. the relation between the material discarded and that used, Which was estimated at 2.7, until now, was substantially more favorable at about 2.1. However, this may still change when other parts of the seam are mined. The average heat rate, ash content, and humidity of the lignite are not too far from the specified values reflecting the average quality expected. Even the most contentious calcium oxide content of the ash (para. 41) is rather near the value anticipated at the design stage. In two respects (one negative, one positive), experience did not confirm expectations at the project stage: (i) hard layers in the overburden and the seams impeded the straightforward utilization of the wheel bucket excavators which form the backbone of the mining equipment and (ii) the groundwater problem was substantially less severe than feared at the design and construction stage. Z. fakt gheel Excavators and Hard Layers in the Mine 45. The preliminary underground investigations had identified in the overburden and the lignite seams layers that were, in part substantially harder than the rest of the material. These layers were lake sediments, inter alia conglomerates and sandstones of uneven quality. The experts (including those sent by the Bank), though according to TKI, somewhat ambivalent about the hardness of the intercalations, considered them no impediment to the use of the bucket wheel excavators. These were of a type with which the German engineers had ample experience. It apparently took until 1981, i.e. more than seven years from the start of construction, for the specialists to agree that the harder parts of the above layers would have to be mined with special equipment. Such equipment is 15 now in use and leads to satisfactory results. The audit found that, even with hindsight, there is little basis for suggesting that the layout of the mine was wrong. F. The Groundwater Issue 46. The groundwater and, in particular its presence in clayey material in the mining area and in the neighborhood of the mine, was, during the opening of the min a one of the main subjects of apprehension. Fortunately, the adverse conditions did not materialize to the extent feared in the worst case scenario. Indeed, in the first eight years of operation, the mine experienced only one landslide of any significance"', as the measures taken proved sufficient and efficient. The main such measures were the construction of 15 main wells and a large number of secondary wells from where TRI pumps about 3 million m3 of water per month (with little variation over the year) to keep the groundwater level in the mine adequately low. From spring to fall, the pumped water is utilized for irrigation and in winter it is released into the nearby river. 0. The Minina Operation 47. The large bucket wheel excavators utilized at Elbistan are specified to mine 3000 m3/h, i.e. 3600 to 3900 t/h' in uniform material. Under the actual conditions prevailing at the TKI mine, their output is planned at 2000 to 3000 m3/h (2400 to 3100 t/h depending on the specific weight of the material). 48. Table 2 below sets forth capacity and utilization data for the mine as a whole and for its main equipment, i.e the bucket wheel excavators. The actual data reflect the performance achieved in 1989 when the power plant reached its highest gross annual output to date, i.e. 5700 GWh or about 70% of the design capacity which is 8160 GWh assuming a utilization of 6000 h per annum. The third column of the table describes the mining operation as planned for the power plant operating at the above design capacity, i.e. it is based on the design parameters which are in part less favorable (especially for the specific weight and the heat rate) than those actually experienced in 1989. 49. In any event, the table clearly documents that, with its six bucket wheel excavators, the mine should still have an ample reserve capacity when the fuel requirements of the power plant will be at their maximum. This statement appears reinforced by the following facts: - According to TKI management and operating staff, the excavators have worked satisfactorily ever since the manufacturers adjusted the motors of the bucket wheels. - The mining equipment appears to be well maintained. 1n The incident took place in 1984 shortly after commissioning of the first two units in the power plant. The specific weight of the material in place was foreseen at 1.2 t/m3. Until now it was nearer to 1.3 t/m. 16 - Contrary to what happens in other public enterprises, TRI management of the Elbistan mine seems to have a reasonable spending authority. The audit understands that the annual order for spare parts to within this authority as well as the maintenance contracts which TRI in part farms out to private enterprises in the region. Table 21 Caoacity of the Excavators and the Mine 1 or Maximum Power Plant ActIal (1989) Utilization Annual Gross Power 5700 GWh 8160 GWh Plant Output Lignite Output and 13.7 million t/a , i.e. 19.6 million tias i.e. at Excavation Equipment at most one excavator for most one excavator for Required 5700 h or two units for 8200 h or two units for 2850 h each 4100 h each Overburden to be Re- 37.0 million tia i.e. at 52.9 million t1a, i.e. at moved and Required most two excavators for most three excavators for Excavation Equipment 7700 h or three units for 7350 h or four units for 5140 h each 5500 h each 50. The above table also shows how much, until now, the mining equipment was under-utilized". It is therefore not surprising that, according to TRI, in mid 1992, the cost to TKI of supplying the lignite to TER and of disposing the ash it receives from the power plant was at some TL 47000 (US$6.80) per t of lignite supplied, hich compares to a contractual price of about TL 42000 (US$6.10) per tons. According to TRI, the actual cost includes depreciation but no overheads for the authority*s central services. Whatever the reliability of the above figures, they clearly Indicate that TKI suffers heavy losses on its Elbistan mining operation. Figures obtained from TRI in Ankara roughly confirm the above information and suggest that the losses are in the order of 152 of revenues. 51. From the point of view of TER, the most problematic aspect of the mining operation is the lack of proper processing of the mined material leading to the exceedingly irregular quality documented in Table 1 which, in turn, is the source of much of the poor performance of the power plant. The audit understands that the mine with its present equipment could produce a lignite of more uniform quality that would conform to the plant specifications, mostly by achieving a better mix of the raw material. Some of the measures that could be envisaged to achieve this would be"s 4 It seems worth mentioning that, at the time of the OED visit to the site, TX was covering the demand working only with one out of the six available bucket wheel excavators. The 01111185 pre-appraisal report (in particular para. 4) for the Elbistan Operation and Maintenance Assistance Project (Loan 2650-TU) already Included proposals for such measures. 