The Brazilian Electric System

This technical note has the purpose of starting studies concerning the Program for Improving the Energy Efficiency of Systems and Equipment in the State of Minas Gerais, to be discussed by the Program Commitees. The aim of the study is to present the present status of the Electric Sector in order to identify the possible improvements regarding the global efficiency, helping the implementation of the pertinent public policies. This Sector underwent a structural reformulation aiming at removing obstacles to development due to its state monopoly nature, according to the official interpretation.

In the present work the restructuring effect on electricity generation cost is discussed. The transmission and distribution costs will be treated in other works of the present series.

Among the large energy systems in operation in the State of Minas Gerais the electric one has a prominent place as the State is a net electricity exporter. As part of the National Interconnected System, the Minas Gerais State System has developed in a similar way as the former, so that the study regarding the national system, of which much detailed and updated data is available, can be applied to the State system without large differences.

Electricity was introduced in Brazil in the 19^th century through the concession of privileges for the exploration of public illumination given by the Emperor D. Pedro II to Thomas Edison. In 1930, the installed power in Brazil reached 350 MW in power plants that are now considered of low power, owned by industries and Municipal Administrations, most of them hydropower plants operating with “FIO D’ÁGUA” or with small reservoirs for daily regulation.

In 1939, in the Vargas Administration, the Waters and Energy National Council was created, a regulatory and inspection organ, later replaced by the Waters and Electric Energy National Department – DNAEE – under the authority of the Ministry of Mines and Energy. The first half of the 20^th century represents the phase that confirms electricity generation as an activity of economic and strategic importance for the country.

From the Second World War on, the Electric System has a large thrust with the construction of the first large power plant, namely Paulo Afonso I, with a power of 180 MW, followed by the Furnas, Três Marias and other power plants with large reservoirs for several years regulation (?). At the end of the sixties, the Interconnected Operation Coordination Group was created, giving birth to the interconnected national system.

During its 100 years of existence, the Brazilian Electric System, predominantly a hydraulic one (88% of power and 94% of energy generated in 1999), has generated 5,000 TWh, a quantity of energy that in terms of exclusively thermal generation corresponds to half of the Brazilian petroleum reserves, evaluated as 20 billion bbl. In this century, the System had different growth rates due either to hydrological regimes or to economic difficulties. The interpretation of the System’s historical path permits to discriminate the effects due to its interaction with other sectors (economic, petroleum, environment, etc.) and its inherent problems so that one can project with more assurance the future evolution, specially its participation in the generation park after the installation of natural gas thermal plants. In the following description we have used data from the National Energy Balance, elaborated since 1974 and containing historical series started in 1970, complemented by data from other sources whenever necessary.

The projection focuses mainly the installed power that, due to its inertia, determined by the relatively long maturation and implementation time, is a relatively “smooth” function of time, and the effective generation (firm (?) generation) or capacity factor, in order to examine the transients.

The data before 1970 are present in the records of DNAEE and in consultants and researchers¹ publications; graphic 1 below summarizes the used data.

The growth rate, with the classic bell-shaped form, suggests that the installed power tends to a maximum value, lower than the accounted and evaluated hydraulic power, 260 GW, showing the existence of a factor resisting the growth of the system. The detailed study, according to the methodology described in the Technical Note “Technological Forecasting” – SECT¹, has shown the limit of 66 GW (graphic 2). Other exercises of the same nature, using data from other time intervals and other grouping techniques furnished results that vary between 70 and 120 GW that show that results depend on the specific treatment method. However, all of them indicate the existence of a limit between ¼ and ½ of the recorded potential. It is interesting to observe that in other countries and regions with areas compared to that of Brazil, the hydraulic potential was not completely explored¹. In the Southeast Region there are few sites that could be used for electricity generation.

Graphic 2 - Installed hydroelectric power.
Adjustment y=65,5/(1+10⁵e^{-0,139 t}). The time scale is zeroed in 1900.

The projection methodology, based on the Theory of Systems, is phenomenological and therefore it does not identifies the nature of the phenomena that conditions the system evolution, which should be investigated by other methods. In the present case, these factors could be of economic (generation cost, for example), social ( preference for other methods of land and water use, territory reserves for indigenous population) or environmental nature (preservation of endemic diseases propagation).

