No 21 -August/September 2000
|e&e No 22
Equivalent Energy Module
Energy for Transport Sector
FUEL CONSUMPTION IN BRAZIL
The fleet of light vehicles in Brazil is a fundamental element to evaluate fuel consumption and its emission. The evolution of the future fleet is important to the dynamic automobile sector and has large repercussion on the downstream and upstream of this sector. Among the repercussions downstream we can mention fuel consumption and consequently the emission of greenhouse effect gases. The existing and future fleets were estimated associated with a reference economic scenario. The consumption in gasoline equivalent energy was estimated as well. This study is part of a project sponsored by the Ministry of Science and Technology and the UNDP for the evaluation of emissions and is associated with a methodology for evaluating the National Energy Matrix and the Emission Matrix associated with the use of energy.
1 - Comments on the Methodology
The online periodical e&e has carried out a study on the fleet behavior for the Ministry of Mines and Energy and UNDP. The objective of the present work is to describe the methodology and the results of a module relative to the behavior of the means of transport to be used in the evaluation of energy consumption and emissions.
In the future this module will be integrated to other ones, an economical and a sectorial modules, that will permit to associate results from different scenarios to energy consumption and emission. For the moment it is possible to establish the relationship between economic growth data and energy consumption using these results in the projection of energy consumption and emission of greenhouse effect gases.
Therefore we aim at getting a model that is sufficiently flexible in order to study different alternatives in the energy area and its repercussions on emissions. We aim as well to find conservative measures or measures with predictable behavior that link energy use to economic activity.
The present work is part of a more general set under construction that intends to connect economic and energy scenarios and to calculate their consequences on emissions. The use of equivalent energy aims at supplying the tools to analyze these alternatives.
The introduction of the equivalent energy concept renders the energy/product ratio less dependent on the development level of the country as shown in ane&e previous article. This makes it possible to compare it with the situation in other countries and serve as a guide for the projections.
2 - Projection Based on Global Data.
Since we are presently dealing with individualized sectors, although comprehensive ones, it is necessary to adopt global consumption in order to be coherent with an economic scenario. However, the approach (which we hope to demonstrate) permits to evaluate the repercussion of specific measures on the global behavior.
Figure 1 shows the evolution of the Equivalent Energy/GNP parameter between 1970 and 1998, that seems adequate for extrapolation, which accumulated a variation inferior to 30 % in the period.
For the projection the program supplies a comparison among values obtained for other countries with different growth levels. Data are those mentioned in the previous article and they consider the GNP values, taking into account the purchasing power parity (PPP). The evaluated data for other countries in equivalent energy were calculated from the reference yield mentioned in the Brazilian BEU as described in the present work. Data were applied to economic sectors with a high aggregating level and also for grouped fuels.
Figure 1- Results concerning Equivalent
Energy/ GNP ratio for Brazil, showing that this is a "well behaved" ratio along
the period and it is favorable for extrapolation.
The projection of this parameter, already integrated with the macroeconomic module, is carried out with the assistance of help screens that guide the choice of parameters to be projected using historical data. In this case, data concerning the Equivalent Energy/GNP ratio for a set of countries whose income level were the most variable possible are also shown on the screen.
In Figure 2 the countries are shown in increasing order of the GNP (PPP) in US$90/inhabitant. In order to avoid conversion problems, values are given relative to the value of this parameter for Brazil in the comparison year (1996).
It can be noticed for all values that, in opposition to the final energy/GNP ratio, there is no systematic behavior where energy intensity is precisely larger in poorer countries. This is due to the overvaluation of traditional energy sources such as firewood, already previously discussed. On the other hand, there is a sharp difference among countries such as the United States and Canada, highly energy-intensive, and Western Europe and Japan that are less intensive in energy use. Brazil is in a relatively comfortable position relative to this analysis.
For global projection, the same screen presents, as shown in Figure 3, the average values for the parameter of some group of countries to be projected. They serve as reference values for maximum energy intensity to be reached in Brazil. In the present case, we have chosen the average value for European countries and Japan (1.15 of that of Brazil in 1996).