17 - To mine the lignite at least in two places simultaneously and to convey the two or more types of raw material to the same stockpile (see Footnote 14); this would provide a first mixing. - To deposit the material at the stockpile in layers along the entire length of the pilel when the audit was at the site, the arriving material was accumulated in cones as high as the ultimate stockpile and only then was the conveyor moved some distance along the stockpile site to build up a new cone. - To mine the layered stockpile in such a way that the removed material represents a cross section of the layers, thus assuring a second mixing of the material. - To mine at the same time at least two stockpiles and to convey the two or more intermediate products to the bunker from where the lignite is transferred to the power plant, the procedure providing a third mixing of the fuel. 52. The audit heard contradicting statements as to the extent to which the above or similar methods are already applied. At the time of the OD visit they obviously were not. However, it is evident that, in view of the large losses it already incurs, TKI has no incentive to apply such methods as they all imply higher costs than those associated with the transfer of the essentially unmixed raw material to TEK. Further, TKI's supply contract, although it includes the specifications of the ranges for the various parameters only foresees bonuses and penalties for deviations from the required average heat rate of the fuel. In this point, TKI was in most cases able to exceed the target and, ironically, to earn a bonus. 53. The improved mix would also help alleviate an operational problem created by TKI's obligation to take over from TER slag and ash produced at the plant and to deposit them in the already exploited parts of the mine. Indeed, the fly-ash which is retained in the electrostatic precipitators and has to be transported to the mine to be dumped there, has to be moistened because otherwise it would be largely carried away by the wind increasing the negative environmental impact. The more water is added, the more the calcium oxide in the ash reacts with the water inter alia increasing the temperature of the mix. If the calcium oxide content is high and enough water is added to obtain a texture adequate for transport, the temperature exceeds 100 degrees Celsius and the heat damages the conveyor belts. If even more water is added to cool down the mixture, it becomes an unmanageable mud. These difficulties do not occur when the calcium oxide content of the ash is lower than the upper limit of 50% specified. This further dramatizes the need to induce TKI, in a first stage, to deliver the lignite according to the specifications. H. Most Urgent Improvements 54. It seems evident that TKI, or whatever entity operates the mine in future, will have to apply more costly methods of extracting, mixing, and transferring the lignite to the point of sale to TEK than those presently used. Such methods, e.g along the lines suggested above, will only be seriously considered, if there 18 is inter alia a financial incentive to do so, something that is lacking now. Therefore, to the audit, it appears that a first step towards improving the operation of the complex is an appropriate adjustment of the lignite supply contract between TER and TKI. The modifications should, in particular, foresee bonuses and penalties for positive and negative deviations of the supplied fuel with respect to the averages and brackets specified for -a1 the main quality indicators. The proper pricing of the lignite would also be in line with the policy set forth in the President** Report (para.88) for the Energy Sector Adjustment Loan supported by Loan 2896-TU and spelling out that TKI shculd "become self financing" 55. Over the period 1985-89 the price TER paid to TKI for the Elbistan lignite deteriorated in real terms by some 202, reflecting a similar deterioration in the Turkish lignite mining sector at large, which was followed by a stagnation at a low price level. Whereas, it seems imperative that the mine operator's revenues from 21bistan be increased, it is likely that the efficiency of the Elbistan mining operation which at present employs 1500 eople (among them 20 with engineering degrees and 770 skilled workers) 6 can be improved, too. Unfortunately, at this stage, when the utilization of the power plant and hence of the mine is between 452 and 70% of capacity, it would be difficult to find a fair and lasting solution, as there are no operational data available reflecting experience with plant and mine operating at a high level of utilization over a reasonably long time. 56. In institutional terms, there are three possibilities to improve the mine operation in order to provide fuel suitable to adequate power plant operation: - TRI operates on the basis of a contract modified along the general lines suggested above; - TEK takes over the operation of the mine, a possibility that apparently has been enviseged off and on; - a private company takes over the operation of the mine; and - the mining operation is privatised outright, i.e. both assets and operation are transferred to the company. 57. The first solution is probably that of least resistance. The second is contrary to the principles enunciated in the 1991 UEK Restructuring project which ais at separating from U2K all operations that are separable and at creating new, if possible private enterprises to take over and operate the separated systems. The third and fourth solutions are likely to be those that at the present juncture are the most desirable, the third being easier to implement in the short or medium term. A change in one of the above directions is the more urgent as TEK envisages to expand the power plant by two units; indeed, it plans to issue the tender documents to potential bidders in 1993. 16 According to the Bank's 06127184 Position Paper the feasibility study for the mine had foreseen a work force of 3600 people I This latter figure seems too high, but the Audit did not investigate this point further. 19 1. Bank Involvement 58. 