Brazil - Paulo Afonso Hydroelectric Project

Ls- WS flp L- 80 This report is restricted to use within the Bank. INTERNATIONAL BANK FOR RECONSTRUCTION AND DEVELOPMENT TECHNICAL REPORT OF THE PAULO AFONSO HYDROELECTRIC PROJECT IN BRAZIL May 24, 1950 Loan Department  PAULO AFONSO HYDROELECTRIC PROJECT - BRAZIL I. Purpose and Scope of this Report This report presents an analysis of the Paulo Afonso hydroelectric nroject wi,rhich h-s been presented by the Brazilian Government as a basis for a loan from the International Bank. The analyrsis is designed to determine the suitability of the project for this purpose. The report also includes the recommendations of the Engineering Staff concerning this project. The report is based on rather complete documentation received from t:e prospective borrower concerning the technical and buisiness features of the project except in the case or the secondary transmission sys'tem. It is based also on a suoplementary study of the power markcet to be served by the project which has been made for the 3Bnk by Mr. Tracy Martin of M14adigan-Hyland, 4ew York, and on recent discussio.'.s in WaslhinZto; with Colonel Carlos Berenhauser, negotiator for the comnoany. This report supersedes a memorandum of june 13th, 1949, prepared by Dr. '. W. Rembert and a ioreliminary renort of January 17, 1950, on the same subject. II. The Borrower The prospective borrower will be Oomnanhia Hidro Eletrica do Sao Francisco (CHESF) which will build and operate the system. This company is legally organIzed as a private corporation but is controlled by the Brazilian CGovernment and has, by law, a fifty-year concession for the exploitation of the power resources of the middle nortion of the San Francisco River. The area in which the company is authorized to operate has a radius of about 450 km. with Paulo Afonso Falls, the location of the power station, as its center. Organization of the company was authorized by law in October 1945 end the comnany was formed in March 1948. Present capital of the company amounts to Cr. 400 million ($21.4 million) wrhich is represented by 200,000 shares of common and 200,000 shares of pre- ferred stock, each class having a par value of Cr. 1,000. Both issues have been fully subscribed. The Brazilian Treasury has subscribed all of the common stock. The nreferred shares have been subscribed by state and munici- pal governments, autarchies, banks and private irnvestors. Detpils of the distribution of the present stockholdings are given in Table I. As a part of the SALTE plan, it is planned that the common issue will be increased by another 400,000 shares (Cr. 400 million), all of which will be subscribed by the Brazilian Treasury. This increase has been authorized by the Chamber of Deputies and the Senate and now requires the approval of the President of Brazil. (Since this Renort was nrepared, the President has approved the capital increase. See also p. 9 and Tnable I.)  -2- Cash payments of 10% are required on all stock at the time of subscrip- tion. In addition, the Brazilian Treasury has raid an additional 301 of its subscription in 1948 and l5-p in 1949, for a total -maid in of Cr. 110 million. The Treasury has been authorized to nay un the balance of its subscription of Cr. 90 millior in 1950. In the case of the preferred shares, payments are at the rate of 15~ per year on subscribed stock. Payments through 1949 have amounted to 25p or Cr. 52.5 million. Total payments up to December 31, 1949, on all issues, therefore, amount to Cr. 162.5 million. The bill authorizing the issue of an additional 400,000 shares auithorizes payments by the Treasury at the rnte of Or. 100 million per year for 4 rears startiiig in 1950. The tota1 raid in capitpl will, therefore, be Cr, 380 million by the end of 1950. The Federal Government guarantees a dividend of 6% on all preferred shares and till pay any deficit in earnings of the comnany required to meet such dividend payments. The government requires reimbursem!ient for such pay- ments from the individual shareholders to the extent to wh-ich dividends of more than 6% may- be raid. By law, voting nower is limited to the common stock and the Treasury must retain 515, of the cormon issue. It is permitted to sell the remaining 49% to other government agencies and institutions. Administratior, of the company- is in the hands of a President and three Directors elected by the shareholders for a four-year term. Re-election is permitted. The representative of the goverament at shareholders meetings is appointed by the President of the Republic. It is obvious, therefore, that for practical purposes the management of the com-any will be controlled by