The program then projects, based on the GNP values calculated in the macroeconomic module, energy consumption in the next 20 years as shown in Figure 4. The growth rates for the reference periods are also summarized. The projections of the demand in equivalent energy and of the GNP are shown in Figure 4 together with historical values from 1970 on. The program indicates on the spreadsheet the growth rate for determined periods.
3 - Projection for the Transport Sector
Projections of energy consumption for the Transport Sector should be coupled with other modules. In this specific sector we have specially used macroeconomic data and data referring to the fleet.
In our methodology we have carried out a global projection for energy similarly with what we have done for the economy. In the economical case, the specific activities are limited to the overall growth of each sector will be handled as a participant in the total growth. In the case of energy projection this is just a first approximation since, for example, global parameters may be influenced by choices of the relative growth of the sectors involved.
The new sectorial scenario (or regional one, according to the approach) will be considered and the input parameters will be modified in the adjustment (for example, future energy intensity) through an iterative process. Based on this new general scenario and on the sectorial participation resulting from the first iterative step the global and sectorial data will be recalculated. A new iterative step will be carried out if it is considered necessary.
In the present stage only two sectors are under treatment: one related to offer (thermoelectric plants) and the other related to demand (road transport, object of the present work). The iterative step must wait for a global run
Due to its past behavior, treatment will be different for the collective and load transport sectors and for individual transport. In the former case, they will be treated as participants in the global consumption and this iteration is not necessary. For most sectors such as that of individual transport there is need of global iterative step that would affect the group. Meanwhile, the participation of this sector in the group could be used as a control item. This could substitute the iterative need previously mentioned.
4 - Projection of Collective and Load Transport
In Figure 5 we are considering the participation of collective and load transport and individual transport in the global energy consumption given in equivalent energy. Transport participation in global consumption will be directly considered in the projections. Individual transport will be used as control item.
Incidentally, it is interesting to notice the important participation of transport in energy use (about 1/3 of equivalent energy). On the other hand it should be emphasized the large participation of individual transport - that serves 20% of the population - and which is responsible for 2/3 of the fuel consumed in the transport of loads and passengers in all other modalities of transport (airline, railway, hydro-ways and road).
As a first hypothesis we are assuming that the participation of the predominant road transport will not be modified. As previously seen this participation grew in the past and reached about 91%. The consequences of a possible restructuring of the transport sector in terms of energy consumption and emission is one of the objectives of the present work and they will be discussed at the end when alternatives will be compared.
5 - Participation of Fuels in other Transport Modalities
The fuel participation in other modalities is examined in a preliminary form. Its complete analysis and that of other sectors will be carried out in the work relative to the Energy Matrix. Even in a preliminary form and assuming that road transport corresponds to 90% of fuel use, it is interesting to have a projection that is not far from reality in what concerns other modalities.
The variety of fuels used in other modalities of transport was reduced in the thirty years followed by the National Energy Balance. Since it seems that in the developed countries there are no notable changes in these modalities, we have consequently opted for a continuation in the observed trends in what regards the participation of energy sources in other modalities of transport.
The participation of several modalities was taken as the average of the last three years. Railway transport absorbed between 1996 and 1998 about 2.1% of the equivalent energy amount of load and passenger transport.
The historical participation in equivalent energy of the different energy sources in the Railway Sector is shown in Figure 6 aa well as the projection corresponding to 40% electricity participation in 2020. It is assumed that competition in the projection is between electricity and diesel oil. The electricity participation value is introduced in an exogenous manner.
Maintenance of the railway transport participation in the total may seem too conservative but it would represent inversion of decreasing trend in energy consumption as can be observed in Figure 7. Resuming may be coherent with some revitalization coming from privatization of the sector that was completely abandoned in the last years.
Figures 8 and 9 show the historical projection and that assumed for the use of equivalent energy in hydro-way and airway.
In order to have an idea of consumption evolution in road transport it is necessary to use a physical module that studies the vehicle fleet behavior that determines fuel consumption.
6 - Collective and Load Road Transport
Collective and load transport, as we have previously commented, was strongly influenced by the "dieselization" of the fleet caused by price differentiation of diesel relative to gasoline in the first petroleum shock. The favoring of diesel was still maintained with the establishment of PROALCOOL.