21bistan has cost some US$2.5 billion in current terms (in excess of 1990 US$4.5 billion). It is a large imaestment, by any standard. Unfortunately, it is also one of the most disappointing experiences the Bank had in the power sector since it started lending for power in the late 1940s, the worst aspect being that some twenty years after construction began, the facility is still not working satisfactorily. Therefore, it seems that the maximization of the benefits to be reaped from the facilities in the shortest term possible is in the interest of Turkey, the Bank, and its co-lenders, who all have been heavily criticized for their involvement in this project. The question is whether it would not make sense for the Bank to follow up again on the optimization of the technical and commercial operation of the Elbistan complex. This could be considered in connection with the preparation of concrete plans for increasing private sector involvement in the energy sector. V. PERFORMANCE OF KARARAYA 59. Table 3 shows that, to date Karakaya operated about as expected, the high output in 1988 being a consequence of the high average flow available in that wet year and the low generation in 1989 being related in part to planned outages for the first revision of the generating units. Table 3: arakaya Rydroplant (1800MW). Generation and Utilization Yar Generation (GWh) Average Annual Utilization (h) 1987 3363 Start of Oper. 1988 7879 4377 1989 5562 3090 1990 7120 3950 Design 7353 4085 60. Except for about 1.5 months in April and May 1986 when, during the filling of the reservoir the average flow was reduced to 420 m3/., until mid 1991, Turkey abided by the "rule of 500" (PAR para. 19) . The Bank always understood this rule to be a step towards a tri-partite agreement between the Euphrates riparian states on the utilization of the river flows. Presently, the situation is unsatisfactory as: (i) such an agreement is not in place and unlikely to be in the foreseeable future, and (ii) in 1991 Turkey has started to fill the AtatUrk reservoir downstream from Karakaya, drastically reducing river flow for a time. Further, since mid-1992, it is operating the Atatirk complex which, in contrast to Karakaya, involves substantial abstractions from the river for irrigation; and 20 (iii) Turkey has refused to provide the Bank with hydrological data inter alia on the operation of the Reban reservoir upstream from Karakaya, thus preventing the Bank from updating the analysis of the adequacy of the arrangements concerning the use of the Euphrates waters (para. 21 and PCR Karakaya, para. 9.03). 61. It is evident from earlier analysis and from the experience in the past 5 years that the appropriate operation of Reban and Karakays allows, in principle, to respect the "Rule of 500". However, it is likely that this is not possible if the Atatfrk complex is included in the system to be operated economically. Although the Bank was never involved in the AtatUrk project, this leaves the Bank in an uncomfortable position from which only a satisfactory tri-partite agreement is able to release it entirely. There is ao evidence, though it is likely, that the Bank was aware of the risk for such an outcome of the riparian issue whereby the Bank could be construed to be guilty by association, if there is a serious breach of the "Rule of 500" in the future. The Bank could only have avoided these risks by not participating in the Karakaya project. The audit tends to conclude that this would have been unjustified in view of the large benefits this facility generates and is likely to continue generating. VI. ENVIROMENTAL ISSUES A. 1biatan 62. The Bank appraised the Elbistan project in 1973, i.e. at a time environmental consciousness was at a relatively early stage of development, both in the countries in which the Bank was operating and in the Bank itself (indeed, the nucleus for the institution's environmental staff was appointed only one or two years earlier). This is reflected in the somewhat perfunctory discussion of the environmental Impact of the project in the SAR (paras. 4.16 to 4.18). The audit found that in 1992, i.e. nineteen years after the appraisal, the project region did not anymore qualify as " sparsely inhabited" (SAR, para. 4.16). This description still applies to the wider surroundings of the plant, but the agglomerations of Blbistan and Afsin , between five and ten km from the plant site represent a population which, in the audits estimate substantially exceeds 20,000. Moreover, Elbistan has in several respects urban character. 63. The SAR emphasized the present main environmental issue, the flyash emissions in the power plant and the mine. There is no reason to contradict the PCR finding that until the PCR mission visited the site (presumably in 1989), the maxtaum pollution recorded was in compliance with the levels permitted by Turkey's and the Bank's environmental standards (PCR, para. 6.05). Yet, in September 1992, when the OED mission visited the site, a dust cloud visible from 20 km afar was a clear indication that the flyash emissions were at the time at an unsatisfactory level. At the present stage, this is the main issue, and it is closely related to the recurrent exceedingly high calcium oxide content of the ash discussed in para. 42 above. above. 21 64. The audit generally concurs with the conclusions of the operating personnel and its advisers, that if the plant received lignite meeting the apecifications, it is likely that its environmental Impact would comply with the regulations. Therefore, working to specifications as required for operational reasons (paras. 41 and 42) is also the first step required to keep adverse environmental impact within acceptable limits. 65. A second stage should involve mainly the careful study of the results provided by the environmental monitoring stations in the plant area. The two stationary installations and the facility that monitors the stack emissions are in place. However, complementary work is still required to bring the installation to the required standardl most Importantly, at least one of the two mobile measuring laboratories purchased in connection with the project and, in late 1992, working in other places should be brought back to Rlbistan, because, as the fixed stations are located relatively near to the plant and not necessarily in the areas of worst impact, the mobile labs are needed for a proper determination of the actual intensity and distribution of the emissions and of the adequacy of the environmental protection measures taken. With lignite meeting the specifications, the solid emissions are likely to be reduced in such a way that the emissions would be at a level compatible with the impact standards adopted at the time of appraisal. Further, the low content of combustible sulfur of the fuel (0.6%) suggests that the SO. emissions would rarely exceed values which could lead to emissions in the plant area and its surroundings higher than the maximum tolerable according to the more recent Turkish and Bank standards. Nevertheless, this needs to be proven and the Bank should follow up on this and on TEK's actions in this point. 