the President of the Republic. Reports of Bank rerresentatives indicate that the present management of the company is competent and is funietioning in an efficient manner. Eiowever, the situation in which the management will always be in the hands of rresidential appointees leaves such to be desired. III, Description of the Pro.et The Sao Francisco i9iver, with a total length of more than 3,000 km., is divided into three -rincioal zones. The upper river flows through molntainous country and, thence, throagh a high nlateau from which it drops through a series of rarids and falls to the coastal plain. Paulo Afonso Falls is a part of the drop from the plateau to the plain and renresents about 25% of the total change in elevation bet-ween these levels. The falls are about 150 miles up- stream from the mouth of the river. The project which has been oroposed as a basis for a loan contemplates the instellation of a generating capacity of 120,000 kv at Paulo Afonso Falls. Thls site, without reservoir storage, has an estimated potential hydroelectric ca;oacity of 540,000 kw. 'ith adequate upstream storage, the full potential of the middle Sao Francisco River is estimated at 900,000 kw, but no reservoir storage is contemnlated in the foreseeable future.  -3- A low concrete dam, built in twro sections and averaging about 10 meters in height, will be built immediately above the falls to divert the flow of the river to the power station. The total. length of this dami will be about 4,4 km. (2.7 miles). The dam will cross a large numiber of channels existing in the river at this point and will be -rovided with gates and spillways for si,illing water not required for power generation into the existing channels. The dam does not provide significant reservoir capacity and is designecd only for diversion of the flow of the river to the intake of the power station. Foundation conditions appear to be excellent for the dam and the rower- house, the rock formation in this area being massive granite. Borings have been made according to conventional practice and are considered adeqaate to assure good foundation conditions. The headworks are relatively simple. A separate intake and nenstoc17 is provided for each 6o,ooo 3,J unit. The penstoc:s are vertic.al tunnels cut in solid gra,nite with short horizontal runs at the bottom into the nowerlh,ox-Lse. The powerhouse will be entirely underground and designed in the first stage for the installation of three 60,000 kw units. The layout is such, however, that fur-ther excavation can provide for the ultimate installation of 9 generat- ing units. The tailrace consists of a short pressure tunnel and surge tank, the tunnel discharging into the canyon below th-e powerho-use. The gross hLead on the turbines has been established at 83.5 meters fo. normal operation. However, this inst311ation is peculiar in that the rise irL the river below the fails during flood neriods is greater tilan the rise abcove the falls and abou.t 305 of the time the station will be operating with a downstream water level considerably above the turbines. This particu.lar aspect of the design of the ple,nt has beer investigated by Professor L. Escandi of' Toulouse University in France. In this investigation, scale models were used and. the design now adooted has been proven to be practical. Generating equirment will consist of two 60,000 kw iunits, each consisting of an 33,000 H.P. Francis type vertical turbine, directly connected to a. generator rated. at 60,000 kw a.t a 905 power factor, The generating voltage will be 13,800 volts with a frequency of 60 cycles. Station auxiliary eq,uin- ment is corventional and has been fully specified. Two ban'cs of single -phase trsnsformers will be used in an o-atdoor SUD- station at the falls. Bach bank will consist of three 22,500 lzva units for stepping up from thae generator voltage of 13,800 volts to the transm2sslon voltage of 220,000 volts. Four smaller 3 phase tran-isforners are to be used for station and local service. The switchgear is conventionel and is fully so,ecified. Plans for the transLmission system in this project have been changed several times but can now be considered to be relatively firm so far as the n,,,rimary transmission lines are concerned. This rart of the systeem consists essentially of' two 220,000 V, single circuit lines originating at the step-up substation at, Paulo Afonso. One of these lines will riun to the east to .iecife via Paquevira, this line having a length of about 400 Im. The other will run to  -4- the south to Salvador via Itabaiana, this line having a length of approximately 440 km. Step-down substations rated at 40,000 kva each