This differentiation made unviable any diesel substitution. The possible alternatives - alcohol and natural gas - are directed to gasoline, considering some incentives. We should point out that diesel motor is more efficient and favored reduction of CO2, as we will see later on. There are limitations in the refining structure that should be considered. This problem almost reached a critical point but the reduction of fuel alcohol from 1986 on eliminated the problem in the near future.
The consumption mentioned as load transport includes all diesel oil and the gasoline and anhydrous alcohol consumption fraction in the mixture used in trucks and buses. In load transport, the effect of remnant fleet was considered in the evaluation of Otto cycle contribution
The contribution of hydrated alcohol was evaluated and it is practically negligible. The use of natural gas fuel in the future will be considered only for passenger vehicles.
7 - Fleet versus Fuel Consumption
Statistics about fleet behavior are precarious and require evaluation based on vehicle sales and scraping. The previous effort helps when projections are made.
From information about the total fleet by category supplied by DENATRAN and the average age of the vehicles, it was possible to suggest differentiated scraping curves by type of vehicle. It was also possible to evaluate the fleet evolution, its composition and age.
The use of fleet data is useful only for energy use and emission projection when it is possible to couple fleet projection with energy consumption. The average consumption by type of vehicle is practically unknown in Brazil. There is also scarce information about consumption variation with time.
We can take advantage of the sudden variations in fleet composition (diesel substituting gasoline and alcohol substituting gasoline and vice versa) in order to obtain some of these parameters.
If we classify vehicles as heavy (trucks and buses) and light (cars and light commercials), as a first approximation, we can consider consumption of diesel cycle vehicles as that of heavy ones and Otto cycle vehicles consumption as that of light ones.
However, evaluation is not that simple because there are a considerable number of remnant commercial light vehicles using diesel and some gasoline load vehicles.
Furthermore, consumption by vehicle has varied with the economic activity that is not immediately followed by corresponding fleet variations. There is also a time variation of vehicles that influences the specific consumption.
Using equivalent energy introduces in principle some facilities since the expected consumption (in kep of equivalent energy by kilometer) of a similar vehicle using diesel, gasoline, alcohol or natural gas should be the same.
Nevertheless, with the introduction of lighter vehicle, the consumption of diesel vehicles remains much higher (about 20 tep/vehicle year) than that of Otto cycle (predominantly light) vehicles, namely 1.25 tep/year.
Changes in the Otto cycle fleet structure are shown in Figure 12 and those of diesel cycle, in Figure 13.
One can establish the correspondence between the consumption of an Otto cycle truck (of medium size) and that of light vehicles. We can imagine that a heavy vehicle consumes annually a fuel volume x times more than a light vehicle. If in the transition period we multiply by x the number of heavy vehicles we would have the fleet in "equivalent light vehicles". This was made for the Otto fleet (gasoline in the transition years). Using the value 9, we obtain the result shown in Figure 14 that indicates consumption by "equivalent light vehicle" approximately constant.
The value of 1.25 tep/year is representative of the consumption of a light vehicle. The consumption of a light truck would be 9 times higher. The consumption could be directly considered in gasoline (or natural gas). For the remainder fuels one should use the indicated equivalence (1 tep of diesel oil is equivalent to 1.57 tep of gasoline and 1 tep of hydrated alcohol is equivalent to 1.37 tep of gasoline).
Figure 14: Considering that a heavy vehicle is equivalent to 9 light ones we obtain a consumption curve by vehicle that better reproduces the expected variations due to the economic activity level.
Using this relationship it was possible to discriminate the consumption of heavy vehicles from that of light ones. We have found that the consumption in equivalent energy of load vehicles was about 5% in 1970. Based on this relationship of consumption we get the results shown in Figure 11. The values obtained from data of Figure 9 correspond to an overestimation of light vehicles consumption since it does not take into account the ascending path of the light fleet in the seventies, retarded by petroleum crisis at the start of the eighties.
It is also possible to extract from data concerning hydrated alcohol consumption the consumption evolution with the fleet's average age. This behavior is shown in Figure 15. There are naturally other factors that influence consumption, including the fleet composition along the years. In the extrapolations the consumption variation behavior was assumed to follow a straight line that was adjusted using alcohol data. A minimum value of 0.5 tep/year per vehicle was considered.