66. In a third stage, the future role of the Elbistan complex in the Turkish power sector will have to be determined concurrently with additional environmental protection measures, if any, that the results from stage two above would suggest. On the one hand, it would seem that there is still a strong case for lignite fired power plants in Turkey; first, the country is heavily dependent on energy imports, of which those of (environmentally benign) natural gas from the former Soviet Union are for a time uncertain to continue in the required quantity; second, there is little prospect for accelerating the construction of hydropower plants. On the other hand, in recent years, the opposition to lignite fired plants has dramatically increased in the affected population, the corresponding state governments, and environmental groups. This opposition has become particularly vehement in Western Turkey, where it has forced TEK to postpone indefinitely the commissioning of the 600 MW Kemerk8y plant. Indeed, TEK may have to shut down, at least temporarily, some of the existing plants along the Aegean and Southern Mediterranean coastline, where, beyond their impact on public health they are seen to unduly affect cultural and tourist areas." As Elbistan is located in a region ecologically less sensitive than the plants further west, it could take up a crucial role in the power sector. But it cannot do this before the three steps above or other actions are taken to ascertain that the plant (1340 MW or more than 2000 MW, if the plans for two more units materialize) will meet the environmental standards of the future, which are a7 See Minutes of the Pre-appraisal Review Meeting on the proposed Coal Pollution Abatement Project, dated June 25, 1992. 22 likely to be more stringent than those on the basis of which the original plant was conceived twenty years ago. 67. Finally, the audit confirmed the PCR's findings (para. 6.05) that the progress on the afforestation of the reclaimed mine area is satisfactory and that the plant and mine effluents are properly disposed of. Indeed, as mentioned in para. 46, a large part of the water pumped from the mine is used for irrigation. 68. Discussing the above issues at TER headquarters in September 1992, the audit found that the creation of a separate environmental department (a condition of the TER Restructuring Project to be fulfilled by June 30, 1992), is apparently underway. But communication between the group to be separated and other units of TER seemed poor and in need of close attention. B. Naka 69. All the evidence the audit collected suggests that DSI and the Bank successfully dealt with the environmental issues related to this large hydroplant with its substantial reservoir. As set forth in paras. 6.04 and 6.05 of the PCR, the main problems were the resettlement of about 17000 people and the lose of modest areas of agricultural land. The loan documents provided the blueprint for the procedure to follow to solve these issues, which seems to have been successfully done, though not without some cases of hardship. The last settlement of compensation claims has taken place in 1992. The Bank also assisted in the successful salvage of archeological objects. The audit considers that enough evidence is available to agree with the PCR (PCR Karakaya, para. 6.04) that for the downstream riparians the project does not have a major adverse Impact as long as the "Rule of 500" is respected. However, with Ataturk in operation this statement is becoming academic. VII. PROJECT COSTS AND TEK INVESTMENT A. Introduction 70. The audit carried out an analysis of the costs of the main elements of the various projects reviewed. In particular, it compared the amounts as estimated at appraisal with the actual expenditures for items that were implemented about as foreseen at appraisal. The quality is tested on the basis of the variation in base cost (including physical contingencies) expressed in constant terms. This comparison allows a judgement on the quality of that part of the cost estimates that is essentially the designer's responsibility (especially the quantities and the basic prices used), B. Base-Coats 71. Table 4 below sets forth the cost over- and underruns in real terms experienced in the various projects under scrutiny. 23 72. The base cost estimate for the TEK II Prolect was adequate. The same applies to the Karakays Project, though it suffered a cost increase of about 14% in real terms, which was mainly attributable to unexpected events such as: (i) the renegotiation of the civil works contract in 1979; (ii) the construction delay of two years forcing e.g. the installation of provisional storage facilities for equipment that was delivered early; and (iii) unexpectedly extensive stabilization work required in the dam abutment areas. Table 4: Increases in Base Costs TotalCost Increases in Base Costs in Terms of in Million Estimated Base Cost inc. Phys. Cont. Proiect 1990 US$ Local Foreign Total TEK Transmission II 93 12% -61 2% Elbistan Power Plant 2465 375% 51 102% Elbistan Mine 1009 275% -11 75% Elbistan Transmission 102 -6% 2201 991 Rarakaya Hydro 1575 7% 201 141 TEK Transmission III 167 -32% -431 -40% 73. The above table also brings home the dramatic cost overruns of the Elbistan Praent. These are due to a large extent to extremely poor project management which inter alia resulted in only a 15% completion by the end of the 1970s i.e. 6 years after project start. Further, the recurring shortage of local funds provided the contractor with strong arguments for compensation often beyond the losses incurred in direct connection with the shortage. The audit was unable to determine whether the base cost was low to start with. The fact that estimated total cost of the complex expressed in US$ doubled in the two years following appraisal tends to suggest so. But the power plant unit costs of 1973 US$350 per kW (corresponding to about 1990 US$1000) were substantially higher than those foreseen in other projects at the time; e.g. appraisal reports prepared a few years later estimated the cost of coal fired units in India of about the same size as those in Elbistan but using lignite of somewhat better quality at the equivalent of about 1990 US$700.18 Therefore, it would appear that the original cost estimate for the power plant, which was based on bid prices, took into account to an extent the fact that the technology to be used was new to Turkey and not widely proven in other countries except Germany. This suggests that, in cases subject to high uncertainty as e.g. when advanced technology is involved, even cost estimates based on bid prices have to be taken with appropriate caution. However, there is a limit to the degree to which such caution can be reflected in the allowances for physical contingencies in the estimates, as large amounts of such contingencies risk to introduce the notion that substantial is Actual costs in India came out at 1990 US$ 700 to 1000. 