Anyway, the importance of hydroelectric generation for Brazil justifies the efforts to clarify the question. In this first approach, the subject studied is the generation cost that could be indicating PROPICIANDO the gradual substitution of hydraulic by thermal generation, as it has happened in the other mentioned countries. However, the Brazilian territorial characteristics, a large area cut by a network of rivers with large flow, induce the consideration of other factors to be examined in other works.

Data concerning electricity generation are related with that of the installed power through the capacity factor, defined as the ratio of the quantity generated and the maximum possible quantity, assuming that the power plants would operate at maximum power during all the time.

Episodes of large generation drop are relatively rare, namely the drop in the fifties, caused by the extremely unfavorable hydrological regime, and the recent 2001 crisis, caused by the combination of the hydrological regime that was moderately unfavorable and the demand growth due to the growth of the economic activity, with restriction to investments in new undertakings as well as the transients of the implantation of the new management model in the Sector¹.

Since the generation crisis of the fifties, the system was conceived to operate with a capacity factor adequate to guaranty the electric energy supply, existing then a certain latitude for exploring the installed power that has been used to accommodate transients of supply and demand. Graphic 3 that follows illustrates the use of this reserve FOLGA.

It is observed that until the start of the Real Plan the capacity factor maintained itself below 0.56 what indicates that the exaggerated (ABUSO) use of this adjustment mechanism might have caused the recent electricity rationing.

The cost of electricity generation is composed of financial charges (investment and interests) and of plant operation and maintenance costs¹. The hydroelectric system is different from the other systems in what regards the high financial charges due to the large construction investments and to the high interest rates applied in Brazil. The operation and maintenance cost includes, besides personnel cost, the cost of peak demand complementation.

Based on the investments foreseen in the last power plant bidding (July/2001), the average unit investment was calculated as US$ 663/kW with dispersion of US$ 177/kW, therefore 27% uncertainty in the average calculation cost of hydroelectric generation for 2001. The calculation method is the one usually used in the Electric Sector. Assuming that the capacity factor is 0.55, interest rate is 12% annually, useful life is 30 years, construction time is 5 years and peak demand complementation is to be made by natural gas in simple cycle, one obtains the values shown below[1].

Marginal is defined, for a given generation status, as the generation cost of one additional unit (Mwh) of electric energy. In the present introductory study, the marginal cost was calculated based on data from the Technical Information n^o 065/85 – DEGE/DPVG of ELETROBRÁS, a forecasting study regarding the competitivity of steam coal plants vis-à-vis hydroelectric plants. The Technical Information shows the evolution of generation cost along the progress of hydraulic potential exploitation (accounted + estimated), using the parameters and costs adopted at the time. The implicit relevant information in this document is the cost evolution law, which is a function of variables not yet financially identified such as the distance between the exploitation site and the consuming centers, the geological constitution of soils, the cost of population displacement, the agricultural production value of the inundated area, etc…

The transfer of information to be used in the present study was made through the calculation of generation cost using data for the year 2001from the Technical Information and by the method used in it, while the fraction of occupied potential in this year was calculated using data of the installed power presented in the National Energy Balance/2000. The calculation results are shown in the graphic 4 below, where it can be seen that the marginal cost (per power plant) tends to US$ 51/Mwh in the year 2050.

The marginal cost per power plant is being considered as a reference for fixing the average tariff in the new structure of the Electric Sector. However, the implementation of the new management model of the Sector, which would change from a state monopoly to a free competition system, seems to be under reconsideration by the Federal Government[2]. The reference cost in the state model would be more appropriately the diluted cost resulting from the distribution of investment differences among the set of power plants or still the amortized diluted cost that takes into account the investment amortization of power plants that keep operating after the useful life which, in this case, has a merely accounting function. Graphic 5 that follows shows the evolution of diluted and amortized investment, retrieved from 1950 on with the help of data from the mentioned Technical Information.

Graphic 5 – Different investment concepts.

For the already amortized power plants, the generation cost is reduced to the peak complementation and the operation and maintenance costs, the notably low value of US$ 4/MWh. Taking into consideration the amortized generation cost in order to fix the tariff would be a form of paying to the tax-payer/consumer dividends for the long years of social investment.

Losses in the Interconnected System.

The interconnection of hydroelectric plants combines (CONCILIA) the hydrologic regimes of the different hydrographic basins, regulating the supply of the covered area’s demand. From the physical economy point of view, the interconnection permits to optimize the use of the potential energy stored in the plant’s reservoirs; on the other hand, energy relative losses in the interconnected system are larger than in the interconnected regional systems due to the long-distance load transfer.