will be installed at Recife and Salvador for service to these cities. A 30,000 kva frequency changer will be installed at Recife to convert from 60 to 50 cycles, in order to tie into the existing distribution system. A 20,000 kwa synchronous con- denrser will be installed at Salvador, in order to improvre the powaer factor of the system. Switchgear and auxiliaries in these substations are conventional. Plans for the secondary transmission system are not yet firm. Furtler design and market stuclies are required but, as presently conceived, this sys- tern will require the installation of 10,000 kva step-down substations at Paquevira and Itabaiana and a 20,000 kva substation at ?ecife in addition to the one *rhich will serve the city. These stations will step down from 220,000 V to 66,ooo V and 33,000 V, the operpting voltages of the secondary lines. Amnroximately 505 km. of 66,0oo v lines are required, 90 km. of which will be double-circuit with steel towers. The remainder will be single-circuit using wood noles. Six step-down substations are required in the 66,ooo V system averaging about 5,000 kva each. The 33,000 V esYstem involves about 800 ici. of single-circuit line, all of which will use wood *ooles, This system will connect into about 40 towns and cities, each of which will require a step-down substation to connect into the distribution systems. The layout of the ertire transmission system as now contemolated is shown on the nap at the end of this report. ITo distribution lines or equipment are included in this rroject. It is contemnlated that CHIESF will sell power at the distribution voltage to the existing: nrivate and municinal companies o-perating in the co.mmunities wrhich rill be served., Furthermore, no rural electrification is provided as a part of this project. The engineering design and estimating on this nroject has been done en- tirely by Brazilian engineers except for the special studies made in France and some now under wayv by Westinghouse on the design of the transmission system. The project has been logically and efficiently developed and the work which has been presented to the Bank indicates th'at the CHESF has a competent tech- nica1 anti business staff. IV. Water Availability There is no -nroblem of water availability in the first stage of the Paulo Afonso development of 120,000 '.t. The water requirement for full load on thlis installation amounts to about 180 cu.m./second. Correlated stream- flow data available for this locatiorn since 1928 establish a minimum flow of the San Francisco hiver at Pa.ulo Afonso Zalls of 800 cu,m./second. The first stnge of the development will, therefore, use less than 25% of the minimum flow.  -5 Silting above the darm does not seem to be an im-oortant consideration in this project. Reservoir capacity is not a factor and if sufficient channels are held open the dam will function indefinitely for a simple diversion of the streinmf low. V. Estimated Cost of the Project The following estimate of the cost of the project is based, on the best iinfor-mation available as of April 1950. All costs have been converted into U.S. dollars using the rate of Cr. 18.72 = $1.00. all in $.000 Cost in Foreign Local Exchange Total Item Currency Cost Cost I. Paulo Afonso Power Station 1. Construction camp 908 - 908 2. Construction equipment 171 375 546 3. Roads and bridges 374 20 394 4. Dam 6,784 - 6,784 5. Cofferdam 535 265 800 6. Intake works 801 - 801 7. Underground works 1,522 - 1,522 8, Tunnel lining 30 42 72 9. Outdoor buildings 107 - 107 1.0. Gates and. screens 23 1,100 1,123 11. Turbines - 620 620 12. Generators - 1,370 1,370 13. Transformers - 412 412 14. Switchgear - 735 735 15. Accessorles & auxiliary equipment 283 100 383 16. Traveling c7anes & elevator - 310 310 17. Erection of eqa-irment _49 80 539 Sub-total for Paulo Afonso Stqtion 11.997 4 1 72..6 II. 220 KV TransmisSion Lines 1. Survey anad 'and rights 468 - 468 2, Towers 4,332 - 4,382 3. Insulators and hardware - 400 400 4. Cables and grouindwires - 2,547 2,547 5. Splices a.nd colmterpoise 2 153 155 6. Labor, food & temnrorary camps 9,250 - 3.950 Sub-total for Trtansmission Lines 8,802 3,100 11,902  -6- all in $1,000 Cost in Foreign local fLxchenge Total Item CurrcLcZ Cost Cost III. Recife and Salvador Substations 1. Land and buildings 117 - 117 2. Structures and svitchgear - 3lO 3940 3. Three-phase transformers - 862 862 4. Shunt reactors 80 80 5. Frequency changers 6715 675 6. Synchronous condenser - 180 i80 7. Auxiliaries and accessories 17 21 33 8. Erection of equipment 47 _ Sub-total for Substations 181 2,168 249 IV. Secondary Transmission System 1. 