24 contract revisions are expected from the very beginning, which lays the ground for a self-fulfilling prophecy. finally, it is important not to read too much into the discrepancies between the breakdowns into local and foreign currencies of the estimated and actual costs, as TSK's accounts are not reliable in this respect. Nevertheless, the skyrocketing of the local costs and the much more modest increases in the foreign exchange component seem to reflect a reality connected to poor project management and to long delays. 74. The most salient result concerning the costs of the TEK III Project is the 402 cost underrun in real terms, which led inter alia to the cancellation of U8$65.2 million from Loan 2322-TU. Two reasons for the low actual cost are those given in para. 5.01 of the PCR, i.e. that: (i) prices obtained under international competitive bidding were much lower than expected because of the economic slump with the associated more intense competition that prevailed at the ttme, and (it) TEK used a substantial quantity of material that was already in stock. A further reason was that the cost estimate was based, in part, on earlier bid prices including financing. Apparently, part of the financial costs were not identifiable as such and not eliminated from the base cost. Table 5: 380kV Transmission Line Costs Project Cost per km in Thousands of 1990 USS Lag"I Forein Total UEK Transmission II estimated 7? 77 ?? (7551) actual 57 66 123 81bistan Transmission estimated 42 46 89 (540km) actual 40 149 189 TE Transmission III estimated 45 118 163 (1500ko) actual 31 67 98 TEU Transmission 1V estimated 94 29 123 (8001M) 75. Table 5 above compares estimated and actual costs of 380kV transmission lines included in the reviewed Bank projects and suggests the following points: The cost of the km of 380kV transmission line seems to be of the order of 1990 US$100,000, which is close to that experienced in other countries. - The somewhat higher costs of the lines in the TEK II project is related to the fact that the project included 601m of high cost lines in the Istanbul urban area including the crossing of the Bosporus. The actual cost of the Elbistan transmission component is most probably misstated. The estimated costs for the lines included in the TEK III project 25 discussed above (para. 74) could have been perceived as out of line with those of other transmission projects implemented at about the same time. 76. The above comparisons are only meaningful because they are carried out on the basis of estimated and actual costs expressed in constant terms. While the audit is aware that costs comparisons in current terms are significant for the discussion of the Beneficiary's finances, it finds that it would be helpful if PCRs would present the cost analysis also in constant terms as these provide a better insight into the quality of project preparation and implementation." C. Balancing TEK's Investment 77. It is evident that the large cost overruns experienced in connection with Elbistan (and other generation projects not reviewed here) imbalanced TEK's investment program at least for a time, as part of the funds earmarked for transmission and distribution had to be allocated to generation with the undesirable effect of slowing down the development of distribution, in particular. However, most recently, the situation seems to have improved. Indeed, investment in generation decreased by one third in real terms over the period 1988-90 and its share in total investment fell from an exceedingly high 80% to a reasonable 48%. In the same period, investment in distribution increased by 401 in real terms and its share in total investment grew from 12% to 201, which figure the successive budgets had projected all along and which seems reasonable, too. VIII. PROJECT MANAGEMENT 78. All the PCRs for the projects reviewed here highlight the fact that project management was a major issue, which only in the case of the Karakaya project was satisfactorily solved and even there, only after protracted discussions between DSI and the Bank. A. Elbistan 79. The SAR for this project deals with this basic issue in one short paragraph (para. 4.15) under the title "Administration of the Project" which fails to convey that the issue was not bureaucratic follow-up but management. It concentrates on an outline of the role of the eleven member coordination committee at the level of the Undersecretary of Energy and the general managers of TEK and TKI for broad guidance, assisted by a sub-committee itself advised by the project consultants, for the day-to-day coordination of the sub-projects. Both units were foreseen to work essentially in Ankara. The setup was certainly not fully satisfactory to the Bank as it violated principles that the Bank usually applies: (i) two entities were responsible for the two integrated parts of the project, and (ii) coordination between the two responsible organizations, 19 The PCR for the TEK II project presents the cost comparison in constant terms. 26 TER and TRI, was at too high a level to have a reasonable chance to be effective at the working level, and (iii) the setup of site management did not foresee an adequate role for the consultants whose expertise was sorely needed. These shortcomings were emphasized by the fact that the setup also failed to take into account several weaknesses of the executing agencies that were very much present in the minds of Bank officials (PCR 8lbistan paras. 4.05 to 4.07): - By 1974, TEK's uastalled capacity was less than 3000 MW, which meant that the Elbistan project was to increase this capacity by nearly 502, making it a large project for the utility. The same applied to TRI, which by the early 1970s produced about 15 million t of lignite and coal per year, an output that the project (20 million t per year) was to increase by more than 100%. - TEX and TKI had major staffing problems both in terms of personnel numbers and quality, as they had to apply rigid Government salary scales entirely inadequate to attract the high caliber staff needed for the implementation of an ambitious and complex project such as 21bistan (see SAR Elbistan paras. 6.08 and 6.20). - Both TEK and TRI were unwilling to delegate authority to the local project winagement and adamantly opposed to giving outsiders a substanti al role in project management. 