Limiting the present discussion to the resistive losses, that are the more important, one can present the problem using the classical equations relative to electric power and losses through the Joule effect, familiar to the engineering students:

The electric power transmitted through a conductor with resistance r is P = i V, where V is the potential difference and i is the current intensity.

Therefore, when r and V are constant, the relative loss is proportional to the transferred power. In order to reduce loss, the long-distance transmission is made at high tension (V), and one also tries to reduce r by using conductors of lower specific resistance (W/m) and associating conductors in parallel or in a network.

Graphic 6 below shows the losses relative to the Electric System since 1970, using data from the National Energy Balance. The last expressive drop in the relative loss occurred around 1984, possibly due to the works prior to the start of operation of the Itaipú Plant. Since then, the relative loss has continuously grown from 11 to 15% of the electric energy offer (generation+ imports-exports) and the

O gráfico 6, abaixo mostra as perdas relativas no Sistema Elétrico desde 1970, com dados do Balanço Energético Nacional/1999. A última queda expressiva na perda relativa ocorreu por volta de l984, possivelmente devido a obras preliminares à entrada em operação da Usina de Itaipu. A partir de então, a perda relativa aumentou continuadamente, de cerca de 11 para 15% da oferta de energia elétrica (geração+importação-exportação) e o risco de interrupção do fornecimento aumentou, causando os chamados “apagões” de curta duração que culminaram no racionamento de 2001, resultado da diminuição de investimentos no Sistema Elétrico (usinas e linhas de transmissão) conjugada com o aumento da demanda, estimulada pelo Plano Real, e pela falta de água nos reservatórios existentes.

Embora muitos especialistas considerem o Plano Real como a principal causa dos problemas vividos pelo Setor Elétrico, é aparente que ele foi apenas a gota d` água, pois o gráfico mostra que a raiz dos mesmos deve ser localizada na crise financeira do início da década dos oitenta, claramente associada ao chamado “choque frio” do petróleo, quando os preços dos produtos de exportação brasileiros caíram de forma acentuada.[3]

O efeito da interligação sobre a economia de energia elétrica fica escondido pelas variações de demanda atribuídas às flutuações da produção econômica no período analisado. Uma leitura conjunta da evolução da eficácia de transmissão e da eficácia de geração (fator de capacidade) está mostrada no gráfico 7, abaixo, vendo-se que a diminuição da eficácia de transmissão praticamente compensou o aumento do fator de capacidade, de forma que a eficácia global ficou quase constante.

Gráfico 7 – Parâmetros do Sistema Hidroelétrico.

O custo da temeridade de forçar o sistema foi a queda de 7,7% na oferta em 2001, em relação a 2000. Na sistemática de cálculo de “custo do déficit”, baseada na Matriz Insumo-Produto, o custo do não-fornecimento de eletricidade é de cerca de 680 R$/MWh[4], o que permite calcular em cerca de R$ 18,3 bilhões, ou cerca de 1,2% do PIB, o prejuízo causado à economia brasileira pelo racionamento de 2001. Para manter a participação da hidroeletricidade na oferta e retornar ao fator de carga de segurança, estimado em 0,55, seria necessário ter 67.000 MW instalados em 2000, quando estavam instalados apenas 61.000 MW. Portanto, o déficit em potência instalada seria de cerca de 6.000 MW e o déficit em investimento na geração seria de R$ 10 bilhões, ou cerca de 55% do prejuízo calculado para 2001.

Quanto ao investimento em transmissão, o Plano Decenal 2001-2010 orçou-o em R$ 10,7 bilhões, ou cerca de US$ 4,6 bilhões, contra R$ 79 bilhões, ou US$ 34 bilhões em geração, estando incluídos naquele montante as interligações regionais.

[1] Generation cost in Brazil is one of the lowest in the world, due to the export of electric energy imbedded in electric-intensive products such as aluminum and ferroalloys.

[2] See, for example, the editorial “Correção de Rota” in PCH Notícias, year 4, nº 12

[3] Ver, por exemplo, “Brasil: o Crescimento Possível”, Carlos Feu Alvim et al. – Ed. Bertrand Brasil-1996.

[4] O custo atualizado do déficit consta no Relatório da Comissão de Análise do Sistema Hidrotérmico de Energia Elétrica.