Figure 16 - Annual consumption of vehicle as a function of the average age for alcohol and gasoline light vehicles.
The relationship can be tested for gasoline vehicles whose average age oscillated in the mentioned decades due to the fleet's renewal or when it did not occur. In this case there were also changes in the fleet's composition. The result can be seen in Figure 16 for the almost 30 years of the series.
It can be historically observed the age decrease and consumption increase of gasoline vehicles when the fleet increase was large in the seventies. In the second half of this decade consumption per vehicle started to decrease (increase in gasoline price and shift of commercial light vehicles of larger tonnage to diesel oil) and this trend continued because vehicles became older due to reduction in sales (alcohol substitution and global drop of sales). In the nineties gasoline car and global sales recovered thus reducing the average age and increasing consumption per vehicle. A certain hysteresis is observed in the curve's behavior. The minimum value of 0.5 tep/year for extrapolation comes from what was observed in what concerns gasoline vehicles.
8 - Fleet Projection
It is expected that the number of cars per inhabitant will be a function or the per capita income. Other factors such as income distribution, vehicle price relative to salaries, financing conditions and fuel price should also be relevant.
Fleet per inhabitant and per capita income grow along time even though in Brazil per capita GNP has had some oscillations in the considered period. As the fleet has inertia larger than that of the GNP one cannot expect a perfect correlation along time. It seems also interesting to notice, besides the time path, the correlation of these two values in other countries for the same year.
For this purpose, given the notorious differences between the purchasing power and the per capita GNP values as well exchange rates, it is necessary to express them in PPP.
The World Bank publishes the per capita GNP values for the different countries while values for the fleet are published by ANFAVEA.
The per capita GNP values are in parity with purchasing power of 1995 American dollars. The GNP/inhabitant for Brazil of 5400 US$/inhabitant was corrected for the other years by the specific GNP deflator. This deflator and the population are supplied by IBGE. Values for Brazil and for some countries are shown in Figures 16, 17 and 18.
The participation of light vehicles in the Brazilian fleet grew dramatically in the period. For the future it was assumed that it would correspond to 93% of the total.
Figure 17: The participation of light vehicles - used mainly in the transport of individual passengers - increased significantly in the sixties and seventies.
9 - Light Fleet Projection and its Consumption
Figure 18 shows the evolution of the light total fleet by inhabitant verified and projected. The fitted data corresponding to the previous figure are shown by applying the straight-line coefficients to the data verified in the GNP/inhabitant. The expected percent growth for the total fleet as a function of GNP/inhabitant growth calculated in the macroeconomic module were extrapolated for the light fleet assuming that its participation in the total fleet is constant.
Figure 19: Evolution of the existing vehicles up to 1999 and its projection based on values for the fleet once the average consumption of the last years is maintained.
With the projected fleet and assuming that the consumption per vehicle is 1.25 tep of equivalent energy (in natural gas or gasoline) and considering the remnant fleet after 1999 and the expected scraping, one obtains the values shown in Figure 19. For the remnant fleet it was considered the vehicles' growing average age effect on consumption
10 - Fuel Consumption in Transport
Fleet consumption is considered in equivalent energy and the effective consumption depends on the sale distribution among the different types of vehicles, besides conjuncture factors. Conversion of old vehicles was not considered even though this is possible in situations when fuel prices are advantageous or when fuel is not available.
For an inertial situation of 1000 alcohol vehicles sold per year, growing but modest participation of natural gas reaching 1% of the total and maintaining the 1998 anhydrous alcohol and gasoline mixture rate, fuel consumption would be that indicated in Figure 20. This is only one of the possible scenarios and it will be used in the comparisons.
Figure 20:Projected historical and "inertial" participation of Otto cycle light vehicles in fuel consumption (used with priority for individual transport).
Adding these values to those already evaluated from hypothesis presented in item 4 of the present work one obtains the fuel consumption in the Transport Sector presented in Figure 21.
The values obtained here correspond to determined vehicle sales and economic scenarios. The methodology permits to analyze easily different economic scenarios and different forms of supplying equivalent energy.