80. The solution Adopted was obviously a compromise to bridge over basic disagreement between the Turkish authorities and the Bank, which took all along the indisputably proper view that such "an integrated project should be carried out under the primary responsibility of a single agency" (OAR Elbistan, para. 6.01). The Bank apparently expected that the dialogue with the Turkish authorities would lead to an improved setup after agreement was reached on the lending operation. However, these expectations proved erroneous. Indeed, Government, TER, and TKI carried out the project by and large ignoring the Bank's and its co-lenders misgivings about project management. Some measures addressing the staffing issues were agreed and in part taken; however, they consistently pro7ed inadequate. Even the 1980 "breakthrough" (para. 16) achieved with the threat of loan suspension by the Bank and its co-lenders, resulted only in an insufficiently increased role for the consultants and in a still inadequate delegation of authority to site management. Nevertheless, the Bank consistently failed to acknowledge the hardness of the Turkish position in this point and even in its 04184 Position Paper expressed the hope to induce TKI to hire a contractor for operating the mine, an expectation that, as so many others, proved futile and, on the basis of the long experience accumulated until then, should have been perceived as such. 81. The above points show that the project management issue in the Elbistan case is a classic example for: (i) an unacknowledged lack of common ground on basic issues (a problem closely related to the "project ownership" issue discussed in paras. 34 to 37), and (ii) lacking realism on the part of the Bank refusing to recognize the Turkish authorities' absolute unwillingness to give in on the delegation of authority issue. 27 B. T IMI and III 82. The SAR for the TER II project is exceedingly terse concerning project anauemnt (SAR TEK II, para. 3.07). Indeed, it only spells out that TER itself would administer (and presumably also manage) the project, which would be carried out by contractors (civil works, towers, etc.), but mostly TEK personnel on force account, an arrangement that the Bank considered satisfactory at the time of appraisal. However, the unit in TEK responsible inter alia for project anagement soon ceased functioning because of staffing problems (PCR TER II, para. 8.6). This experience led, in connection with the TER III project) to the creation within TEK of a project management unit assisted by a Turkish consulting firm, legally private, but owned by the main electricity utilities of the country (SAR TER III, para. 3.10). This outfit teamed up with Kennedy and Donkin of UK and provided excellent services. However, this arrangement did not work to full satisfaction, as liaison between TEK's departments involved in the project remained poor and the consultants could not play the central role they were expected to assume (PCR TER III, para. 5.01). C. Karakays 83. The Bank made its loan for the Karakaya project after some eight years of discussions with the Turkish authorities involved. Originally, a setup was adopted that was similar to that created for Elbistan with a coordinating committee chaired by the Undersecretary of Energy to make the broad decisions. However, as only one executing agency, DSI, an organization quite experienced in planning and construction of large hydroplants was involved day-to-day coordination was a priori easier than in the .se of Elbistan. The original setup nevertheless suffered from too little delegation of authority to the site management. In this case, however, the Bank's and its co-lenders' views prevailed and Turkey agreed to set up a strong site organization in which the engineering consultants had an appropriate major role. This arrangement proved successful. D. Conclusions 84. The Bank, in general, insists on the delegation of as much project management authority as possible to a site organization in which the engineering consultants play a role by and large defined by the capability of the executing agency. The positive outcome of Karakays and, to some extent, of the physical component of the TEK III project as well as the negative experience in Elbistan and TER II confirm the validity of this approach. Further, the quality of the consultants' work heavily depends on the quality of the executing agency's counterpart personnel, to the point that a weak executing agency may need another consultant to manage a large and complex consultants contract. Such an additional assistance would probably have been desirable throughout the execution of Elbistan, where the Bank was mistaken to consider the shortfalls in project management perceived at the appraisal stage as corrigible through stopgap measures and further actions to follow. Experience in Turkey should have told that these actions went against the local authorities' convictions and normal procedures, and therefore had little chance of materializing. 28 IX. TEK'S OPERATIONS 85. The present chapter sets forth a few indicators of TEK's performance as background and support of the findings of the present audit. It concentrates on the development of the utility since 1983, when it incorporated the distribution utilities in the large municipalities and thus became the entity it is now. A. Power and EnerM Table 6: TK: Development of Demand, Capacity, and Sales 1980 1984 1990 Peak Demand MW 8670 Installed MW 4050 7190 14725 Capacity Growth 15.4% 12.7% Sales TWh 18.6 25.5 44.0 Growth 8.2% 9.4% 86. From Table 6 above, it would appear that the installed capacity is substantially too high for meeting the va=dm= demand. This is related to the fact that after the energy crises of the 1970s Turkey tried in a hurry to develop its installed capacity using local resources (such as lignite and water power). However, most of these long term developments fell substantially behind schedule (see Elbistan and to a lesser degree Karakaya) and had to be complemented by projects that could be carried out in a shorter time and could avail themselves of the gas supply from Eastern European countries, which, in the meantime, had become quite economic. Further, demand grew less fast than anticipated in the 1970s. Finally, the installed capacity figure for 1990 is somewhat misleading as, in November of that year, the actually available capacity in TER's system was only about 10,500 MW. Some 1500 4W of the difference with the installed capacity were due to planned overhauls and to seasonal water shortages in the hydroplants, the rest was related to forced outages of plants (e.g. low availability of Blbistan, see para. 38) much of it in connection with improper operating and maintenance practices (SAR TEK Restructuring Project, para. 3.05). 