66 EV Lines and Substations (a) 505 Ion. lines 3,631 3,631 (b) Substations 657 2,C10 4,288 1,353 5,641 2. 33 KV Lines and Substations (a) 801 km. lines 1,523 - 1,523 (b) Substations 8 _ .373. 1,603 293 1,896 Sub-total for Secondary System 5,891 1,646 J2,S G. General Plant 1. Structures and im'orovements 801 - 801 2. Office furniture & equipment 155 - 155 3. Transportation equinoment 641 200 841 4. Stores eauipment 64 - 64 5. Iaboratory equir,ment 22 - 22 6. Tools and eaui ment 641 350 991 7. Com=unications equipment 69 - 69 8. iiiscellaneous equipment 320 150 470 9. Other nrorerty 91 - 91 10. G-en. expenses during construction1?2Z7 50 1,327 Sub-total for General Plant 4,081 750 4.831 VI. Continsencies, Freight & Overhead 1. Freight and insurance 3,860 205 4,o65 2. Overhead and. contingencies 31993 879 4,872 3. Eng4neering and suaervision 1,574 50 1,624 4. Interest during construction - 1,042 1,o42 Sub-total for Contingencies, Freight and Overhead 9,2 11,6 TOtTAL FSTIYATED COST 40.379 15,269 5_,648  -7- In the original estimates submitted by CPESF dated March 1949, the total cost of the -oroject was estimated at about $43.2 million, of which the foreign exchange requirement was nbout $15 million. This estimate has since been revised three times taking into account changes in -olans for the transmission system and changes in -prices on those items wl.hich were firm from a design standpoixit. On a com-oarable basis, the trend in estimated costs has been doun from the standopoint of foreign exchange requirements anid up substantial'y in the case of local currency expenditures. In the current estimate, account has been ta'len of changes resulting from discussions between the Bank and the Brazilian negotiators which took place in February and May nnd, also, revisions in the estimate which have been found necessary due to changes in designs. Furthermore, the complete secondary trans- mission system has been included in the nroject to be considered by the Bank for the first time. At present, the total estimated cost of the -oroject amounts to the eouivalent of $55.7 million, of which $15.3 million is estimated as the foreign exci;ange reouir-ement. This includes the cost of the secondary trens- mission system, estimated to cost the equivalent of $9.00 million, of which $1.80 million is required as foreign exchange. In the above estimate, all freight is assumed to be poaid in local currency which is in 'Line with instr-uctions which have been received by CHESF from the Brazilian Government. Payments for insurance, storage, U.S. inland freight and Sirlilar items, to tse entent of $205,000 will be made in foreign currency. Nqo interest during construction has been included on the estirmiated local currencyr exnenditures, since all local currency will be iorovided through pay- ments on stock subscriptions. The CM_SF estimate as submitted to the Bank in February included the cost of the trsnsmission towers as a foreign exchange item amounting to about $4.3 million. Originally, it was contemplated that these towers would be produced in Brazil, since the Brazilian steel industry can manufacture the types and sizes of steel required and adequate fabricating facilities are avail- able in the country. Recent argumrents 4ae been presented by CHESF in favor of purchasing the towers abroad, alf 'b"s'e'&'ro%taining lower prices, but this noint will not be definitely established until bids obtained in April have been opened and analyzed. In the estimate as presented above, therefore, the cost of the towers has been shown as a local currency item, If the towers are bought abroad, the Government of Brazil has undertaken to nrovide the foreign exchange from its own resources. Overhead and contingencies are now included in the estimate at about 8,5% of the total cost of the project, whlaich is a reasonable figare at this stage of development. The cost of the general plant estinated at the equivalent of $4.8 million is correctly included as a capital item but, on the basis of cur- rent estimates, the cost is considered to be high for a project of this type. The cost of the Paulo Afonso power plant and substation, including freight, contingencies and interest during construction, is now estimated at the equiva- lent of about $22,1 million and this estimate apnears to be reasonable. On  this basis, the cost is $184 per installed kw, which comperes !iith $300 to M6Oo for reservoir type plants. Previous estimates covering the 1)rimary transmission system w^nere of the order of' $45,000 per mile. This figure has been reduced to about t$30,000 noer mile in the current estimate but this cost is still consid.ered to be excessively high. It i s believed that the cost of the tren-smi.s