29 B. Personnel 87. During the period 1984-90, TEK's personnel increased by some 32 per year to 65,000 (excluding temporary staff), i.e. at a pace substantially slower than sales, mainly thanks to a 1989 3% reduction in staff. Accordingly, the number of MWh sold per permanent staff employed increased from 480 to 670, indicating progress in the right direction. Personnel and administrative cost per kWh sold and per permanent staff employed, however, increased at an average annual rate of 7% and 142 respectively, which, by itself would be cause for concern. But, in this case, it seems to reflect in part a long overdue substantial improvement in TEK's salary policies. The graphs in Attachment 2 set forth the development of cash operating costs per kWh in constant terms which suggested the above conclusions. C. Revenues and Tariff. 88. During the first half of the 1980s, in the context of the structural adjustment programs supported by the Bank, Government permitted TEK to adjust its tariffs by over 15% per annum in real terms. After TEK started operating in its present setup, the increases continued for two years (see graph in Attachment 2). In 1986, the adjustments allowed to maintain the average revenue in real terms. But with the recurrence of high inflation and the relaxation of the pricing policies in connection with elections in three consecutive years, tariffs quickly deteriorated from some 80% of long run marginal cost (about 90% for high voltage sales and 70% for low voltage supplies) in 1986 to some 50% in 1990. In 1990 the deterioration stopped and subsequently the trend seems to have changed for the better. D. Finances 89. The above development of revenues was the main factor affecting TEK's finances, which for the most recent past are characterized by the indicators in Table 7 most of them extracted from paras. 3.07 to 3.09 of the SAR for the 1991 TEK Restructuring Project. 90. In terms of compliance with the financial covenants agreed with the Bank, Government and TEK have a poor record: - In the period 1975-80, TEK never achieved the covenanted 6% (for 1976) and 8% rate of return, as the actual figures were between 2% and 42; even these figures have to be considered with caution in view of the uncertainty about the valuation of the asset base (PCR TEK II, Annex 7). - In the 1980s, when a contribution to investment covenant with a 20% threshold replaced the rate of return stipulation, TEK complied between 1982 and 1986 when revenues per kWh increased fast. But performance returned even more quickly to dismal (-51% contribution in 1989) when revenues per kWh worsened in the second half of the 1980s. Though the decline stopped in 1990-91, TEK was not yet able to generate again the agreed contribution to investment. 30 - In the 1970s, Government and TER never complied with the covenant limiting accounts receivable from municipalities to tree months' billings of its bulk supplies mostly to municipalities. Indeed, the indicator was consistently in excess of 6 months throughout 1982. Since the integration of the municipal distribution entities into TER, receivables have improved to about 4 to 5 months' billings to ultimate consumers, still short of the agreed maximum value, which remained at three months after 1982. - Except by the end of 1984, TEK, in the past 10 years was never able to achieve the 1.0 current ratio called for in the TEK III loan documents. Table 7: TEK: Financial Indicators Year 1985 1987 1989 1991 Operating Ratio 612 85% 1062 Current Ratio 0.9 0.8 0.6 Return on Assets 262 112 -4% Contribution of Internal 32% 7% -512 Cash Generation to Investment Debt Service Coverage 1.6 1.0 0.4 Debt/Equity Ratio 35/65 58/42 52/48 Accounts Receivable in 7.0 6.0 4.3 Months of Billings 91. The 1991 TEK Restructuring Project is addressing these financial as well as the accounting (in particular asset valuation) and financial management issues already addressed in one way or another in the reviewed projects as well as subsequent operations. Indeed, this most recent operation includes a comprehensive and ambitious Financial Management Improvement Program (0MIP), whose chances of success will substantially depend on the degree to which Government and TER are committed to this operation and consider themselves as "co-owners" (paras. 34 to 37). I. CONCLUSIONS AND LESSONS 92. The Bank contributed over US$430 million to the financing of the five projects audited. If the question is, whether it got its money's worth, the audit concludes that it did not mostly because of the disappointing outcome of 31 the US$2.5 billion lbistan development and the lack of progress on the institutional strengthening of the power sector in general and of TEK in particular. A brief review of the positive and negative aspects of the operations is likely to provide a more nuanced view. A. Accomplishments 93. The two generation projects, Elbistan and Karakaya, increased Turkey's g3neration capacity by some 2300 MW (2500 MW, when the full capacity of the Elbistan units will be available). In 1990, this represented about 15% of the country's installed capacity with a possible contribution of some 25% to the instantaneous peak in the national grid. The projects further included about one third of the 8500 km of 380 kV transmission lines operating in 1990. The generation facilities achieved one of Government's main objectives, i.e. to increase the use of domestic non-exportable resources for power generation. Elbistan has the potential to become one of the pillars of power generation in Turkey, the more so as it is situated in an environmentally less sensitive area than some other lignite-fired plants located further west and whose future is questionable. The successful completion of Karakays resulted in one of the last major additions to Turkey's hydro capacity in the Euphrates basin that was possible without substantially affecting the flow available to Syria and Iraq. B. Rfrawbacks 94. On the physical side, the main shortfall was that about 20 years after project start, 81bistan: (i) is not able to work at full capacity until all the turbogenerators are modified, which is presently expected to be the case in 1994, and (ii) is operating with lignite of adequate average but highly irregular quality, which results in improper operation of the power plant, loss of plant availability, reduced overall efficiency, and possible excessive adverse impact on the environment. A better mixing of the lignite is likely to go a long way towards eliminating the inadequacies of present plant operation. To that end the mine operator, at present TKI, needs appropriate incentives, in particular an adequate remuneration, which it presently does not get. 95. The Bank had supported the Karakaya project and agreed with Turkey that TEK would operate Keban and Karakaya in such a way as to guarantee an appropriate flow of the Euphrates at the border with Syria. Until 1991 the rule was for any practical purpose respected. The Bank expected that, when AtatUrk would be completed, an agreement between Turkey, Syria, and Iraq would have replaced the agrecment with the Bank. However, the tripartite agreement did not materialize, leaving the Bank in a somewhat uncomfortable position. 96. The instftutional improvements in the power sector and in TEK, in particular, which the Bank (though not Turkey) considered a major objective of the projects have been achieved only to a minimal degree, even though they were taken over in an ever adjusted form from project to project, even beyond those audited here. C. Bank Performance 97. As the reviewed projects were implemented over nearly two decades in a 32 changing, mostly difficult, environment, it is not surprising that the Bank's performance had its highs and lows, too. During the periods when Turkey strongly co-owned the policies sponsored by the Bank, especially the early 1980s, when the structural adjustment program was implemented, the Bank's approach to the power sector was quite forceful and successful. During most of the rest of the period, however, the Bank tended to react weakly to Turkey's many shortfalls from agreed action and to underestimate the significance of the lack of ownership by the government and the borrowers of the measures needed to improve sector performance in general and project implementation in particular. D. Ratings 98. The audit concurs with all the PCRs in rating the institutional progress achieved through the audited projects as negligible. As to the overall rating of the individual project, there is no doubt that the Elbistan project was unsatisfactory. The Karakaya project, in turn, despite the substantial shortfalls (also mostly in the institutional aspects) is rated as satisfactory, a judgement that the PCR does not contradict. In the transmission projects, the physical objectives were by and large achieved at reasonable cost but much later than anticipated at appraisal. The rating therefore mainly rests on the weight attached to the institutional improvements aimed at and achieved through the projects. As this weight was rather high and the achievements negligible, the audit rates both projects as unsatisfactory. Though the PCRs do not set forth an overall judgement to this effect, they show in the annexes discussing Government's and TEK's compliance with Bank covenants that with one exception the Turkish side fell short, and in most cases far short, of compliance. E. Lessons 99. The main lessons emerging from the operations discussed here are closely related to the issues covered in the recent report of the Bank's Portfolio Management Task Force under the need for: (i) project ownership by Government, Borrowers, and Beneficiaries, and (ii) realism and candidness in the Bank's evaluations and reporting - "Ownership" by the country authorities has to cover (i) the physical component, its execution and operation and (ii) the institutional objectives and the way to reach them. - Deeply held convictions such as those reflected in Turkey's extreme centralism and the refusal to relinquish, for the sake of efficiency, a reasonable measure of authority to executing organizations (the Government to the utility, the utility to site management and consultants, etc.) cannot be eliminated by a contract between local authorities and the Bank. Thus, the Bank has to take such foresee small steps providing pressure towards a change of the efficiency inhibiting attitudes. - Institutional reform cannot be divorced from the political and economic environment. Therefore, the Bank, in an environment hostile to some of its broad sector policies should either not lend, or, continuing the dialogue, frame the project in such a way that 33 vhatever institutional progress it envisages is realistically achievable and represents a significant step towards its long term goals. (see PCR TEK II, para. 8.4). - In a project like Elbistan, where sustainability of benefits is achievable but still in doubt at project completion, the Bank could consider a close follow-up of the project, in particular when the Bank's and its co-lenders' performance is under criticism. - A reasonably detailed plan for project management to which the Borrower is fully committed should be an integral part of a Bank project involving major physical facilities. - Supervision should always include interviews with people directly involved in the projects. Frequent site visits of complex and/or difficult projects with the corresponding interviews of site management, consultants, and operating staff are a must, as they often provide a valuable perspective, entirely different from that set forth at the higher levels of Government and the utility. 100. To conclude, the audit should like to refer to three observations made earlier in the present report. They refer to: - the gaps found in the project files (para. 9); - the usefulness of preparing the PCRe of particularly important projects (successful or unsuccessful) as early as possible after loan closing, irrespective of ongoing related operations (para. 11); and - the desirability of setting forth in PCRs cost comparisons in real terms, whenever this makes sense. TURKEY: BANK OPERATIONS IN THE POWER SECTOR Loan/ Project Name Beneficiary Loan Amount Month Month Month Month Credit Amount Canceled Ap- Si- Effec- Closed Number MUS$ MUS$ proved gued tive L.0063 Flood Control, Irrig. & Government 25.2 2.4 6/52 6152 9/52 6/61 Power (Seyhan Project) C.0034 Cukurova Power ;EAS 2.1 0.0 1/63 2/63 4163 12165 C.0059 Second Cukurova Power 9EAS 29.0 0.0 7/64 7/64 10/64 12/71 L.0568 Keban Transmission ETIBANK 24.0 0.0 10/68 10/68 1/69 8/74 L.0623 Third Cukurova Power QEAS 11.5 0.0 6/69 6/69 12/69 12/73 L.0763 Power Transmission TEK 24.0 0.0 6/71 6/71 10/71 6/76 L.0775 Fourth Cukurova Power 9EAS 7.0 0.0 6/71 6/71 10171 6175 L.0892 Istanbul Power Distri- 1ETT 14.0 0.0 4/73 5/73 9/73 10180 bution L.1023 Elbistan TRKITKI 148.0 0.0 6/74 6/74 6/76 6/83 L.1194 Second TEK Transmission TEK 56.0 0.6 11/75 6/76 4/78 12/81 L.1844 Karakays Hydropower DSI/TEK 120.0 0.0 5/80 5/80 10/80 12/87 L.2322 Third TEK Transmission TEK 163.0 65.2 6/83 6/83 11/83 9/89 L.2586 Fourth TEK Transmission TEK 142.0 0.0 6/85 6/85 10/85 12/91 L.2602 Power System Operation TEK 140.0 0.0 7/85 7/85 10/85 6/92 Assistance L.2650 Elbietan 04 TEK/TKI 10.0 0.1 2/86 2/86 5/86 12/88 L.2655 Kayraktope Hydropower TEK 200.0 0.0 2/86 3/86 3/86 12/94 L.2750 Sit Hydropower 9EAS 132.0 0.0 8/86 9/86 12/86 12/92 L.3345 TER Restructuring TEK 300.0 0.0 6/91 9/91 - 6/95 L.3476 Berke Hydropower 9EAS 270.0 0.0 5/92 6/92 * 12/97 The following operation is closely related to the above ones in the power sector L.2856 Energy Sector Adjust. Goverment 325.0 0.0 6/87 6/87 7/87 9/92 TURKEY: TEK Revenues and Costs per kWh Sold 280 260 - 240 - 220 - 200- 180 160 14 12 100 80 60 40 20 mms too 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 O Rev. + Fuel AOM COC